==> picture [215 x 22] intentionally omitted <==

Single Technology Appraisal

Tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

Committee Papers

==> picture [215 x 22] intentionally omitted <==

NATIONAL INSTITUTE FOR HEALTH AND CARE EXCELLENCE

SINGLE TECHNOLOGY APPRAISAL

Tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

Contents:

1. Pre-Meeting Briefing

2. Company submission from Novartis

3. Clarification letters

• NICE request to the company for clarification on their submission

• Company response to NICE’s request for clarification

4. Patient group, professional group and NHS organisation submission from:  Bloodwise

• Leukaemia CARE

• Royal College of Pathologists - British Society for Haematology (joint submission)

5. Expert statements from:

• Professor David Marks – clinical expert, nominated by Novartis

• Mr Michael Brandon – patient expert nominated by Leukaemia CARE

• Clare Foreman – commissioning expert, nominated by NHS England

• Peter Clark – Cancer Drugs Fund clinical lead

6. Evidence Review Group report prepared by Centre for Reviews and Dissemination and Centre for Health Economics – York (updated following factual accuracy check)

7. Evidence Review Group report - addendum

8. Evidence Review Group report – factual accuracy check

9. Technical engagement document prepared by NICE technical team and committee chair

10. Technical engagement responses from:

• Novartis (company) including appendix

• NHS England

Any information supplied to NICE which has been marked as confidential, has been redacted. All personal information has also been redacted.

==> picture [413 x 293] intentionally omitted <==

1

National Institute for Health and Care Excellence Pre-meeting briefing 2

==> picture [413 x 293] intentionally omitted <==

National Institute for Health and Care Excellence Pre-meeting briefing 3

==> picture [413 x 293] intentionally omitted <==

==> picture [413 x 293] intentionally omitted <==

4

==> picture [413 x 293] intentionally omitted <==

5

==> picture [413 x 293] intentionally omitted <==

6

==> picture [413 x 293] intentionally omitted <==

No NICE guidelines in ALL. EMSO Guidelines:

Ann Oncol. 2016 Apr 7. pii: mdw025 Acute lymphoblastic leukaemia in adult patients: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Hoelzer D, Bassan R, Dombret H, Fielding A, Ribera JM, Buske C; ESMO Guidelines Committee.

Full ponatinib recommendation:

Ponatinib is recommended, within its marketing authorisation, as an option for treating Philadelphia-chromosome-positive acute lymphoblastic leukaemia in adults when:

the disease is resistant to dasatinib or

they cannot tolerate dasatinib and for whom subsequent treatment with imatinib is not clinically appropriate or

the T315I gene mutation is present.

7

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission table 2, p16-19

8

==> picture [413 x 293] intentionally omitted <==

CAR T-cell therapy:

Engineered Autologous Cell Therapy is a process by which a patient’s own T-cells are collected and genetically altered to recognise and target antigens expressed on the cell surface of specific malignancies. (Source: Kochenderfer JN, Dudley ME, Kassim SH, et al. Chemotherapy-refractory diffuse large B-cell lymphoma and indolent B-cell malignancies can be effectively treated with autologous T cells expressing an anti-CD19 chimeric antigen receptor. J Clin Oncol . 2015; 33(6):540-9.)

National Institute for Health and Care Excellence Pre-meeting briefing

9

==> picture [413 x 293] intentionally omitted <==

Source; company submission, table 2 page 18:

Step 1: The patient is admitted to hospital and their mononuclear cells are obtained by a process known as leukapheresis.

Step 2: The patient’s cells are then cryopreserved in the vapour phase of liquid nitrogen and shipped to a manufacturing facility in the US using a dedicated courier service.

Step 3: At the manufacturing facility, the mononuclear cells are thawed and enriched for T cells. The T cells are activated with antibody-coated beads and genetically transduced using a lentiviral vector (inactive virus) containing the anti-CD19 CAR transgene.

Step 4: The CAR-T-cells then undergo ex vivo expansion on antibody-coated beads

Step 5: The CAR-T-cells undergo formulation and a strict quality assessment before being released, cryopreserved and shipped. As an autologous (patient specific) product entering the EU, each individual therapy then requires a separate certification and batch release appropriate to the European regulations governing genetically modified advanced therapy medicinal products. Following certification, the product is sent back to the hospital where it can be stored in liquid nitrogen for up to 9 months until

National Institute for Health and Care Excellence Pre-meeting briefing

10

the treating centre / staff are ready to administer, and the patient is ready to receive treatment.

Step 6: The patient can receive bridging chemotherapy between leukapheresis and infusion at the discretion of the treating physician. Prior to tisagenlecleucel infusion the patient receives a preparative low dose lymphodepleting regimen. The CAR-T-cells are then thawed and reinfused into the patient as a one-time single-dose of tisagenlecleucel.

10

==> picture [413 x 293] intentionally omitted <==

11

==> picture [413 x 293] intentionally omitted <==

National Institute for Health and Care Excellence Pre-meeting briefing 12

==> picture [413 x 293] intentionally omitted <==

National Institute for Health and Care Excellence Pre-meeting briefing 13

==> picture [413 x 293] intentionally omitted <==

Source: statement from NHSE specialised commissioning

National Institute for Health and Care Excellence Pre-meeting briefing

14

==> picture [413 x 293] intentionally omitted <==

Source: statement from clinical lead for CDF

National Institute for Health and Care Excellence Pre-meeting briefing

15

==> picture [413 x 293] intentionally omitted <==

Source: statement from clinical lead for CDF

National Institute for Health and Care Excellence Pre-meeting briefing

16

==> picture [413 x 293] intentionally omitted <==

Source: Company submission, table 1, p12-15

17

==> picture [413 x 293] intentionally omitted <==

Source: Company submission, table 1, p12-15

18

==> picture [413 x 293] intentionally omitted <==

Sources: company submission, p24-25, figure 6, ERG report pages 28-29

Notes:

Company submission states: No NICE clinical guideline, ESMO guideline doesn’t specify a specific choice of treatment for relapsed/refractory ALL. ERG report notes that there are (unpublished) guidelines by the Childhood Leukaemia Clinicians Network (CLCN).

Company states in submission that

Approx 20% of patients experience relapse after first-line chemotherapy

Clofarabine is licensed for second relapse – company states feedback from clinicians suggest it is rarely used.

Pathway for Ph +ve ALL not shown. Patients treated with tyrosine kinase inhibitors. Patients had to have failed 2 lines of TKI to be eligible for tisagenlecleucel trials. ESMO guidelines for treatment of

19

Ph+ ALL are available at https://academic.oup.com/view-large/35557775

19

==> picture [413 x 293] intentionally omitted <==

20

==> picture [413 x 293] intentionally omitted <==

Source, company submission table 3, p28, and p66-69

Notes:

Matching indirect comparison used as scenario analysis (naïve comparison used in model

Alternative studies identified for comparators were used in scenario analyses

21

==> picture [413 x 293] intentionally omitted <==

Source: Company submission page 29-32 (study design diagram from figure 7)

Notes:

Other secondary endpoints in ELIANA include:

ORR (best overall response [BOR] of CR or CRi) with MRD negative bone marrow

DOR

RFS

Patient-reported outcomes

ORR determined by IRC assessment (defined as a BOR of either CR and CRi within 3 months of tisagenlecleucel administration) (US manufacturing facility only)

BOR of CR or CRi with MRD negative bone marrow (US manufacturing facility only)

Percentage of patients who achieve CR or CRi at Month 6 without allo-SCT between tisagenlecleucel infusion and Month 6 response

22

assessment

Percentage of patients who achieve CR or CRi and proceed to allo-SCT while in remission before Month 6 response assessment

Disease response at Day 28±4 days

Impact of Baseline tumour burden on response

Quality of response using MRD disease assessments before treatment and at Day 28±4 days after treatment

22

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: company submission p27, p33-37, table 4

23

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Adapted from company submission table 6, p40-41, ERG report p37

Ph+ve proportion from company response to clarification, p5-6

24

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission table 11, p47-48

Notes:

MRD negative status defined as MRD less than 0.01%. MRD status is a prognostic factor for relapse and MRD negative status a marker of deep remission.

As with baseline characteristics ERG notes that these results are on infused population only and not the entire intention-to-treat population

25

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission page 64-65, figure adapted from figure 21

26

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission page 64-65, figure adapted from figure 21. ERG report p50

27

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: company submission p52-53, figure 13

Notes:

EQ-5D-3L questionnaire used for patients aged 13 and above, EQ5D-3Y (adapted questionnaire for younger patients) used for patients between 8 and 12 years old

28

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p27, p73-74 (table 21), p83-84, ERG report p52

29

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

ERG report p37-38, 41-51

30

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source, ERG report p37-39

31

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p66, 70. ERG report p52-53

32

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p67-70. ERG report p53-55

33

==> picture [413 x 293] intentionally omitted <==

Source:. ERG report p54-55

34

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: ERG report p82, table 7, final limitation identified by company in factual accuracy check

35

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source, company submission p66-71 (table 20, figures 23 and 24). ERG report p56

Notes: Shaded regions represent 95% confidence intervals

Clofarabine efficacy used as a proxy for FLA-IDA salvage chemotherapy efficacy

Hazard ratios from indirect treatment comparisons are not used in the economic model. Scenario analysis conducted by the company using the matched-adjusted population had a relatively small effect on the incremental cost effectiveness ratios compared with the company base case (increases ICER vs salvage chemo by around £2,000, decreases ICER vs blinatumomab by £3,000).

36

==> picture [413 x 293] intentionally omitted <==

37

==> picture [413 x 293] intentionally omitted <==

38

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p93-98 (figures 25 and 26, table 26))

39

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p93-98 (figures 25 and 26, table 26). ERG report p69-70

40

==> picture [413 x 293] intentionally omitted <==

Source: Company submission page 106-109

41

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: company submission p109-110. figure 29 and table 30. ERG report p84

Notes:

ERG report states “uncertainty regarding the need to consolidate tisagenlecleucel-T response with SCT for patients to achieve longterm remission” as a further reason not to exclude lognormal and generalised gamma

42

==> picture [413 x 293] intentionally omitted <==

Source: company submission p114-118. figure 29 and table 30, ERG report p86

43

==> picture [413 x 293] intentionally omitted <==

Source: company submission p110-114. figure 31, ERG report p88

44

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission, figure 36, p119

45

==> picture [413 x 293] intentionally omitted <==

Source: ERG report p84-85, 88

46

==> picture [413 x 293] intentionally omitted <==

Source: ERG report, p120-121, table 19, company factual accuracy check response.

Figure not part of company submission or ERG report. Generated for illustrative purposes by NICE technical team from Markov trace in the company and ERG models (‘Trace-Blinatumomab’ sheet, column M).

47

==> picture [413 x 293] intentionally omitted <==

Source: ERG report, p127-129, table 23

48

==> picture [413 x 293] intentionally omitted <==

Source: ERG report, p127-129, table 23

Notes: As discussed in clinical section ERG notes that differences in baseline characteristic between Kuhlen lead to an underestimate in overall survival for FLA-IDA and therefore favour tisagenlecleucelT

49

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Figure not part of ERG report or company submission. Generated by NICE technical team from overall survival data in ‘Efficacy’ sheet of company and ERG models (truncated at 10 years).

50

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: company submission p122, figure 38 and table 40. ERG report p89-90

51

==> picture [413 x 293] intentionally omitted <==

Source: company submission p122-123. ERG report p90. p127128, 1p35-136.

52

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

53

==> picture [413 x 293] intentionally omitted <==

Source: company submission p128-129, table 43. ERG report p9293, p134

54

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p124-125, p127-128. ERG report p90-94

ERG preferred a Lower disutility applied from 3 – 12 months postSCT from Felder-Puig et al. Scenario analysis showed this change had minimal effect on the company’s base case ICERs .

55

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p131-137 (tables 44-46)

56

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p138-139 (table 47). ERG report, p97-98

57

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p138-139 (table 47). ERG report p104-106, p133.

58

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p138-139 (table 47). ERG report, p104-106, p129-130, p133

59

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

60

==> picture [413 x 293] intentionally omitted <==

Source: Erg report p132, p135

Notes:

NICE TA 450 (blinatumomab) section 4.16 states that “If the disease responded after 2 cycles but a suitable donor match was not immediately available, or if stem cell transplantation was not appropriate, they might have more than 2 cycles.”

61

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission, p154-155 table 59

National Institute for Health and Care Excellence Pre-meeting briefing

62

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p157-158, table 61 and 62

63

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: ERG report, p120-121, table 19, company factual accuracy check response

64

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission p163, table 64 and 6

65

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: company submission p163-166 (figures 42 and 45)

66

==> picture [413 x 293] intentionally omitted <==

67

==> picture [413 x 293] intentionally omitted <==

Source: Company submission, p171-176, tables 67-73

Notes: Alternative time points for application of standardised mortality ratios were also examined.

68

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: company submission table 71, page 175

Scenario analyses with discount rate of 1.5% are available in ERG addendum and the associated confidential appendix

69

==> picture [413 x 293] intentionally omitted <==

Source: ERG report p134-137, table 29

National Institute for Health and Care Excellence Pre-meeting briefing

70

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: ERG report p136, table 29

71

==> picture [413 x 293] intentionally omitted <==

ERG report p137-139, table 30.

Notes:

ERG consider rate of SCT after tis-T being 0% to be a plausible scenario in clinical practice. Use as bridge to SCT is not intended use of tis-T but is in line with use of other CAR-T therapies as bridge to SCT.

72

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission table 24, p87-88, ERG report p141

Figure not part of company submission or ERG report. Generated by NICE technical team for illustrative purposes from Markov trace in the company and ERG models (‘Trace-Blinatumomab’ and ‘TraceChemo’ sheets, column M).

73

CONFIDENTIAL

==> picture [413 x 293] intentionally omitted <==

Source: Company submission table 24, p87-88, ERG report p141

74

==> picture [413 x 293] intentionally omitted <==

75

==> picture [413 x 293] intentionally omitted <==

76

==> picture [413 x 293] intentionally omitted <==

77

==> picture [270 x 48] intentionally omitted <==

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167]

Thank you for agreeing to give us your comments and feedback as part of the technical engagement step to assist us in identifying the most plausible assumptions in the clinical and cost-effectiveness for this technology.

As a technical engagement stakeholder for this appraisal step, we highly appreciate your input, comment and ongoing support for this appraisal.

To help you give your views, please use this questionnaire. You do not have to answer every question. The text boxes will expand as you type. Please read the checklist for submitting comments at the end of this form. We cannot accept forms that are not filled in correctly.

Information on completing this technical engagement response

• Prior to completing this response table please see the technical engagement document which summarises the background, and submitted evidence for this appraisal. This will provide you with context and outline the questions below in greater detail for which we require your comments and feedback.

• Please do not embed documents (such as a PDF) in a submission because this may lead to the information being mislaid or make the submission unreadable

• We are committed to meeting the requirements of copyright legislation. If you intend to include journal articles in your submission you must have copyright clearance for these articles. We can accept journal articles in NICE Docs.

Please note that comments from the technical engagement will be collated and summarised as part of the committee pre-meeting briefing document, which will be made available to all stakeholders with a signed confidentiality agreement as part of the committee papers accompanying the post committee documentation (ACD or FAD) following the meeting on 22 August 2018

Deadline for comments 12pm Monday 13 August 2018 email: tacommc@nice.org.uk /NICE DOCS

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167]

1 of 7

Issue Date: August 2018

==> picture [270 x 48] intentionally omitted <==

About you

Questions for engagement

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

2 of 7

==> picture [270 x 48] intentionally omitted <==

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

3 of 7

==> picture [270 x 48] intentionally omitted <==

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

4 of 7

==> picture [270 x 48] intentionally omitted <==

Blinatumomab is an appropriate comparator for patients both <18 years and >18 years as it is a treatment option offered to patients at an equivalent point in the treatment pathway to where tisagenlecleucel is anticipated to be placed. This is supported by feedback from UK clinical experts who confirmed that blinatumomab, along with salvage chemotherapy (FLA-IDA), were the Is blinatumomab an appropriate comparator to current standards of care for paediatric and young adult patients with a second or later relapse of tisagenlecleucel-T for relapsed disease: ALL in both age groups.  for people younger than 18 years of age? Although blinatumomab may be offered to some patients following a first relapse, this does not  for people aged 18-25 years? preclude its use at a later treatment line. Therefore, blinatumomab remains a comparator to tisagenlecleucel at second or later relapse. Furthermore, feedback from UK clinical experts was that the use of one or two cycles of blinatumomab following a first relapse would be highly unlikely to result in a CD19-negative relapse, and therefore the use of blinatumomab at this stage would not preclude the use of tisagenlecleucel following a second relapse.

Question 4: Is it appropriate to use clofarabine as a proxy for the efficacy of FLA-IDA (that is salvage chemotherapy)?

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

5 of 7

==> picture [270 x 48] intentionally omitted <==

Stackelberg et al. 2011, Kantarjian et al. 2017 (both of which investigated a mixture of chemotherapy regimens) and Hijiya et al. 2011 (clofarabine, etoposide and cyclophosphamide). These scenarios were not associated with significant changes to the base case results of the economic model, with ICERs (with PAS) for tisagenlecleucel versus salvage chemotherapy of £20,890, £26,743 and £27,615, respectively, compared to the base case ICER of £25,404. Therefore, Novartis believe we have made every effort to produce as robust a comparison versus salvage chemotherapy as was possible, and have accompanied this with several scenarios to explore any potential uncertainty. Based on these results, the ICERs versus salvage chemotherapy when using the various sources of efficacy data consistently remained below a cost-effectiveness threshold of £30,000 per QALY gained. Finally, it is important to note that the ERG’s preferred source of efficacy data for salvage chemotherapy (Kuhlen et al. 2017) is associated with several limitations:  The proportion of patients with a previous allo-SCT was 100% in Kuhlen versus 54.2% in the tisagenlecleucel trials, hence it has only been conducted in a subset of the population potentially eligible for tisagenlecleucel and the Jeha 2006 study is therefore more inclusive of the overall population. However, 26.3% of patients received a further subsequent SCT. Second SCTs are extremely rare in the UK in this patient population, which raises questions about the representativeness of this study to UK practice. The high rate of SCT is biasing results against tisagenlecleucel as SCT is a curative option and therefore OS in this study is a clear overestimate.  Patients with extramedullary relapse (which are shown to have statistically significantly better outcomes for both OS and EFS) were excluded from the tisagenlecleucel trials, but represent 19.7% of the patient population in the Kuhlen et al. paper.  Finally, the HR for OS and EFS for T-ALL versus B-ALL (Table II in the Kuhlen study) was also not statistically significant and therefore it is misleading to state that there is a difference in outcomes between these groups. It is important to acknowledge when differences are not significant as this prevents misinterpretations of data. Non-significant data may result from chance rather than an actual observed difference. Taken together, Novartis believe these limitations discredit the Kuhlen study from being a more appropriate source of efficacy data than the Jeha 2006 study used within the company submission.

