Single Technology Appraisal

Bulevirtide for treating chronic hepatitis D [ID3732]

Committee Papers

NATIONAL INSTITUTE FOR HEALTH AND CARE EXCELLENCE

SINGLE TECHNOLOGY APPRAISAL

Bulevirtide for treating chronic hepatitis D [ID3732]

Contents:

The following documents are made available to stakeholders:

1. Response to consultee, commentator and public comments on the Draft Guidance

2. Comments on the Draft Guidance from Gilead

3. Consultee and commentator comments on the Draft Guidance from:

• a. NHS England

4. External Assessment Group critique of company comments on the Draft Guidance

5. External Assessment Group review of PAS price change

6. Company response to External Assessment Group critique

7. External Assessment Group critique of company response on the Draft Guidance

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

Appraisal title

Single Technology Appraisal

Response to consultee, commentator and public comments on the Appraisal Consultation Document (ACD)

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

Type of stakeholder:

Consultees – Organisations that accept an invitation to participate in the appraisal including the companies, national professional organisations, national patient organisations, the Department of Health and Social Care and the Welsh Government and relevant NHS organisations in England. Consultees can make a submission and participate in the consultation on the appraisal consultation document (ACD; if produced). All non-company consultees can nominate clinical experts and/or patient experts to verbally present their personal views to the Appraisal Committee. Company consultees can also nominate clinical experts. Representatives from NHS England and clinical commissioning groups invited to participate in the appraisal may also attend the Appraisal Committee as NHS commissioning experts. All consultees have the opportunity to consider an appeal against the final recommendations, or report any factual errors, within the final appraisal document (FAD).

Clinical and patient experts and NHS commissioning experts – The Chair of the Appraisal Committee and the NICE project team select clinical experts and patient experts from nominations by consultees and commentators. They attend the Appraisal Committee meeting as individuals to answer questions to help clarify issues about the submitted evidence and to provide their views and experiences of the technology and/or condition. Before they attend the meeting, all experts must either submit a written statement (using a template) or indicate they agree with the submission made by their nominating organisation..

Commentators – Commentators can participate in the consultation on the ACD (if produced), but NICE does not ask them to make any submission for the appraisal. Non-company commentator organisations can nominate clinical experts and patient experts to verbally present their personal views to the Appraisal Committee. Commentator organisations representing relevant comparator technology companies can also nominate clinical experts. These organisations receive the FAD and have opportunity to report any factual errors. These organisations include comparator technology companies, Healthcare Improvement Scotland any relevant National Collaborating Centre (a group commissioned by NICE to develop clinical guidelines), other related research groups where appropriate (for example, the Medical Research Council and National Cancer Research Institute); other groups such as the NHS Confederation, the NHS Commercial Medicines Unit, the Scottish Medicines Consortium, the Medicines and Healthcare Products Regulatory Agency, the Department of Health and Social Care, Social Services and Public Safety for Northern Ireland).

Public – Members of the public have the opportunity to comment on the ACD when it is posted on the Institute’s web site 5 days after it is sent to consultees and commentators. These comments are usually presented to the appraisal committee in full, but NICE reserves the right to summarise and edit comments received during consultations, or not to publish them at all, where in the reasonable opinion of NICE, the comments are voluminous, publication would be unlawful or publication would be otherwise inappropriate.

2 of 42

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

Please note: Comments received in the course of consultations carried out by NICE are published in the interests of openness and transparency, and to promote understanding of how recommendations are developed. The comments are published as a record of the submissions that NICE has received, and are not endorsed by NICE, its officers or advisory committees.

1 Based on a transient elastography of ≥8.0 kPA.

3 of 42

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

2 The 1.2x QALY modifier threshold in the EAG’s model is 36.49 years.

4 of 42

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

5 of 42

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

6 of 42

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

3 Please note that BSC results remain unchanged.

7 of 42

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

8 of 42

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

9 of 42

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

10 of 42

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

11 of 42

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

12 of 42

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

13 of 42

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

14 of 42

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

15 of 42

NICE Response Please respond to each comment

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

Comment Type of Organisation number stakeholder name

Stakeholder comment Please insert each new comment in a new row These rates of regression are likely generalisable to a patient with CHD experiencing a combined response. We have nevertheless, as per the committee’s request, carried out a scenario analysis where regression rates in the model are reduced. While this has a larger magnitude of impact on the ICER than reducing progression rates, we do not consider this to be a realistic scenario given that our original modelled estimates are obtained from real-world evidence of regression in viral hepatitis.

3.2 Progression through fibrosis stage for combined responders

The committee considered that the risk of progression through fibrosis stage for the combined responders should be low but not zero. We have explored this by assuming that progression is reduced relative to that of partial responders instead of assuming that it is zero. We have carried out an analysis where we assume that the hazard ratio for progression in combined responders is 20% of that applied to partial responders. That is, 0.08 in combined responders vs. 0.42 in partial responders for fibrosis states Fx->Fx+1 and 0.05 in combined responders vs. 0.26 in partial responders for fibrosis states F4->decompensated cirrhosis (DCC). Results are reported in Error! Reference source not found. .

Table 8: Scenario of progression through fibrosis stage for combined responders

Note : we were unable to set the model to the EAG base case of £48,097 reported in the appraisal consultation document. The EAG base case ICERs therefore represent the closest result we were able to obtain.

3.3 Progression to hepatocellular carcinoma for combined responders

The committee considered that the risk of progression to hepatocellular carcinoma (HCC) for combined responders should be low but not zero. We have explored this by assuming that progression is reduced relative to that of partial responders instead of assuming that it is zero. We have carried out an analysis where we assume that the hazard ratio for progression in combined responders is 20% of that applied to partial responders. That is, 0.07 in complete responders vs. 0.34 in partial responders for all transitions to HCC. Results are reported in Error! Reference source not found. .

Table 9: Scenario of progression to hepatocellular carcinoma for combined responders

16 of 42

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

17 of 42

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

18 of 42

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

19 of 42

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

Comment Type of Organisation Stakeholder comment number stakeholder name Please insert each new comment in a new row document. The EAG base case ICERs therefore represent the closest result we were able to obtain. 4.4 Stopping treatment with virus eradication

20 of 42

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

21 of 42

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

22 of 42

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

23 of 42

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

24 of 42

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

25 of 42

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

26 of 42

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

27 of 42

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

28 of 42

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

29 of 42

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

30 of 42

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

31 of 42

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

32 of 42

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

33 of 42

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

34 of 42

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

35 of 42

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

36 of 42

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

37 of 42

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

38 of 42

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

39 of 42

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

40 of 42

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

41 of 42

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

42 of 42

Bulevirtide for treating chronic hepatitis D [ID3732]

Company response to NICE appraisal consultation document

November 2022

Version 2

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

1

Contents

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

3

Abbreviations

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

List of tables

Table 1: Company’s ACD revised base case cost-effectiveness estimates .............. 12 Table 2: Results of the subgroup analyses for virologic response ............................ 18 Table 3: Results of the subgroup analyses for combined response .......................... 19 Table 4: Scenario applying responses from METAVIR score ≥F2 ............................... 20 Table 5: Scenario applying responses from FibroScan score ≥7.25 kPA ................. 20 Table 6: Scenario responses from FibroScan score ≥8.0 kPA .................................. 20 Table 7: Scenario where baseline age is set to UKHSA mean vs. median ............... 22 Table 8: Scenario of progression through fibrosis stage for combined responders 24 Table 9: Scenario of progression to HCC for combined responders ......................... 25 Table 10: Scenario of regression in combined responders ....................................... 25 Table 11: Scenario of treatment continuation with HCC ........................................... 27 Table 12: Scenario with treatment continuation for virologic responders (Week 48) 28 Table 13: Scenario with treatment continuation for virologic responders (Week 72) 28 Table 14: Scenario of continuing treatment despite virus eradication ....................... 29 Table 15: Scenario of treatment continuation for virologic responders combined with stopping treatment with virus eradication .................................................................. 30 Table 16: Utility gains for patients with SVR in NICE TAs ......................................... 34 Table 17: Utility gains for patients with SVR in the literature ..................................... 34 Table 18: Scenario of assuming 50% of the current utility gain for responders 35 Table 19: Scenario of assuming 75% of the current utility gain for responders 35

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

5

List of figures

Figure 1: EQ-5D scores at Week 48 overall and treatment by arm ........................... 32 Figure 2: Cumulative incidence of compensated cirrhosis in Romeo et al., (2009) and Roulot et al ., (2002) studies ...................................................................................... 38 Figure 3: Cumulative incidence of compensated cirrhosis in Romeo 2009 and model Romeo predicted 2002 studies ................................................................................ 39 Figure 4: Kaplan-Meier decompensation-free survival curves based on HDV RNA status from Kamal et al ., (2020)................................................................................ 40 Figure 5: Comparison of decompensation-free survival of patients from Kamal et al ., (2020) vs. predictions from economic model ............................................................ 40 Figure 6: Kaplan-Meier hepatocellular carcinoma-free survival curve from Yurdaydin et al. , (2018) in patients with and without MVR ......................................................... 41 Figure 7: Comparison of survival of patients from Yurdaydin et al ., (2018) vs. predictions from economic model ............................................................................. 42 Figure 8: Kaplan-Meier survival curve of patients with compensated HBV-HDV cirrhosis from Gheorge et al ., (2005) ........................................................................ 43 Figure 9: Comparison of survival of compensated cirrhosis patients from Gheorge et al ., (2005) vs. predictions from economic model ....................................................... 44 Figure 10: Survival without liver transplantation according to persistent HDV viremia before endpoint from Roulot et al ., (2020) ................................................................ 45 Figure 11: Comparison of survival of F0-F4 HDV RNA+ patients from Roulot et al ., (2020) vs. predictions from economic model ............................................................ 46 Figure 12: Liver-related mortality stratified by MVR status in Yurdaydin et al., (2018) . ................................................................................................................................ 47 Figure 13: Comparison of survival of F0-F4 patients without MVR vs. predictions from economic model ....................................................................................................... 47 Figure 14: Health state occupancy, Bulevirtide arm .................................................. 48 Figure 15: Health state occupancy, BSC arm ........................................................... 49 Figure 16: Overall survival ........................................................................................ 49

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

Executive summary

The Company is grateful for the opportunity to respond to the appraisal consultation document (ACD). While we are disappointed that bulevirtide did not receive an initial positive recommendation for treating chronic hepatitis D (CHD) with compensated liver disease in adults[1] , we are pleased that the appraisal committee recognizes that there is a significant unmet need for effective treatments in this population because the current treatment options are limited.

