CADTH Reimbursement Review

Pemigatinib (Pemazyre)

Sponsor: Incyte Biosciences Canada Corporation

Therapeutic area: Cholangiocarcinoma

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Stakeholder Input

Clinical Review

Executive Summary

An overview of the submission details for the drug under review is provided in Table 1.

Introduction

Gallbladder cancer and cholangiocarcinoma (CCA) are known as biliary tract cancers (BTCs) accounting for 10% to 15% of all primary liver cancer.^1,2 CCAs are most commonly adenocarcinomas and comprise 2 main subtypes: intrahepatic (iCCA), initiating from the biliary tree within the liver, and extrahepatic (eCCA), initiating outside the liver parenchyma.^2,3 In Canada and the US, respectively, there are approximately 400 and 5,000 new cases of CCA diagnosed each year.⁴ The median age at diagnosis is 65 years in Western industrialized nations.⁵ The 5-year relative survival rates for iCCA and eCCA, respectively, are 9% and 10%.⁶ Diagnosis of CCA is most commonly made in advanced stages (70% of patients are diagnosed with unresectable, locally advanced or metastatic disease)⁷ due to an absence of symptoms until later in the course of the disease.⁸ The rate of recurrence is high in the minority of patients who are able to undergo potentially curative surgery.⁹ Symptoms commonly appear when a bile duct is blocked and include jaundice, itching, light-coloured and greasy stools, dark urine, abdominal pain, loss of appetite or weight loss, fever, and nausea and vomiting.⁸

One of the most frequent genetic alterations in patients with iCCA involve the fibroblast growth factor receptor (FGFR) 2 (FGFR2).⁷ The FGFR2 fusions or rearrangements are found in 10% to 20%¹⁰ of patients with iCCA, while they rarely occur in eCCA. Alterations involving other members of the FGFR are rare, with an incidence below 0.5%.¹¹ While there is strong genetic and functional evidence that FGFR genetic alterations can drive the formation of tumours,⁷ it is currently not known whether patients whose disease is positive for an FGFR2 alteration represent a distinct prognostic subgroup.¹¹

For patients with advanced-stage or unresectable CCA and a good Eastern Cooperative Oncology Group Performance Status (ECOG PS) score of 0 or 1, standard-of-care first-line treatment is gemcitabine and cisplatin.⁹ If there are concerns about a patient’s renal function, oxaliplatin may be substituted for cisplatin.² For patients with an ECOG PS of 2, gemcitabine monotherapy may be considered as first-line therapy.² The median overall survival (OS), median progression-free survival (PFS), and objective response rate (ORR) in patients with BTCs treated with standard-care first-line palliative treatment with gemcitabine and cisplatin ranges from 11.2 to 11.7 months, 5.8 to 8.0 months, and 19.5% to 26.1%, respectively.¹² The clinical experts consulted by CADTH noted there are currently no funded standard treatment options for patients in the second-line setting once the disease has progressed on first-line treatment. In the absence of proven treatment options in the second-line setting, participation in a clinical trial and best supportive care are recommended, including alleviating biliary obstruction and full access to palliative care and symptom management.² According to the clinical experts consulted by CADTH, the second-line therapies used in Canadian clinical practice include folinic acid, fluorouracil, and oxaliplatin (FOLFOX); folinic acid, fluorouracil, and irinotecan hydrochloride (FOLFIRI); fluorouracil (alone or in combination with cisplatin or oxaliplatin); and capecitabine (alone or in combination with cisplatin or oxaliplatin). Second-line treatment with FOLFOX is currently the only drug based on phase III trial data in this setting.⁵ The ABC-06 trial¹³ evaluated the efficacy and safety of modified folinic acid, fluorouracil, and oxaliplatin (mFOLFOX) plus active symptom control (ASC) compared with ASC alone in patients with locally advanced or metastatic BTC (including CCA and gallbladder or ampullary carcinoma) whose disease had progressed on first-line cisplatin and gemcitabine therapy. At the median follow-up time of 21.7 months, median OS was 6.2 months in the FOLFOX group and 5.3 months in the control group (hazard ratio [HR] = 0.69; 95% confidence interval [CI], 0.50 to 0.97; P = 0.031); median PFS was 4 months in the FOLFOX group, and an objective response was observed in 5% of patients in the FOLFOX group. The clinical experts consulted by CADTH agreed there is an unmet need for effective therapies with an acceptable toxicity profile that achieve disease control, delay worsening of symptoms, maintain health-related quality of life (HRQoL), delay disease progression, and prolong survival.

Pemigatinib is a molecule kinase inhibitor with antitumour activity that inhibits FGFRs. FGFRs are receptor tyrosine kinases that activate signalling pathways in tumour cells.¹⁴ On September 17, 2021, pemigatinib was approved by Health Canada for the treatment of adults with previously treated, unresectable, locally advanced, or metastatic CCA with an FGFR2 fusion or other rearrangement. The sponsor’s requested reimbursement criteria for pemigatinib are per the Health Canada–approved indication. Pemigatinib underwent review by Health Canada through a standard review pathway. Pemigatinib has not previously been reviewed by CADTH. Oral pemigatinib is available as 4.5 mg, 9 mg, and 13.5 mg tablets. The recommended starting dose is 13.5 mg administered orally for 14 consecutive days followed by 7 days off therapy, in 21-day cycles. The product monograph states that treatment is to be continued until disease progression or unacceptable toxicity. Furthermore, it is recommended that a low-phosphate diet be initiated when the phosphate level is greater than 5.5 mg/dL, and that adding a phosphate-lowering therapy should be considered when the level is greater than 7 mg/dL. The dose of phosphate-lowering therapy is to be adjusted until the phosphate level returns to less than 7 mg/dL. It is recommended that discontinuation of phosphate-lowering therapy be considered during pemigatinib treatment breaks or if the phosphate level falls below normal.¹⁵ The objective of this CADTH review is to perform a systematic review of the beneficial and harmful effects of pemigatinib for the treatment of adult patients with previously treated, unresectable, locally advanced, or metastatic CCA with an FGFR2 fusion or other rearrangement.

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient groups who responded to CADTH’s call for patient input and from the clinical experts consulted by CADTH for the purpose of this review.