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

6 of 7

==> picture [270 x 48] intentionally omitted <==

Question 5: Long term usage and costs of IVIG treatment - real world experience Would people younger than 18 years of age require Consensus from several UK clinical experts consulted in response to this question was that a continued IVIG treatment and for how long? lifetime duration of IVIG is clinically implausible and the duration of IVIG treatment in patients <18 years of age would be aligned with the duration of B-cell aplasia; the estimate of 11.4 months used in the base case of our submission (which was based on the time to B-cell recovery), was therefore validated by UK clinical experts and is considered appropriate. Clinical experts also stated that when paediatric patients transition to the adult population (i.e. >18 years of age), they would be treated according to the adult protocol (see below). This involves the receipt of IVIG only if a patient has B-cell aplasia alongside a severe infection or severe cytomegalovirus (CMV) reactivation. This only occurs in approximately 20% adult patients, and patients would be treated with IVIG for 6-12 months only. Would people aged 18-25 years require continued As highlighted above, patients with r/r B-cell ALL aged 18–25 would not receive continued IVIG IVIG treatment and for how long? treatment following infusion with tisagenlecleucel. Feedback from UK clinical experts sought in response to this question was that patients will only receive treatment with IVIG if they have B-cell aplasia alongside a severe infection or severe CMV reactivation. This only occurs in approximately 20% adult patients, and patients would be treated with IVIG for 6-12 months only. It should be noted that this feedback was received after the company submission to NICE, and therefore the assumptions made within the company base case with regards to the administration of IVIG were conservative.

Thank you for your time.

Please log in to your NICE Docs account to upload your completed response form

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167]

7 of 7

Issue Date: August 2018

Frimley Business Park, Frimley, Camberley, Surrey, GU16 7SR, UK

==> picture [217 x 69] intentionally omitted <==

Appendix

Re: Tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia (ALL) in people aged up to 25 years [ID1167] – updated data cut-off from the ELIANA clinical trial

Latest data cut-off from the ELIANA clinical trial (13[th] Apr 2018)

The latest data cut-off from ELIANA (13[th] Apr 2018) is the second presented to the Committee, following the initial data cut-off (31[st] Dec 2017), which were presented in the initial company submission. A summary of the latest data cut-off (13[th] Apr 2018) compared to that presented in the company submission (31[st] Dec 2017) is presented below in Table 1. As is evident in Table 1, the results from the updated data cut-off (13[th] Apr 2018) are consistent with those from the data cut-off presented within the company submission (31[st] Dec 2017). These data highlight the robustness of the data presented initially, and continue to support the clinical benefits of tisagenlecleucel in paediatric and young adult patients with ALL and the assumptions upon which the economic analysis within the company submission were based.

All data from the 31[st] Dec 2017 and 13[th] Apr 2018 data cut-offs for the ELIANA clinical trial are academic in confidence and should remain confidential.

Table 1: Overview of clinical effectiveness results from the ELIANA clinical trial

© Novartis Pharmaceuticals Limited. All rights reserved.

1 of 4

aORR and DoR from the 31st Dec 2017 data cut-off for the ELIANA clinical trial were assessed in patients at least 3 months post-tisagenlecleucel infusion only (efficacy analysis set).[b] BOR is reported within 3 months for the ELIANA clinical trial.[c] No formal significance testing was conducted as the endpoint was met at the interim analysis. Nominal p-value is presented.[d] ‘Unknown’ is assigned in case the Baseline assessment of the response assessment is not done, incomplete, indeterminate, or not performed within the respective time frame. Abbreviations: BOR: best overall response; CI: confidence interval; CR: complete remission; CRi: CR with incomplete blood count recovery; DoR: duration of remission; MRD: minimum residual disease; NE: not estimable; NR: nonresponder/no remission; ORR: overall remission rate

Source: ELIANA CSR (31[st] Dec 2017);[1] ELIANA CSR (13[th] Apr 2018).[2]

Event-free survival

At the latest data cut-off (13[th] Apr 2018), in the full analysis set (FAS), xx of the xx patients (39.2%) per IRC review reported treatment failure, relapse or death due to any cause after remission prior to the data cut-off. The median EFS was xxxxxxxxxxx. The estimated event-free probability was xxxxx (95% CI: xxxxxxxxxx) at Month 6 and xxxxx (95% CI: xxxxxxxxxx) at Month 12 and Month 18. The Kaplan-Meier plot for EFS per IRC assessment is presented in Figure 1 .

Figure 1. Kaplan-Meier Plot for EFS censoring allo-SCT by IRC assessment in the ELIANA clinical trial (13[th] Apr 2018; FAS)

==> picture [463 x 151] intentionally omitted <==

Abbreviations: allo-SCT: allogeneic stem cell transplantation; CI: confidence interval; EFS: event-free survival; FAS: full analysis set; IRC: Independent Review Committee; NE: not estimable. Source: ELIANA CSR (13[th] Apr 2018).[2]

Overall survival

At the latest data cut-off (13[th] Apr 2018), in the FAS, xxxxx patients (xxxxx) died after tisagenlecleucel infusion and the estimated probability of survival was xxxxx (95% CI: xxxxxxxxxx) at Month 6, xxxxx (95% CI: xxxxxxxxxx) at Month 12 and xxxxx (95% CI: xxxxxxxxxx) at Month 18. Median OS was xxxxxxxxxxx. The Kaplan-Meier plot for OS is presented in Figure 2.

© Novartis Pharmaceuticals Limited. All rights reserved.

2 of 4

Figure 2: Kaplan-Meier plot for OS in the ELIANA clinical trial (13[th] Apr 2018; FAS)

==> picture [462 x 151] intentionally omitted <==

Abbreviations: CI: confidence interval; FAS: full analysis set; NE: not estimable; OS: overall survival. Source: ELIANA CSR (13[th] Apr 2018).[2]

© Novartis Pharmaceuticals Limited. All rights reserved.

3 of 4

References

1. Novartis Pharmaceuticals Ltd. ELIANA: A Phase II, single arm, multicenter trial to determine the efficacy and safety of CTL019 in pediatric patients with relapsed and refractory B-cell acute lymphoblastic leukemia. Clinical Study Report (31st December 2017 data cut-off). Data on File. 2017.

2. Novartis Pharmaceuticals Ltd. ELIANA: A Phase II, single arm, multicenter trial to determine the efficacy and safety of CTL019 in pediatric patients with relapsed and refractory B-cell acute lymphoblastic leukemia. Clinical Study Report (13th April 2018 data cut-off). Data on File. 2018.

© Novartis Pharmaceuticals Limited. All rights reserved.

4 of 4

==> picture [270 x 48] intentionally omitted <==

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167]

Thank you for agreeing to give us your comments and feedback as part of the technical engagement step to assist us in identifying the most plausible assumptions in the clinical and cost-effectiveness for this technology.

As a technical engagement stakeholder for this appraisal step, we highly appreciate your input, comment and ongoing support for this appraisal.

To help you give your views, please use this questionnaire. You do not have to answer every question. The text boxes will expand as you type. Please read the checklist for submitting comments at the end of this form. We cannot accept forms that are not filled in correctly.

Information on completing this technical engagement response

• Prior to completing this response table please see the technical engagement document which summarises the background, and submitted evidence for this appraisal. This will provide you with context and outline the questions below in greater detail for which we require your comments and feedback.

• Please do not embed documents (such as a PDF) in a submission because this may lead to the information being mislaid or make the submission unreadable

• We are committed to meeting the requirements of copyright legislation. If you intend to include journal articles in your submission you must have copyright clearance for these articles. We can accept journal articles in NICE Docs.

Please note that comments from the technical engagement will be collated and summarised as part of the committee pre-meeting briefing document, which will be made available to all stakeholders with a signed confidentiality agreement as part of the committee papers accompanying the post committee documentation (ACD or FAD) following the meeting on 22 August 2018

Deadline for comments 12pm Monday 13 August 2018 email: tacommc@nice.org.uk /NICE DOCS

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167]

1 of 8

Issue Date: August 2018

==> picture [270 x 48] intentionally omitted <==

About you

Questions for engagement

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

2 of 8

==> picture [270 x 48] intentionally omitted <==

Question 2: What is the treatment pathway for people younger than 18 years of age with primary refractory B-cell ALL?

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

3 of 8

==> picture [270 x 48] intentionally omitted <==

option in adults with relapsed/refractory Philadelphia chromosome negative ALL. Blinatumomab access has been extended to the non-adult ALL population by NHS England. The administration of blinatumomab is inconvenient and demanding for patients and clinical staff. Of note too is that approximately 22% of patients who relapse post blinatumomab do so with ALL which no longer expresses CD19. T-L CAR T cell therapy also targets CD19 and as a consequence there is therefore a concern that patients previously treated with blinatumomab and who then relapse may have clones of B cells which do not express CD19. In such circumstances, treatment with T-L would therefore not be expected to have any significant chance of curing the patient. The 3 T-L trials excluded patients previously treated with blinatumomab and thus there is no evidence of the efficacy of T-L in patients previously treated with blinatumomab. As a consequence of the biological plausibility of prior blinatumomab reducing the benefits of CAR T cell treatment directed at CD19 plus the exclusion of patients with prior blinatumomab exposure in the T-L trials, there will be wariness by haematologists in the use of blinatumomab if CAR T cell therapy with T-L is a potential salvage therapy later in the treatment pathway. Although combination chemotherapy and blinatumomab were commissioned options for relapsed/refractory ALL at the times of the NICE scope and the Novartis and ERG submissions, inotuzumab ozogamicin is now NICE-recommended in adults with relapsed/refractory ALL and funding has been extended to children by NHS England. Inotuzumab is directed against CD22 and thus does not carry any biological plausibility in potentially reducing the benefits of subsequent T-L therapy. In addition, it is a much more convenient drug to receive and deliver than blinatumomab. Hence it is likely to rapidly displace much use of blinatumomab and especially so in the relapsed/refractory ALL population in which CAR T cell therapy with T-L could be an option later in the treatment pathway. The administration costs of inotuzumab are much less than for blinatumomab and it is likely that drug procurement costs (based on the list prices of the two drugs) will also result in inotuzumab costing less than blinatumomab. As inotuzumab results in

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

4 of 8

==> picture [270 x 48] intentionally omitted <==

higher rates of CR and SCT than combination chemotherapy at 1[st] relapse, it is likely to become the treatment of choice at this place in the treatment pathway. NHS England notes that that at the time of the NICE scope, NICE stated that the comparators for T-L should be ‘established clinical management without T-L’. NICE did list the inotuzumab appraisal in the March 2018 scope as an appraisal in development. Although NHS England recognises that inotuzumab is not yet in August 2018 a part of ‘established clinical management’, it will become so in the very near future given its obvious practical advantages. For the much larger T-L eligible populations of relapsed post-SCT and in 2[nd] or further relapse that have not had SCT, the comparator options are currently the same treatments in these 2 places in the treatment pathway and depend on what has been used previously – if chemotherapy is used at 1[st] relapse, then the comparator at 2[nd] relapse would be blinatumomab (though shortly to be inotuzumab); if blinatumomab is used at 1[st] relapse (and shortly to be replaced by inotuzumab), then the comparator for 2[nd] relapse would be chemotherapy, the most commonly used regimen being FLA(G)-IDA or the ALLR3 protocol (which is similar to FLA-IDA although given for longer) or the combination of clofarabine, cyclophosphamide and etoposide. As has been stated above, treatment for 1[st] line relapse is likely to become inotuzumab in the near future and hence these same 2 options of blinatumomab and FLA(G)-IDA apply as comparators for T-L. There is little data on the use of blinatumomab after previous inotuzumab although there is no biologically plausible reason as to why blinatumomab should not be active. However this lack of evidence may affect the choice of treatment. Is blinatumomab an appropriate comparator to tisagenlecleucel-T for relapsed disease: See above  for people younger than 18 years of age?  for people aged 18-25 years?

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

5 of 8

==> picture [270 x 48] intentionally omitted <==

Question 4: Is it appropriate to use clofarabine as a proxy for the efficacy of FLA-IDA (that is salvage chemotherapy)?

Is clofarabine used in clinical practice in the NHS in Clofarabine is used but in combination chemotherapy ie not as monotherapy. England?

Is there any evidence to support the equivalence of FLA-IDA and clofarabine?

NHS England notes that Novartis used clofarabine monotherapy data as the proxy for combination chemotherapy with FLA-IDA. The clofarabine data was use of clofarabine monotherapy, not combination treatment (single-agent cytotoxic chemotherapy is very rarely used in acute leukaemia). The clofarabine monotherapy data was old, the first patient being treated in 2002 and the data cut off was in September 2004. Supportive care has changed much since 2002-2004 with significantly improved outcomes, including in the access to and the speed of access to SCT donors. This therefore means that the outcomes in the clofarabine monotherapy dataset are likely to be inferior to those of the combination FLA-IDA given in in a more contemporaneous time.

The indirect comparison of the pooled T-L studies with old clofarabine monotherapy data used as a proxy for FLA-IDA is inappropriate as there is more contemporaneous data for FLA-IDA (according to the ERG) with greater numbers of patients and longer median duration of follow-up. The heterogeneity of the data in any indirect comparisons of T-L with chemotherapy and also with blinatumomab is noteworthy.

Question 5: Long term usage and costs of IVIG treatment - real world experience

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

6 of 8

==> picture [270 x 48] intentionally omitted <==

responding to CAR-T cells developed B-cell aplasia and most of these 75 patients (exact number not specified) received IVIg. The probability of B cell recovery was *** at 12 months but NHS England notes that this figure did not change at ** months (albeit based on small numbers). From the point of view of a clinician looking after these highly immunosuppressed patients who all undergo conditioning chemotherapy prior to CAR-T cell treatment, there is bound to be considerable anxiety associated with merely observing a patient with no circulating B cells and Ig, as opposed to intervening with prophylactic Ig. Until there is solid longitudinal data on the infection risks associated with CAR-T cell associated agammaglobulinaemia, there is bound to be great and clinically justifiable pressure to use prophylactic Ig. Whilst it is not expected that every patient who receives a B-cell directed CAR-T cell treatment will require IVIg, it is predicted that the majority of responders to CAR-T cells will do so. For the purposes of costing IVIg requirements, long term follow up data on the proportion of patients who developed B-cell aplasia and low Igs as a consequence of CAR-T cell therapy is required. Until that is known, a pragmatic estimate of that up to 50% of responders will require IVIg (until B cell aplasia recovers) for a period of 12-24 months would not be unreasonable. As regards route of delivery, both intravenous Ig (IVIg) and subcutaneous Ig (SCIg) would be equally efficacious. Given that CAR-T cell therapy will be limited to major haematology centres, it is expected that the majority of those patients requiring Ig will be able to undergo training for home administration of SCIg. IVIg and SCIg are costly interventions and thus could have a significant impact on the mean cost of the supportive care that has to be wrapped around each patient who responds to T-L.

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167] Issue Date: August 2018

7 of 8

==> picture [270 x 48] intentionally omitted <==

Would people aged 18-25 years require continued IVIG treatment and for how long?

Thank you for your time.

Please log in to your NICE Docs account to upload your completed response form

Technical engagement response form

Tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged 3 to 25 years [ID1167]

Issue Date: August 2018

8 of 8

NATIONAL INSTITUTE FOR HEALTH AND CARE EXCELLENCE

Single technology appraisal

Tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

Document B Company evidence submission

==> picture [217 x 69] intentionally omitted <==

May 2018

Company evidence submission template for tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved. Page 1 of 185

Instructions for companies

This is the template for submission of evidence to the National Institute for Health and Care Excellence (NICE) as part of the single technology appraisal (STA) process. Please note that the information requirements for submissions are summarised in this template; full details of the requirements for pharmaceuticals and devices are in the user guide.

This submission must not be longer than 150 pages, excluding appendices and the pages covered by this template. If it is too long it will not be accepted.

Companies making evidence submissions to NICE should also refer to the NICE guide to the methods of technology appraisal and the NICE guide to the processes of technology appraisal.

Company evidence submission template for tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

Contents

Instructions for companies ............................................................................................................. 2 Contents ........................................................................................................................................ 3 List of tables .................................................................................................................................. 5 List of figures ................................................................................................................................. 7 List of abbreviations ....................................................................................................................... 9 B.1 Decision problem, description of the technology and clinical care pathway .......................... 12 Decision problem .............................................................................................................. 12 Description of the technology being appraised .................................................................. 16 Health condition and position of the technology in the treatment pathway ........................ 20 Disease overview ....................................................................................................... 21 Clinical pathway of care ............................................................................................. 23 Equality considerations ..................................................................................................... 25 B.2 Clinical effectiveness ............................................................................................................. 26 Identification and selection of relevant studies .................................................................. 27 List of relevant clinical effectiveness evidence .................................................................. 27 Summary of methodology of the relevant clinical effectiveness evidence ......................... 29 Trial design ................................................................................................................ 29 Trial methodology ...................................................................................................... 32 Baseline characteristics ............................................................................................. 39 Statistical analysis and definition of study groups in the relevant clinical effectiveness evidence ..................................................................................................................................... 41 Quality assessment of the relevant clinical effectiveness evidence .................................. 45 Clinical effectiveness results of the relevant trials ............................................................. 46 Clinical effectiveness results overview ....................................................................... 47 ELIANA ...................................................................................................................... 48 ENSIGN ..................................................................................................................... 53 B2101J ....................................................................................................................... 57 Subgroup analysis ............................................................................................................. 60 Meta-analysis .................................................................................................................... 61 Indirect and mixed treatment comparisons........................................................................ 65 Adverse reactions ........................................................................................................... 70 Treatment duration and dosage ............................................................................... 71 Safety analysis in the relevant clinical trials ............................................................. 72 Ongoing studies .............................................................................................................. 85 Innovation ........................................................................................................................ 85 Interpretation of clinical effectiveness and safety evidence ............................................. 86 Principal findings from the clinical evidence base .................................................... 86 End-of-life criteria ............................................................................................................ 88 B.3 Cost-effectiveness ................................................................................................................. 90 Published cost-effectiveness studies ................................................................................ 91 Economic analysis ............................................................................................................ 94 Patient population ...................................................................................................... 94 Model structure .......................................................................................................... 94 Intervention technology and comparators ................................................................ 100 Clinical parameters and variables ................................................................................... 104 Baseline characteristics ........................................................................................... 104 Clinical efficacy inputs .............................................................................................. 105

Company evidence submission template for tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

Survival inputs and assumptions .............................................................................. 106 Adverse events ........................................................................................................ 124 Measurement and valuation of health effects .................................................................. 126 Health-related quality-of-life data from clinical trials ................................................. 126 Mapping ................................................................................................................... 127 Health-related quality-of-life studies ......................................................................... 127 Adverse reactions .................................................................................................... 127 Health-related quality-of-life data used in the cost-effectiveness analysis ............... 128 Cost and healthcare resource use identification, measurement and valuation ............... 131 Intervention and comparators’ costs and resource use ............................................ 132 Health-state unit costs and resource use ................................................................. 145 Adverse reaction unit costs and resource use ......................................................... 149 Miscellaneous unit costs and resource use .............................................................. 155 Summary of base-case analysis inputs and assumptions ............................................... 155 Summary of base case analysis inputs .................................................................... 155 Assumptions ............................................................................................................ 156 Base case results ............................................................................................................ 158 Base case incremental cost-effectiveness analysis results ...................................... 158 Sensitivity analyses ......................................................................................................... 159 Probabilistic sensitivity analysis ............................................................................... 159 Deterministic sensitivity analysis .............................................................................. 166 Scenario analyses .................................................................................................... 171 Summary of sensitivity analyses results ................................................................... 176 Subgroup analysis ........................................................................................................... 176 Validation ...................................................................................................................... 176 Validation of cost-effectiveness analysis ................................................................ 176 Interpretation and conclusions of economic evidence ................................................... 177 References ................................................................................................................................ 179