In the ACD, the Committee noted that the clinical evidence collected from the pivotal MYR 301 clinical study showed bulevirtide is effective compared with current standard of care and acknowledged the large benefit for patients who had treatment with bulevirtide at Week 48. However, the Committee also considered that there are uncertainties surrounding the duration of on-treatment effect for bulevirtide, as well as uncertainties relating to how eligible patients would be identified using METAVIR staging. The Committee noted that this translated into uncertainty around the estimates of cost-effectiveness.

In Section 3.16 of the ACD (pages 14-16), the Committee sets out its preferred assumptions and notes specific areas of uncertainty. The Company has addressed these topics in our response to the ACD in the analyses and discussion that follows. The Committee also requested a range of scenario analyses along with additional data. Where feasible, the Company has addressed each of the areas of uncertainty and has carried out additional scenario analyses, the results of which are reported within the associated topics of this response.

A patient access scheme (PAS) has been approved by PASLU for NHSE&I. The PAS involves a simple discount from list price. In response to the ACD, the confidential net price has been decreased from £ to £ per pack.

1 Bulevirtide is indicated for the treatment of chronic HDV infection in plasma (or serum) HDV-RNA positive adult patients with compensated liver disease

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

7

Summary conclusions are as follows:

1. The company’s proposed positioning in people with METAVIR stage ≥F2

(Topic 1)

• The Company’s proposed positioning for bulevirtide aligns with existing National Institute for Health and Care Excellence (NICE) clinical guideline (CG) 165.

• Subgroup analysis of patients in MYR 301 in METAVIR fibrosis stages ≥F2 according to equivalent FibroScan cut-offs were conducted; the majority of MYR 301 patients ( %)[2 ] had a liver stiffness measurement consistent with METAVIR ≥F2 at baseline. The subgroup analysis and overall population (full analysis set) is therefore relevant to the decision problem. Response rates in the subgroup analysis are consistent with response rates in the overall population (presented in the Company’s original submission) further demonstrating the efficacy of bulevirtide in patients with a high unmet need.

• Cost-effectiveness analysis using response data from the subgroups defined by liver stiffness demonstrate the incremental cost-effectiveness ratio (ICER) to be robust and the ICER for bulevirtide ranging from £22,228 to £24,298 per quality-adjusted life year (QALY) gained when considering the 1.2 severity modifier.

2. The mean age of people diagnosed with hepatitis D in the UK (Topic 2)

• The mean age of people diagnosed with hepatitis D in the UK determined by the UK Health Security Agency (UKHSA) does not alter eligibility for a severity weight when applied to the company base-case. The mean age of people with hepatitis delta virus (HDV) is estimated to be 36.9 years, however the data is skewed meaning that the median age (35.0 years) is a more appropriate measure of average age and is consistent with the average age used in the Company’s base case analysis (35.1 years, as per Spaan et al., 2020) (1).

• UKHSA data demonstrates that the majority of patients (55%) with hepatitis D are young, being 36 years of age or less[3 ] (2).

2 Based on a transient elastography of ≥8.0 kPA.

3 The 1.2x QALY modifier threshold in the EAG’s model is years.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

8

• Cost-effectiveness analysis using the UKHSA data further demonstrates that there is a significant QALY shortfall amongst people living with CHD compared to the expected future health without the condition over the remaining lifetime of the patient; meaning bulevirtide qualifies for the 1.2 severity modifier.

• Bulevirtide is a cost-effective treatment option compared to best supportive care (BSC); the ICER ranging from £24,061 to £24,433 per QALY gained when considering the 1.2 severity modifier and the UKHSA median vs. mean age.

3. Progression/regression rates of combined responders (Topic 3)

• A low but not zero risk of progression through fibrosis stages for combined responders was explored in the model, resulting in a negligible impact on the Company’s base-case ICER (with severity modifier of 1.2) (+£463).

• A low but not zero risk of progression to hepatocellular carcinoma (HCC) for combined responders was explored in the model, resulting in a negligible impact on the Company’s base-case ICER (with severity modifier of 1.2) (+£458).

• A lower probability of fibrosis regression for combined responders in the model had a relatively modest impact on the Company’s base-case ICER (with severity modifier of 1.2) (+£807). Regression rates in the model were informed by observed responses to antiviral treatments in chronic hepatitis B (CHB) and chronic hepatitis delta (CHD) patients in the real-world. Given that these rates of regression were informed by real-world evidence in patients with viral hepatitis, we do not consider scenario analyses for lower regression rates requested by committee to be realistic.

• The Company’s base-case ICERs range from £24,519 to £24,868 per QALY gained across these scenario analyses (including the 1.2 severity modifier).

4. Treatment duration beyond 48 weeks in MYR 301 (Topic 4)

• The Company’s revised base case now includes the EAG’s preferred assumption of continuing treatment in patients who develop HCC. The impact of the ICER is negligible, increasing by £6 from £24,055 to £24,061 per QALY gained (including the 1.2 severity modifier).

• While assuming the same continuation rules for virologic and combined responders increases the ICER, we note that clinicians were unsure that

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

9

virologic responders would have the same continuation rules and that this may only apply to specific cases. The associated ICERs (including the 1.2 severity modifier) were £28,957 to £29,332.

• As requested by the Committee, the impact of stopping treatment in patients with virus eradication was explored. Note, the Company’s revised base-case include this assumption. Continuing treatment in patients with evidence of virus eradication increases the ICER (including the 1.2 severity modifier) from £24,061 (company base-case) to £27,165 per QALY gained.

5. The size of utility gain for combined responders (Topic 5)

• NICE methods guidance stipulates that where possible utilities from the technology’s clinical trials should be used in the economic model. The Company’s choice of regression model for generating utilities was underpinned by observed ceiling effects from the MYR 301 data.

• Utility gains for patients with sustained virologic response (SVR) in other hepatitis technology appraisals has been slightly lower (different by ≤0.02) than that observed for combined responders in MYR 301, however this may be explained by differences in the type of hepatitis infection and population.

• The impact on the ICER of varying the size of the utility gain for combined responders was explored by assuming 75% and 50% of the current MYR 301 responder utility gain; the associated ICER were £24,605 and £25,175 respectively (including the 1.2 severity modifier).

6. The long-term survival for people on standard care, in the absence of bulevirtide (Topic 6)

• The Company has validated the progression and survival rates predicted by the economic model with those in the published CHD literature by comparing Kaplan-Meier plots with survival plots extracted from the model.

• There was close alignment between the model predictions and the published literature, with divergence only observed at later timepoints where costs and outcomes are discounted.

• The analysis demonstrates that that Company’s base-case assumptions associated with the natural history of CHD are appropriate.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

10

The NICE health technology evaluation manual (2022) notes that a higher degree of uncertainty regarding estimates of cost-effectiveness is acceptable for health technologies for which evidence generation is particularly difficult, specifically for rare diseases and innovative technologies (see section 6.2.34, pp. 157).

The majority of the scenarios requested by the Committee have a relatively modest impact on the ICER. In the majority of cases the ICER associated with bulevirtide remains with the £20,000 to £30,000 per QALY gained range normally considered a cost-effective use of NHS resources. This is particularly compelling given that CHD is a rare disease and bulevirtide is an innovative first-in-class treatment with GB orphan drug designation.

The Company accepts a number of the External Assessment Group (EAG) preferred assumptions (the majority of which were not discussed in part 1 of the appraisal committee meeting) and is willing to adjust its preferred base case accordingly.

The Company has therefore updated its current base case assumptions with the following changes:

• i. Average age of people diagnosed with hepatitis D in the UK based on UKHSA median age of 35.0 years, further supported by Spaan et al. (2020).

• ii. Treatment stopped for those with convincing evidence of virus eradication (see section 3.11 of the ACD). Clinical experts confirmed that patients with convincing evidence of virus eradication, treatment would likely be stopped. This scenario had previously been conservatively excluded from the Company base-case, but as the Company is aware that this is a feasible scenario in UK clinical practice, this has now been included in the post-committee base-case.