Patient Input

Three patient groups, the Canadian Liver Foundation, the Canadian Organization for Rare Disorders, and the Cholangiocarcinoma Foundation, co-created 1 patient input for this review. This input was based on an online survey and a virtual focus group with a total of 27 respondents: 15 were patients diagnosed with CCA (4 of whom had CCA with FGFR2 fusions), and 2 were patients who had symptoms of CCA but did not have a diagnosis of CCA; 10 respondents were caregivers or family members of patients with CCA.

Respondents indicated a varying range of CCA symptoms affecting patients’ daily activities (including their social, work, and school lives and their relationships) causing detrimental effects on patients’ quality of life (QoL). Respondents highlighted problems with intimacy or sexual desire, fatigue, and anxiety. Other commonly experienced symptoms indicated by respondents included unintended weight loss, insomnia, gastrointestinal problems, abdominal pain, constipation, depression, and neuropathy. According to the 3 patient groups, delayed diagnosis, misdiagnosis, and a lack of specialists and treatment options available for this rare cancer significantly contribute to patients’ feelings of stress and anxiety and may delay or eliminate treatment options.

According to the patient input received, respondents reported they expect the following key outcomes to be improved from any new drug or treatment: QoL, tumour response, delay in disease progression, and additional treatment choice. Additionally, it was highlighted by the 3 patient groups that the identification of gene mutations and the development of targeted therapies was perceived by respondents to be very important and would spur hope for curable options. Four respondents indicated they had direct experience with taking pemigatinib. Respondents indicated little overall challenge dealing with the side effects from pemigatinib.

Clinician Input

Input From the Clinical Experts Consulted by CADTH

The clinical experts consulted by CADTH indicated there are currently no standard publicly funded second-line treatment options. Palliative therapy (e.g., FOLFOX, FOLFIRI, fluorouracil, and capecitabine) and best supportive care are recommended for patients in the present target setting. The clinical experts identified an unmet need for effective therapies with acceptable toxicity profiles that achieve disease control, delay worsening of symptoms, maintain HRQoL, delay disease progression, and prolong survival. The clinical experts consulted by CADTH stated that pemigatinib was to be used in adult patients with previously treated, unresectable, locally advanced, or metastatic CCA with an FGFR2 fusion or other rearrangement, as per the FIGHT-202 trial. Among patients enrolled in cohort A of the FIGHT-202 trial, the clinical experts did not identify any subgroups of patients who would potentially be best suited for, or benefit the least from, pemigatinib. The clinical experts consulted by CADTH felt it would be reasonable to generalize the results from cohort A to patients with FGFR2 alterations who are intolerant to first-line therapy.

The clinical experts agreed that patients would be identified as possible candidates for pemigatinib if they had the FGFR2 alteration. Clinical assessment to evaluate response to treatment with pemigatinib would include regular radiological imaging (i.e., CT or MRI) and a CA19 to 9 biomarker test every 2 to 3 months to determine whether a patient has experienced disease progression. In addition, patients would be seen by an oncologist every 3 to 4 weeks for clinical assessment (i.e., to assess disease symptoms and a patient’s ECOG PS). The clinical experts indicated that the most clinically meaningful responses to treatment include disease control (i.e., disease stability or response), improvement in disease-related symptoms, better pain control, weight gain, regaining a more active lifestyle, maintenance of HRQoL, and prolonged PFS and OS. Acceptable drug-related toxicity was also noted as a clinically meaningful outcome.

In the opinion of the clinical experts consulted by CADTH, treatment with pemigatinib should be discontinued if a patient experiences disease progression, has a worsening ECOG PS, is intolerant to or experiences unacceptable toxicity from pemigatinib (which cannot be improved with dose delays or reductions), or the patient is not interested in continuing treatment.

Clinician Group Input

Two clinician group inputs were provided, 1 from the Ontario Health (Cancer Care Ontario) Gastrointestinal Cancer Drug Advisory Committee and 1 from the Canadian Gastrointestinal Oncology Evidence Network (CGOEN) and other physicians who treat CCA. The views of the clinician groups overall were consistent with the clinical experts consulted by CADTH, indicating that the most important treatment goals are achieving disease control, delaying worsening of symptoms, maintaining HRQoL, delaying disease progression, and prolonging survival. It also important that the drug have an acceptable safety profile. The clinicians from the CGOEN also highlighted that the convenient oral route of administration of pemigatinib would contribute to improvements in QoL for patients, as fewer visits to a cancer centre and less chair time would be required compared with alternative treatment options. The clinicians from the CGOEN further suggested it would be reasonable to consider pemigatinib upfront for patients deemed unsuitable for standard first-line chemotherapy. This clinician group also noted that patients with compromised hepatic function or significant hyperbilirubinemia would be least suitable for treatment with pemigatinib. The clinicians from both inputs anticipated that pemigatinib would offer a clinically meaningful benefit and improved efficacy to patients with the potential for improved QoL.

Drug Program Input

The Provincial Advisory Group identified the following jurisdictional implementation issues: relevant comparators, consideration for initiation of therapy, consideration for discontinuation of therapy, generalizability, care provision, system issues, and economic considerations. The clinical experts consulted by CADTH weighed evidence from the FIGHT-202 trial and other clinical considerations to provide responses to the Provincial Advisory Group’s drug program implementation questions. Refer to Table 4 for more details.

Clinical Evidence

Pivotal Studies and Protocol Selected Studies

The FIGHT-202 trial is an ongoing, multi-centre, open-label, single-arm phase II trial evaluating the efficacy and safety of pemigatinib in patients with advanced, metastatic, or surgically unresectable CCA with FGFR2 alterations, other fibroblast growth factor (FGF) or FGFR alterations, or no FGF/FGFR alterations, whose disease did not respond to previous therapy. Patients were assigned to 3 cohorts, depending on the patient’s FGF/FGFR status (cohort A: FGFR2 fusions or rearrangements; cohort B: FGF/FGFR alterations other than FGFR2 fusions or rearrangements; cohort C: negative for FGF/FGFR alterations). This CADTH review focuses on cohort A, as cohorts B and C were not part of the requested reimbursement criteria submitted to CADTH and not approved in the Health Canada Notification of Compliance with conditions; therefore, they are beyond the scope of this review. Selected results for cohorts B and C have been included in Appendix 3. A total of 147 patients were enrolled to received oral pemigatinib (13.5 mg orally once daily on a schedule of 2 weeks on and 1 week off for each 21-day cycle). The primary outcome was ORR in cohort A, and secondary outcomes included ORR in cohorts B, A plus B, and C; PFS; duration of response (DOR); disease control rate (DCR); OS; and safety assessed in all 3 cohorts, respectively. Exploratory end points included HRQoL and symptom severity.