Company evidence submission template for tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

List of tables

Table 1: The decision problem ....................................................................................................... 12 Table 2: Technology being appraised ............................................................................................. 16 Table 3: Clinical effectiveness evidence ......................................................................................... 27 Table 4: Summary of methodology of studies................................................................................. 33 Table 5: Outcome definitions in ELIANA, ENSIGN and B2101J ..................................................... 38 Table 6: Baseline characteristics (full analysis set) ........................................................................ 40 Table 7: Analysis set definitions ..................................................................................................... 42 Table 8: Trial populations used for the analysis of outcomes of relevant clinical trials ................... 42 Table 9: Statistical methods for the primary analysis of relevant clinical trials ................................ 43 Table 10: Overview of the quality assessment of ELIANA, ENSIGN and B2101J based on the GRACE checklist ............................................................................................................................ 45 Table 11: Summary of the clinical effectiveness results in ELIANA, ENSIGN and B2101J ............ 47 Table 12: Summary of IRC-assessed ORR within three months post-tisagenlecleucel infusion in ELIANA (full analysis set) ............................................................................................................... 49 Table 13: Summary of IRC-assessed bone marrow MRD status within three months posttisagenlecleucel infusion in ELIANA (full analysis set) .................................................................... 50 Table 14: Summary of IRC-assessed ORR within six months post-tisagenlecleucel infusion in ENSIGN (efficacy analysis set) ....................................................................................................... 54 Table 15: Summary of IRC-assessed bone marrow MRD status within six months posttisagenlecleucel infusion in ENSIGN (efficacy analysis set) ........................................................... 54 Table 16: Summary of ORR at Day 28 in B2101J .......................................................................... 58 Table 17: Summary of bone marrow MRD status in B2101J (full analysis set) .............................. 58 Table 18: Key patient baseline characteristics from the pooled analysis ........................................ 62 Table 19: Clinical evidence identified to be used as a proxy for the efficacy of FLA-IDA ............... 67 Table 20: Overall survival hazard ratios ......................................................................................... 69 Table 21: Overall summary of AEs in ELIANA, ENSIGN and B2101J (safety set) ......................... 72 Table 22: Summary of AEs reported in ≥10% of patients post-tisagenlecleucel infusion, regardless of study drug relationship (safety set) ............................................................................................. 75 Table 23: Summary of SAEs reported in at least two patients post-tisagenlecleucel infusion, regardless of study drug relationship (safety set) ........................................................................... 81 Table 24: End-of-life criteria............................................................................................................ 88 Table 25: Summary list of published cost-effectiveness studies ..................................................... 92 Table 26: Features of the economic analysis ................................................................................. 98 Table 27: Proportion of patients receiving subsequent allo-SCT in the model ............................. 104 Table 28: Patient baseline characteristics of the base case economic analysis ........................... 104 Table 29: Summary of goodness-of-fit data for tisagenlecleucel overall survival – standard parametric and spline models ....................................................................................................... 109 Table 30: Summary of goodness-of-fit data for tisagenlecleucel overall survival – mixture cure models .......................................................................................................................................... 110 Table 31: Summary of goodness-of-fit data for salvage chemotherapy (FLA-IDA) overall survival – standard parametric and spline models ........................................................................................ 111 Table 32: Summary of goodness-of-fit data for salvage chemotherapy (FLA-IDA) overall survival – mixture cure models ..................................................................................................................... 112 Table 33: Summary of survival projections for salvage chemotherapy (FLA-IDA) survival models ..................................................................................................................................................... 113 Table 34: Summary of goodness-of-fit data for blinatumomab overall survival – standard parametric and spline models ....................................................................................................... 114 Table 35: Summary of goodness-of-fit data for blinatumomab overall survival – mixture cure models .......................................................................................................................................... 117 Table 36: Assessment of the plausibility of the exponential, Weibull and Gompertz mixture cure models to model blinatumomab OS .............................................................................................. 118 Table 37: Summary of OS projections for blinatumomab survival models .................................... 119 Table 38: Long-term survival input sources .................................................................................. 120

Company evidence submission template for tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

Table 39: Summary of goodness-of-fit data for tisagenlecleucel event-free survival –standard parametric and spline models ....................................................................................................... 122 Table 40: Summary of goodness-of-fit data for tisagenlecleucel event-free survival – mixture cure models .......................................................................................................................................... 122 Table 41: Incidence of Grade 3 or 4 adverse events included in the model ................................. 125 Table 42: Descriptive statistics on EQ-5D utility values in ELIANA trial ........................................ 127 Table 43: Utility values used within the economic model .............................................................. 130 Table 44: Tisagenlecleucel pre-treatment costs (drug/procedure costs) ...................................... 135 Table 45: Tisagenlecleucel pre-treatment costs (outpatient administration costs)........................ 137 Table 46: Tisagenlecleucel pre-treatment costs (hospitalisation costs) ........................................ 137 Table 47: Infusion costs (and other hospitalisation costs) with tisagenlecleucel .......................... 139 Table 48: Salvage chemotherapy (FLA-IDA) drug costs .............................................................. 142 Table 49: Blinatumomab drug costs for the paediatric dose ......................................................... 142 Table 50: Blinatumomab drug costs for the adult dose ................................................................. 143 Table 51: Subsequent allo-SCT costs .......................................................................................... 144 Table 52: Subsequent allo-SCT follow-up cost breakdown .......................................................... 144 Table 53: Follow-up schedule and unit cost inputs for patients in the EFS health state ............... 146 Table 54: Follow-up schedule and unit cost inputs for patients in the PD health state ................. 148 Table 55: Follow-up cost inputs summary (monthly cost by treatment) ........................................ 148 Table 56: Detailed resource use inputs for CRS cost estimation .................................................. 150 Table 57: Associated AE costs for B-cell aplasia .......................................................................... 150 Table 58: Rates and unit costs of the AEs included in the economic model ................................. 151 Table 59: Summary of variables applied in the economic model .................................................. 155 Table 60: List of assumptions for the base case analysis ............................................................. 156 Table 61: Deterministic base case results (tisagenlecleucel list price) ......................................... 158 Table 62: Deterministic base-case results (tisagenlecleucel PAS price) ...................................... 159 Table 63: PSA inputs .................................................................................................................... 160 Table 64: Probabilistic results (tisagenlecleucel list price) ............................................................ 164 Table 65: Probabilistic results (tisagenlecleucel PAS price: xxx discount) ................................... 164 Table 66: DSA inputs .................................................................................................................... 167 Table 67: Cure model approach scenarios ................................................................................... 171 Table 68: Standard parametric survival model (plus ALL-adjusted mortality) scenarios ............... 171 Table 69: Alternative efficacy input scenarios............................................................................... 173 Table 70: Utility values scenarios ................................................................................................. 174 Table 71: Time horizon and discounting scenarios ....................................................................... 174 Table 72: Cost scenarios .............................................................................................................. 175 Table 73: Decision tree scenarios ................................................................................................ 175

Company evidence submission template for tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

List of figures

Figure 1: Domains of the chimeric antigen receptor construct of tisagenlecleucel ......................... 16 Figure 2: Cytotoxic mechanism of CAR-T therapy .......................................................................... 17 Figure 3: Summary of the administration process for tisagenlecleucel ........................................... 18 Figure 4: Tisagenlecleucel infusion process ................................................................................... 19 Figure 5: Average number of new cases of ALL per year and age-specific incidence rates in the UK (2013-2015) .............................................................................................................................. 23 Figure 6: Treatment pathway for ALL in the UK with the potential positioning of tisagenlecleucel . 24 Figure 7: ELIANA trial design ......................................................................................................... 30 Figure 8: ENSIGN trial design ........................................................................................................ 31 Figure 9: B2101J trial design .......................................................................................................... 32 Figure 10: Kaplan-Meier plot for IRC-assessed DoR (censoring for allo-SCT) in ELIANA (full analysis set)[a] .................................................................................................................................. 51 Figure 11: Kaplan-Meier plot for IRC-assessed EFS (censoring for allo-SCT) in ELIANA (full analysis set)[a] .................................................................................................................................. 51 Figure 12: Kaplan-Meier plot for OS in ELIANA (full analysis set)[a] ................................................. 52 Figure 13: Summary of PedsQL and EQ VAS scores in ELIANA (patients ≥8 years old achieving CR/CRi) .......................................................................................................................................... 53 Figure 14: Kaplan-Meier plot for IRC-assessed DoR (censoring for allo-SCT) in ENSIGN (efficacy analysis set)[a] .................................................................................................................................. 55 Figure 15: Kaplan-Meier plot for EFS (censoring for allo-SCT) in ENSIGN (full analysis set)[a] ....... 56 Figure 16: Kaplan-Meier plot for OS in ENSIGN (full analysis set)[a] ................................................ 57 Figure 17: Kaplan-Meier plot for DoR in B2101J (full analysis set)[a] ............................................... 59 Figure 18: Kaplan-Meier plot for EFS in B2101J (full analysis set)[a] ............................................... 60 Figure 19: Kaplan-Meier plot for OS in B2101J (full analysis set)[a] ................................................. 60 Figure 20: ORR within 3 months post-tisagenlecleucel infusion by IRC assessment. Forest plot for subgroups from ELIANA (full analysis set) ..................................................................................... 61 Figure 21: Kaplan-Meier curves for EFS in ELIANA, ENSIGN and B2101J and the pooled analysis ....................................................................................................................................................... 64 Figure 22: OS Kaplan-Meier curves for ELIANA, ENSIGN and B2101J and the pooled analysis .. 65 Figure 23: Overall survival for tisagenlecleucel versus blinatumomab ........................................... 70 Figure 24: Overall survival for tisagenlecleucel versus salvage chemotherapy .............................. 70 Figure 25: Decision tree structure for tisagenlecleucel cohort ........................................................ 95 Figure 26: Partitioned survival modelling approach ........................................................................ 95 Figure 27: Summary of the manufacturing and administration process for tisagenlecleucel ........ 100 Figure 28: Extrapolation of tisagenlecleucel overall survival – standard parametric and spline models .......................................................................................................................................... 109 Figure 29: Extrapolation of tisagenlecleucel overall survival – mixture cure models .................... 110 Figure 30: Extrapolation of salvage chemotherapy (FLA-IDA) overall survival – standard parametric and spline models ....................................................................................................... 112 Figure 31: Extrapolation of salvage chemotherapy (FLA-IDA) overall survival – standard parametric and spline models (top five curves according to AIC) ................................................. 112 Figure 32: Extrapolation of salvage chemotherapy (FLA-IDA) overall survival – mixture cure models .......................................................................................................................................... 113 Figure 33: Extrapolation of blinatumomab overall survival – standard parametric and spline models ..................................................................................................................................................... 115 Figure 34: Extrapolation of blinatumomab overall survival – standard parametric and spline models (top five curves according to AIC) ................................................................................................. 115 Figure 35: Extrapolation of blinatumomab overall survival – mixture cure models ....................... 117 Figure 36: Base case OS extrapolations ...................................................................................... 120 Figure 37: Extrapolation of tisagenlecleucel event-free survival – standard parametric and spline models .......................................................................................................................................... 122 Figure 38: Extrapolation of tisagenlecleucel event-free survival – mixture cure models ............... 123 Figure 39: Base case EFS extrapolations .................................................................................... 124

Company evidence submission template for tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

Figure 40: Cost-effectiveness plane: tisagenlecleucel (list price) versus salvage chemotherapy (FLA-IDA) ..................................................................................................................................... 164 Figure 41: Cost-effectiveness plane: tisagenlecleucel (list price) versus blinatumomab .............. 164 Figure 42: Cost-effectiveness plane: tisagenlecleucel (PAS price: xxx discount) versus salvage chemotherapy (FLA-IDA) .............................................................................................................. 165 Figure 43: Cost-effectiveness plane: tisagenlecleucel (PAS price: xxx discount) versus blinatumomab ............................................................................................................................... 165 Figure 44: Cost-effectiveness acceptability curve for all comparators: tisagenlecleucel (list price) ..................................................................................................................................................... 166 Figure 45: Cost-effectiveness acceptability curve for all comparators: tisagenlecleucel (PAS price: xxx) ............................................................................................................................................... 166 Figure 46: Tornado diagram of the twenty most influential parameters from the DSA: tisagenlecleucel (list price) versus salvage chemotherapy (FLA-IDA) .......................................... 169 Figure 47: Tornado diagram of the twenty most influential parameters from the DSA: tisagenlecleucel (list price) versus blinatumomab ......................................................................... 169 Figure 48: Tornado diagram of the twenty most influential parameters from the DSA: tisagenlecleucel (PAS price) versus salvage chemotherapy (FLA-IDA) ....................................... 170 Figure 49: Tornado diagram of the twenty most influential parameters from the DSA: tisagenlecleucel (PAS price) versus blinatumomab ...................................................................... 170

Company evidence submission template for tisagenlecleucel-T for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved. Page 8 of 185

List of abbreviations

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

B.1 Decision problem, description of the technology and clinical care pathway

==> picture [34 x 11] intentionally omitted <==

Decision problem

This submission covers the full anticipated marketing authorisation for the technology tisagenlecleucel (Kymriah™) for the treatment of paediatric and young adult patients up to 25 years of age with B-cell acute lymphoblastic leukaemia (ALL) that is refractory, in relapse post-transplant, or in second or later relapse, hereafter referred to as relapsed/refractory (r/r) B-cell ALL.

The decision problem addressed within this submission is consistent with the NICE final scope for this appraisal as outlined in Table 1.

Table 1: The decision problem

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

==> picture [85 x 38] intentionally omitted <==

or comparator technologies will be taken into account

Abbreviations: ALL: acute lymphoblastic leukaemia; allo-SCT: allogeneic stem cell transplantation; EQ-5D-3L: EuroQoL 5-dimensions 3-levels; FLA-IDA: fludarabine, cytarabine, idarubicin; NICE: National Institute for Health and Care Excellence; NHS: National Health Service; PAS: Patient access scheme; PedsQL: Paediatric Quality of Life questionnaire; Ph+ve: Philadelphia Chromosome positive; PSS: Personal and Social Services; QALY: quality-adjusted life year; SmPC: Summary of Product Characteristics; TKI: tyrosine kinase inhibitor; UK: United Kingdom.

Source: NICE Final Scope for ID1167.[4]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals UK Ltd. 2018. All rights reserved.

==> picture [35 x 12] intentionally omitted <==

Description of the technology being appraised

A summary of the mechanism of action, marketing authorisation status, costs and administration requirements associated with the technology, tisagenlecleucel, for the treatment of paediatric and young adult patients with r/r B-cell ALL is presented in Table 2.

Table 2: Technology being appraised

UK approved name and brand Tisagenlecleucel (Kymriah[TM] ). name Tisagenlecleucel is a genetically modified chimeric antigen receptor (CAR)-based autologous immunocellular therapy administered as a single intravenous (iv) infusion for the treatment of r/r B-cell ALL that utilises similar mechanisms to that of cytotoxic T-cells to kill leukaemic cells and thereafter maintain ongoing anti-tumour surveillance. A patient’s own T-cells are genetically engineered to express a CAR construct, which contains an external target-binding domain responsible for recognising leukaemic cells, and an internal activating domain which initiates T-cell activation (see Figure 1), allowing the induction of leukaemic cell death. As a second-generation CAR, tisagenlecleucel not only comprises the T-cell CD3ζ signalling domain, but has a costimulatory domain (4-1BB), in order to increase T-cell activation, anti-leukaemia activity and CAR-T-cell persistence.[5] Figure 1: Domains of the chimeric antigen receptor construct of tisagenlecleucel Mechanism of action Source: Novartis Pharmaceuticals UK Ltd.

The underlying mechanism of action of tisagenlecleucel involves preferentially targeting the CD19 antigen, a glycoprotein with near-universal expression on B-cell precursors and B-cells.[6] Expression of CD19 is largely restricted to B lineage cells;[5] therefore, CD19 represents an attractive immunotherapy target in ALL as it is present on leukaemic B-cells but is not found on bone marrow stem cells or other healthy tissues. Tisagenlecleucel is therefore able to target tumour cells whilst largely sparing non-cancerous cells from cytotoxicity, consequently limiting systemic effects.[7] Once tisagenlecleucel binds to CD19-positive leukaemic cells,[5] the CAR-T-cell becomes activated and the cytotoxic potential of these cells is realised (see Figure 2). Death of leukaemic cells is primarily induced through CAR-mediated cytolysis (where target cells are killed due to destruction of the cell membrane), and the release of cytokines from the CAR-T-cell.[8] Ligation of the CAR-T receptor also leads to CAR-Tcell proliferation.[8]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 16 of 185

==> picture [92 x 311] intentionally omitted <==

Figure 2: Cytotoxic mechanism of CAR-T therapy

==> picture [359 x 141] intentionally omitted <==

Source: Novartis Pharmaceuticals UK Ltd.

In contrast to the therapies currently available for patients with r/r B-cell ALL, the mechanism of action of tisagenlecleucel is entirely novel. CAR T cells have so far been detected in the peripheral blood of patients up to 784 days post infusion.[1] By using the patients’ own T-cells and their capacity for memory and surveillance, tisagenlecleucel acts as a ‘living drug’ that can provide an enduring response potentially over the course of a lifetime. As a patient-specific, single-dose, immunocellular gene-transfer therapy produced using pioneering technology, tisagenlecleucel is the first in this class of CAR-T therapy for the treatment of r/r B- cell ALL, representing a paradigm-shift in the treatment approach for this aggressive and potentially fatal disease that offers paediatric and young adult patients the potential for a cure with just a single infusion.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 17 of 185

Method of administration

In contrast to typical small molecule or biologic products, each dose of tisagenlecleucel is specifically tailored to, and manufactured for, each patient using the patient’s own blood cells, representing an entirely novel and personalised approach to the manufacturing, logistics and administration of treatment for paediatric and young adult patients with r/r B-cell ALL. The multistep supply chain process is summarised below (and in Figure 3):[5, 10, 11]

• Step 1: The patient is admitted to hospital and their mononuclear cells are obtained by a process known as leukapheresis.

• Step 2: The patient’s cells are then cryopreserved in the vapour phase of liquid nitrogen and shipped to a manufacturing facility in the US using a dedicated courier service.

• Step 3: At the manufacturing facility, the mononuclear cells are thawed and enriched for T cells. The T cells are activated with antibody-coated beads and genetically transduced using a lentiviral vector (inactive virus) containing the antiCD19 CAR transgene.

• Step 4: The CAR-T-cells then undergo ex vivo expansion on antibody-coated beads

• Step 5: The CAR-T-cells undergo formulation and a strict quality assessment before being released, cryopreserved and shipped. As an autologous (patient specific) product entering the EU, each individual therapy then requires a separate certification and batch release appropriate to the European regulations governing genetically modified advanced therapy medicinal products. Following certification, the product is sent back to the hospital where it can be stored in liquid nitrogen for up to 9 months until the treating centre / staff are ready to administer, and the patient is ready to receive treatment.