• iii. MYR 301 data not extrapolated beyond Week 48; note this is a conservative assumption considering the observed trend of increasing response rates across the first 48 weeks of treatment with bulevirtide.

• iv. Age-related utility decrements included.

• v. Health state utilities based on MYR 301.

• vi. Post-liver transplant utility value set to MYR 301.

• vii. Assuming that patients who develop HCC remain on bulevirtide.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

11

The revised Company base case cost effectiveness results are presented in Table 1. It can be seen that the severity-weighted ICER lies below the willingness-to-pay threshold of £30,000 per QALY gained using a severity modifier of 1.2x, demonstrating that bulevirtide not only provides significant benefits in terms of improved patient prognosis, but is a cost-effective treatment for the NHS.

Table 1: Company’s ACD revised base case cost-effectiveness estimates

The ICER does not reflect additional benefits associated with bulevirtide

The Committee noted in the ACD that there are several benefits of bulevirtide that were not captured by the QALY calculation.

Hepatitis D is a rare disease and bulevirtide is the first licensed treatment option for the treatment of CHD. The Committee were of the opinion that bulevirtide is an innovative treatment option which is well tolerated, and which addresses this unmet need. Bulevirtide is a first-in-class medicine with GB orphan designation (PLGB 50662/0002/OD1) and promising innovative medicines (PIM) designation. The Committee heard from clinical experts that bulevirtide represents a step change in the management of the CHD.

The Committee acknowledged that bulevirtide reduces the viral load in infected people thereby preventing the spread of infection; a significant benefit not captured by the QALY calculation. Additionally, the Committee heard from a patient advocacy group representative and clinicians concerning the stigma associated with blood-borne viruses. Clinical advisers have advised the Company that the introduction of a licenced treatment option such as bulevirtide is expected to bring about healthcare systemwide benefits as clinicians and patients will now have an effective treatment option such as encouraging HDV testing and reducing regional variation in practice.

4 Please note that BSC results remain unchanged.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

12

Early diagnosis and treatment of HDV is expected to positively impact health outcomes and reduce costs associated with healthcare resource use. Reducing and/or delaying the progression to more severe disease including decompensated cirrhosis (DCC), liver cancer, and liver transplant, is expected to result in a significant reduction to healthcare costs and a substantial QALY gain. In the appraisal committee meeting, the Committee acknowledged that HDV disproportionately affects some groups of people such as Black African people and economic migrants.

The Company recognises that the pivotal registrational phase III study of bulevirtide, MYR 301, is ongoing. However, the Company is of the strong opinion that based on the totality of evidence, the uncertainty associated with an innovative treatment for a rare disease is proportionate; bulevirtide represents a cost-effective treatment option and should therefore be recommended.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

13

Topic 1 The Company’s proposed positioning in people with METAVIR stage ≥F2

ACD section 3.3 “The company presented data from the full analysis set from MYR 301, which included people with all METAVIR fibrosis stages (F0 to F4), so it was unclear why the company positioned bulevirtide only for METAVIR stage F2 and above…”

ACD section 3.4 “METAVIR staging is done using a liver biopsy, which is invasive and carries a morbidity and mortality risk. Therefore, many people refuse this procedure… i t would be useful for the company to present data using transient elastography rather than liver biopsy (METAVIR staging) to assess fibrosis”

Company response:

• Existing NICE guidelines CG165 advise that peginterferon alfa-2a (PEG-IFN) should only be initiated in patients co-infected with hepatitis D “ who have evidence of significant fibrosis (METAVIR stage greater than or equal to F2 or Ishak stage greater than or equal to 3) ”. The Company’s proposed positioning for bulevirtide aligns with this.

• Transient elastography (FibroScan) scores were collected for all patients in the MYR 301 study. We were able to identify subgroups of patients in MYR 301 in fibrosis stages greater than or equal to F2 (≥F2) according to published FibroScan cut-offs (≥7.25 kPa or ≥8.0 kPa) and found that % of MYR 301 patients were ≥F2.

• Scenario analyses demonstrate the ICER to be robust in the subgroup of patients with FibroScan scores considered to align with fibrosis stage ≥F2 in UK clinical practice.

1.1 Summary

The committee queried the Company’s proposed positioning of bulevirtide for the treatment of CHD in adult patients with compensated liver disease and evidence of significant fibrosis (METAVIR stage ≥F2), given the METAVIR F-stage distribution of patients recruited to MYR 301. They further queried how patients with METAVIR stage ≥F2 would be identified in clinical practice. In section 1.2, we explain that the

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

14

identification of patients with significant fibrosis is already required to identify patients eligible for treatment with peginterferon alpha-2a (PEG-IFN) in the UK and that this is based on transient elastography (FibroScan) score cut-off points.

Liver stiffness, as measured by transient elastography (FibroScan) was collected for all patients enrolled onto the pivotal MYR 301 clinical study. As a result, it was possible to assign patients in MYR 301 to METAVIR fibrosis stages based on their FibroScan score in a manner aligned with UK clinical practice, and identify subgroups for patients with METAVIR stage ≥F2 according to their FibroScan score. When using this approach, we found that an overwhelming majority (≥ %) of MYR 301 patients at baseline in the bulevirtide 2 mg and delayed treatment arms would be deemed to be ≥F2 according to UK clinical practice. In summary, we have demonstrated that the proposed positioning of bulevirtide is in line with the existing NICE guideline CG165, that identification of patients in METAVIR fibrosis stage ≥F2 is feasible in clinical practice and that the MYR 301 data are generalisable to that population.

Furthermore, in section 1.3 we undertake a scenario analysis in the model where we incorporate the responder data for MYR 301 patients identified as being ≥F2 according to both FibroScan and METAVIR score. The Company’s ICERs remain largely unchanged in these scenarios, varying only slightly compared with the base case.

1.2 Identification of patients in METAVIR fibrosis stage ≥F2 in clinical practice The treatment options for patients with chronic hepatitis delta (CHD) are limited. NICE guideline CG165 recommends that adults with chronic hepatitis B (CHB) and hepatitis delta infection, who have evidence of significant fibrosis (METAVIR stage ≥F2 or Ishak stage ≥3), should be offered a 48-week course of off-label PEG-IFN (3). Similarly, bulevirtide is proposed to treat adults with CHD with METAVIR stage ≥F2, albeit patients are required to have compensated liver disease and should not have responded well enough to a prior course of interferon-based therapy (hereafter referred to as IFN-based therapy), should have an intolerance to IFN-based therapy, or should have a contraindication to IFN-based therapy.

To determine METAVIR stage, and thus level of liver fibrosis, patients are required to undergo a liver biopsy. However, as discussed in the draft guidance consultation, clinical experts highlighted that liver biopsy is an invasive procedure which carries a morbidity and mortality risk to the patient. As such, many patients refuse to undergo

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

15

this procedure. Clinical experts explained that in clinical practice, transient elastography (FibroScan), a non-invasive test recommended in NICE guideline CG165 to assess liver disease in all adults diagnosed with chronic hepatitis B (CHB), is widely used to assess eligibility for PEG-IFN in CHD patients. Experts highlighted that they would like to use transient elastography to determine the patient eligibility for treatment with bulevirtide, as opposed to undertaking liver biopsy.

FibroScan can identify significant fibrosis in patients with viral hepatitis and advanced liver disease to a good degree of accuracy (4, 5). A FibroScan threshold that corresponds to significant fibrosis (METAVIR stage ≥F2) in CHD has not yet been published, perhaps due to the orphan nature of the disease, and consultation with clinical experts confirms that heterogeneity exists regarding the appropriate cut-off value. In the absence of a clinical expert consensus, we conducted a literature search to identify potential cut-offs that could be applied to the Company’s FibroScan data. Based on this literature search, we identified two thresholds. A meta-analysis by Qi et al ., (2018) assessing the diagnostic accuracy of transient elastography in 7,808 CHB patients identified an optimal threshold value of ≥7.25 kPa at baseline to determine the presence of significant fibrosis (6). In addition, 2021 European Association for the Study of the Liver (EASL) clinical practice guidelines on non-invasive tests for evaluation of liver disease and severity and prognosis strongly recommend a FibroScan score of ≥8.0 kPa to confirm METAVIR fibrosis score ≥F2 (4).

FibroScan scores were collected in the pivotal MYR 301 study in all patients at baseline and Week 48. We have utilised both literature cut-offs to define subgroups of patients with a METAVIR fibrosis score of ≥F2 given their FibroScan score. % and % of patients in the bulevirtide 2 mg and delayed treatment arms in MYR 301 had a FibroScan score of ≥7.25 kPa and ≥8.0 kPa, respectively, indicating that an overwhelming majority of patients in MYR 301 study were in line with the proposed positioning in patients with a METAVIR fibrosis score of ≥F2.

Furthermore, we carried out a scenario analysis in the model using the response rates (both combined and virologic) of these subgroups of patients in fibrosis stage ≥F2. The response rates of the subgroups, compared with the overall population, are shown in Table 2. Results of the scenario analyses in the subgroups of patients with a transient elastography score aligned with METAVIR stage ≥F2 in UK clinical practice can be

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

16

found below. We also present, alongside this, the results of the subgroup analysis based on the METAVIR score collected in MYR 301 (noting that biopsy data were unavailable for of patients at baseline across the delayed treatment and bulevirtide 2 mg arms). The METAVIR scores may not be missing at random, given that more advanced patients may have been more likely to have had a liver biopsy in the past and may therefore have been less willing to have one as part of MYR 301. The results of both scenario analyses show the ICER to be robust in the subgroup of patients estimated to be in ≥F2 according to their FibroScan score.