Adults diagnosed with locally advanced, metastatic, or surgically unresectable CCA with FGFR2-positive disease who had documented disease progression after at least 1 line of prior systemic therapy were enrolled into cohort A of the Fight-202 trial. At baseline, 107 patients were identified as having FGFR2 fusions or rearrangements and were grouped into cohort A. Cohort B included 20 patients with other FGF/ FGFR alterations other than FGFR2, and cohort C included 18 patients with no identified FGF/FGFR alterations. One patient who was placed into an “undetermined” group was not assigned to any of the 3 cohorts, as their FGF and FGFR status results could not be confirmed by the central genomics laboratory. For patients in cohort A, the mean age was 55.3 years (standard deviation [SD] = 12.02). Most patients were female (60.7%) and enrolled in trial sites in North America (59.8%) or Europe (29.9%). Almost all patients (89% of patients overall and 98.1% of patients in cohort A) had iCCA. The majority of patients in cohort A had metastatic disease (82.2%) ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Median time since diagnosis was 1.28 years (range, 0.03 to 11.1 years) for patients in cohort A. The majority of patients in cohort A had an ECOG PS of 1 (53.3%), and all patients had received at least 1 line of prior systemic therapy for advanced or metastatic disease (60.7%, 27.1%, and 12.1% of patients received 1, 2, and ≥ 3 prior lines, respectively). Renal and hepatic impairment grades were normal or mild for most patients in cohort A (39.3% and 43.9% normal and mild renal impairment grades, respectively; 44.9% and 48.6% normal and mild hepatic grades, respectively).¹⁶

The futility analysis, which was performed on October 12, 2017,¹⁷ was pre-specified a priori in the statistical analysis plan. The timing of the subsequent analysis (March 22, 2019), at which point the predetermined threshold (i.e., lower limit of the 95% CI for an ORR > 15%) would be assessed, was not pre-specified a priori in the statistical analysis plan; however, the sponsor’s proposed timing was agreed upon by the FDA during its review process for pemigatinib. Two additional updated analyses occurred at the August 2019 and April 2020 data cut-off dates; the former was a 4-month safety update required for the FDA New Drug Application (NDA), the latter was performed to support the safety data summaries for another indication outside of Canada.¹⁷ The trial is still ongoing, with an estimated completion date in the first quarter of 2022.¹⁷

Efficacy Results

The key efficacy results from cohort A of the FIGHT-202 trial are summarized in Table 2. At the latest data cut-off date (April 7, 2020), the median duration of follow-up was 27.9 months in cohort A. Median OS was 17.48 (95% CI, 14.42 to 22.93) months. The survival probabilities of patients surviving for 6 and 12 months were ||||||||||||||||||||||||||||||||||||||| and 67.3% (95% CI, 57.4 to 75.4), respectively. Median PFS was 7.03 months (95% CI, 6.08 to 10.48). The PFS probabilities at 6 and 12 months were |||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||. The PFS results for the subgroup of interest, as pre-specified a priori in the protocol for this CADTH review, suggested that the treatment effects on PFS for the subgroups of patients with an ECOG PS of 0 and 1 plus 2 were generally consistent with the overall population in cohort A.

As of the April 7, 2020, data cut-off date, the proportion of patients who achieved an objective response was 37.0% (N = 40) (95% CI, 27.94 to 46.86), including 4 patients (3.7%) with a complete response (CR) and 36 patients (33.3%) with a partial response (PR). The ORR results for the subgroup of interest suggested that the treatment effects on ORR for the subgroups of patients with an ECOG PS of 0 and 1 plus 2 were generally consistent with the overall population in cohort A. Among the 40 patients who achieved an objective response, the median DOR was 8.08 months (95% CI, 5.65 to 13.14). The probabilities of maintaining a response for at least 6 and 12 months were ||||||||||||||||||||||||||||||||| and |||||||||||||||||||||||||||||||||||||, respectively.

The results for DCR were reported only for the March 22, 2019, data cut-off date. The proportion of patients with a best response of CR, PR, or stable disease was 82.2% (N = 88) (95% CI, 73.7 to 89.0), including 3 patients (2.8%) with a CR, 35 patients (32.7%) with a PR, and 50 patients (46.7%) with stable disease for 39 or more days since the first pemigatinib dose. The DCR results based on investigator assessment showed generally consistent results with those of an independent review committee (IRC); DCR was ||||||||||||||||||||||||||||||||||||||||||.

The descriptive summary statistics of observed scores for the European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) and the EORTC Quality of Life Questionnaire Cholangiocarcinomas and Gallbladder Cancer Module 21 (EORTC QLQ-BIL21) from baseline to ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶ A definition for what constituted a clinically meaningful change from baseline in the present target population was not provided. A post-hoc analysis assessed observed mean changes from baseline to week 16 by subgroups of patients (i.e., patients with a CR or PR, stable disease, or progressive disease [PD]). Results suggested that changes from baseline appeared directionally more favourable in patients with a CR or PR, or stable disease than in patients with PD.

Harms Results

Key harms reported in cohort A of the FIGHT-202 trial are summarized in Table 2. All patients in cohort A experienced at least 1 treatment-emergent adverse event (TEAE) (100.0%). The most commonly reported TEAEs were alopecia (|||||||), hyperphosphatemia (|||||||), diarrhea (|||||||), dysgeusia (|||||||), fatigue (|||||||), and nausea (|||||||). The percentage of patients experiencing serious TEAEs was ||||||| in cohort A. The most common serious TEAEs were |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Adverse events (AEs) led to discontinuation of study treatment in ||||||| of patients in cohort A. None of the patients withdrew from the FIGHT-202 study due to an AE as primary reason. TEAEs leading to treatment discontinuation included |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. TEAEs leading to death occurred relatively rarely in cohort A (N = 3; 2.8%) and included failure to thrive and bile duct obstruction. None of the TEAEs leading to death were considered treatment-related.¹⁶

The percentage of patients experiencing nail toxicity TEAEs was ||||||| in patients in cohort A. The most commonly reported nail toxicity included ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