• Step 6: The patient can receive bridging chemotherapy between leukapheresis

• Method of and infusion at the discretion of the treating physician. Prior to tisagenlecleucel

• administration infusion the patient receives a preparative low dose lymphodepleting regimen. and dosage The CAR-T-cells are then thawed and reinfused into the patient as a one-time single-dose of tisagenlecleucel.

According to the SmPC, tisagenlecleucel infusions should be administered by a healthcare provider experienced with immunosuppressed patients and trained for administration of tisagenlecleucel and management of patients treated with tisagenlecleucel. Tocilizumab and emergency equipment must be available prior to infusion and during the recovery period.

Figure 3: Summary of the administration process for tisagenlecleucel

==> picture [357 x 110] intentionally omitted <==

==> picture [357 x 110] intentionally omitted <==

Source: Novartis Pharmaceuticals UK Ltd.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 18 of 185

Abbreviations: ALL: acute lymphoblastic leukaemia; allo-SCT: allogeneic stem-cell transplant; CAR: chimeric antigen receptor; CD3: cluster of differentiation 3; CD19: cluster of differentiation 19; CHMP: Committee for Medicinal Products for Human Use; EMA: European Medicines Agency; EU: European Union; FDA: Food and Drug Administration; GVHD: graft versus host disease; HBV: hepatitis B virus; HCV: hepatitis C virus; HIV: juman immunodeficiency virus; iv: intravenous; MHC: major histocompatibility complex; r/r: relapsed/refractory; SmPC: summary of product characteristics; TCR: T-cell receptor; UK: United Kingdom.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 19 of 185

Health condition and position of the technology in the treatment pathway

Disease overview

• ALL is a rare but aggressive haematological malignancy and yet one of the most common cancers to affect children and young adults.[12] It is characterised by the overproduction and accumulation of immature white blood cells (lymphoblasts) which causes the inhibition of normal blood cell production and function and eventually leads to the infiltration of lymphoblasts to other organs.[12]

• Whilst remission rates to conventional first-line chemotherapy are high, approximately 20% of patients will experience disease relapse, and a further 36% of patients will experience a second relapse, with disease prognosis worsening with each subsequent relapse.

• For paediatric and young adult patients experiencing a second or greater relapse the prognosis is dismal; median OS ranges from 3–7.5 months and current treatment options are associated with poor remission rates, reduced HRQoL, as well as medical and psychosocial consequences.[13, 14] For these patients, there is a critical unmet need for a novel therapy that can provide improved remission rates and the potential for a cure.

• As a disease that affects children and young adults, who in some cases are very young, r/r B- cell ALL has a substantial impact on parents and caregivers, who can experience significant psychological distress, depression, anxiety, stress, emotional and financial pressures.[15, 16]

Epidemiology

• ALL is a rare disease overall, with only 832 new cases diagnosed in the UK in 2015.[17] However, in contrast to many other cancer types, ALL has the highest incidence in children and young adults; approximately half of these cases are in patients aged<25 years and as such, ALL represents a major contribution to the burden of paediatric cancer in the UK.

Clinical pathway of care

• There are currently no paediatric or young-adult specific national clinical guidelines for the treatment of ALL in the UK.

• Based on feedback from UK clinical experts, ALL patients <18 years who experience a first relapse in the UK are typically treated according to the ALLR3 trial protocol, with older patients (aged 18–25 years) generally receiving blinatumomab with the aim of bridging to allo-SCT (if patients are eligible).

• If a second relapse occurs, treatment options are severely limited and prognosis is extremely poor, particularly if relapse occurs following allo-SCT. The vast majority of patients are treated with either salvage chemotherapy (specifically FLA-IDA) or blinatumomab (if not received previously) in this setting, with a minority of patients enrolling on to investigational clinical trials.

• Tisagenlecleucel is anticipated to be licensed as a treatment option for paediatric and young adult patients up to 25 years of age with B-cell acute lymphoblastic leukaemia (ALL) that is refractory, in relapse post-transplant, or in second or later relapse. In this setting, tisagenlecleucel offers a revolutionary and individualised approach to meet the critical unmet need in r/r B-cell ALL with just a single infusion, providing paediatric and young adult patients with r/r B-cell ALL the potential for a cure.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 20 of 185

==> picture [42 x 11] intentionally omitted <==

Disease overview

ALL (also called acute lymphocytic leukaemia) is a rare haematological malignancy characterised by the overproduction and accumulation of cancerous, immature white blood cells (lymphoblasts) that originate within the bone marrow.[12] As an acute leukaemia, ALL is an aggressive disease that develops rapidly (within months) and is one of the most common cancers to affect children and young adults. This is in contrast to chronic lymphocytic leukaemia which develops more slowly (over years) and rarely affects children and young adults.[18, 19]

Disease categorisation

ALL can be further categorised according to the type of lymphocytes affected (B or T-cell) and the presence or absence of the Philadelphia (Ph) chromosome.[12, 20] B-cell ALL is considerably more common than T-cell ALL, representing 80–85% of cases in children.[21] In addition, the vast majority of patients have Ph-ve ALL, with just 3% of children suffering from Ph+ve disease, which is associated with a poorer prognosis and is notoriously harder to treat than Ph-ve ALL.[20, 22] The anticipated licence for tisagenlecleucel covers all B-cell ALL patients regardless of Ph chromosome disease status.

Pathophysiology

The proliferation of lymphoblasts in patients with ALL causes the inhibition of normal blood cell production and function (red cells, white cells and platelets) and may eventually lead to the spread and infiltration of lymphoblasts to other organs, including the lymph nodes, liver, spleen, central nervous system (CNS) and testicles.[23] This rapid increase in cancerous lymphoblasts leads to the presentation of many non-specific symptoms indicative of reduced functional blood cell production, including fatigue, bruising, bone pain, fever, lymphadenopathy (swollen lymph nodes), infection and unusual and frequent bleeding.[24, 25] As an aggressive disease, if left untreated, ALL is usually fatal within a few weeks or months.[23]

Aim of treatment

The aim of treatment for paediatric and young adult patients with ALL at any stage of disease is to induce complete remission (CR).[26] For children and young adult patients who are diagnosed with ALL and are able to be treated, CR rates with conventional first-line chemotherapy are as high as 80–85%.[27, 28] However, despite these high remission rates, approximately 20% of patients will subsequently experience disease relapse, and the majority of relapses occur within two years of first-line treatment.[25, 29]

The aim of treatment for children and young adults who experience a first relapse is to achieve a second CR with the aim of, in most cases, enabling patients to receive an allogeneic stem cell transplant (allo-SCT) if they are eligible.[2] Second CR rates with chemotherapy for patients who experience a first relapse are still reasonably high, and can range from 71–93%.[3] However, the chances of a patient achieving CR are substantially reduced with every subsequent relapse: CR rates for second, third and fourth or later relapse are reported to be 44%, 27% and 12% respectively, demonstrating a substantial decrease in responsiveness with every treatment failure.[30] The proportion of patients estimated to experience a second relapse is 36%. This highlights the clinical burden in the relapsed setting, emphasising the urgent need for treatment options for patients who experience more than one disease relapse following conventional chemotherapy.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 21 of 185

In addition to the morbidity and mortality associated with relapsed disease, a small proportion of patients may experience refractory disease, which can either be defined by a lack of CR after primary induction therapy for newly-diagnosed ALL (primary refractory) or a lack of CR after chemotherapy received in the relapsed setting (chemo-refractory).[31] Although rare (primary induction failure typically occurs in only 2–3% of patients), primary-refractory patients are severely limited in their options for successful treatment and remain a therapeutic challenge.[3]

Burden of disease

The burden of disease for ALL is associated with significant patient and parent/caregiver impact.[15, ] 16, 32, 33 Patients with r/r B-cell ALL have an extremely poor prognosis and this is exacerbated further with each subsequent relapse.[30] Median OS with current treatment in the r/r setting ranges from less than 3 months to 7.5 months.[34, 35] Therefore, and unsurprisingly, r/r B-cell ALL survivors are even more likely to report poor general health, functional impairment, and activity limitations, respectively, compared with non-relapsed survivors.[32] The burden of disease is made worse by the fact that current treatments for r/r B-cell ALL are associated with poor clinical outcomes, poor HRQoL, and medical and psychosocial consequences.[32, 33]

As a disease that affects children and young adults, who in some cases are very young, r/r B-cell ALL has a substantial impact on parents and caregivers, who can experience significant psychological distress, depression, anxiety, stress and emotional pressures.[15, 16] Moreover, the economic burden of ALL can also be a major source of anxiety as regular inpatient and outpatient visits often disrupt parent and caregivers’ employment and diminish their productivity.[36] The burden of disease is therefore not only felt by patients themselves, but has a dramatic and widespread impact on entire families and their wider support networks.

The provision of a more effective treatment for r/r B-cell ALL that can offer substantial life extension and the potential for a cure, will therefore help to alleviate this parent and caregiver burden, improving the quality of life of children and young adults affected by ALL.

Incidence of ALL in children and young adults

ALL is considered a rare disease, with just 832 new cases of ALL diagnosed in the UK in 2015, accounting for less than 1% of all new cancer diagnoses in adults and children in the UK.[17] However, the incidence of ALL is strongly related to age and, in stark contrast to most other cancers, ALL has the highest incidence in children and young adults, with the peak incidence in children aged 0–4 years old (see Figure 5).[37] Of the average 811 new cases of ALL diagnosed in the UK each year between 2013–2015, 520 cases (64%) were in patients aged 0 to 24 years.[17] As such, although ALL is rare overall, the disease represents a major contribution to the burden of paediatric cancer in the UK, and accounts for almost 80% of all childhood leukaemias and 25% of all childhood cancers.[12]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 22 of 185

Figure 5: Average number of new cases of ALL per year and age-specific incidence rates in the UK (2013-2015)

==> picture [447 x 97] intentionally omitted <==

==> picture [447 x 97] intentionally omitted <==

==> picture [447 x 96] intentionally omitted <==

Abbreviations: ALL: acute lymphoblastic leukaemia; UK: United Kingdom. Source: Cancer Research UK.[37]

==> picture [43 x 11] intentionally omitted <==

Clinical pathway of care

Tisagenlecleucel is anticipated to be licensed for the treatment of paediatric and young adult patients up to 25 years of age with B-cell acute lymphoblastic leukaemia (ALL) that is refractory, in relapse post-transplant, or in second or later relapse.

Whilst clinical guidelines are available for adults from the European Society for Medical Oncology (ESMO) and for paediatric and young adult patients from the US National Comprehensive Cancer Network (NCCN) in the US, there are currently no paediatric or young-adult specific national clinical guidelines for the treatment of ALL in the UK.[38, 39]

The NCCN guidelines are not followed by UK clinical experts and many paediatric and young adult patients with r/r B-cell ALL in the UK are typically entered into experimental clinical trials if possible; for patients who do not enter a clinical trial, treatment is guided by clinical trial protocols, where available, and by clinician choice.[2]

The current treatment pathway for paediatric and young adult patients with B-cell ALL in the UK together with the potential positioning of tisagenlecleucel is summarised in Figure 6 based on feedback from several clinical experts in the UK.[2]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 23 of 185

Figure 6: Treatment pathway for ALL in the UK with the potential positioning of tisagenlecleucel

==> picture [457 x 91] intentionally omitted <==

==> picture [457 x 90] intentionally omitted <==

==> picture [457 x 90] intentionally omitted <==

Abbreviations: ALL: acute lymphoblastic leukaemia; allo-SCT: stem cell transplantation; FLA-IDA: fludarabine, cytarabine and idarubicin. Source: UK expert clinician feedback.[2]

Newly-diagnosed ALL

The aim of any treatment for paediatric and young adult patients with ALL at any stage of disease is to induce CR.[26] Standard first-line treatment for newly-diagnosed ALL in paediatric and young adult patients in the UK is multi-drug chemotherapy, which typically comprises a combination of cyclophosphamide, vincristine, doxorubicin, dexamethasone, methotrexate and cytarabine.[40]

Relapsed disease

Despite high CR rates that can be achieved with first-line chemotherapy, approximately 20% of patients will experience relapsed disease following CR from first-line chemotherapy.[28] Patients under the age of 18 years who experience a first relapse or refractory disease in the UK are typically treated according to the ALLR3 protocol, an international collaborative clinical trial protocol developed by the Childhood Leukaemia Working Party in the UK.[41] The ALLR3 protocol varies according to patient risk and contains three phases; induction, consolidation and intensification. For patients who achieve a CR to ALLR3 induction therapy, some patients will receive maintenance chemotherapy and some will go on to receive an allo-SCT (if eligible) (see Figure 6).[2, 41]

Patients over the age of 18 who experience a first relapse in the UK are typically treated with blinatumomab, with the aim of bridging to allo-SCT (if patients are eligible).[2]

For patients who then experience a second relapse following maintenance chemotherapy or alloSCT, or relapse before being able to receive an allo-SCT, treatment options are severely limited, and an established protocol of care does not exist. The only therapy licensed by the EMA for the

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 24 of 185

treatment of r/r B-cell ALL in paediatric/young adult patients who have received at least two prior regimens is clofarabine; however, the consensus from UK clinical experts is that clofarabine is rarely used, if at all, in the UK due to toxicity.[2, 42] The vast majority of these patients typically receive treatment with either salvage chemotherapy (specifically the FLA-IDA regimen: fludarabine, cytarabine and idarubicin) or blinatumomab (if not received previously) (see Figure 6).

Blinatumomab is licensed by the EMA and NICE approved for the treatment of adults with r/r B-cell ALL.[43, 44] Whilst it is yet to be licensed in the paediatric population, due to the NHS England national commissioning policy, blinatumomab is currently also available for paediatric patients with r/r B-cell ALL in England.[45] However, feedback from UK clinical experts is that many patients are in fact treated with blinatumomab earlier on in the pathway following a first relapse with the aim of bridging to allo-SCT.[2] This is also in line with the recommendations from NICE for blinatumomab in the adult population.[44] As such, given the earlier use of blinatumomab, some clinical experts would view salvage chemotherapy (FLA-IDA) as their preferred treatment option following a second relapse, or relapse post allo-SCT.[2]

Proposed positioning of tisagenlecleucel

Tisagenlecleucel is anticipated to be positioned as a treatment option for paediatric and young adult patients up to 25 years of age with B-cell acute lymphoblastic leukaemia (ALL) that is refractory, in relapse post-transplant, or in second or later relapse. Therefore, within the context of this appraisal, salvage chemotherapy (FLA-IDA) and blinatumomab represent the most relevant comparators to tisagenlecleucel within the treatment pathway for paediatric and young adult patients who have r/r B-cell ALL.

Despite its use, no clinical evidence exists for the efficacy of FLA-IDA in paediatric and young adult patients with r/r B-cell ALL. Consensus from four UK clinical experts was that expected median survival outcomes with FLA-IDA are poor, and can be considered comparable to those observed with clofarabine monotherapy, which has been shown to be less than 3 months in this patient population and the rate of CR was 30%.[2, 34] The efficacy of blinatumomab has been studied in both paediatric patients (<18 years) and adults (>18 years) with r/r B-cell ALL; CR rates and median OS were very similar between the two populations.[35, 46] In paediatric patients, the CR rate for blinatumomab was 39% (95% CI: 27, 51%), with median OS only 7.5 months (95% CI: 4.0, 11.8 months); in adults, the rate of CR was 34% (95% CI: 28, 40%) with median OS 7.7 months (95% CI: 5.6, 9.6).[35, 46]

Therefore, there is a critical unmet need for a novel therapy that can provide improved remission rates and extended survival for paediatric and young adult patients with r/r B-cell ALL. Tisagenlecleucel offers a revolutionary and individualised approach to meet this unmet need, providing children and young adults with r/r B-cell ALL the potential for a cure after only a single infusion. The clinical evidence for tisagenlecleucel in this patient population is compelling, and derives from three clinical trials with a total sample size of xxx patients. Across all three clinical trials, tisagenlecleucel has demonstrated consistent, clinically meaningful efficacy with high remission rates, deep molecular responses, and durable remissions. Full details of the results from all three tisagenlecleucel clinical trials are presented in Section B.2 of this submission.

==> picture [35 x 12] intentionally omitted <==

Equality considerations

No equality issues related to the use of tisagenlecleucel are foreseen.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 25 of 185

B.2 Clinical effectiveness

Summary of clinical effectiveness

 The efficacy of tisagenlecleucel has been established in three clinical trials (ELIANA, ENSIGN and B2101J) involving xxx paediatric and young adult patients with r/r B-cell ALL.[47-] 49  The primary efficacy endpoint of ELIANA was met, with an independent review committee (IRC)-assessed overall remission rate (ORR) of xxxxx (95% confidence Interval [CI]: xxxxxxxxxx) within 3 months after infusion.[47] Similarly high remission rates were achieved in ENSIGN (ORR of 69.0%) and B2101J (ORR of xxxxx).[48, 49]  A key secondary efficacy endpoint in the ELIANA trial was bone marrow minimal residual disease (MRD) negative complete remission/complete remission with incomplete blood count recovery (CR/CRi) within 3 months post infusion.[31] Results for this endpoint (xxxxx; 95% CI: xxxxxxxxxx) demonstrated deep remissions in xxxxx of patients, and confirmed the observed clinical benefit seen with the primary endpoint.[47] Similarly high ORRs with MRDnegative bone marrow remissions were reported in ENSIGN (64.3%), and B2101J (xxxxx).[48, ] 49  In the majority of patients, durable remissions were observed across all three trials.[47-49] The rate of event-free survival (EFS) at 12 months was xxxxx in ELIANA, xxxxx in ENSIGN and xxxxx in B2101J.[47-49]  Promising survival outcomes were observed across all three trials.[47-49] The probability of survival at Month 12 was xxxxx in ELIANA, 62.6% in ENSIGN, and xxxxx in B2101J.[47-49] Median OS in B2101J, the trial with the longest follow-up, was xxxx months.[49]  In ELIANA, two different HRQoL tools demonstrated improvements in patient-reported outcomes at 3 and 6 months following infusion further supporting the clinical benefit of tisagenlecleucel.[50] Summary of the results from the indirect treatment comparison  Given the absence of a head-to-head clinical trial versus the relevant comparators to this appraisal, a matched-adjusted indirect comparison (MAIC) was conducted for OS versus salvage chemotherapy (using clofarabine monotherapy as a proxy for the efficacy of FLA-IDA) and blinatumomab.  After adjusting for population differences via the MAIC, tisagenlecleucel was estimated to have statistically superior OS over both clofarabine monotherapy (a proxy for salvage chemotherapy) and blinatumomab. Full details of the MAIC are presented in Section B.2.8. Summary of safety results for tisagenlecleucel  The safety profile of tisagenlecleucel has been well characterised and was consistent across all three trials.[47-49] Full details of the safety profile of tisagenlecleucel in paediatric and young adult patients with r/r B-cell ALL are presented in Section B.2.10.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 26 of 185

==> picture [34 x 12] intentionally omitted <==

Identification and selection of relevant studies

An SLR was conducted in March 2018 to identify relevant clinical evidence on the efficacy and safety of tisagenlecleucel for the treatment of paediatric patients with r/r B-cell ALL. Full details of the SLR search strategy, study selection process and results can be found in Appendix D.