1.3 Scenario using transient elastography (FibroScan)

Transient elastography is the most widely available and validated non-invasive test to assess the level of fibrosis in patients with advanced liver disease and is recommended by both the American Association for the Study of Liver Diseases (AASLD) and EASL (4, 7). EASL clinical practice guidelines on non-invasive tests for the evaluation of liver disease severity and prognosis stipulate that the diagnostic accuracy of transient elastography for detecting significant fibrosis in patients with advanced liver disease is good, with diagnostic accuracy (area under the receiver operating characteristic curve [AUROC]) around 0.85 (4). Similar AUROCs were detected in meta-analyses describing the diagnostic accuracy of transient elastography for predicting liver fibrosis in patients with CHB. Meta-analyses of studies comprising CHB patients define an optimal cut-off value of transient elastography for diagnosing METAVIR stage ≥F2 as 7.2 kPa (5, 6). However, differences in the cut-off value for METAVIR stage ≥F2 were present in the studies analysed in the metaanalyses, supporting claims by clinical experts that there is some heterogeneity regarding the appropriate cut-off values.

As a result, we have explored three subgroup analyses based on combining response at Week 24 and 48 by METAVIR Score (≥F2) with two published corresponding baseline liver stiffness measures (FibroScan ≥8.0kPA and FibroScan ≥7.25 kPA) (4, 6). As no extrapolations could be carried out for the subgroups in time for this response, the extrapolations are derived using the odds ratios of response rates between the extrapolated and observed data in the overall population. The results of these analyses are presented in Table 4 to Table 6.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

17

Table 2: Results of the subgroup analyses for virologic response

• Note that % of patients had missing data for METAVIR fibrosis score at baseline, hence the much smaller sample size. TE = transient elastography (e.g., FibroScan).

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

18

Table 3: Results of the subgroup analyses for combined response

• Note that % of patients had missing data for METAVIR fibrosis score at baseline, hence the much smaller sample size.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

19

Table 4: Scenario applying responses from METAVIR score ≥F2

Table 5: Scenario applying responses from FibroScan score ≥7.25 kPA

Table 6: Scenario responses from FibroScan score ≥8.0 kPA

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

20

Topic 2 The mean age and cirrhosis status of people diagnosed with hepatitis D in the UK

ACD section 3.7 “People in Spaan et al. had a baseline age of 35 years and 60% had cirrhosis. In MYR 301 the baseline age was 42 years and 47% had cirrhosis... The company also presented data published by Public Health England (now the UK Health Security Agency [UKHSA]) on routine blood-borne virus testing. The median age between 2011 to 2020 was around 36 years ACD section 3.16 “The committee noted that it would like to see the mean age and cirrhosis status of UK patients at diagnosis based on UKHSA data”

Company response:

• The mean age of people diagnosed with hepatitis D in the UK determined by UKHSA does not alter the eligibility for a severity weight when applied to the company base-case.

• A significant proportion of patients with hepatitis D are young, and the inability to apply a severity modifier due to a fraction of patients being older is discriminatory towards this patient population with a large unmet need.

2.1 Summary

In the absence of UK clinical study sites in the pivotal MYR 301 clinical study, the Company felt it was most appropriate to use baseline characteristics from Spaan et al. (2020), a retrospective analysis of 107 patients with CHD in the UK (4), in the economic model. Data on the median age of patients with CHD, collected by UKHSA from 2011 to 2020, was also presented to the committee, however the committee considered that UKHSA data on mean (rather than median) age at baseline would be more helpful to increase certainty in the cost-effectiveness results and the severity weighting applied to bulevirtide, despite the age distribution being skewed to the left e.g., the majority of patients being young.

UKHSA confirmed that the median age was 35.0 years (n= patients) for patients when HDV ribonucleic acid (RNA) was first detected with a mean age years. When considering the cohort of patients who are currently alive, the median age remained at 35 years (n= patients) with the mean age for patients when HDV

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

21

RNA was first detected decreased to years (5). This data supports the Company’s original base case analysis which assumed an average age of 35.1 years as per Spaan et al. (2020) (4). The data on mean age of patients with CHD at baseline in the UK supplied by UKHSA has a minor impact on the ICER (see Table 7 below) and does not alter the eligibility for a severity weighting of 1.2 when applied to the Company base-case.

UKHSA data on baseline age at diagnosis implies that the majority of CHD patients in the UK are young, approx. 55% being age years or younger which corresponds to the threshold for the severity modifier in the EAG’s model. The Company is therefore of the opinion that the severity modifier should be applied. The Committee noted in the draft guidance the high disease burden of chronic hepatitis D, therefore we believe it would be unreasonable to characterise CHD as a non-severe condition. Failing to apply a severity weighting as a result of a fraction of patients being older is potentially discriminatory towards the large cohort (55%) of young patients diagnosed with hepatitis D who have a large unmet need.

2.2 Scenario based on mean age from UKHSA

The committee considered basing the baseline age for patients with cirrhosis on diagnosis in the UK with CHD on UKHSA data. Data on cirrhosis status were not provided by UKHSA however the mean age of all patients when they first tested positive for HDV RNA has been explored in the model by changing the baseline age

to years. Table 7 shows that the ICER (including the 1.2 severity modifier) increases from £24,061 to £24,433 noting that both analyses qualify for the 1.2 severity modifier.

Table 7: Scenario where baseline age is set to UKHSA mean vs. median

*Average age based on UKHSA (2022) mean of years.

†Average age based on UKHSA (2022) median of 35.0 years.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

22

Topic 3 Progression/regression rates of combined responders

ACD section 3.9 “Clinical experts agreed with the company that combined responders would have a low risk of progression through fibrosis stages, but argued that this would not be zero because this group could still have detectable levels of virus. They added that even combined responders may still be at risk of hepatocellular carcinoma. Clinical experts further explained that it is plausible that fibrosis regression could occur in combined responders, but added that the company’s assumed transition probabilities for fibrosis regression seemed high.”

Company response:

• Including a small but non-zero risk of progression for combined responders in the model had a minor impact on the company’s base case ICER.

• Reducing regression rates in the model had a more substantial impact, but given these rates of regression were informed by real-world evidence in patients with viral hepatitis, we do not consider the requested scenario analysis to be realistic.

3.1 Summary

Data from MYR 301 are currently too immature to reliably inform the impact of bulevirtide on rate of progression. However, it is reasonable to assume that a patient who has had a ≥99% (≥2-log10) reduction in their viral RNA load and alanine aminostransferase (ALT) normalisation will have a substantial benefit approaching zero progression if their hepatitis B infection is well controlled. However, we have carried out scenario analyses in the following sections that demonstrate that assuming a low rate of progression relative to virologic-only responders has a low impact on the ICER compared with the Company base case.

Similarly, the estimates for regression in the model were based on the published literature in viral hepatitis. The source informing regression from the compensated cirrhosis (CC) (F4) to F3 health states was Farci et al., (2004) (8), a longitudinal study of 41 CHD patients based in Italy. Thirty-six patients with CHD who participated in a randomised controlled trial of a 48-week course of high (9 million units) or low (3 million units) doses of PEG-IFN or no treatment were followed for an additional 2 to 14 years. The mean follow-up time was 10.8 years. The regression rates in the model were

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

23

informed by the regression rates observed in patients who had sustained biochemical response in Farci et al., (2004). For F3 to F2, regression rates were sourced from Marcellin et al., (2013) (9), a 5-year follow-up study of HBV mono-infected patients who had been enrolled in a 48-week randomised clinical trial where they had been treated with tenofovir disoproxil fumarate (TDF) or adefovir dipivoxil. In the follow-up study, patients received treatment with TDF. Six hundred and forty-one patients were recruited from 80 different study locations including the US, Canada, France, Turkey and the UK. Five hundred and eighty-five (91%) of these patients entered the openlabel phase (follow-up study) and 489 patients (76%) completed 240 weeks of the study. Regression rates in the model were informed by the proportion of patients who experienced viral suppression while on treatment with TDF.

These rates of regression are likely generalisable to a patient with CHD experiencing a combined response. We have nevertheless, as per the committee’s request, carried out a scenario analysis where regression rates in the model are reduced. While this has a larger magnitude of impact on the ICER than reducing progression rates, we do not consider this to be a realistic scenario given that our original modelled estimates are obtained from real-world evidence of regression in viral hepatitis.

3.2 Progression through fibrosis stage for combined responders

The committee considered that the risk of progression through fibrosis stage for the combined responders should be low but not zero. We have explored this by assuming that progression is reduced relative to that of partial responders instead of assuming that it is zero. We have carried out an analysis where we assume that the hazard ratio for progression in combined responders is 20% of that applied to partial responders. That is, 0.08 in combined responders vs. 0.42 in partial responders for fibrosis states Fx->Fx+1 and 0.05 in combined responders vs. 0.26 in partial responders for fibrosis states F4->decompensated cirrhosis (DCC). Results are reported in Table 8.