The percentage of patients experiencing serous retinal detachment TEAEs in cohort A was |||||||. The most commonly reported serous retinal detachment was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

The percentage of patients experiencing hyperphosphatemia TEAEs in cohort A was |||||||. The most commonly reported hyperphosphatemia events were ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

The percentage of patients experiencing hypophosphatemia TEAEs in cohort A was |||||||. The most commonly reported hypophosphatemia events were ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

Critical Appraisal

The primary objective of phase II (randomized or non-randomized) trials is to document the safety outcomes and investigate whether the estimate of effect for a new drug is large enough to use it in confirmatory phase III trials. Phase II trials may not accurately predict the harm and/or effectiveness of treatments. The clinical experts consulted by CADTH noted that, despite the high unmet need, conducting a randomized controlled trial (RCT) in this setting with a targeted therapy, such as pemigatinib, compared with currently available therapies in the second line in Canadian clinical practice would likely not be feasible. The FIGHT-202 trial included no formal statistical significance and hypotheses testing and point estimates with 95% CIs were reported to estimate the magnitude of treatment effect. A greater than 95% probability of having a 95% CI for ORR in cohort A with a lower limit larger than 15% was the basis for the sample size determination and was regarded as the threshold for a positive study outcome. The subgroup analyses were non-inferential; wide CIs reflected uncertainty in the effect estimates, and small sample sizes limited the generalizability to a broader population. Interpretation of time‐to‐event end points such as OS and PFS is limited in single‐arm studies; since all patients in cohort A received the same treatment, the extent to which the observed survival is due to the natural history of the tumour or the intervention remains unclear. While there is strong genetic and functional evidence that FGFR genetic alterations can drive the formation of tumours,⁷ it is currently not known whether patients with an FGFR2 alteration represent a distinct prognostic subgroup.¹¹ The clinical experts agreed that progression on prior systematic therapy is a major prognostic factor in the present target population, and they did not anticipate that patients would derive any substantial benefit from their underlying disease biology at the time they were enrolled into the FIGHT-202 trial. The results for patient-reported outcomes were inconclusive, given the non-comparative, open-label design of the trial, the lack of a pre-specified analysis of the patient-reported outcomes data, the substantial decline in the number of patients available to provide assessments over time, and the lack of a definition for what constituted a clinically meaningful change from baseline in the target population.

In the absence of a direct comparison of pemigatinib with relevant treatment options, the sponsor submitted an indirect treatment comparison (ITC) in the form of a matching-adjusted indirect comparison (MAIC) comparing the efficacy of pemigatinib (cohort A of the FIGHT-202 trial) with each of the 2 treatment groups in the ABC-06 study. The results of the ITC favoured pemigatinib for PFS and OS in comparison with FOLFOX plus ASC as well as compared with ASC alone. The clinical experts agreed with the CADTH clinical review team that, given the absence of robust comparative data on PFS and OS, the ability to interpret the relative treatment effects observed between pemigatinib and FOLFOX plus ASC and ASC alone was limited, and no firm conclusions could be drawn on how pemigatinib compared with other relevant treatment options. The clinical experts consulted by CADTH anticipated, however, that based on the FIGHT-202 results and on the poor results with existing treatment options in clinical practice, pemigatinib appeared to offer at least similar or improved clinical benefits compared with current therapies, with better tolerability.

Indirect Comparisons

Description of Studies

Two studies, the FIGHT-202 trial and the ABC-06 study, were included in the sponsor’s ITC. The sponsor submitted an ITC in the form of an MAIC between cohort A of the FIGHT-202 study and each of the 2 treatment groups in the ABC-06 study. The ABC-06 study compared mFOLFOX plus ASC versus ASC alone in patients with locally advanced or metastatic BTC. Cohort A of the FIGHT trial included only patients with unresectable, locally advanced, or metastatic CCA who had the FGFR2 mutation.

Efficacy Results

OS: Pemigatinib Versus mFOLFOX Plus ASC

The results of the ITC favoured pemigatinib for PFS and OS in comparison with mFOLFOX plus ASC as well as with ASC alone. Median OS was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| for the pemigatinib group versus ||||||||||||||||||||||||||||||||||||||||||||||||||||||| for the mFOLFOX plus ASC group, based on the March 22, 2019, data cut-off for the FIGHT-202 study. The corresponding HR was |||||||||||||||||||||||||||||||||||||||| and the HR using the results from the April 7, 2020, data cut-off was |||||||||||||||||||||||||||||||||||||

|||||||||||||||||||||||||||||||||||||||||||||||||||||.

OS: Pemigatinib Versus ASC Alone

Median OS was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| for the pemigatinib group versus ||||||||||||||||||||||||||||||||||||||| months for the ASC alone group, based on the March 22, 2019, data cut-off for the FIGHT-202 study. The corresponding HR was |||||||||||||||||||||||||||| ||||||||||||||||||||||||||| and the HR using the results from the April 7, 2020, data cut-off was ||||||||||||||||||||||||||||||||||||||||||||||.

PFS: Pemigatinib Versus mFOLFOX Plus ASC

Median PFS was ||||||||||||||||||||||||||||||||||||||| months versus ||||||||||||||||||||||||||||||||||||||||| months for the pemigatinib versus mFOLFOX plus ASC groups, based on the March 22, 2019, data cut-off for the FIGHT-202 study. The corresponding HR was |||||||||||||||||||||||||| ||||||||||||||||||||||||||||| and the HR using the results from the April 7, 2020, data cut-off was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.

PFS for pemigatinib versus ASC alone was not assessed.

Harms Results

No comparisons for harms or safety were incorporated in the sponsor’s ITC.

Critical Appraisal

There were potentially important underlying differences between the FIGHT-202 and ABC-06 studies. In particular, the FGFR2 alterations were not reported in the ABC-06 trial. Given that FGFR2 alterations occur almost exclusively in iCCA and that the prevalence of FGFR2 alterations is less than 20%⁸ in patients with iCCA, there is likely a large disparity in FGFR2 mutation status between the study populations. While the FIGHT-202 study included only patients with CCA, the ABC-06 study included patients with BTC, which encompasses gallbladder cancer and ampullary cancer in addition to CCA. Ninety-eight percent of patients in cohort A of the FIGHT-202 study had iCCA compared with 42% and 47% in the mFOLFOX plus ASC and ASC alone groups, respectively. Since disease type and FGFR2 status were more restricted in the FIGHT-202 study, these differences could not be addressed through the weighting of patients in the pemigatinib group.