The SLR included a total of 77 publications, reporting on 66 unique clinical trials. Of these, six publications reporting on three clinical trials were identified that investigated tisagenlecleucel in the patient population of interest for this appraisal: ELIANA [NCT02435849], ENSIGN [NCT02228096] and B2101J [NCT01626495]; see Section B.2.2).[51-56]

==> picture [35 x 12] intentionally omitted <==

List of relevant clinical effectiveness evidence

Three clinical trials were identified in the SLR that provide clinical evidence for the efficacy and safety of tisagenlecleucel for the treatment of paediatric and young adult patients with r/r B-cell ALL: ELIANA (NCT02435849), ENSIGN (NCT02228096) and B2101J (NCT01626495).

ELIANA is an ongoing, international, multicentre, phase II, single-arm, open-label study to determine the efficacy, safety and patient-reported outcomes of tisagenlecleucel in paediatric and young adult patients with r/r B-cell ALL.[57] Data from ELIANA have been published by Maude et al. (2018) based on a median 13.1 months of follow up;[54] however, as the publication does not present the most recent data cut from this trial, the data presented within this submission are taken from the ELIANA Clinical Study Report (CSR; data cut-off 31[st] Dec 2017 representing a median xxxxxxxxxxx follow-up).[47]

ENSIGN is an ongoing, US-based, multicentre, phase II, single-arm, open-label study to determine the efficacy and safety of tisagenlecleucel in paediatric and young adult patients with r/r B-cell ALL.[58] Data from ENSIGN have been published by Maude et al. (2016) representing a median 6.4 months of follow up;[55] however, as the publication does not present the most recent data cut from this trial, the data presented within this submission are taken from the ENSIGN CSR (data cut-off 6th Oct 2017 representing a median xxxxxxxxxxx of follow up).[48]

B2101J was the first trial to be conducted in tisagenlecleucel and is an ongoing, US-based, singlecentre, phase I/IIa, single-arm, open-label study to determine the safety, tolerability and engraftment potential of tisagenlecleucel in patients with r/r B-cell ALL.[59] Data from B2101J have been published by Maude et al. (2014) representing a median 7 months of follow up;[60] however, as the publication does not present the most recent data cut from this trial, the data presented within this submission are taken from the B2101J CSR (data cut-off 30th Jan 2017 representing a median xxxxxxxxxxx of follow up).[49]

An overview of the three tisagenlecleucel clinical trials ELIANA, ENSIGN and B2101J is provided in Table 3 below.

Table 3: Clinical effectiveness evidence

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 27 of 185

aNote as of the respective data cuts presented within this submission, no patients with lymphoma had been infused with tisagenlecleucel and therefore the populations treated and subsequently analysed within this submission exclusively include patients with r/r B-cell ALL. b Reference to the patients in B2101J refers to the non-CNS3 cohort only and data for the non-CNS3 cohort only are presented within this submission. cA target per-protocol dose of CTL019 transduced cells for paediatric patients consists of a single infusion of 2.0 to 5.0×10[6] transduced cells per kg body weight (for patients ≤50 kg) and 1.0 to 2.5×10[8] CTL019 transduced viable T cells (for patients >50 kg). The following cell dose ranges here were infused if all other safety release criteria were met.

Abbreviations: ALL: acute lymphoblastic leukaemia; CD19: cluster of differentiation 19; CNS: central nervous system; DoR: duration of remission; EFS: event-free survival, EQ-5D-3L: EuroQol 5-Dimensions 3-Levels; iv: intravenous; MRD: minimal residual disease; N/A: not applicable; ORR: overall remission rate; OS: overall survival; RFS: relapse-free survival; r/r: relapsed/refractory; US: United States. Source: ELIANA CSR (31[st] Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 28 of 185

Summary of methodology of the relevant clinical effectiveness evidence

==> picture [42 x 11] intentionally omitted <==

Trial design

All three tisagenlecleucel clinical trials followed a similar trial design, with sequential phases of screening, enrolment, treatment (including apheresis, bridging chemotherapy, lymphodepleting chemotherapy and tisagenlecleucel administration) and follow-up.[47-49]

ELIANA trial design

ELIANA is an ongoing, international, multicentre, phase II, single-arm, open-label study.[57] Paediatric and young adult patients with r/r B-cell ALL who were primary refractory, chemorefractory, in 2[nd] or greater bone marrow relapse, relapsed after allogeneic allo-SCT, or otherwise ineligible for allogeneic allo-SCT were enrolled in the trial.[57]

A schematic of the ELIANA trial design is presented in Figure 7. The trial consists of several sequential phases: screening, pre-treatment, treatment and primary follow-up, secondary follow-up and survival follow-up.[47]

Screening and pre-treatment: Patients were screened for eligibility following leukapheresis. Eligible patients were then enrolled in the trial, and treated with bridging chemotherapy (where appropriate) followed by lymphodepleting chemotherapy 2–14 days prior to tisagenlecleucel infusion.

Treatment and primary follow-up: After tisagenlecleucel infusion, patients entered the primary follow-up period, during which efficacy was assessed monthly for the first six months, and then quarterly for up to 2 years and bi-annually for up to 5 years, or patient relapse.

Secondary follow-up: Patients could discontinue from primary follow-up due to reasons such as treatment failure, relapse after remission, pursuing allo-SCT while in remission or voluntary withdrawal. Patients who discontinued from the primary follow-up period before Month 60 continue to be followed in the secondary follow-up period for the collection of safety and survival data (every 3 months) for up to 5 years.

Survival and long-term safety follow-up: The survival follow-up period is to collect survival data (every 3 months) on patients who have completed the study up to 5 years post-tisagenlecleucel infusion. Patients will then continue to be followed as part of the long-term safety follow-up until 15 years post-tisagenlecleucel infusion.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 29 of 185

Figure 7: ELIANA trial design

==> picture [446 x 87] intentionally omitted <==

==> picture [446 x 87] intentionally omitted <==

1Performed prior to study entry; 2As indicated per protocol; 3Only for patients who drop out of the primary follow-up before Month 60;[4] Patients will be followed for survival until the end of trial, or until they are enrolled in the longterm follow-up;[5] Long-term safety follow-up conducted under a separate protocol;[6] To be completed 2–14 days prior to tisagenlecleucel infusion: fludarabine (30 mg/m[2] iv daily for 4 doses) plus cyclophosphamide (500 mg/m[2] iv daily for 2 doses). Source: ELIANA CSR (31[st] Dec 2017).[47]

ENSIGN trial design

ENSIGN is an ongoing, international, multicentre, phase II, single-arm, open-label study to determine the efficacy and safety of tisagenlecleucel in paediatric and young adult patients with r/r B-cell ALL.[58] Paediatric and young adult patients with r/r B-cell ALL or lymphoblastic lymphoma who were primary refractory, chemo-refractory, in 2[nd] or greater bone marrow relapse, relapsed after allogeneic allo-SCT, or otherwise ineligible for allogeneic allo-SCT were enrolled in the trial.[48] As of the data cut-off date presented within this submission (6th Oct 2017), no patients with lymphoblastic lymphoma had been infused with tisagenlecleucel and therefore the population treated and subsequently analysed within this submission exclusively includes patients with r/r B- cell ALL.

A schematic of the ENSIGN trial design is presented in Figure 8. The trial consists of several sequential phases: screening, pre-treatment, treatment and primary follow-up, secondary follow-up and survival follow-up.[48]

Screening and pre-treatment: Patients were screened for eligibility following leukapheresis. Eligible patients were then enrolled in the trial, and treated with bridging chemotherapy (where appropriate) followed by lymphodepleting chemotherapy 2–14 days prior to tisagenlecleucel infusion.

Treatment and primary follow-up: After tisagenlecleucel infusion, patients entered the primary follow-up period, during which efficacy was assessed monthly for the first six months, and then quarterly for up to 2 years and bi-annually for up to 5 years, or patient relapse.

Secondary follow-up: Patients could discontinue from primary follow-up due to reasons such as treatment failure, relapse after remission, pursuing allo-SCT while in remission or voluntary withdrawal. Patients who discontinued from the primary follow-up period before Month 60 continue to be followed in the secondary follow-up period for the collection of safety and survival data for up to 5 years.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 30 of 185

Survival and long-term safety follow-up: The survival follow-up period is to collect survival data on patients who have completed the study up to 5 years post-tisagenlecleucel infusion. Patients will then continue to be followed as part of the long-term safety follow-up until 15 years posttisagenlecleucel infusion.

Figure 8: ENSIGN trial design

==> picture [441 x 98] intentionally omitted <==

==> picture [441 x 97] intentionally omitted <==

==> picture [213 x 19] intentionally omitted <==

----- Start of picture text -----
Abbreviations: LD: lymphodepleting chemotherapy.
Source: ENSIGN CSR (6 [th] Oct 2017). [48]
----- End of picture text -----

B2101J trial

B2101J is an ongoing, single centre, phase I/IIa, single-arm, open-label study.[59] Paediatric and young adult patients with r/r B-cell ALL who were treatment refractory, relapsed after allogeneic allo-SCT, or were otherwise ineligible for allogeneic allo-SCT were enrolled in the trial.[59] The data presented within this submission only includes individuals from the B2101J cohort with non-CNS3 ALL who were analysed separately i.e. non-lymphoma patients and those without CNS relapse (<5 white blood cells [WBCs] per mL with leukaemic blast cells after cytocentrifugation following traumatic lumbar puncture), in line with the patient populations of ELIANA and ENSIGN.[61]

A schematic of the B2101J trial design is presented in Figure 9. The trial consists of several sequential phases: screening, treatment (consisting of apheresis, cytoreductive chemotherapy and tisagenlecleucel administration) and follow-up.[49]

Screening and pre-treatment: Patients were screened for eligibility and eligible patients were then enrolled in the trial. Leukapheresis could occur prior to, or after enrolment. Patients were then treated with bridging chemotherapy (where appropriate) followed by lymphodepleting chemotherapy approximately one week prior to tisagenlecleucel infusion.

Treatment and primary follow-up: In B2101J, tisagenlecleucel infusion was administered in a dose-escalated manner, a minimum of 1–5 days after the completion of cytoreductive chemotherapy. After tisagenlecleucel infusion, patients entered the primary follow-up period, during which efficacy was assessed monthly for the first six months, and then quarterly for up to 2 years post-infusion.

Secondary follow-up: For patients who completed or prematurely discontinued from the primary follow-up phase while in remission, follow-up attempts were made to assess the patient’s relapse,

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 31 of 185

post-treatment antineoplastic therapy, and survival status until two years post the last patient infusion.

Survival and long-term safety follow-up: Once patients relapsed, they were followed for survival only. Patients completing or prematurely discontinuing participation in this study will then continue to be followed as part of the long-term safety follow-up until 15 years post-tisagenlecleucel infusion.

Figure 9: B2101J trial design

==> picture [453 x 114] intentionally omitted <==

Abbreviations: AE: adverse event; DP: destination protocol; FDA: Food and Drug Administration; PBMC: peripheral blood mononuclear cells; Wk: week. Source: B2101J CSR (30[th] Jan 2017).[49]

==> picture [43 x 12] intentionally omitted <==

Trial methodology

A summary of the methodology of ELIANA, ENSIGN and B2101J is presented in Table 4. All three trials had a very similar study design, ALL patient population and methodology with the exception of the following minor differences. B2101J was a single-site study whereas ENSIGN and ELIANA were conducted across multiple sites.[57-59] The inclusion criteria for each trial were similar and although ENSIGN and B2101J allowed the inclusion of patients with lymphoma, the data presented within this submission are for patients with r/r B-cell ALL only.[48, 50, 59] The same target dose for tisagenlecleucel was followed in ELIANA and ENSIGN; as the first study of tisagenlecleucel in this indication, B2101J followed a dose-escalation regimen with a broader target dose range.[48-50]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 32 of 185

Table 4: Summary of methodology of studies

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals Ltd. 2018. All rights reserved

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals Ltd. 2018. All rights reserved

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals Ltd. 2018. All rights reserved

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals Ltd. 2018. All rights reserved

aAs of the respective data cuts presented within this submission, no patients with lymphoma had been infused with tisagenlecleucel and therefore the ENSIGN population treated and subsequently analysed within this submission exclusively includes patients with r/r B-cell ALL.[b] Data for B2101J presented in this submission refer to the non-CNS3 ALL cohort only. Abbreviations: ALL: acute lymphoblastic leukaemia; BCR/ABL: breakpoint cluster region Abelson; BM: bone marrow; BOR: best overall response; CAR: chimeric antigen receptor; CD3: cluster of differentiation 3; CD19: cluster of differentiation 19; CLL: chronic lymphoblastic leukaemia; CNS: central nervous system; CR: complete remission; CRi: complete remission with incomplete blood count recovery; CRF: case report form; DLBCL: diffuse large B-cell lymphoma; DoR: duration of remission; EFS: event-free survival; EU: European Union; FDA: Food and Drug Administration; GVHD: graft-versus-host disease; IRC: Independent Review Committee; LD: lymphodepleting; MLL: mixed lineage leukaemia; MRD: minimal residual disease; NHL: non-Hodgkin’s lymphoma; ORR: overall remission rate; OS: overall survival; Ph+ve: Philadelphia chromosome positive; PLL: prolymphocytic leukaemia; RFS: relapse-free survival; RT-PCR: reverse transcription polymerase chain reaction; allo-SCT: stem cell transplantation; TBI: total body irradiation; TCR: T-cell receptor; TKI: tyrosine kinase inhibitor; US: United States. Source: ELIANA CSR (31[st] Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017);[49] ClinicalTrials.gov.[59]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals Ltd. 2018. All rights reserved

Description of outcomes reported in ELIANA, ENSIGN and B2101J

Definitions of the primary and key secondary outcomes assessed in ELIANA, ENSIGN and B2101J are provided in Table 5. ORR was the primary endpoint in ELIANA and ENSIGN, and was also assessed in B2101J. Key secondary outcomes reported across all three trials include EFS, DoR, RFS and OS.[47-49]

Table 5: Outcome definitions in ELIANA, ENSIGN and B2101J

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 38 of 185

Abbreviations: AE: adverse event; BOR: best overall response; CR: complete remission; CRi: complete remission with incomplete blood count recovery; DoR: duration of remission; EFS: event-free survival; ORR: overall remission rate; OS: overall survival; RFS: relapse-free survival; allo-SCT: stem cell transplantation. Source: ELIANA CSR (31[st] Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

==> picture [43 x 11] intentionally omitted <==

Baseline characteristics

The ELIANA trial was initiated on the 8[th] April 2015 (first patient first visit) and is ongoing. At the time of the data cut-off date presented within this submission (31[st] Dec 2017), xxx patients had been screened, xx patients were enrolled, and xx patients had been treated with tisagenlecleucel.[47] The xx patients who received tisagenlecleucel infusion were aged between xxxxxxx years of age (mean xxxx years), with fairly equal gender distribution. The vast majority of patients had a Karnofsky/Lanksy performance status of greater than 70, with a median of x prior therapies of which xxxxx of patients had failed prior allo-SCT. The majority of patients had relapsed disease (xxxxx) and xxxx patients had primary refractory ALL.[47]

The ENSIGN trial was initiated on the 14[th] August 2014 (first patient first visit) and is ongoing. At the time of the data cut-off date presented within this submission (6[th] Oct 2017), 85 patients had been screened, 73 patients were enrolled, and 58 patients had been treated with tisagenlecleucel.[48] The 58 patients who received tisagenlecleucel infusion were aged between 3 to 25 years of age (mean xxxx years), with fairly equal gender distribution. All patients had a Karnofsky/Lanksy performance status of at least 50, with a median of 3 prior therapies of which 44.8% of patients had failed prior allo-SCT. The majority of patients had relapsed disease (91.4%) and 8.6% patients had primary refractory ALL.[48]

The B2101J trial was initiated on the 15[th] March 2012 (first patient first visit) and is ongoing. At the time of the data cut-off date presented within this submission (30[th] Jan 2017), xx patients were enrolled, and xx patients had been treated with tisagenlecleucel.[49] The xx patients who received tisagenlecleucel infusion were aged between xxxxxxx years of age (mean xxxx years), with fairly equal gender distribution. All patients had a Karnofsky/Lanksy performance status of at least 80, xxxxx of patients had failed prior allo-SCT, and xxxxx patients had Ph+ve disease. The majority of patients had relapsed disease (xxxxx) and xxxx had primary refractory ALL.[49]

Baseline demographics, disease characteristics and a summary of disease history for the patients treated with tisagenlecleucel in ELIANA, ENSIGN and B2101J are presented in Table 6. The patient populations of each trial can be considered very similar and feedback from clinical experts in the treatment of ALL in the UK was that the study populations of each of the trials are reflective of the clinical population of paediatric and young adults patients with r/r B-cell ALL that would be candidates for tisagenlecleucel in the UK.[2] Although these trials did not include centres within the UK specifically, other European centres were included, and there was consensus from UK clinical

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 39 of 185

experts that the patients in all three trials can be considered comparable to the patients likely to be treated with tisagenlecleucel in the UK and this should not affect the applicability of tisagenlecleucel to patients in the UK setting.[2]

Table 6: Baseline characteristics (full analysis set)

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 40 of 185

a Data for B2101J presented in this submission refer to the non-CNS3 ALL cohort only.

b Data not available for all patients, hence why n numbers are less than the total full analysis set. c This value for B2101J is for patients receiving >1 prior allo-SCT, rather than exactly two. d Calculated for relapsed patients only

Abbreviations: ALL: acute lymphoblastic leukaemia; CNS: central nervous system; MRD: minimal residual disease; N/A: not applicable; NR: not reported; allo-SCT: stem cell transplantation; SD: standard deviation. Source: ELIANA CSR (31[st] Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

Statistical analysis and definition of study groups in the relevant clinical effectiveness evidence

Definitions of the key study populations analysed from ELIANA, ENSIGN and B2101J are presented in Table 7. Assessments of all endpoints were based on data from patients who received tisagenlecleucel (i.e. the full analysis set for efficacy endpoints, and the safety set for safety endpoints). The numbers of patients in each analysis set are presented in Table 8.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 41 of 185

Table 7: Analysis set definitions

aThere was no requirement for an efficacy analysis set in B2101J, hence the FAS was used for all outcomes. Abbreviations: ALL: acute lymphoblastic leukaemia; CNS: central nervous system. Source: ELIANA CSR (31[st] Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017);[49] B2101J CSR supplementary appendix (30[th] Jan 2017).[62]

Table 8: Trial populations used for the analysis of outcomes of relevant clinical trials

aData for B2101J presented in this submission refer to the non-CNS3 ALL cohort only. bThe FAS for ORR and DOR in ELIANA includes xx patients to allow for at least 3 months between infusion and the data cut-off (31[st] Dec 2017).[c] The efficacy analysis set was used only for outcomes related to ORR in ENSIGN. There was no requirement for an efficacy analysis set in B2101J, hence the FAS was used for all outcomes. Abbreviations: ALL: acute lymphoblastic leukaemia; CNS: central nervous system; NR: not reported. Source: ELIANA CSR (31[st] Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

The statistical analyses used for the primary endpoints of the ELIANA, ENSIGN and B2101J trials alongside sample size calculations and methods for handling missing data, are presented in Table 9.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 42 of 185

Table 9: Statistical methods for the primary analysis of relevant clinical trials