Table 8: Scenario of progression through fibrosis stage for combined responders

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

24

3.3 Progression to HCC for combined responders

The committee considered that the risk of progression to hepatocellular carcinoma (HCC) for combined responders should be low but not zero. We have explored this by assuming that progression is reduced relative to that of partial responders instead of assuming that it is zero. We have carried out an analysis where we assume that the hazard ratio for progression in combined responders is 20% of that applied to partial responders. That is, 0.07 in complete responders vs. 0.34 in partial responders for all transitions to HCC. Results are reported in Table 9.

Table 9: Scenario of progression to HCC for combined responders

3.4 Regression in combined responders

The committee considered that the Company’s assumed transition probabilities for fibrosis regression appeared high. We have explored this by carrying out an analysis where we reduced the probability of fibrosis regression for combined responders by 50% (4.41% for CC (F4) ➔ F3; and 6.65% for F3 ➔ F2 in the scenario). Results are reported in Table 10. It can be seen that the impact on the Company’s base case ICER is modest, resulting in an increase of £807 however the ICER remains well below the ICER threshold of £30,000 per QALY gained.

Table 10: Scenario of regression in combined responders

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

25

Topic 4 Treatment duration beyond 48 weeks in MYR 301

ACD section 3.11 “Clinical experts broadly agreed with the company’s model assumptions for combined responders and non-responders but were less sure of what would happen for virological responders…Clinical experts added that treatment would also likely continue for combined or virological responders who develop hepatocellular carcinoma, and that for people with convincing evidence of virus eradication, treatment would likely be stopped.

The committee agreed with the clinical experts’ assumptions but noted that there is remaining uncertainty around whether the stopping rules assumed by the company are aligned with those used in MYR 301 until data beyond 48 weeks becomes available”

Company response:

• The Company has carried out a scenario analyses which explores the impact of continuing treatment in patients who develop HCC and relaxing the treatment continuation criteria to achievement of virologic response only.

• While assuming the same continuation rules for virologic and combined responders increases the ICER, we note that clinicians were unsure that virologic responders would have the same continuation rules and that this may only apply to specific cases.

• The more realistic scenario that patients with virus eradication could discontinue treatment substantially offsets any effect from a minority of virologic responders continuing treatment.

4.1 Summary

The Company has included a range of scenarios to explore how treatment continuation criteria impact on estimates of cost-effectiveness. These include:

• Continuing treatment with bulevirtide in patients who develop HCC.

• Assuming the same continuation rules for virologic responders as combined responders, either at Week 48 or Week 72.

• Assuming that patients with undetectable HDV RNA discontinue bulevirtide.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

26

Continued treatment for HCC patients is a scenario previously explored by the EAG and included in their base case, with negligible impact on the ICER.

With respect to the scenarios applying the same continuation rules for virologic-only responders, we note from the ACD that clinicians appeared unsure about this assumption, and that it might apply only in specific circumstances e.g., “ if a patient had a virological response but high ALT for reasons other than hepatitis, for example fatty liver disease or alcohol use .” We therefore consider that the results of this scenario analysis represent a situation where all patients with a virological response remain on treatment.

Finally, the scenario where patients with undetectable HDV RNA discontinue treatment is a realistic one given this approach is taken for other viral hepatitis treatments. In isolation, this scenario leads to a substantial reduction in the ICER as can be seen in Table 14. Applying this clinically realistic scenario to the more unrealistic scenario that all virologic-only responders continue treatment substantially reduces the unfavourable impact of the latter assumption.

4.2 Treatment continuation with hepatocellular carcinoma

The committee considered that treatment should be continued for people who develop HCC. We have explored this by carrying out an analysis where we captured the cost of treatment for HCC patients in the model. Note that this was a scenario that had already previously been introduced by the EAG to which the model is insensitive. Furthermore, the Company has now adopted this as part of their new base case. For comparison purposes, we therefore compare the results of excluding costs of HCC patients. Table 11 shows that excluding treatment costs for patients who develop HCC has a negligible impact on the ICER which decreases from £24,061 to £24,055.

Table 11: Scenario of treatment continuation with HCC

*Impact of no treatment costs in patients who develop HCC.

†Impact of including treatment costs in patients who develop HCC.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

27

4.3 Treatment continuation for virologic responders

The committee considered having the same treatment continuation assumptions for virologic responders as for combined responders. We have explored this in the model by changing the response criteria in the existing model Settings sheet. That is, at Week 48 non-responders discontinue treatment but virologic responders remain on treatment without any further assessment at Week 72. We then carry out a further scenario where the assessment for non-response is instead carried out at Week 72 instead of Week 48 (which uses the model extrapolations in the Company’s scenario). Results are reported in Table 12 and Table 13.

This scenario leads to the largest increase in the ICER, but we note that the committee also discussed the potential for patients with virus eradication to discontinue treatment, modelled in the next scenario. We also explore the impact of combining the present scenario with that of discontinuation of treatment with virus eradication.

Table 12: Scenario with treatment continuation for virologic responders (Week 48)

Table 13: Scenario with treatment continuation for virologic responders (Week 72)

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

28

4.4 Stopping treatment with virus eradication

The committee considered having the treatment stopped for those with convincing evidence of virus eradication. In the absence of specific feedback from clinicians regarding what constitutes “convincing evidence”, we have explored this by assuming that patients with a combined response who had undetectable HDV RNA at 48 weeks discontinue treatment 52 weeks later. This is achieved by removing the costs of bulevirtide from Week 120 onward for the proportion of combined response patients with undetectable HDV RNA at 48 weeks while maintaining

them in the complete response health states. Note that this will still overpredict costs given that, for simplicity, we do not remove the costs of hepatitis D monitoring. The results of this scenario are presented in Table 14. As the Company has now incorporated this into their updated base case, for comparison purposes, we therefore compare the results of excluding the assumption of treatment discontinuation. It can be seen that continuing treatment for all patients increases the ICER (with severity modifier of 1.2) from £24,061 to £27,165 per QALY gained.

Table 14: Scenario of continuing treatment despite virus eradication

*Impact of continuing treatment for all patients including those with convincing evidence of virus eradication.

†Impact of stopping treatment for patients with convincing evidence of virus eradication.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

29

4.5 Treatment continuation for virologic responders combined with stopping treatment with virus eradication

In section 4.3 we explored changing the response criteria to virologic. That is, at Week 48 non-responders discontinue treatment but virologic responders remain on treatment without any further assessment at Week 72. Below we combine that scenario with that in section 0 where patients with virus eradication are assumed to discontinue treatment with bulevirtide. As previously, we deduct the costs of bulevirtide from Week 120 onward for the proportion of patients with undetectable HDV RNA at 48 weeks while maintaining them in the complete response health states (now assumed within our Company base case). In this scenario, the proportion of patients with undetectable RNA is , as the definition of a complete

responder is virologic whereas in section 0 it was combined response. The results of this scenario are presented in Table 15.

Table 15: Scenario of treatment continuation for virologic responders combined with stopping treatment with virus eradication

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

30

Topic 5 The size of the utility gain for combined responders

ACD section 3.12 “The committee was less certain about the size of the utility gain that should be applied. It noted the lack of justification for the Tobit approach and highlighted that the resulting utility gain from the regression model was not statistically significant. It recalled that in previous appraisals of hepatitis C, combined response was associated with a smaller utility gain than assumed by the company. The committee concluded that the size of the utility benefit for combined responders was uncertain.”

Company response:

• NICE methods guidance stipulates that where possible utilities from the technology’s clinical trials should be used in the economic model. Our choice of regression model for generating utilities was underpinned by observed ceiling effects in the MYR 301 data.

• Utility gain for patients with sustained virologic response (SVR) in other hepatitis appraisals has been slightly lower (different by ≤0.02) than that observed for combined responders in MYR 301, however this may be explained by differences in the type of hepatitis infection and population.

5.1 Summary

The committee queried the method used for deriving the utility gain of responders from the MYR 301 data and requested a comparison with the utility values used in other relevant technology appraisals (TAs). In section 5.2 we provide more information regarding the observed ceiling effect that justifies the use of the Tobit regression model.

In section 5.3 we present the results of a literature search covering both prior NICE TAs and the published literature. The results of this search showed high heterogeneity, highlighting that utility gain can vary significantly by population sampled. The values used for SVR in other NICE TAs were broadly in line with those obtained from MYR 301 using our Tobit regression model. However, NICE methods stipulate that the preferred source of utility in an economic model is to use “ EQ-5D reported by patients/carers in a relevant study ” (10). As the MYR 301 study is the only study that

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

31

reports utility gain from the relevant population of CHD patients, then the data from MYR 301 are the appropriate values to be used in the model.

Notwithstanding this, we have explored in section 5.4 the impact of utility gain in a scenario analysis by assuming 50% or 75% of the current utility gain for responders. Results for both cases are reported in Table 18 and Table 19, in which it can be seen that this has a modest impact on the Company’s base case cost-effectiveness results.

5.2 Justification for use of the Company’s Tobit model for utility gain

The committee noted the lack of justification for the Company’s Tobit model around median (CLAD [censored least absolute deviations] regression) approach to deriving utility gain for responders.

EuroQol 5 Dimension (EQ-5D) index scores from MYR 301 are presented in Figure 1, which show that the distribution of the EQ-5D index was significantly skewed to the left, with the peak of the histogram on the right; approximately 40-50% of the subjects obtaining the highest score. This strongly suggests that a ceiling effect was present in the data, which occurs when a high proportions of subjects in a study have maximum scores on the observed variables.