The covariates chosen for adjustment were based on age, sex, ECOG PS, and serum albumin. The following baseline characteristics were also available for both studies and did not appear to be considered: disease stage, percentage of patients with prior surgery for cancer, and number of lines of prior systemic therapy for advanced or metastatic cancer. The clinical experts consulted by CADTH for this review were of the opinion that the number of lines of previous therapy was of key importance in terms of prognosis. The clinical experts were not aware of any additional prognostic factors and/or effect modifiers that were not reported in both studies and should have been considered.

While there are retrospective studies suggesting that the presence of FGFR2 mutations in CCA may be associated with a better prognosis,^18,19 the clinical experts consulted by CADTH were of the opinion that FGFR2 mutation status was not an important prognostic factor in the indicated patient population. The clinical experts considered the fact that patients in both the FIGHT-202 and ABC-06 trials had progressed on prior systemic therapy to be of greater importance in terms of prognosis. The clinical experts expected patients in the FIGHT-202 study to have more advanced disease than patients in the ABC-06 study because the FIGHT-202 study population was more heavily pre-treated overall. It is unclear whether the pemigatinib group was more or less similar to the ASC-06 groups in this respect following weighting, as the weighting process did not take the number of prior lines of systemic therapy into account.

The effective sample size of the pemigatinib group was reduced by approximately 50% after weighting to the mFOLFOX plus ASC and ASC alone groups, and it is unclear how representative the post-weighting pemigatinib groups are of cohort A of the FIGHT-202 study.

Comparisons of pemigatinib with other relevant comparators (FOLFIRI, fluorouracil alone or in combination with cisplatin or oxaliplatin, and capecitabine alone or in combination with cisplatin or oxaliplatin) were not available. Given that mFOLFOX plus ASC is the only therapy beyond the first-line setting with RCT evidence of an OS benefit, the clinical experts consulted by CADTH expected that mFOLFOX plus ASC would have the greatest efficacy out of all the relevant comparators.

In summary, for the unanchored MAIC to produce unbiased treatment effect estimates, both the effect modifiers and prognostic variables need to be adjusted for in the analysis. Residual confounding remains the major limitation of the MAIC despite adjusting for age, sex, ECOG PS, and serum albumin in the comparisons of pemigatinib with mFOLFOX plus ASC and ASC alone. While any bias introduced by the differences between the FIGHT-202 and ASC-06 studies in the number of prior lines of systemic therapy may have been against pemigatinib, the substantial differences in FGFR2 mutation status and tumour site between trials introduce a high degree of uncertainty in the OS and PFS results. Furthermore, MAICs cannot account for unknown cross-trial differences; thus, the MAIC estimates are susceptible to bias from unknown confounding. An evaluation of potential bias from residual confounding was not reported; therefore, the magnitude of this bias in the relative treatment effect estimates is unclear. Overall, it remains uncertain whether pemigatinib provides additional OS or PFS benefit versus mFOLFOX plus ASC or ASC alone.

Other Relevant Evidence

One additional relevant report was summarized that was included in the sponsor’s submission to CADTH. FIGHT-101 is an ongoing, open-label phase I/II dose-escalation and expansion study of pemigatinib among participants with previously treated advanced malignancies with and without FGF/FGFR alterations. As of February 2019, FIGHT-101 had enrolled 160 participants from 14 study sites in the US and Denmark. Of these, 116 received at least 1 dose of pemigatinib monotherapy. Sixteen participants who were treated with pemigatinib monotherapy had CCA; 6 of these patients had FGFR2 rearrangements or fusions and received pemigatinib 13.5 mg orally once daily on a schedule of 2 weeks on, 1 week off for each 21-day cycle. All 6 of these patients had a reduction in tumour volume during the study. The best overall response of PR was observed in 1 of the 6 aforementioned patients with a DOR of 101 days, and the remaining 5 participants had a best overall response of stable disease. In terms of safety outcomes, 1 TEAE of non-serous grade 2 retinal edema occurred in 1 participant with CCA who was receiving 13.5 mg of pemigatinib monotherapy on an intermittent schedule. However, due to the open-label design and the limited data on the efficacy of pemigatinib on CCA within the report, the ability to interpret these results is considerably limited.

Conclusions

One phase II, singe-arm, open-label trial (FIGHT-202) provided evidence regarding the efficacy and safety of pemigatinib in patients with advanced, metastatic, or surgically unresectable CCA with FGFR2 alterations (cohort A) whose disease did not respond to previous therapy. The FIGHT-202 trial achieved the predetermined threshold for a positive outcome (a lower limit of the 95% CI for ORR > 15%) in cohort A. The clinical experts consulted by CADTH felt that the achieved ORR of 37% (April 7, 2020, data cut-off date) was clinically meaningful for the target population and durable (median of 8.08 months; range, 5.65 to 13.14). In the opinion of the clinical experts, the observed responses appeared higher than what is seen with the therapies currently used in the second line in this setting. There was uncertainty around the magnitude of the clinical benefit, given the limitations in the evidence from the non-comparative phase II clinical trial. The clinical experts consulted by CADTH noted that, despite the high unmet need, conducting an RCT with a targeted therapy such as pemigatinib, compared with the therapies that are currently available in the second line in Canadian clinical practice, would likely not be feasible. While the secondary efficacy outcomes, OS and PFS, appeared supportive of the observed ORR achievements, the non-randomized design of the FIGHT-202 trial made interpreting PFS and OS events attributable to pemigatinib challenging. In the absence of a direct comparison of pemigatinib with relevant treatment options, the sponsor submitted an ITC. However, the CADTH critical assessment identified limitations with the sponsor’s submitted unanchored MAIC (including heterogeneity across study designs and populations and the inability to adjust for all potential confounders and prognostic variables), which limited the ability to interpret the relative treatments effects observed between pemigatinib and other treatments. The results for the HRQoL and symptom severity exploratory outcomes remained inconclusive due to a number of important limitations. The toxicity profile of pemigatinib was considered manageable by the clinical experts consulted by CADTH and appeared favourable compared with currently available chemotherapy options. However, the non-comparative design of the FIGHT-202 trial made interpreting the safety events attributable to pemigatinib challenging, since all patients in cohort A received the same treatment.