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals Ltd. 2018. All rights reserved

Abbreviations: CR: complete remission; CRi: complete remission with incomplete blood count recovery, IRC: Independent Review Committee; ORR: overall remission rate. Source: ELIANA CSR (31[st] Dec 2017);[47] Maude et al. (2018) trial protocol;[63] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 44 of 185

Quality assessment of the relevant clinical effectiveness evidence

An overview of the quality assessments of ELIANA, ENSIGN and B2101J is presented below. These quality assessments were performed based on the GRACE checklist for the methodological quality of randomised and non-randomised studies of health care interventions, and indicate that all three trials can be considered to be of good quality.[64]

Table 10: Overview of the quality assessment of ELIANA, ENSIGN and B2101J based on the GRACE checklist

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 45 of 185

aN/A as ELIANA, ENSIGN and B2101J are all single-arm clinical trials. Abbreviations: N/A: not applicable. Source: ELIANA CSR (31st Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

==> picture [35 x 12] intentionally omitted <==

Clinical effectiveness results of the relevant trials

Summary of the clinical effectiveness results of the relevant trials

ELIANA

 The study met its primary objective at the first interim analysis (data cut-off date: 17[th] Aug 2016), with an ORR of xxxxx (95% CI: xxxxxxxxxx) within 3 months of tisagenlecleucel infusion.[47]

 As of the latest data cut-off (31st Dec 2017), the ORR remained consistent with the interim analysis; of the xx patients in the efficacy analysis set, the ORR within 3 months of tisagenlecleucel infusion was xxxxx (95% CI: xxxxxxxxxx). and xx patients (xxxxx, 95% CI: xxxx, xxxx) achieved BOR of CR or CRi with bone marrow MRD negative disease (i.e. MRD <0.01%) demonstrating the depth and quality of the remissions achieved.[47]

 Median DoR has xxxxxxxxxxxxxxxxxxxx, with six-month DoR of xxxxx (95% CI: xxxxxxxxxx). Median EFS and OS were also not reached, with six-month EFS of xxxxx (95% CI: xxxxxxxxxx) and 12-month OS of xxxxx (95% CI: xxxxxxxxxx).[47] These results support the durability of remission for the majority of patients and are comparable to the results from both ENSIGN and B2101J.[48, 49]

 Patient reported quality of life assessments xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx at three and six months post-tisagenlecleucel infusion compared to baseline, further supporting the benefit of treatment with tisagenlecleucel.[50]

ENSIGN

 The study met its primary objective at the first interim analysis (data cut-off 1[st] Feb 2016), with an ORR of 69.0% (95% CI: 52.9, 82.4) within 6 months of tisagenlecleucel infusion.[65]  As of the latest data cut-off (6[th] Oct 2017), the ORR remained consistent with the interim analysis; of the 42 patients in the efficacy analysis set, the ORR within 6 months of tisagenlecleucel infusion was 69.0% (95% CI: 52.9, 82.4) and 27 patients (64.3%, 95% CI: 48.0, 78.4) achieved BOR of CR or CRi with bone marrow MRD negative disease.[48]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 46 of 185

==> picture [460 x 372] intentionally omitted <==

----- Start of picture text -----
 Median DoR was not reached, with 6-month DoR of 71.4%. Median EFS was xxx months,
with six-month EFS of xxxxx. 12-month OS was 62.6%. [48]
B2101J
 The efficacy data from B2101J are consistent with that seen in ELIANA and ENSIGN. As of
the data cut-off date of 30th Jan 2017, the ORR at Day 28 was xxxxx (xxxxx patients). [49] Of
the xx patients who achieved a CR or CRi, xx patients (xxxxx, 95% CI: xxxx, xxxx) had bone
marrow MRD negative disease within 28 days of tisagenlecleucel infusion. [49]
 Median DoR was xxxx months (95% CI: xxx, NE) and in patients who had a BOR of CR or
CRi, median EFS was xxxx months (95% CI: xxx, NE). [49] Median OS was xxxx months, and
the estimated probability of being alive was xxxxx at Month 12 and xxxxx at two years. [49]
 With the longest follow-up of xxxx months (median xxxx months), B2101J demonstrates the
sustained duration of remission possible with tisagenlecleucel [49]
Pooled analysis
 A pooled analysis of all three tisagenlecleucel clinical trials was also conducted. The
individual patient-level data (IPD) from each of the latest data cut-offs for all three clinical
trials ELIANA (xxxx), ENSIGN (n=58) and B2101J (xxxx) were combined directly without
adjustment to derive a pooled estimate of EFS and OS for tisagenlecleucel.
 In the pooled analysis, the probability of being event-free was xxxxx (95% CI: xxxxxxxxxx) at
one year, xxxxx (95% CI: xxxxxxxxxx) at two years and xxxxx at (95% CI: xxxxxxxxxx) at
three years. [47]
 In terms of OS, the probability of survival at one year was xxxxx (95% CI: xxxxxxxxxx),
xxxxx (95% CI: xxxxxxxxxx) at two years and xxxxx (95% CI: xxxxxxxxxx) at three years,
indicating durable remission and a high probability of survival up to 3 years after infusion. [47]
----- End of picture text -----

==> picture [42 x 11] intentionally omitted <==

Clinical effectiveness results overview

An overview of the clinical effectiveness results from all three trials is provided in Table 11 below.

Table 11: Summary of the clinical effectiveness results in ELIANA, ENSIGN and B2101J

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 47 of 185

aORR and DoR from the latest data cut of ELIANA were assessed in patients with 3 months post-tisagenlecleucel infusion only. In ENSIGN these outcomes were assessed in patients with 6 months post-tisagenlecleucel infusion only (efficacy analysis set).[b] Data for B2101J presented in this submission refer to the non-CNS3 ALL cohort only. cBOR is reported within 3 months, 6 months and 28 days of tisagenlecleucel respectively for ELIANA, ENSIGN and B2101J, respectively.[d] ’Unknown’ is assigned in case the Baseline assessment of the response assessment is not done, incomplete, indeterminate, or not performed within the respective time frame. *No formal significance testing was conducted as the endpoint was met at the interim analysis. Nominal p-value is presented. Abbreviations: BOR: best overall response; CI: confidence interval; CR: complete remission; CRi: CR with incomplete blood count recovery; DOR: duration of remission; FAS: full analysis set; MRD: minimum residual disease; NE: not estimable; NR: non-responder/no remission; ORR: overall remission rate Source: ELIANA CSR (31[st] Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

==> picture [43 x 12] intentionally omitted <==

ELIANA

Data from the ELIANA trial are presented from the latest data cut-off date of 31st Dec 2017. All outcomes are presented for the full analysis set (all patients who received infusion of tisagenlecleucel; xxxx), with the exception of ORR and DoR/RFS which were analysed in the efficacy analysis set (all patients treated with tisagenlecleucel at least 3 months prior to the data cut-off; xxxx).

Primary outcome: ORR

ELIANA met its primary endpoint with an ORR of xxxxx (95% CI: xxxxxxxxxxx

The primary outcome of the ELIANA trial was ORR within 3 months of tisagenlecleucel administration as determined by IRC assessment; the primary endpoint was an ORR of >20% (the null hypothesis). ORR was defined as the proportion of patients with a best overall disease response of CR or CRi on the basis of the results of laboratory testing of blood, bone marrow, and cerebrospinal fluid (CSF), as well as physical examination. Responses were required to be maintained for 28 days.[50]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 48 of 185

An interim analysis was performed on the first 50 patients infused with tisagenlecleucel (data cutoff date: 17[th] Aug 2016). The primary endpoint was met and the ORR by IRC assessment during the 3 months post-tisagenlecleucel infusion in the full analysis set (all patients who received infusion of tisagenlecleucel) was 82.0% (41/50) (95% CI: 68.6, 91.4), with xx patients (xxxxx) achieving a CR and xxxxx patients (xxx) achieving a CRi.[50]

As of the data cut-off date of 31st Dec 2017, the ORR in the full analysis set was xxxxx xxxxxxx (95% CI: xxxxxxxxxx). CR was achieved in xx patients xxxxxxx and xx patients xxxxxxx achieved a CRi.[47] Full results of the ORR analyses in the full analysis set are summarised in Table 12.

Table 12: Summary of IRC-assessed ORR within three months post-tisagenlecleucel infusion in ELIANA (full analysis set)

aORR from the latest data cut of ELIANA (31st Dec 2017) was assessed in patients with 3 months posttisagenlecleucel infusion only.[b] ’Unknown’ is assigned in case the Baseline assessment of the response assessment is not done, incomplete, indeterminate, or not performed within the respective time frame. Abbreviations: BOR: best overall response; CI: confidence interval; CR: complete remission; CRi: complete remission with incomplete blood count recovery; ORR: overall remission rate. Source: Maude et al. (2018);[54] ELIANA CSR (31[st] Dec 2017).[47]

Bone marrow MRD status

Of patients who achieved an ORR of CR or CRi, xxxxx had MRD negative disease, a key prognostic factor and marker of deep remission

Bone marrow MRD status by IRC assessment during the 3 months post-tisagenlecleucel infusion as determined by IRC assessment was a key secondary endpoint outcome of the ELIANA trial. Bone marrow MRD status was assessed by flow cytometry, and was defined as the minimum MRD percentage during the corresponding time frame. MRD status can be used to assess early treatment response and to detect relapse in a precise manner.[66] An MRD of less than 0.01% was defined as ‘MRD negative disease’.[63]

As of the latest data cut-off date of 31st Dec 2017, xx of the xx patients infused with tisagenlecleucel (xxxxxxxxxxxxxxxxx) achieved an ORR of CR or CRi during the three months post-tisagenlecleucel infusion, of which xxxxx (xxxxx) of patients were bone marrow MRD negative (i.e. MRD <0.01%).[47] Similar results were also achieved at the interim analysis performed on the first 50 patients infused with tisagenlecleucel (data cut-off date: 17[th ] Aug 2016), and in the results of the data cut-off of 25th Apr 2017.[50, 54]

Given the prognostic association with MRD status and its use as a robust indicator of relapse, the results for this key secondary outcome demonstrate the depth and quality of the response achieved by tisagenlecleucel as assessed by bone marrow MRD negative remission rate in

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 49 of 185

patients with an ORR of CR/CRi.[47] Full results of the bone marrow MRD status analyses are summarised in Table 13.

Table 13: Summary of IRC-assessed bone marrow MRD status within three months posttisagenlecleucel infusion in ELIANA (full analysis set)

aORR from the latest data cut of ELIANA was assessed in patients with 3 months post-tisagenlecleucel infusion only. Abbreviations: BOR: best overall response; CI: confidence interval; CR: complete remission; CRi: complete remission with incomplete blood count recovery; FAS: full analysis set; MRD: minimal residual disease. Source: ELIANA CSR (31[st] Dec 2017).[47]

DoR

Remissions were durable; as of the latest data cut-off median DoR and RFS had not been reached

DoR was defined as the time from the date of achievement of CR or CRi to the date of relapse or death due to underlying cancer, as determined by IRC assessment. RFS was defined as the time from achievement of CR or CRi whichever occurred first, to relapse or death due to any cause. As of the latest data cut-off date (31st Dec 2017), among patients with a BOR of CR or CRi, there were no deaths due to reasons other than the underlying cancer, and thus RFS was the same as DOR.[50]

As of the latest data cut-off date (31st Dec 2017), xxxxxxxxxxxxx of patients who had achieved a BOR of CR or CRi had not relapsed and median DoR had not been reached (95% CI: xxxxxxxx).[47] The estimated rate of RFS after onset of remission was xxxxx (95% CI: xxxxxxxxxx) at Month 6 and xxxxx (95% CI: xxxxxxxxxxx) at Month 12.[47] The Kaplan-Meier plot for the analysis of DoR (data cut-off date of 31st Dec 2017) is presented in Figure 10.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 50 of 185

Figure 10: Kaplan-Meier plot for IRC-assessed DoR (censoring for allo-SCT) in ELIANA (full analysis set)[a]

==> picture [442 x 158] intentionally omitted <==

aAs of data cut-off 31st Dec 2017.

Abbreviations: CI: confidence interval; DoR: duration of response; NE: not estimable. Source: ELIANA CSR (31[st] Dec 2017).[47]

EFS

Median EFS had not been reached at the time of the latest data cut-off

EFS was defined as the time from the date of first tisagenlecleucel infusion to the earliest date of either death due to any cause after remission, relapse, or treatment failure, as determined by IRC assessment. Treatment failure was defined as no response in the study and discontinuation from the study due to death, AE, lack of efficacy, or new anticancer therapy. In case of treatment failure, the event date was set to study Day 1.[63]

At the time of the latest data cut-off (31st Dec 2017), median EFS had not yet been reached (95% CI: xxxxxxx) and only xx of the xx patients infused with tisagenlecleucel (xxxxx) had experienced an EFS event (death due to any cause after remission, relapse, or treatment failure). The probability of being event-free was xxxxx (95% CI: xxxxxxxxxx) at Month 6 and xxxxx at Month 12 (95% CI: xxxxxxxxxx).[47] The Kaplan-Meier plot for the analysis of EFS at the data cut-off date of 31st Dec 2017 is presented in Figure 11.

Figure 11: Kaplan-Meier plot for IRC-assessed EFS (censoring for allo-SCT) in ELIANA (full analysis set)[a]

==> picture [442 x 157] intentionally omitted <==

aAs of data cut-off 31st Dec 2017.

Abbreviations: CI: confidence interval; EFS: event-free survival; NE: not estimable. Source: ELIANA CSR (31[st] Dec 2017).[47]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 51 of 185

OS

OS data demonstrate durable remissions and a high probability of long-term survival

OS was defined as the time from first tisagenlecleucel infusion to the date of death due to any cause.[63] As of the latest data cut-off of 31st Dec 2017, median OS had not yet been reached and only xxxxx patients (xxxxx) had died following tisagenlecleucel infusion. The probability of survival at Month 6 was xxxxx (95% CI: xxxxxxxxxx), and at Month 12 was xxxxx (95% CI: xxxxxxxxxx), indicating durable remission and a high probability of survival up to 12 months after infusion.[47] The Kaplan-Meier plot for the analysis of OS at the data cut-off date of 31st Dec 2017 is presented in Figure 12.

Figure 12: Kaplan-Meier plot for OS in ELIANA (full analysis set)[a]

==> picture [442 x 157] intentionally omitted <==

aAs of data cut-off 31st Dec 2017.

Abbreviations: CI: confidence interval; OS: overall survival; NE: not estimable. Source: ELIANA CSR (31st Dec 2017).[47]

Patient-reported outcomes

Patient-reported outcomes (PROs) were assessed via the paediatric quality of life questionnaire (PedsQL) and the EQ-5D-3L questionnaire in patients ≥8 years old only.[67, 68]

The PedsQL is a generic instrument that is commonly used to measure HRQoL in children. The EQ-5D is a widely used, self-administered questionnaire designed to assess health status in adults and in adolescents aged 12 to 18 years. A child-friendly version, the EQ-5D-Y, has been developed for use in children aged 8 years and older.[69] In the ELIANA trial, EQ-5D-3L was used for patients aged 13 and above at study entry and EQ-5D-Y was used for patients between the ages of 8 and 12 years at study entry. Each patient completed the questionnaire(s) at each scheduled visit before interacting with the Investigator or undergoing other clinical assessments.

PedsQL questionnaire

For patients ≥8 years old who achieved CR/CRi following tisagenlecleucel infusion, higher mean scores on the PedsQL questionnaire for emotional, social, school, physical, and psychosocial health subscales were reported at Month 3, 6, 9, 12, 18 and 24 compared to Baseline, indicating consistent improvement of HRQoL up to two years following tisagenlecleucel infusion. However, results beyond Month 12 should be interpreted with caution, as the number of patients with PRO results after this timepoint were limited (see Figure 13).[47]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 52 of 185

In the full analysis set (data cut-off 31st Dec 2017), the mean change from Baseline in the PedsQL total score was xxxx at Month 6, xxxx at Month 12, xxxx at Month 18 and xxxx at Month 24, indicating an overall improvement in HRQoL after tisagenlecleucel infusion.[47] The minimal clinically important difference for the PedsQL total scores has been estimated using distribution-based methods to be 4.4 for self-report.[70] Thus, the observed changes from Baseline in the PedsQL total score and in each PedsQL subscale at each time point appear to represent clinically meaningful improvements in HRQoL.

EQ-5D-3L

For patients ≥8 years old who achieved CR/CRi following tisagenlecleucel infusion, the mean change from Baseline in the European quality of life visual analogue scale (EQ VAS) was xxxx at Month 6, xxxx at Month 12, xxxx at Month 18, and xxxx at Month 24, again indicating an overall improvement in HRQoL following tisagenlecleucel infusion (see Figure 13).[47] The number of patients are relatively small at the later timepoints and therefore interpretation should again be conducted with caution.

Given that minimally important differences for the EQ VAS among cancer patients were estimated to range from 7–10 using anchor-based categories from the FACT-G (Pickard et al. [2007]), the observed changes from Baseline in EQ VAS at each timepoint appear to represent meaningful improvements in HRQoL.[71] Additionally, while the mean EQ VAS score at Baseline xxxxxx was comparable to that of patients sampled from cancers of various aetiologies (Pickard et al . [2007]), the mean scores at Month 6 (xxxx), Month 12 (xxxx), Month 18 (xxxx), and Month 24 (xxxx) were comparable to normative means of general populations.[47, 71, 72]

Figure 13: Summary of PedsQL and EQ VAS scores in ELIANA (patients ≥8 years old achieving CR/CRi)

==> picture [442 x 157] intentionally omitted <==

Mean change from baseline in patients who had both baseline and post-baseline score. Only patients 8 years or older were required to complete the assessments. Abbreviations: CR: complete remission; CRi: complete remission with incomplete blood count recovery; EQ VAS: EuroQol-visual analogue scales; HRQoL: health-related quality of life; PedsQL: Pediatric quality of life Source:[a] Varni et al. (2003);[70][b] Pickard et al. (2007);[71][c] Varni et al. (2001);[73][d] Janssen et al. (2014).[72]

ENSIGN

==> picture [43 x 11] intentionally omitted <==

Data from the ENSIGN trial are presented from the latest data cut-off date of 6th Oct 2017, which provides median follow-up of xxxx months and a maximum follow-up of xxxx months.[48] All outcomes are presented for the full analysis set (all patients who received infusion of tisagenlecleucel; n=58) with the exception of ORR and DoR/RFS which were analysed in the efficacy analysis set (all patients treated with tisagenlecleucel at least 6 months prior to the data cut-off; n=42).

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 53 of 185

Primary outcome: ORR

The primary outcome of the ENSIGN trial was ORR within 6 months of tisagenlecleucel administration as determined by IRC assessment. ORR was defined as the proportion of patients with a best overall disease response of CR or CRi on the basis of the results of laboratory testing of blood, bone marrow, and cerebrospinal fluid (CSF), as well as physical examination. Responses were required to be maintained for 28 days.[48]

As of the latest data cut-off date of 6[th] Oct 2017, the study met its primary endpoint with an ORR of 69.0% (29/42) (95% CI: 52.9, 82.4). CR was achieved in 27 patients (64.3%) and two patients (4.8%) achieved a CRi.[48] In sensitivity analyses performed using local Investigator assessment, there was xxxx concordance between IRC assessment and local Investigator assessment of ORR.[48] Full results of the ORR analysis are summarised in Table 14.