Figure 1: EQ-5D scores at Week 48 overall and treatment by arm

==> picture [499 x 152] intentionally omitted <==

When data have pronounced ceiling effects, the use of ordinary least squares (OLS)

regression violates the statistical requirement for linearity of conditional expectation, leading to inaccurate predictions of preference-based scores and inaccurate identification of predictor variables (11).

The Tobit model is preferable over OLS regression when a ceiling effect is present or the dependent variable is censored. The CLAD approach is based on an assumption that the median will be more robust to ceiling effects than the mean. The coefficients

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

32

are estimated so as to minimize the sum of the absolute deviations from the regression line (12, 13). The Company therefore considered that the Tobit CLAD regression comprised the most appropriate approach to deriving utility gain of responders from the MYR 301 data.

5.3 Utility gain for responders from past technology appraisals

The committee considered having alternative estimates of utility gain for combined responders, based on previous hepatitis technology appraisals (TAs). The Company has explored this issue by reviewing previous TAs in hepatitis B and C, supplemented with data from the literature by re-examining the papers retried in our health-related quality of life systematic literature review. In hepatitis B TAs, no utility gain was assumed for virologic responders who were not in SVR and patients with SVR were assumed to have a small utility decrement (1%) relative to the general population. In TAs for hepatitis C we could only identify utility gains for patients with SVR. These are provided in Table 16 below.

It can be seen that these utility gains ranged from 0.03 to 0.05. However, we note the committee’s comment from TA413 that “ where available, it prefers utility values collected from the clinical trials used to inform the effectiveness of the intervention under evaluation to those estimated from other sources ." Given that the value used in the economic model was directly from the MYR 301 trial data, it is therefore appropriate to use this value in the model as it was obtained from a hepatitis D-infected population in which combined response might not unreasonably lead to greater utility gains.

Values in the literature show high heterogeneity, ranging from 0.053 to 0.2 in various regression models (Table 17), demonstrating how much this can vary between different cohorts of patients. This variability further supports the need to utilise the utility gain from responders collected in the MYR 301 study, as it will be representative of a cohort of patients receiving treatment with bulevirtide for CHD.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

33

Table 16: Utility gains for patients with SVR in NICE TAs

Table 17: Utility gains for patients with SVR in the literature

5.4 Alternative scenarios for responder utility gains

The committee requested to see alternative estimates of utility gain for combined responders, based on previous hepatitis appraisals. As explained in the previous section, utility gains for patients in SVR have been in the same range as those observed for combined responders in MYR 301. However, recognizing that the patients in MYR 301 were not in SVR, we explore two scenarios where the current utility gain in the model ( ) is reduced by 50% and 75% in Table 18 and Table 19 below. In both scenario analyses it can be seen that bulevirtide is associated with

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

34

substantially higher QALYs compared to BSC, with the Company’s base case ICER being below £30,000 per QALY gained with the severity modifier.

Table 18: Scenario of assuming 50% of the current utility gain for responders

Table 19: Scenario of assuming 75% of the current utility gain for responders

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

35

Topic 6 The long-term survival for people on standard care, in the absence of bulevirtide

ACD section 3.15 “The committee added that many of the EAG’s preferred assumptions around the natural history modelling of chronic hepatitis D may also affect the severity weighting calculations because they affect QALYs accrued by people having standard care…It added that validation of the model predictions for people on standard care using external literature sources would be helpful, along with graphical representations of health state occupation over time.”

Company response:

• As requested, we have validated the progression and survival rates predicted by the economic model with those in the published CHD literature by comparing Kaplan-Meier plots with survival plots extracted from the model.

• There was close alignment between the model predictions and the published literature, with divergence only observed at later timepoints where costs and outcomes are discounted

6.1 Summary

The committee requested validation of the predictions of disease progression for patients on best supportive care from the economic model using the external published literature. The Company has carried this out for a number of model transitions, including those to more severe fibrosis states, to HCC and for overall survival. It can be seen from the superimposed Kaplan-Meier/survival plots that the model predictions align closely with those observed in the CHD literature, with the curve shapes only diverging at later timepoints. This is likely a feature of time-changing hazards of progression and death in the real-world, whereas the model uses constant transition probabilities conditional on response. Introducing time-dependent transition probabilities would require substantial reworking of the model and increase model complexity. Furthermore, in many cases divergence from the observed data only occurred after many years, by which time outcomes in the model are highly discounted and thus less likely to impact the ICER.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

36

We have furthermore, in section 6.3, produced plots of model health-state occupancy over time as well as an overall survival plot for the two arms to facilitate decisionmaking.

6.2 Progression/survival rates in the CHD literature

The validation of the long-term survival for people on BSC in the absence of bulevirtide was conducted based on the following factors:

1. Availability of Kaplan-Meier survival data on HDV.

2. Availability of information on granular fibrosis stages and/or data specific to compensated cirrhosis patients.

3. Data specific to untreated patients or patients without treatment response (as these would be assumed to be most similar to BSC patients in the model).

Kaplan-Meier curves from the selected natural history studies were digitized using the program Plot Digitizer. Baseline demographics (e.g., baseline fibrosis distribution, patient age, sex distribution) were aligned with the natural history studies based on available data. In several studies, fibrosis stage data were only available for compensated cirrhotic (F4) vs. non-cirrhotic (F0-F3) health states. In these cases, the relative distribution of patients across F0-F3 was based on Romeo et al., (2009) (19). Model outcomes for advanced liver-disease events were compared against the digitized Kaplan-Meier curves from these natural history studies based on visual inspection. We have explored this in several studies, described in detail below.

6.2.1 Compensated cirrhosis

Given the availability of granular information regarding the distribution of patients from F0-F4 health states in the Romeo et al., (2009) study, this study was selected for the validation of the economic model outcomes as compared to HDV natural history. In Romeo et al., 2009, the cumulative probability of cirrhosis at 20 years was 55% with an incidence rate of 4% per year in the overall (F0-F3) non-cirrhotic patients (Figure 1) (19). Given that the values were not reported granularly (i.e., F2-F3, F3-F4), individual transition probabilities could not be derived.

Notably, these estimates from Romeo et al., (2009) may even be conservative, as the Roulot et al., (2020) study of a French retrospective cohort of HDV patients estimated

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

37

a 5-year risk of cirrhosis of 49.4% in non-cirrhotic patients (notably including both treated and untreated patients; Figure 2) (20). Further, in this study, where 407

(36.6%) patients had significant or severe fibrosis (i.e., METAVIR score ≥F2) at baseline, among new cirrhotic patients after a median follow-up of 3.0 years, 166 out of 174 (95.4%) patients had been classified as having METAVIR score ≥F2 at referral

(20). These data support a fast rate of progression in patients with late-stage fibrosis.

Figure 2: Cumulative incidence of compensated cirrhosis in Romeo et al., (2009) and Roulot et al ., (2002) studies

==> picture [446 x 304] intentionally omitted <==

As shown in Figure 3 below, the model predictions for the incidence of compensated cirrhosis amongst the patients who are F0-F3 at model start is generally in alignment with findings from the Romeo et al ., (2009) study. The incidence is slightly higher overall which may be supported given the results observed in Roulot et al ., (2020) (20).

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

38

Figure 3: Cumulative incidence of compensated cirrhosis in Romeo 2009 and model Romeo predicted 2002 studies

==> picture [417 x 241] intentionally omitted <==

----- Start of picture text -----
Natural history Model predicted
----- End of picture text -----

6.2.2 Decompensated cirrhosis

In the Kamal et al., (2020) study, 337 patients with anti-HDV positivity, including 233 patients with HDV RNA viremia were retrospectively studied with a mean follow-up time of 6.5 years (range 0.5-33.1). In patients with HDV RNA positivity, 29.6% of patients had liver cirrhosis at baseline. 39.1% of patients with HDV RNA positivity and cirrhosis at baseline experienced hepatic decompensation and 3% of patients with HDV RNA positivity and without cirrhosis at baseline experienced hepatic decompensation (21). Cumulative decompensation-free survival is shown below in Figure 4.

Predictions from the model are similar to those from the Kamal et al., (2020) study (Figure 5) (21). Further, the rate of hepatic decompensation in patients with cirrhosis at baseline was 10.2% per person-year, similar to the rate estimated for use in the economic model (10.67%).

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

39

Figure 4: Kaplan-Meier decompensation-free survival curves based on HDV RNA status from Kamal et al ., (2020)

==> picture [399 x 258] intentionally omitted <==

Figure 5: Comparison of decompensation-free survival of patients from Kamal et al ., (2020) vs. predictions from economic model

==> picture [422 x 266] intentionally omitted <==

----- Start of picture text -----
Natural history Model predicted
----- End of picture text -----

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

40

6.2.3 Hepatocellular carcinoma (HCC)

In the Yurdaydin et al., (2018) study, a hepatitis delta database was analysed for the effects of treatment duration on virologic response and clinical outcomes. 99 chronic hepatitis delta patients who received at least 6 months of interferon treatment were selected. Post-treatment median follow-up was 55 months (24-225 months). Of these patients, 35 achieved a maintained virologic response (MVR). In the non-responder patients, 22% (14/64) had cirrhosis present at baseline. HCC-free survival outcomes for patients without MVR, assumed to be most appropriate for comparison with BSC, are shown in Figure 6.