Introduction

Disease Background

Hepatobiliary cancers are highly lethal cancers and refer to cancers arising in the liver (hepatocellular carcinoma), gall bladder, and bile ducts (iCCA and eCCA). Gallbladder cancer and CCA are known as BTC.³ The most common type of liver cancer is hepatocellular carcinoma, followed by BTC, accounting for 70% to 85% and 10% to 15% of all primary liver cancers, respectively.^1,2 Gallbladder cancer is the most common type of BTC.³

CCAs are most commonly adenocarcinomas³ and comprise 2 main subtypes: iCCA, initiating from the biliary tree within the liver, and eCCA, initiating outside the liver parenchyma; eCCA is subdivided into perihilar CCA or Klatskin tumour and distal CCA.² With eCCA accounting for 80% to 90% of all CCA, iCCA is the least frequently reported subtype.² The incidence of CCA is generally low (0.3 to 3.5 per 100,000) in Europe, the US, and Australia, but is higher in other parts of the world where certain parasite infections are common (e.g., Thailand, China, and Korea); northeast Thailand has the highest incidence (90 per 100,000) of CCA in the world.² In Canada and the US, respectively, there are approximately 400 and 5,000 new cases of CCA diagnosed each year.⁴ The median age at diagnosis is 65 years in Western industrialized nations.⁵ The 5-year relative survival rates for iCCA and eCCA, respectively, are 9% and 10%. The 5-year relative survival rates broken down by stages of disease are 25%, 8%, and 2%, respectively, for localized, regional, and distant iCCA; and 15%, 16%, and 2% for localized, regional, and distant eCCA, respectively.⁶

While most CCAs arise spontaneously, without any known risk factors,⁷ established risk factors for CCAs include primary sclerosing cholangitis, chronic ulcerative colitis, cysts in the bile ducts, and infection with a Chinese liver fluke parasite.²⁰ Additionally, in Western countries, hepatitis C and liver cirrhosis have been identified as risk factors for iCCA, while obesity, diabetes mellitus, metabolic disease, and certain substances (alcohol, tobacco, oral contraceptive pills, dioxin, and asbestos) have also been suggested as risk factor for CCA.⁷

Diagnosis of CCA is most commonly made in advanced stages (70% of patients are diagnosed with unresectable, locally advanced or metastatic disease) due to an absence of symptoms until later in the course of the disease.⁷ The rate of recurrence is high among the minority of patients who are able to undergo potentially curative surgery.⁹ Symptoms commonly appear when a bile duct is blocked and include jaundice, itching, light-coloured and greasy stools, dark urine, abdominal pain, loss of appetite or weight loss, fever, and nausea and vomiting.⁸

Different genetic alterations in BTC with oncogenic properties have been identified in recent years. Nearly 40% of patients harbour genetic alterations (e.g., isocitrate dehydrogenase 1 [IDH1] or isocitrate dehydrogenase 2 [IDH2], FGFR2, BRAF, and HE2/neu)⁹; however, evaluation of targeted treatment options is hampered by the low overall patient numbers.¹¹ One of the most frequent genetic alterations in patients with iCCA involves FGFR2.⁷ FGFR2 fusions or rearrangements are found in 10% to 20%¹⁰ of patients with iCCA, while they rarely occur in eCCA.¹¹ Alterations involving other members of the FGFR are rare, with an incidence below 0.5%.¹¹ While there is strong genetic and functional evidence that FGFR genetic alterations can drive the formation of tumours,⁷ it is currently not known whether FGFR2 alteration–positive patients represent a distinct prognostic subgroup¹¹ and/or respond differently to chemotherapy compared with patients with unselected CCA.¹² Retrospective studies^19,21-23 in the first-line setting in patients with CCA suggest that patients with FGFR alterations appear to have a better prognosis compared with patients with unselected CCA, and that FGFR alterations occur more frequently in young women.⁹ A number of phase II studies have been published that report the results of FGFR2-directed therapies. Notably, 2 non-comparative phase II trials have reported very similar results for patients with iCCA that is positive for FGFR2 fusions or rearrangements: objective responses were 35.5% (38 out of 107), with an estimated median PFS of 6.9 months for pemigatinib, and 31% (22 out of 71) with a median PFS of 5.8 months in patients treated with infigratinib.¹¹ Pemigatinib was approved by the FDA in April 2020 for the treatment of adults with previously treated, unresectable, locally advanced, or metastatic CCA with an FGFR2 fusion or other rearrangement as detected by an FDA-approved test.²⁴ Since March 2021, pemigatinib has been authorized in the European Union²⁵ as monotherapy and is indicated for the treatment of adults with locally advanced or metastatic CCA with an FGFR2 fusion or rearrangement whose cancer has progressed after at least 1 prior line of systemic therapy.²⁶ The FDA has granted accelerated approval to infigratinib for the treatment of adult patients with previously treated, unresectable, locally advanced, or metastatic CCA with an FGFR2 fusion or other rearrangement as detected by an FDA-approved test.²⁷ Currently, 3 phase III RCTs are recruiting patients with CCA that is positive for FGFR2 fusions or rearrangements that compare first-line standard of care (gemcitabine plus cisplatin) with the following treatments:

• pemigatinib (FIGHT-302 [NCT03656536], which has an estimated completion date of June 28, 2026),²⁸ which has a frequency of administration that is different than that in FIGHT-202 (where it is administered continuously)

• infigratinib (PROOF [NCT03773302],²⁹ which has an estimated completion date of September 30, 2024)

• futibatinib (FOENIX-CCA3 [NCT04093362],³⁰ which has an estimated completion date of February 2026).

All 3 trials have PFS as the primary end point.