Table 14: Summary of IRC-assessed ORR within six months post-tisagenlecleucel infusion in ENSIGN (efficacy analysis set)

a’Unknown’ is assigned in case the Baseline assessment or the response assessment is not done, incomplete, indeterminate, or not performed within the respective time frame. bNo formal significance testing was conducted as the endpoint was met at the interim analysis. Nominal p-value is presented. Abbreviations: BOR: best overall response; CI: confidence interval; CR: complete remission; CRi: complete remission with incomplete blood count recovery; ORR: overall remission rate. Source: ENSIGN CSR (6[th] Oct 2017).[48]

Bone marrow MRD status

Bone marrow MRD status by IRC assessment during the six months post-tisagenlecleucel infusion was a key secondary outcome for patients who received tisagenlecleucel in the ENSIGN trial. MRD status was assessed by flow cytometry, and was defined as the minimum MRD percentage during the corresponding time frame. An MRD of less than 0.01% is defined as ‘MRD negative disease’.[48]

As of the latest data cut-off date of 6th Oct 2017, 29/42 patients (69.0%; 95% CI: 52.9, 82.4) achieved an ORR of CR or CRi during the six months post-tisagenlecleucel infusion, of which 27 patients (64.3%; 95% CI: 48.0, 78.4) were bone marrow MRD negative (i.e. MRD <0.01%) and therefore achieved bone marrow MRD negative remission.[48] Full results of the bone marrow MRD status analysis within six months post-tisagenlecleucel infusion are summarised in Table 15.

Table 15: Summary of IRC-assessed bone marrow MRD status within six months posttisagenlecleucel infusion in ENSIGN (efficacy analysis set)

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 54 of 185

Abbreviations: BOR: best overall response; CI: confidence interval; CR: complete remission; CRi: complete remission with incomplete blood count recovery; MRD: minimal residual disease. Source: ENSIGN CSR (6[th] Oct 2017).[48]

DoR

DoR was defined as the time from the date of achievement of CR or CRi to the date of relapse or death due to underlying cancer, as determined by IRC assessment. RFS was defined as the time from achievement of CR or CRi whichever occurred first, to relapse or death due to any cause. As of the latest data cut-off data (6th Oct 2017), among patients with a BOR of CR or CRi, there were no deaths due to reasons other than the underlying cancer, and thus RFS was the same as DoR.[48] 20/29 patients (69.0%) who achieved a BOR of CR or CRi had not relapsed, and median DoR had not been reached. The estimated rate of RFS after onset of remission was 71.4% (95% CI: 48.5, 85.5) at Month 6 and 61.2% (95% CI: 37.8, 78.0) at Month 12. A pre-planned sensitivity analysis of DoR without censoring at time of allo-SCT was conducted and the results were similar to the main analysis (full results not shown).[48] The Kaplan-Meier plot for the analysis of DoR at the data cut-off date of 6[th] Oct 2017 is presented in Figure 14.

Figure 14: Kaplan-Meier plot for IRC-assessed DoR (censoring for allo-SCT) in ENSIGN (efficacy analysis set)[a]

==> picture [302 x 278] intentionally omitted <==

----- Start of picture text -----
100
80
60
40 Censoring Times
All patients (N= 29)
Number of Events (n)
All patients: 9
20 Kaplan-Meier medians
All patients: NE months, 95 % CI [5.9, NE]
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Number of patients still at risk Time (months)
All patients 29 27 25 14 11 9 5 5 5 5 5 5 3 2 2 0
Probability (%) of event free
----- End of picture text -----

aAs of data cut-off 6th Oct 2017. Only patients who achieved CR or CRi are included. Time is relative to onset of remission, 1 month = 30.4375 days. Abbreviations: CI: confidence interval; DoR: duration of response; NE: not estimable. Source: ENSIGN CSR (6[th] Oct 2017).[48]

EFS

EFS was defined as the time from the date of first tisagenlecleucel infusion to the earliest date of either death due to any cause after remission, relapse, or treatment failure, as determined by IRC assessment. Treatment failure was defined as no response in the study and discontinuation from

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 55 of 185

the study due to death, AE, lack of efficacy, or new anticancer therapy. In case of treatment failure, the event date was set to study Day 1.[48]

At the time of the latest data cut-off (6[th] Oct 2017), median EFS was xxx months (95% CI: xxxxxxx) and xx of the xx patients (xxxxx) reported treatment failure or relapse, as determined by IRC assessment.[48] The estimated probability of being event-free was xxxxx (95% CI: xxxxxxxxxx) at Month 6 and xxxx% (95% CI: xxxxxxxxxxx at Month 12.[48] The Kaplan-Meier plot for the analysis of EFS at the data cut-off date of 6[th] Oct 2017 is presented in Figure 15.

Figure 15: Kaplan-Meier plot for EFS (censoring for allo-SCT) in ENSIGN (full analysis set)[a]

==> picture [442 x 157] intentionally omitted <==

aAs of data cut-off 6th Oct 2017. Time is relative to first tisagenlecleucel infusion date, 1 month = 30.4375 days. EFS of treatment failure patient is set to Day 1. Abbreviations: CI: confidence interval; EFS: event-free survival; NE: not estimable. Source: ENSIGN CSR (6[th] Oct 2017).[48]

OS

OS was defined as the time from first tisagenlecleucel infusion to the date of death due to any cause. As of the latest data cut-off of 6[th] Oct 2017, median OS was 23.8 months (95% CI: 8.8, NE) and 19/58 patients (32.8%) had died following tisagenlecleucel infusion; however, median OS should be interpreted with caution because only seven patients were at risk beyond that point. The probability of survival at Month 6 was 79.3% (95% CI: 64.9, 88.4), and at Month 12 was 62.6% (95% CI: 45.8, 75.6).[48] The Kaplan-Meier plot for the analysis of OS at the data cut-off date of 6th Oct 2017 is presented in Figure 16.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 56 of 185

Figure 16: Kaplan-Meier plot for OS in ENSIGN (full analysis set)[a]

==> picture [340 x 311] intentionally omitted <==

----- Start of picture text -----
100
80
60
40 Censoring Times
All patients (N= 58)
Number of Events (n)
All patients: 19
20 Kaplan-Meier medians
All patients: 23.8 months, 95 % CI [8.8, NE]
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34
Number of patients still at risk Time (months)
All patients 58 46 39 31 25 20 14 12 10 10 10 10 7 6 4 3 2 0
Probability (%) of event free
----- End of picture text -----

aAs of data cut-off 6th Oct 2017. Time is relative to first tisagenlecleucel infusion date, 1 month = 30.4375 days. Abbreviations: CI: confidence interval; OS: overall survival; NE: not estimable. Source: ENSIGN CSR (6[th] Oct 2017).[48]

==> picture [43 x 11] intentionally omitted <==

B2101J

Data from the B2101J trial are presented from the latest data cut-off date of 30th Jan 2017, which provides a median follow-up of xxxx months and a maximum follow-up of xxxx months.[49] All outcomes are presented for the full analysis set (all patients who received infusion of tisagenlecleucel) in the xx patients with non-CNS3 ALL only.

ORR

A key outcome of the B2101J trial was the ORR at Day 28 after tisagenlecleucel administration, as determined by Investigator assessment. ORR was defined as the proportion of patients with a best overall disease response of CR or CRi at the Day 28 visit. Disease assessment performed between study Day 2 to Day 59 and prior to the receipt of any new therapy was considered within the window.[49]

As of the latest data cut-off date of 30th Jan 2017, the ORR was xxxxxxxxxxxxx (95% CI: xxxxxxxxxx), with xx patients (xxxxx) achieving a CR and xx patients (xxxxx) achieving CRi by Day 28 post-tisagenlecleucel infusion. In the analysis of BOR at any time, ORR was xxxxxxxxxxxxx (95% CI: xxxxxxxxxx), with xx patients (xxxxx) achieving CR and xx patients (xxxxx) with CRi. Only xxxxx patients (xxxx) did not respond to treatment with tisagenlecleucel.[49] Full results of the ORR analysis are summarised in Table 16.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 57 of 185

Table 16: Summary of ORR at Day 28 in B2101J

aData for B2101J presented in this submission refer to the non-CNS3 ALL cohort only. Abbreviations: ALL: acute lymphoblastic leukaemia; BOR: best overall response; CI: confidence interval; CNS: central nervous system; CR: complete remission; CRi: complete remission with incomplete blood count recovery; MRD: minimal residual disease; ORR: overall remission rate. Source: B2101J CSR (30[th] Jan 2017).[49]

Bone marrow MRD status

Bone marrow MRD status post-tisagenlecleucel infusion at Day 28, as determined by Investigator assessment, was a key secondary outcome for patients who received tisagenlecleucel in the B2101J trial. Bone marrow MRD status was assessed by flow cytometry, and was defined as the percentage of patients achieving MRD negative bone marrow post-tisagenlecleucel infusion; an MRD of less than 0.01% was defined as ‘MRD negative disease’.[49]

As of the latest data cut-off of 30th Jan 2017, xx of the xx patients infused with tisagenlecleucel (xxxxx; 95% CI: xxxxxxxxxx) achieved a BOR of CR or CRi both at Day 28, or any time posttisagenlecleucel infusion. At Day 28, negative bone marrow MRD status was achieved in xx patients (xxxxx), and at any time following infusion, xxxxxxxxxx patients achieved bone marrow MRD negative remission.[49] Full results for the analysis of remission with MRD negative bone marrow are summarised in Table 17.

Table 17: Summary of bone marrow MRD status in B2101J (full analysis set)

aData for B2101J presented in this submission refer to the non-CNS3 ALL cohort only. Abbreviations: ALL: acute lymphoblastic leukaemia; CI: confidence interval; CNS: central nervous system; CR: complete remission; CRi: complete remission with incomplete blood count recovery; MRD: minimal residual disease.

Source: B2110J CSR (30[th] Jan 2017).[49]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 58 of 185

DoR

DoR was defined as the duration from the date when the response criteria of CR or CRi was first met to the date of relapse or death due to underlying cancer. RFS was measured by the time from achievement of CR or CRi whichever occurred first, to relapse or death due to any cause during CR or CRi. RFS was assessed only in patients with a BOR of CR or CRi. As of the latest data cutoff (30th Jan 2017), among patients with a BOR of CR or CRi, there were no deaths due to reasons other than the underlying cancer, and thus RFS was the same as DoR.[49]

As of the latest data cut-off date (30th Jan 2017), xxxxx (xxxxx) of patients who achieved a BOR of CR or CRi had not suffered an event (relapse or death due to underlying cancer), with median DoR of xxxx months. The estimated relapse-free rate after onset of remission was xxxxx (95% CI: xxxxxxxxxx) at Month 12, xxxxx (95% CI: xxxxxxxxxx) at Month 24, and xxxxx (95% CI: xxxxxxxxxx) at Month 36.[49] The Kaplan-Meier plot for the analysis of DoR at the data cut-off date of 30th Jan 2017 is presented in Figure 17.

Figure 17: Kaplan-Meier plot for DoR in B2101J (full analysis set)[a]

==> picture [442 x 158] intentionally omitted <==

aAs of data cut-off 30th Jan 2017.

Abbreviations: ALL: acute lymphoblastic leukaemia; CI: confidence interval; CNS: central nervous system; CR: complete remission; CRi: complete remission with incomplete blood count recovery; DoR: duration of response; NE: not estimable.

Source: B2101J CSR (30[th] Jan 2017).[49]

EFS

EFS was defined as the time from date of first tisagenlecleucel infusion to the earliest date of death due to any cause after remission, relapse, or treatment failure. Treatment failure was defined as no response in the study and discontinuation from the study due to death, AE, lack of efficacy, or new anticancer therapy. In case of treatment failure, the event date was set to study Day 1.[49]

As of the latest data cut-off date (30th Jan 2017), median EFS was xxxx months (95% CI: xxxxxxx), and xxxxx patients (xxxxx) had experienced an event. The estimated probability of being event-free was xxxxx (95% CI: xxxxxxxxxx) at Month 12, xxxxx (95% CI: xxxxxxxxxx) at Month 24, and xxxxx (95% CI: xxxxxxxxxx) at Month 36.[49]

The Kaplan-Meier plot for the analysis of EFS at the data cut-off date of 30th Jan 2017 is presented in Figure 18.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 59 of 185

Figure 18: Kaplan-Meier plot for EFS in B2101J (full analysis set)[a]

==> picture [442 x 158] intentionally omitted <==

aAs of data cut-off 30th Jan 2017.

Abbreviations: CI: confidence interval; EFS: event-free survival; NE: not estimable. Source: B2101J CSR (30[th] Jan 2017).[49]

OS

OS was defined as the time from date of randomisation or first tisagenlecleucel infusion to the date of death due to any cause. As of the latest data cut-off of 30th Jan 2017, median OS was xxxx months (95% CI: xxxxxxxx) and xxxxx patients (xxxxx) had died following tisagenlecleucel infusion. The probability of survival was xxxxx (95% CI: xxxxxxxxxx) at Month 12, xxxxx (95% CI: xxxxxxxxxx) at Month 24, and xxxxx (95% CI: xxxxxxxxxx) at Month 36, demonstrating durable

remission and a high probability of survival up to three years post-tisagenlecleucel infusion.[49] The Kaplan-Meier plot for the analysis of OS at the data cut-off date of 30th Jan 2017 is presented in Figure 19.

Figure 19: Kaplan-Meier plot for OS in B2101J (full analysis set)[a]

==> picture [442 x 157] intentionally omitted <==

aAs of data cut-off 30th Jan 2017.

Abbreviations: ALL: acute lymphoblastic leukaemia; CI: confidence interval; CNS: central nervous system; FAS: full analysis set; OS: overall survival; NE: not estimable. Source: B2101J CSR (30[th] Jan 2017).[49]

Subgroup analysis

==> picture [35 x 12] intentionally omitted <==

In all three trials, the robustness and consistency of the primary analysis was confirmed by a series of pre-specified subgroup analyses for the ORR based on pre-determined baseline variables, including age, gender, baseline bone marrow tumour burden (an indicator of overall disease burden) and prior allo-SCT.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 60 of 185

In the ELIANA trial, data cut-off 25th Apr 2017 (subgroup analyses are not yet available for the latest data cut-off of 31st Dec 2017), subgroup ORR analyses were conducted for subgroups with at least 5 patients. The ORR by IRC assessment was consistently ≥20% across all subgroups evaluated (hence the null hypothesis that the ORR was ≤20% could be rejected; see Figure 20). In the ENSIGN trial (data cut-off 6th Oct 2017), the ORR by IRC assessment across the various subgroups with at least five patients was consistently ≥55% (ranging from 55.6% to 93.3%) and in B2101J (data cut-off 30th Jan 2017), the ORR by IRC assessment at Day 28 was similarly high in all pre-specified subgroups analysed (≥80% in all subgroups). Forest plots for the subgroup analyses from ENSIGN and B2101J are presented in Appendix E.

Figure 20: ORR within 3 months post-tisagenlecleucel infusion by IRC assessment. Forest plot for subgroups from ELIANA (full analysis set)

==> picture [434 x 273] intentionally omitted <==

*Low disease burden, <50% lymphoblasts in bone marrow; high disease burden, ≥50% lymphoblasts in bone marrow. † ≥5 unrelated abnormalities. ‡ BCR-ABL1, MLL rearrangement, hypoploidy, lesions associated with BCR-ABL1–like gene signature, or complex karyotype.

The area of each box is proportional to the number of patients in the particular grouping. 95% CIs are exact Clopper-Pearson CIs calculated for each subgroup. Abbreviations : CI: confidence interval; IRC: independent review committee; ORR: overall remission rate; alloSCT: stem cell transplantation. Source: Maude et al. (2018) supplementary appendix.[31]

==> picture [35 x 12] intentionally omitted <==

Meta-analysis

For the purposes of increasing the overall available sample size for tisagenlecleucel and allowing the use of the longest-term follow-up data available within the economic analysis, data for EFS and OS from all three tisagenlecleucel clinical trials were pooled as part of a meta-analysis. The feasibility of pooling all three trials was assessed by taking into consideration the study design, definitions of outcomes, and patient baseline characteristics of all three tisagenlecleucel clinical trials and further details are presented in Appendix D.

Study design

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 61 of 185

All three trials followed almost identical study designs. The only difference for B2101J was the dosing regimen: in ELIANA and ENSIGN, patients received a single infusion with a narrower target dose range whereas in the B2101J study, patients were treated according to a dose escalation protocol, with a wider target dose range, and could therefore receive multiple infusions. Whilst this difference in dosing between the trials is noted, the median dose received across all three trials was of the same magnitude and therefore this difference was not expected to bias the pooled estimate of efficacy for tisagenlecleucel.

Outcome definitions

The definitions of EFS and OS, the key outcome measures informing the economic analysis, were identical across all three trials (see Table 5).

Patient baseline characteristics

The eligibility criteria of the B2101J trial were broader, and allowed the inclusion of patients with prior anti-CD19 therapy, CNS3 disease, and patients up to the age of 24 at diagnosis (compared with 21 years in ELIANA and ENSIGN). However, only 4 patients in the B2101J trial had received prior anti-CD19 therapy and therefore this minority is not expected to have a large impact on the results of the trial. Furthermore, the analyses of the B2101J trial presented in this submission are for patients without CNS3 disease, hence this difference can be considered accounted for. In terms of age, the mean age across all three trials was very similar (xxxx, xxxx and xxxx in ELIANA, ENSIGN, and B2101J, respectively) and the age range across all three trials was between 1 and 25.[47-49]

Key patient baseline characteristics can be found in Table 18 below. Overall, it was considered that any differences between baseline characteristics were minor, and therefore it was considered appropriate to pool the data from all three trials. Furthermore, the eligibility criteria of all three trials match the intended patient population for tisagenlecleucel in UK clinical practice.[74] Therefore, taken together, the pooling of all three trials generates a larger sample size of a group of patients that can be considered, overall, to be representative of the “true” population likely to be treated with tisagenlecleucel in UK clinical practice.

Table 18: Key patient baseline characteristics from the pooled analysis

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 62 of 185

aData for B2101J presented in this submission refer to the non-CNS3 ALL cohort only. bFor B2101J, this value refers to patients who have received >1 prior allo-SCT. Abbreviations: SD: standard deviation.

Source: ELIANA CSR (31[st] Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

The individual patient-level data (IPD) from each of the latest data cut-offs for all three clinical trials ELIANA (xxxx), ENSIGN (n=58) and B2101J (xxxx) were combined directly without adjustment to derive a pooled estimate of EFS and OS for tisagenlecleucel. A total of xxx patients were therefore included in the pooled analysis.

EFS

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 63 of 185

EFS was defined as the time from the date of first tisagenlecleucel infusion to the earliest date of either death due to any cause after remission, relapse, or treatment failure, as determined by IRC assessment. Treatment failure was defined as no response in the study and discontinuation from the study due to death, AE, lack of efficacy, or new anticancer therapy. In case of treatment failure, the event date was set to study Day 1.[63]

In the pooled analysis, median EFS is 28.9 months (95% CI: xxxxxxx) and xx of the xxx patients infused with tisagenlecleucel (xxxxx) had experienced an EFS event (death due to any cause after remission, relapse, or treatment failure). The probability of being event-free was xxxxx (95% CI: xxxxxxxxxx) at one year, xxxxx (95% CI: xxxxxxxxxx) at two years and xxxxx at (95% CI: xxxxxxxxxx) at three years.[47]

The Kaplan-Meier plot showing the EFS curve for each trial separately, together with the pooled EFS curve is presented in Figure 21 below.