Given the lack of data regarding the distribution of patients from F0-F3, a similar distribution of non-cirrhotic patients was assumed based on data from Romeo et al ., (2009) (19). The model showed generally similar results for the cumulative incidence of hepatocellular carcinoma for BSC compared to those without MVR from the Yurdaydin et al., (2018) study (Figure 7) (22).

Figure 6: Kaplan-Meier hepatocellular carcinoma-free survival curve from Yurdaydin et al. , (2018) in patients with and without MVR

==> picture [366 x 263] intentionally omitted <==

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

41

Figure 7: Comparison of survival of patients from Yurdaydin et al ., (2018) vs. predictions from economic model

==> picture [422 x 280] intentionally omitted <==

6.2.4 Liver-related mortality

Three natural history studies were evaluated to compare the projections from the economic model regarding mortality.

6.2.4.1 Survival of compensated cirrh osis (F4) patients

The first study included 166 patients with compensated HDV-related cirrhosis diagnosed since 1994 and followed until death or 31[st ] December 2004. Patients had a mean age of 40.7±7.9 years. The median survival was 58.3 months since the diagnosis of compensated cirrhosis, with a probability of survival after the diagnosis of compensated cirrhosis of 94.3%, 82.5%, and 51.5% at 1, 2, and 5 years, respectively (Figure 8) (23). Predictions from a purely compensated cirrhotic (i.e., 100% F4) population in the model demonstrated strong alignment with those projected from this study (Figure 9).

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

42

Figure 8: Kaplan-Meier survival curve of patients with compensated HBV-HDV cirrhosis from Gheorge et al ., (2005)

==> picture [408 x 272] intentionally omitted <==

==> picture [251 x 12] intentionally omitted <==

----- Start of picture text -----
6.2.4.2 Survival in broad F0-F4 population
----- End of picture text -----

To determine whether projections from the combined non-cirrhotic and cirrhotic populations aligned with natural history studies regarding mortality, two studies were selected based on availability of data for patients without HDV RNA - (Roulot et al., 2020) and for those without MVR due to treatment (Yurdaydin et al ., 2018) (20, 22).

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

43

Figure 9: Comparison of survival of compensated cirrhosis patients from Gheorge et al ., (2005) vs. predictions from economic model

==> picture [432 x 239] intentionally omitted <==

In the Roulot et al., (2020) study, at referral, 28.1% of patients had cirrhosis, 36.6% had significant or severe fibrosis (≥F2), and 16.8% had no or minimal fibrosis (F0-F1). The 5-year risk of death in the entire population, including patients who may have received treatment, was 20.2%. Survival according to HDV RNA status at the end of follow-up showed that patients with positive HDV viral load had a higher chance of death (hazard ratio 3.30, p<0.001; Figure 10). Projections from the model are generally similar to those from the Roulot et al, (2020) study, though a low number of events towards the end of follow-up creates uncertainty about survival outcomes beyond 8 years (Figure 11).

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

44

Figure 10: Survival without liver transplantation according to persistent HDV viremia before endpoint from Roulot et al ., (2020)

==> picture [400 x 345] intentionally omitted <==

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

45

Figure 11: Comparison of survival of F0-F4 HDV RNA+ patients from Roulot et al ., (2020) vs. predictions from economic model

==> picture [439 x 217] intentionally omitted <==

Survival outcomes in the Yurdaydin et al ., (2018) study, for patients without MVR, assumed to be the most appropriate for comparison with BSC, are shown in Figure 12.

Given the lack of data regarding the distribution of patients from F0-F3, a similar distribution of non-cirrhotic patients was assumed based on Romeo et al. , (2009). Given the relatively low number of observations (Figure 12) after 10 years, these first ten years were analysed to compare model survival outcomes vs. the study. The model showed generally similar results for liver-related mortality for BSC compared to those without MVR from the Yurdaydin et al ., (2018) study (Figure 13).

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

46

Figure 12: Liver-related mortality stratified by MVR status in Yurdaydin et al., (2018)

==> picture [427 x 312] intentionally omitted <==

Figure 13: Comparison of survival of F0-F4 patients without MVR vs. predictions from economic model

==> picture [435 x 215] intentionally omitted <==

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

47

6.3 Health state occupancy

The committee requested graphical representations of fibrosis health state occupation over time from the economic model. We have created these graphs for both arms, in addition to a graph which shows the survival over the time horizon. Results are reported in Figure 14 to Figure 16. These can also be found in the RESULTS sheet of the executable model.

Figure 14: Health state occupancy, Bulevirtide arm

==> picture [448 x 277] intentionally omitted <==

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

48

Figure 15: Health state occupancy, BSC arm

==> picture [451 x 280] intentionally omitted <==

Figure 16: Overall survival

==> picture [450 x 259] intentionally omitted <==

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

49

References

1. Spaan M, Carey I, Bruce M, Shang D, Horner M, Dusheiko G, et al. Hepatitis delta genotype 5 is associated with favourable disease outcome and better response to treatment compared to genotype 1. J Hepatol. 2020;72(6):1097-104.

2. UK Health Security Agency. Mean age of HDV RNA positive patients at baseline. Data on File.

3. National Institute for Health and Care Excellence. Hepatitis B (chronic): diagnosis and management. 2013 26 June 2013.

4. European Association for the Study of the Liver. EASL Clinical Practice Guidelines on non-invasive tests for evaluation of liver disease severity and prognosis - 2021 update. J Hepatol. 2021;75(3):659-89.

5. Li Y, Huang YS, Wang ZZ, Yang ZR, Sun F, Zhan SY, et al. Systematic review with meta-analysis: the diagnostic accuracy of transient elastography for the staging of liver fibrosis in patients with chronic hepatitis B. Aliment Pharmacol Ther. 2016;43(4):458-69.

6. Qi X, An M, Wu T, Jiang D, Peng M, Wang W, et al. Transient Elastography for Significant Liver Fibrosis and Cirrhosis in Chronic Hepatitis B: A Meta-Analysis. Can J Gastroenterol Hepatol. 2018;2018:3406789.

7. Terrault NA, Lok ASF, McMahon BJ, Chang KM, Hwang JP, Jonas MM, et al. Update on prevention, diagnosis, and treatment of chronic hepatitis B: AASLD 2018 hepatitis B guidance. Hepatology. 2018;67(4):1560-99.

8. Farci P, Roskams T, Chessa L, Peddis G, Mazzoleni AP, Scioscia R, et al. Long-term benefit of interferon alpha therapy of chronic hepatitis D: regression of advanced hepatic fibrosis. Gastroenterology. 2004;126(7):1740-9.

9. Marcellin P, Gane E, Buti M, Afdhal N, Sievert W, Jacobson IM, et al. Regression of cirrhosis during treatment with tenofovir disoproxil fumarate for chronic hepatitis B: a 5-year open-label follow-up study. Lancet. 2013;381(9865):468-75.

10. National Institute for Health and Care Excellence. Centre for Health Technology Evaluation (CHTE) Methods Review: Health-Related Quality of Life. 2020.

11. Kennedy P. Guide to Econometrics. Cambridge MA: MIT Press; 1998.

12. Arabmazar A, Schmidt P. Further evidence on the robustness of the Tobit estimator to heteroskedasticity. Journal of Econometrics. 1981;17(2):253-8.

13. Johnston J. Econometrics Methods. New York: McGraw-Hill; 1997.

14. Wright M, Grieve R, Roberts J, Main J, Thomas HC. Health benefits of antiviral therapy for mild chronic hepatitis C: randomised controlled trial and economic evaluation. Health Technol Assess. 2006;10(21):1-113, iii.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

50

15. Vera-Llonch M, Martin M, Aggarwal J, Donepudi M, Bayliss M, Goss T, et al. Health-related quality of life in genotype 1 treatment-naive chronic hepatitis C patients receiving telaprevir combination treatment in the ADVANCE study. Aliment Pharmacol Ther. 2013;38(2):124-33.

16. Stepanova M, Nader F, Cure S, Bourhis F, Hunt S, Younossi ZM. Patients' preferences and health utility assessment with SF-6D and EQ-5D in patients with chronic hepatitis C treated with sofosbuvir regimens. Aliment Pharmacol Ther. 2014;40(6):676-85.

17. Samp JC, Perry R, Piercy J, Wood R, Baran RW. Patient health utility, work productivity, and lifestyle impairment in chronic hepatitis C patients in France. Clin Res Hepatol Gastroenterol. 2015;39(3):307-14.

18. Juanbeltz R, Goñi-Esarte S, Martínez-Baz I, San Miguel R, Zozaya J, Rivero M, et al. 1ISG-001 Health utilities in chronic hepatitis C patients one year after successful treatment with direct-acting antivirals. European Journal of Hospital Pharmacy. 2019;26.

19. Romeo R, Del Ninno E, Rumi M, Russo A, Sangiovanni A, de Franchis R, et al. A 28-year study of the course of hepatitis Delta infection: a risk factor for cirrhosis and hepatocellular carcinoma. Gastroenterology. 2009;136(5):1629-38.

20. Roulot D, Brichler S, Layese R, BenAbdesselam Z, Zoulim F, Thibault V, et al. Origin, HDV genotype and persistent viremia determine outcome and treatment response in patients with chronic hepatitis delta. J Hepatol. 2020;73(5):1046-62.