Standards of Therapy

Most patients with CCA have advanced-stage disease at the time of diagnosis and although surgery is the preferred treatment option, only 35% of patients are eligible for surgical resection with curative intent.⁵ For patients with advanced-stage or unresectable CCA and a good ECOG PS (0 or 1), standard-of-care first-line treatment is gemcitabine and cisplatin.⁹ If there are concerns about a patient’s renal function, oxaliplatin may be substituted for cisplatin.² For patients with an ECOG PS of 2, gemcitabine monotherapy may be considered as first-line therapy. The median OS, median PFS, and ORR in patients with BTCs treated with standard-care, first-line palliative treatment with gemcitabine and cisplatin range from 11.2 to 11.7 months, 5.8 to 8.0 months, and 19.5% to 26.1%, respectively.¹² Patients with molecularly unselected iCCA treated with standard first-line therapy were shown to have median PFS and median OS of 8.4 and 15.4 months, respectively.¹² There are currently no funded standard treatment options for patients in the second line once the disease has progressed on first-line treatment.⁹ In the absence of proven treatment options in the second-line setting for patients with CCA, participation in a clinical trial and best supportive care are recommended, including alleviating biliary obstruction and full access to palliative care and symptom management.² According to the clinical experts consulted by CADTH, the second-line therapies used in Canadian clinical practice include FOLFOX, FOLFIRI, fluorouracil (alone or in combination with cisplatin or oxaliplatin), and capecitabine (alone or in combination with cisplatin or oxaliplatin). A systematic review,³¹ including 761 patients participating in case reports, retrospective analyses, or phase II trials of second-line therapies for advanced BTC, reported a mean OS of 7.2 months (95% CI, 6.2 to 8.2), a mean PFS of 3.2 months (95% CI, 2.7 to 3.7), and a response rate of 7.7% (95% CI, 4.6% to 10.9%).³¹

Second-line treatment with FOLFOX is currently the only drug based on phase III trial data in this setting.⁵ The ABC-06 trial¹³ evaluated the efficacy and safety of FOLFOX plus ASC compared with ASC alone in patients with locally advanced or metastatic BTC (including CCA and gallbladder or ampullary carcinoma) whose disease had progressed on first-line cisplatin and gemcitabine therapy. At the median follow-up time of 21.7 months, median OS was 6.2 months in the FOLFOX group and 5.3 months in the control group (HR = 0.69; 95% CI, 0.50 to 0.97; P = 0.031); in the FOLFOX group, median PFS was 4 months and an objective response was observed in 5% of patients.¹³

In Canada, there are currently no funded standard targeted treatment options for patients with CCA whose disease harbours generic alterations that have been identified for targeted therapeutics. Another common genetic alteration in iCCA are IDH1 and IDH2 mutations, which are found in 10% to 23% of iCCA, and a targeted treatment, ivosidentib,³² has received priority review by the FDA for the treatment of patients with previously treated, IDH1-mutant CCA.³³

There was consensus among the clinicians that there is an unmet need for effective therapies with an acceptable toxicity profile that achieve disease control, delay worsening of symptoms, maintain HRQoL, delay disease progression, and prolong survival. It was also mentioned by the clinical experts that there are currently no biomarker-directed regimens specifically for patients with CCA that is FGFR2-positive. The experts anticipated more promising benefit with a targeted therapy option in later lines than with chemotherapy with a disease that is steadily growing more resistant.

Drug

Pemigatinib is a molecule kinase inhibitor with antitumour activity that inhibits FGFRs. FGFRs are receptor tyrosine kinases that activate signalling pathways in tumour cells.¹⁴

On September 17, 2021, pemigatinib was issued market authorization with conditions by Health Canada for the treatment of adults with previously treated, unresectable, locally advanced, or metastatic CCA with an FGFR2 fusion or other rearrangement. The sponsor’s requested reimbursement criteria for pemigatinib are as per the Health Canada–approved indication. In addition, the Health Canada indication states that the “clinical effectiveness of pemigatinib is based on ORR and DOR from a single-arm phase II trial in patients with specific FGFR2 rearrangements. Treatment with pemigatinib should be initiated following confirmation of a susceptible genetic alteration using a validated test.”¹⁵ Pemigatinib underwent review by Health Canada through a standard review pathway. Pemigatinib has no other Health Canada–approved indication and has not previously been reviewed by CADTH.

After being granted priority review and breakthrough therapy and orphan drug designations, pemigatinib received an accelerated approval by the FDA in April 2020 for the treatment of adults with previously treated, unresectable, locally advanced, or metastatic CCA with an FGFR2 fusion or other rearrangement, as detected by an FDA-approved test.²⁴ Since March 2021, pemigatinib has been authorized in the European Union²⁵ as monotherapy and is indicated for the treatment of adults with locally advanced or metastatic CCA with an FGFR2 fusion or rearrangement whose cancer has progressed after at least 1 prior line of systemic therapy.²⁶ Pemigatinib has also been approved in Japan.³⁴

Oral pemigatinib is available as 4.5 mg, 9 mg, and 13.5 mg tablets. The recommended starting dose is 13.5 mg administered orally for 14 consecutive days followed by 7 days off therapy, in 21-day cycles. The product monograph states that treatment is to be continued until disease progression or unacceptable toxicity. Furthermore, it is recommended that a low-phosphate diet be initiated when the phosphate level is greater than 5.5 mg/dL, and that consideration be given to adding a phosphate-lowering therapy when the level is greater than 7 mg/dL. The dose of phosphate-lowering therapy is to be adjusted until the phosphate level returns to less than 7 mg/dL. It is recommended that consideration be given to discontinuing phosphate-lowering therapy during pemigatinib treatment breaks or if the phosphate level falls below normal.¹⁵

Stakeholder Perspectives

Patient Group Input

The information in this section is a summary of input provided by the patient groups that responded to CADTH’s call for patient input for the purpose of this review. The full patient input received is included in Appendix 5.

Three patient groups, the Canadian Liver Foundation, the Canadian Organization for Rare Disorders, and the Cholangiocarcinoma Foundation, co-created 1 patient input for this review. The input was based on an online survey and a virtual focus group. A total of 27 respondents were included in the patient input. Of the respondents, 15 were patients who had been diagnosed with CCA (4 patients had CCA with FGFR2 fusions), 2 were patients who had symptoms of CCA but no diagnosis, and 10 were caregivers or family members of patients with CCA.

Respondents indicated a varying range of CCA symptoms affecting patients’ daily activities (including their social, work, and school lives and their relationships) and causing detrimental effects on patients’ QoL. Respondents highlighted problems with intimacy or sexual desire, fatigue, and anxiety. Other commonly experienced symptoms indicated by respondents included unintended weight loss, insomnia, gastrointestinal problems, abdominal pain, constipation, depression, and neuropathy. According to the 3 patient groups, delayed diagnosis, misdiagnosis, and a lack of specialists and treatment options available for this rare cancer significantly contribute to patients’ feelings of stress and anxiety and may delay or eliminate treatment options.