Figure 21: Kaplan-Meier curves for EFS in ELIANA, ENSIGN and B2101J and the pooled analysis

==> picture [442 x 157] intentionally omitted <==

Abbreviations: EFS: event-free survival.

Source: ELIANA CSR (31[st] Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

OS

OS was defined as the time from first tisagenlecleucel infusion to the date of death due to any cause.[63] In the pooled analysis, median OS is xxxx months (95% CI: xxxxxxxx) and xxxxxx patients (xxxxx) had died following tisagenlecleucel infusion. The probability of survival at one year was xxxxx (95% CI: xxxxxxxxxx), xxxxx (95% CI: xxxxxxxxxx) at two years and xxxxx (95% CI: xxxxxxxxxx) at three years, indicating durable remission and a high probability of survival up to 3 years after infusion.[47]

The Kaplan-Meier plot showing the OS curve for each trial separately, together with the pooled OS curve is presented in Figure 22 below.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 64 of 185

Figure 22: OS Kaplan-Meier curves for ELIANA, ENSIGN and B2101J and the pooled analysis

==> picture [442 x 157] intentionally omitted <==

Abbreviations: OS: overall survival.

Source: ELIANA CSR (31[st] Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

==> picture [35 x 12] intentionally omitted <==

Indirect and mixed treatment comparisons

In the absence of head-to-head clinical trial evidence of tisagenlecleucel versus either blinatumomab or salvage chemotherapy (FLA-IDA), an SLR was conducted to identify relevant evidence on the comparator treatments for the purposes of conducting a possible indirect treatment comparison. Full details of the methodology and results of the SLR are presented in Appendix D.

Blinatumomab

Of the 66 studies ultimately identified in the SLR, two trials were identified that investigated the use of blinatumomab in paediatric patients aged up to 18 years with r/r B-cell ALL: a phase II clinical trial (n=70) published by von Stackelberg et al. (2016) and an expanded open-access study (n=40) published as a poster at the American Society of Clinical Oncology conference 2017 (the RIALTO study).[35, 75] The eligibility criteria of the RIALTO study permitted patients previously treated with blinatumomab, and therefore it was considered that some patients may have overlapped between the von Stackelberg et al. (2016) and RIALTO studies. For this reason, the RIALTO study was not considered further for inclusion within an indirect treatment comparison, nor was it considered appropriate to explore a pooling of the von Stackelberg et al. (2016) and RIALTO studies. As such, and given the von Stackelberg et al. (2016) study represents the pivotal clinical trial for blinatumomab in paediatric patients with r/r/ B-cell ALL and the larger of the two identified clinical trials for blinatumomab, the von Stackelberg et al. (2016) study alone was considered for further inclusion within an indirect treatment comparison. Further details of this study are presented in Appendix D. The use of the RIALTO study was explored within the economic analysis.

Salvage chemotherapy (FLA-IDA)

No trials were identified for FLA-IDA in paediatric patients with r/r B-cell ALL. As such, an assessment of the included studies was performed to identify efficacy data that could be used as a proxy for the efficacy of FLA-IDA.

The 66 included studies were first assessed based on the following elements: 1) comparable patient population to the three tisagenlecleucel clinical trials; 2) relevant EFS and OS measures reported in the form of Kaplan-Meier curves.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 65 of 185

Full details of the eligibility criteria for ELIANA, ENSIGN and B2101J can be found in Appendix L. Additional study population criteria were applied to the 66 studies included in the SLR to systematically select the studies with populations comparable to the tisagenlecleucel clinical trials. A total of 10 studies in first relapse patients only, first relapse/primary refractory patients only, or unclear populations were subsequently excluded. The second step was to review the remaining studies in terms of the availability of relevant EFS and OS measures reported in the form of Kaplan-Meier curves; of these, 8 studies without an OS Kaplan-Meier curve were excluded. As a last step, any studies conducted in Japan were also excluded from consideration, and they were not deemed to be conducted in patient populations similar enough to the UK. Finally, the studies of blinatumomab were not considered to represent the efficacy of FLA-IDA and were removed.

The remaining 6 trials are presented in Table 28. These were presented for review by UK clinical experts, who advised on whether the efficacy outcomes could be considered comparable to the outcomes expected with FLA-IDA. Feedback from UK clinical experts was that median OS with FLA-IDA would be around 3 months. Median OS for all 6 trials ranged from 11 weeks to 9 months, and therefore the trials with median OS of 9 months were further excluded, given these survival outcomes did not align with the clinical expert feedback. Based on this, only 4 possible trials remained, which investigated the use of clofarabine combination therapy (Hijaya et al. [2011; Miano et al. [2012]; Cooper et al. [2013] ) or clofarabine monotherapy (Jeha et al. [2006]). Feedback from UK clinical experts was that efficacy with clofarabine monotherapy or clofarabine combination therapy could be considered appropriate for use as a proxy for the clinical efficacy of FLA-IDA. Given the fact that clofarabine monotherapy is licensed in the UK for paediatric/young adult patients who have received at least two prior regimens, and the data were also used as part of the NICE mock appraisal, the study by Jeha et al. (2006) was ultimately considered to be the most appropriate source of clinical data for the salvage chemotherapy (FLA-IDA) comparator within this submission. The other clofarabine studies (Hijaya et al. [2011; Miano et al. [2012]) relate to combination therapy and were therefore excluded on this basis.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 66 of 185

Table 19: Clinical evidence identified to be used as a proxy for the efficacy of FLA-IDA

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals Ltd. 2018. All rights reserved

Abbreviations: EFS: event-free survival; NR: not reported; OS: overall survival; allo-SCT: stem-cell transplant.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B-cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167]

© Novartis Pharmaceuticals Ltd. 2018. All rights reserved

Matched-adjusted indirect treatment comparison

Due to the single-arm nature of the clinical trials investigating tisagenlecleucel and the relevant comparators (identified above), the conduct of a conventional indirect treatment comparison was not possible. As such, the use of a MAIC approach was explored as part of a scenario analysis in order to explore adjustments of the pooled tisagenlecleucel population to more closely match that of the von Stackelberg et al. (2016) and Jeha et al. (2006) population, respectively, and hence account for any impact of population differences on OS estimates.[34, 35] Versus the blinatumomab von Stackelberg et al. (2016) study, feedback from UK clinical experts was that it would be reasonable to conclude that patients in the blinatumomab trial were fitter based on the proportion refractory and those with >3 lines of prior therapy.

An overview of the MAIC results for OS are presented below. Full details of the methodology of this approach are presented in Appendix D. The OS benefit observed for tisagenlecleucel in the naïve comparison remained consistent and statistically significant at the 95% confidence level in the MAIC. The resulting HRs are presented in Table 20.

Table 20: Overall survival hazard ratios

Abbreviations: CI: confidence interval; HR: hazard ratio; MAIC: matching-adjusted indirect comparison. Source: ELIANA CSR (31[st] Dec 2017);[47] ; ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30th Jan 2017);[49] von Stackelberg et al. (2016)[35] ; Jeha et al. (2006).[34]

The Kaplan-Meier plot of OS for the matched tisagenlecleucel cohort versus blinatumomab is presented in Figure 23 and for the matched tisagenlecleucel cohort versus salvage chemotherapy (using clofarabine monotherapy as a proxy for FLA-IDA) is presented in Figure 24. In the comparison to blinatumomab, the matched and unmatched curves were seen to be very similar, with the matched curve associated with a slightly higher plateau than the unmatched curve. In the comparison to salvage chemotherapy the matched curve was seen to be associated with slightly lower survival for earlier timepoints, but a higher plateau at the later timepoints than the unmatched curve. In both comparisons, the 95% confidence intervals of the matched and unmatched curves overlapped, indicating that differences between the unmatched and unmatched curves may reflect uncertainty inherent in the sample estimates rather than a true difference in efficacy.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 69 of 185

Figure 23: Overall survival for tisagenlecleucel versus blinatumomab

==> picture [442 x 158] intentionally omitted <==

Shaded regions represent 95% CIs.

Abbreviations: CI: confidence interval; HR: hazard ratio; MAIC: matching-adjusted indirect comparison; OS: overall survival.

Source: Blinatumomab: von Stackelberg et al. (2016);[35] tisagenlecleucel: ELIANA CSR (31st Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

Figure 24: Overall survival for tisagenlecleucel versus salvage chemotherapy

==> picture [442 x 158] intentionally omitted <==

Shaded regions represent 95% CIs. Abbreviations: CI: confidence interval; HR: hazard ratio; MAIC: matching-adjusted indirect comparison; OS: overall survival.

Source: Salvage chemotherapy (using clofarabine monotherapy as a proxy for FLA-IDA): Jeha et al. (2006); tisagenlecleucel: ELIANA CSR (31st Dec 2017);[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

==> picture [42 x 12] intentionally omitted <==

Adverse reactions

Summary of clinical trial safety analysis  As of the latest data cuts reported in this submission, a total of xxx patients had received infusion with tisagenlecleucel and were analysed as part of the safety sets of ELIANA, ENSIGN and B2101J. Based on these safety analyses, the safety profile of tisagenlecleucel is well characterised and was consistent across all three clinical trials.[47-49]  The most frequently occurring AE, regardless of study drug relationship, in all three trials was cytokine release syndrome (CRS), which was reported in xxxxxxxxx), 47 (81.0%) and xxxxxxxxxx patients in the ELIANA, ENSIGN and B2101J trials, respectively.[47-49] CRS is an expected AE in patients infused with tisagenlecleucel, as a class-effect of T-cell directed therapies and an on-target mechanism of action effect. CRS is generally reversible, and can be effectively managed with treatment guidelines.[31]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 70 of 185

 In the ELIANA trial, AEs were reported primarily within eight weeks post-tisagenlecleucel infusion and the incidence of all AEs decreased substantially after this time point. Of note, no patients had CRS after this eight-week period and only xxxxxxxxxxxxxxxxxxxxxx had AEs more than one year post-tisagenlecleucel infusion.[50]  In terms of SAEs, CRS was the most commonly reported SAE regardless of study drug relationship, occurring in xxxxx, xxxxx and xxxxx patients in the ELIANA, ENSIGN and B2101J trials, respectively. Febrile neutropenia was consistently the next most common (xxxxx, xxxxx and xxxxx, respectively), followed by hypotension (xxxxx, xxxxx and xxxxx, respectively).[47-49]  Across all three trials, a total of xx deaths were reported post-tisagenlecleucel infusion. Deaths occurring within 30 days post-tisagenlecleucel infusion were reported for xxx patients (xxxx) and xxx patients (xxxx) in the ELIANA and ENSIGN trials respectively. None of these deaths were due to the tisagenlecleucel infusion.  In B2101J, no deaths occurred within 30 days of the first tisagenlecleucel infusion and deaths were reported for xxxxx patients within 30 days of the last tisagenlecleucel infusion (patients in B2101J could receive more than one infusion). In the period more than 30 days post-tisagenlecleucel infusion, xxxxxxxxxx and xxxxxxxxxx patients died in the ELIANA and ENSIGN trials, respectively. In B2101J, xx patients xxxxxxx died more than 30 days after the last tisagenlecleucel infusion.[47-49]

The safety and tolerability of tisagenlecleucel for the treatment of paediatric and young adult patients with r/r B-cell ALL was evaluated as a secondary outcome in both ELIANA and ENSIGN, and as part of the primary outcome of B2101J.

In all three trials, the safety population included all patients who received at least one infusion of tisagenlecleucel. The assessment of safety was based mainly on the proportion of patients reporting AEs, serious AEs (SAEs), AE of special interest (AESI), deaths, pregnancies and immunogenicity.

Safety in the ELIANA trial was assessed by monitoring and recording potential AEs using MedDRA version 20.0 and the Common Terminology Criteria for Adverse Events (CTCAE) version 4.03. In the ENSIGN trial, reporting of AEs was based on MedDRA version 19.0 and CTCAE version 4.03. In B2101J, AEs were reported using MedDRA version 19.1 and CTCAE version 3.0. In all trials, the grading of cytokine release syndrome (CRS) and graft-versus-host disease (GVHD) was based on protocol-specific grading scales.

==> picture [50 x 11] intentionally omitted <==

Treatment duration and dosage

In ELIANA, xx patients (xxxxx) out of the xx patients enrolled were able to receive tisagenlecleucel which was administered as a single intravenous infusion with a target dose range of 2.0 to 5.0×10[6] tisagenlecleucel cells per kg (for patients ≤50 kg) or of 1.0 to 2.5×10[8] tisagenlecleucel cells (for patients >50 kg). Of the 79 patients that received tisagenlecleucel, xx patients (xxxxx) received tisagenlecleucel doses within the protocol-specified target dose range. xxx patients (xxxx) received a dose above the target range and xxxxx patients xxxxxx received a dose below the target range. The median total tisagenlecleucel dose infused was xxxxxxx[x] cells (range xxxxxxxxxxxxxxx[x] ) and the median weight-adjusted tisagenlecleucel dose infused was xxxxxxx[x] cells/kg (range xxxxxxxxxxxxx[x] ).[47]

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 71 of 185

In ENSIGN, 58 patients (79.5%) out of the 73 patients enrolled were able to receive tisagenlecleucel which was administered as a single intravenous infusion with a target dose equivalent to the ELIANA trial with a range of 2.0 to 5.0×10[6] tisagenlecleucel cells per kg (for patients ≤50 kg) or of 1.0 to 2.5×10[8] tisagenlecleucel cells (for patients >50 kg). Of the 58 patients that received tisagenlecleucel, 49 patients (84.5%) received tisagenlecleucel doses within the protocol-specified target dose range as specified above and nine patients (15.5%) received a below target dose range. The median total tisagenlecleucel dose infused was xxxxxxx[x] cells (range xxxxxxx[x] xxxxxxxxxxx[x] ) and the median weight-adjusted tisagenlecleucel dose infused was 3.55×10[6] cells/kg (range 0.2×10[6] to 5.0×10[6] ).[48]

In the B2101J trial, xx patients of the xx patients enrolled were able to receive tisagenlecleucel which was administered according to a dose escalation schedule (10% on Day 0, 30% on Day 1, and 60% on Day 14 or later, with necessary protocol-specified adjustments where appropriate) with a maximum total dose target range of 1.5×10[7] to 5×10[9] total cells. Full details of any protocolspecific dose adjustments received can be found in the CSR for B2101J.[49] The median total tisagenlecleucel dose infused during the overall study was xxxxxx[x] cells (range xxxxxx[x] xxxxxxxxxxx[x] ) and the median weight-adjusted tisagenlecleucel dose infused during the overall study was xxxxxx[x] cells/kg (range xxxxxx[x] xxxxxxxxxxx[x] ).[49]

==> picture [51 x 11] intentionally omitted <==

Safety analysis in the relevant clinical trials

A summary of the safety results from the ELIANA, ENSIGN and B2101J clinical trials is presented in Table 21 below. Across all three trials, as of the latest data cuts reported in this submission, a total of xx, 58 and xx patients had received infusion with tisagenlecleucel and were analysed in the safety sets of ELIANA, ENSIGN and B2101J, respectively.[47-49]

Table 21: Overall summary of AEs in ELIANA, ENSIGN and B2101J (safety set)

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 72 of 185

aData for B2101J presented in this submission refer to the non-CNS3 ALL cohort only.

*In the B2101J trial, this refers to deaths within 30 days of the last infusion of tisagenlecleucel (no deaths occurred within 30 days of the first infusion). **Value calculated based on 22 deaths post-tisagenlecleucel infusion, and three of these occurring within 30 days of the last infusion. Therefore, 19 occurred 30 days after the last infusion (33.9% of 56).

All deaths during both study follow-up and survival follow-up are summarised. Abbreviations: AE: adverse event; SAE: serious adverse event. Source: ELIANA CSR (31[st] Dec 2017),[47] ENSIGN CSR (6[th] Oct 2017);[48] B2101J CSR (30[th] Jan 2017).[49]

Deaths

A total of xx deaths (xxxxx) occurred in the ELIANA trial post-tisagenlecleucel infusion (data cut-off 31st Dec 2017). xxx patients died within 30 days post-tisagenlecleucel infusion; xxx due to cerebral haemorrhage and xxx due to underlying disease progression. The other 21 deaths occurred more than 30 days post-tisagenlecleucel infusion. Of these, xx deaths (xxxxx) were attributed to underlying disease progression, xxx was due to viral encephalitis, xxx due to systematic mycosis, xxx due to lower respiratory tract infection, xxx due to hepatobiliary disease and xxx due to an unknown reason.[47]

A total of 19 deaths (32.8%) occurred in the ENSIGN trial post-tisagenlecleucel infusion (data cutoff 6th Oct 2017). xxx patients died within 30 days post-tisagenlecleucel infusion; xxx due to underlying disease progression and xxx due to embolic stroke. The other 17 deaths occurred more than 30 days post-tisagenlecleucel infusion. Of these, 15 were attributed to underlying disease progression, one due to acute respiratory failure and one due to complications of transplant surgery.[48]

A total of xx deaths (xxxxx) occurred in the B2101J trial post-tisagenlecleucel infusion (data cut-off 30th Jan 2017). No deaths were reported within 30 days after the first tisagenlecleucel infusion, whereas xxxxx patients (xxxx) died within 30 days after the final tisagenlecleucel infusion, and xx patients xxxxxx) died more than 30 days after the last tisagenlecleucel infusion.[49]

AEs post-tisagenlecleucel infusion, regardless of study drug relationship

AEs regardless of study drug relationship occurred in xxxx patients in ELIANA and B2101J, and xxxxx patients in ENSIGN.[47-49] In the ELIANA trial, regardless of study drug relationship, the most frequent AEs overall were CRS, pyrexia, decreased appetite and hypogammaglobulinaemia, which occurred in at any grade in xxxxx, xxxxx, xxxxx and xxxxx patients, respectively.[47] CRS was also the most common AE regardless of study drug relationship in ENSIGN, followed by hypogammaglobulinaemia, a decreased white blood cell count and anaemia. These AEs occurred in 81.0%, xxxxx, xxxxx and xxxxx patients, respectively.[48] Lastly, in B2101J, decreased white

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 73 of 185

blood cell count was the most common AE regardless of study drug relationship, occurring in xxxxx patients. The next most common AEs were a decrease in haemoglobin, a decreased neutrophil count and CRS, occurring in xxxxx, xxxxx and xxxxx patient, respectively.[49]

A summary of frequently reported AEs post-tisagenlecleucel infusion, regardless of study drug relationship for all three trials is presented in Table 22.

Company evidence submission template for tisagenlecleucel for treating relapsed or refractory B- cell acute lymphoblastic leukaemia in people aged up to 25 years [ID1167] © Novartis Pharmaceuticals Ltd. 2018. All rights reserved Page 74 of 185

Table 22: Summary of AEs reported in ≥10% of patients post-tisagenlecleucel infusion, regardless of study drug relationship (safety set)