21. Kamal H, Westman G, Falconer K, Duberg AS, Weiland O, Haverinen S, et al. Long-Term Study of Hepatitis Delta Virus Infection at Secondary Care Centers: The Impact of Viremia on Liver-Related Outcomes. Hepatology. 2020;72(4):1177-90.

22. Yurdaydin C, Keskin O, Kalkan Ç, Karakaya F, Çalişkan A, Kabaçam G, et al. Interferon Treatment Duration in Patients With Chronic Delta Hepatitis and its Effect on the Natural Course of the Disease. The Journal of Infectious Diseases. 2018;217(8):1184-92.

23. Gheorghe L, Iacob S, Simionov I, Vadan R, Gheorghe C, Iacob R, et al. Natural history of compensated viral B and D cirrhosis. Rom J Gastroenterol. 2005;14(4):32935.

Company response to draft guidance consultation [ID3732], issued November 2022. © Gilead Sciences Ltd (2022) All rights reserved.

51

Bulevirtide for treating chronic hepatitis D [ID3732]

Consultation on the appraisal consultation document – deadline for comments 5pm on 25 November 2022. Please submit via NICE Docs.

Please return to: NICE DOCS

Bulevirtide for treating chronic hepatitis D [ID3732]

Consultation on the appraisal consultation document – deadline for comments 5pm on 25 November 2022. Please submit via NICE Docs.

Insert extra rows as needed

• Checklist for submitting comments • Use this comment form and submit it as a Word document (not a PDF). • Complete the disclosure about links with, or funding from, the tobacco industry. • Combine all comments from your organisation into 1 response. We cannot accept more than 1 set of comments from each organisation.

• • Do not paste other tables into this table – type directly into the table. • Please underline all confidential information, and separately highlight information that is submitted under ‘commercial in confidence’ in turquoise and all information submitted under ‘academic in confidence’ in yellow. If confidential information is submitted, please also send a 2[nd] version of your comment with that information replaced with the following text: ‘academic / commercial in confidence information removed’. See the Guide to the processes of technology appraisal (section 3.1.23 to 3.1.29) for more information.

• • Do not include medical information about yourself or another person from which you or the person could be identified.

• • Do not use abbreviations • Do not include attachments such as research articles, letters or leaflets. For copyright reasons, we will have to return comments forms that have attachments without reading them. You can resubmit your comments form without attachments, it must send it by the deadline.

• • If you have received agreement from NICE to submit additional evidence with your comments on the appraisal consultation document, please submit these separately.

• Note: We reserve the right to summarise and edit comments received during consultations, or not to publish them at all, if we consider the comments are too long, or publication would be unlawful or otherwise inappropriate.

Comments received during our consultations are published in the interests of openness and transparency, and to promote understanding of how recommendations are developed. The comments are published as a record of the comments we received, and are not endorsed by NICE, its officers or advisory committees.

Please return to: NICE DOCS

==> picture [219 x 81] intentionally omitted <==

Bulevirtide for treating chronic hepatitis D

Review of the company’s response

January 2023

Source of funding

This report was commissioned by the NIHR Evidence Synthesis Programme as project number 13/37/00.

1 Introduction

This document provides the EAG’s review of the company’s response.

2 Company updated results

The company updated the patient access scheme (PAS) discount for bulevirtide from xxxx to xxxx. The company maintained its view that the severity modifier of 1.2 should be used in the base case analysis. Additionally, the company made the following changes to their base case analysis (in relation to the base case results presented at the second Appraisal Committee Meeting [ACM]):

1. The response rates in the model were changed to reflect those of the subgroup of patients from MYR301 with a FibroScan® score greater than or equal to 8 kPa.

2. The baseline age in the model was updated to 37 years as per the UKHSA study.

3. Combined responders (CR) were assumed to be at risk of developing hepatocellular carcinoma (HCC), estimated as 20% of that assumed for partial responders (PRs).

4. It was assumed that CRs had a probability of experiencing fibrosis progression, estimated as 20% of that assumed for PRs.

5. It was assumed that 30% of patients with HCC will be cured and accrue a utility of 0.81 afterwards.

6. The company reports assuming that CR and PRs have the same probability of developing HCC.

7. Assuming that fibrosis regression only starts occurring from cycle 4 onwards (i.e. 96 weeks) in the model.

8. Assuming that the baseline proportion of patients with METAVIR fibrosis stage F4 is 47 (based on MYR301 data) and aggregating the baseline proportion of patients in the F2 and F3 states (53% of patients) occupying these health state at baseline.

9. Assuming a utility gain of 0.05 for CR vs PR and non-responders (NRs), as per TA330.

The company’s updated deterministic ICER is reported in Table 1. As the company did not provide an updated model with its response, the EAG used the previous (most recent) version of the company’s model in order to try to replicate and validate the company’s updated results. The EAG could not fully replicate the company’s updated base case ICER in the previous version of the model. The company’s updated base case ICER without the severity weighting is £29,083 (whereas the EAGreproduced ICER in the company’s previous model is £29,086), and the ICER with a 1.2. severity

==> picture [80 x 28] intentionally omitted <==

PAGE 2

weighting included is £24,236 (whereas the EAG-reproduced ICER in the company’s model is £24,238). The biggest uncertainty in the company’s implementation of their updated base case assumptions is around the use of the baseline distribution of fibrosis in the model (as discussed in Table 2).

In Table 2, the EAG summarises the changes made by the company in the economic model after the second ACM, together with the EAG’s critique of the latter.

Table 1. Company’s deterministic base case results post technical engagement

==> picture [80 x 28] intentionally omitted <==

PAGE 3

Table 2. Changes in company’s model post second ACM

==> picture [81 x 28] intentionally omitted <==

PAGE 4

==> picture [81 x 28] intentionally omitted <==

PAGE 5

==> picture [81 x 28] intentionally omitted <==

PAGE 6

3 EAG preferred assumptions

Table 3 reports the EAG’s deterministic ICERs, including all EAG’s preferred assumptions (as detailed in Table 1). The EAG notes that some of the EAG-preferred assumptions were not incorporated by the company in their update dbase case. These were the following:

• Estimation of the probability of HCC from the F2-F4 states according to Romeo[1] and Kushner[2] – Table 25, Section 4.2.5.3.1 of the EAG report.

• Estimation of the probability of fibrosis progression from the F2-F4 states according to Romeo[1] – Table 26, Section 4.2.5.3.1 of the EAG report.

These changes, in combination with the ones reported in Table 1, are included in the ICER reported in Table 3 and Table 4, for the deterministic, and probabilistic ICERs. The results include the company’s updated PAS.

The EAG-preferred deterministic ICER amounts to £36,027. The EAG notes that the economic results are not eligible for the use of a severity weighting given that with a baseline age of 37 years, the total QALYs associated with BSC would have to be below 6.6 in order for the severity weighting to be above 1.

Finally, the EAG notes that all the ICERs herein provided remain highly uncertain with regards to the duration of treatment response and duration of treatment in the economic model. In the MYR 301 trial, participants were reportedly scheduled to continued bulevirtide treatment up to 144 weeks. However, in the economic model, the company assumed that:

• Partial responders who have not achieved a complete response continue treatment up to week 72 but then discontinue treatment (if they don’t achieve a complete response). The EAG notes that in MYR 301, treatment is likely to have carried on for a longer period of time for these patients.

• Non-responders to treatment at week 48 discontinue treatment; however, the company did not provide a clear justification for this assumption and the EAG is unclear if 48 weeks was chosen due to this being the same data cut-off period

==> picture [80 x 28] intentionally omitted <==

PAGE 7

available for MYR 301; or for any other reason. The EAG remains unclear if in MYR 301 non-responders discontinued treatment at 48 weeks.

Crucially, the EAG notes that the duration of complete response and duration of treatment in the economic model would need careful re-assessment when the 96-week follow-up data are available for MYR 301. At the second ACM, the committee, “ recognised there remained some ambiguity in how long treatment should continue treatment and when people who do not respond to treatment should stop having bulevirtide […] The committee concluded there was still uncertainty surrounding treatment duration and stopping rules .”

Table 3. EAG’s deterministic base case results post technical engagement

Table 4. EAG’s probabilistic base case results post technical engagement

==> picture [80 x 28] intentionally omitted <==

PAGE 8

4 References

1. Romeo R, Del Ninno E, Rumi M, et al. A 28-Year Study of the Course of Hepatitis Δ Infection: A Risk Factor for Cirrhosis and Hepatocellular Carcinoma. Gastroenterology 2009;136(5):1629-38. doi: 10.1053/j.gastro.2009.01.052

2. Tatyana Kushner MS, David Kaplan. Delta Hepatitis within the Veterans Affairs Medical System in the United States: Prevalence, Risk Factors, and Outcomes. Journal of Hepatology 2015;63(3):586-92.

==> picture [80 x 28] intentionally omitted <==

PAGE 9

==> picture [219 x 81] intentionally omitted <==

Bulevirtide for treating chronic hepatitis D

PAS update

April 2023

Source of funding

This report was commissioned by the NIHR Evidence Synthesis Programme as project number 133700.

1 Company updated results and EAG critique

In April 2023, the company updated the patient access scheme (PAS) discount for bulevirtide from

xxxxxx to xxxxxx

The company’s revised base case is provided in Table 1.

Table 1. Company’s deterministic base case results in its third response after 2nd ACM (deterministic)