According to the patient input received, respondents reported they expect the following key outcomes to be improved from any new drug or treatment: QoL, tumour response, delay in disease progression, and additional treatment choice. Additionally, it was highlighted by the 3 patient groups that the identification of gene mutations and the development of targeted therapies was perceived to be very important by respondents and would spur the hope for curable options. Four respondents indicated they had direct experience with taking pemigatinib. Respondents indicated little overall challenge dealing with the side effects from pemigatinib.

Clinician Input

Input From the Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise regarding the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, and providing guidance on the potential place in therapy). The following input was provided by 2 clinical specialists with expertise in the diagnosis and management of CCA.

Unmet Needs

The clinical experts consulted by CADTH noted there are currently no funded standard second-line treatment options for patients with unresectable, locally advanced, or metastatic CCA with FGFR2 alterations. Until recently, no high-quality evidence was available in this setting. The first randomized phase III RCT that included patients with CCA in the second line evaluated the chemotherapy option mFOLFOX plus ASC compared with ASC alone in patients with BTC after standard first-line chemotherapy. mFOLFOX was associated with high toxicity, low activity (the ORR achieved with the FOLFOX regimen was 5%; CR = 1%; PR = 4%; stable disease = 28%), and low efficacy (median OS of 6.2 months with the mFOLFOX regimen versus 5.3 months with ASC alone). The clinical expert noted that, currently, mFOLFOX is the most commonly used therapy in the target population in Canadian clinical practice. It was emphasized by the clinical experts that patients with disease that has progressed on first-line chemotherapy often have a rapidly declining ECOG PS, and only a small proportion of patients may be suitable for further systemic treatment. There was consensus among the clinicians that there is an unmet need for effective therapies with acceptable toxicity profiles that achieve disease control, delay worsening of symptoms, maintain HRQoL, delay disease progression, and prolong survival. It was also mentioned by the clinical experts that there are currently no biomarker-directed regimens specifically for patients with CCA whose disease is FGFR2-positive. The experts anticipated more promising benefit with a targeted therapy option in later lines than with another line of chemotherapy with a disease that is steadily growing more resistant.

Place in Therapy

The clinical experts consulted by CADTH stated that pemigatinib was to be used in adult patients with previously treated, unresectable, locally advanced, or metastatic CCA with an FGFR2 fusion or other rearrangement as per the FIGHT-202 trial. It was agreed that oral pemigatinib would likely shift the current treatment paradigm. It was also noted that in the absence of robust comparative data, no firm conclusions could be drawn on how pemigatinib compared with other relevant treatment options; however, the clinical experts anticipated that, based on the FIGHT-202 results and on poor results with existing treatment options in clinical practice, pemigatinib appeared to offer at least similar or improved clinical benefits compared with current therapies, with better tolerability. The FIGHT-202 trial excluded patients who would be intolerant to standard first-line therapy without experiencing PD. The clinical experts consulted by CADTH felt it would be reasonable to generalize the results from cohort A to patients with FGFR2 alterations who are intolerant to first-line therapy, given the favourable safety profile of oral pemigatinib. Furthermore, the clinical experts anticipated seeing the benefit of treatment with pemigatinib regardless of the number of previous lines of systemic therapy, as long as the patients’ cancer has the FGFR2 alteration. However, the clinical experts agreed that patients should not have been previously treated with an FGFR2-targeted therapy.

Patient Population

Overall, the clinical experts consulted by CADTH agreed that patients, as selected per the inclusion and exclusion criteria for cohort A of the FIGHT-202 trial, should be eligible for pemigatinib therapy. Among patients enrolled in cohort A of the FIGHT-202 trial, the clinical experts did not identify any subgroups of patients who would potentially be either best suited for or benefit the least from oral pemigatinib. It was emphasized that patients with CCA that does not have the FGFR2 alteration would not be expected to derive any benefit from pemigatinib. The clinical experts noted that patients with disease that has progressed on first-line chemotherapy often have a rapidly declining ECOG PS and only a small proportion of patients may be suitable for further systemic chemotherapy treatment; for example, platinum-based therapies can lead to neuropathy (after approximately 6 months), and patients who receive standard first-line platinum-based chemotherapy may not be able to be treated with second-line FOLFOX due to this toxicity.

The clinical experts considered that the determination of the presence of FGFR2 fusions or other rearrangements is required before the initiation of treatment with pemigatinib. The clinical experts noted it would be ideal to have molecular FGFR2 testing results available after commencing first-line therapy and before the patients’ disease progresses. A valid test would involve next-generation sequencing, which was used in the FIGHT-202 trial to enrol patients into cohort A. FGFR2 testing is currently not routinely available nor funded in jurisdictions in Canada.

Assessing Response to Treatment

In the opinion of the clinical experts consulted by CADTH, assessments to evaluate the response to treatment include regular radiological imaging (i.e., CT or MRI) and a CA19 to 9 biomarker test every 2 to 3 months to determine whether a patient’s disease has progressed. In addition, patients would be seen by an oncologist every 3 to 4 weeks for clinical assessment (i.e., to assess disease symptoms and the patient’s ECOG PS).

The clinical experts indicated that the most clinically meaningful responses to treatment include disease control (i.e., disease stability or response), improvement in disease-related symptoms, better pain control, weight gain, regaining a more active lifestyle, maintenance of HRQoL, and prolonged PFS and OS. Acceptable drug-related toxicities was also noted as a clinically meaningful outcome.

Discontinuing Treatment

In the opinion of the clinical experts consulted by CADTH, treatment with pemigatinib should be discontinued if a patient experiences disease progression, has a worsening ECOG PS, is intolerant to or experiences unacceptable toxicity from pemigatinib (which cannot be improved with dose delays or reductions), or the patient is not interested in continuing treatment.

Prescribing Conditions

The clinical experts consulted by CADTH noted that because pemigatinib is an oral drug that is self-administered at home, patients should have regular access to outpatient oncology clinics to confirm treatment tolerance and that the disease has not progressed.

Clinician Group Input

The information in this section is a summary of 2 inputs provided by the registered clinician groups that responded to CADTH’s call for clinician input for the purpose of this review.

Two clinician group inputs were provided, 1 from the Ontario Health (Cancer Care Ontario) Gastrointestinal Cancer Drug Advisory Committee and 1 from the CGOEN and other physicians who treat CCA. The views of the clinician groups overall were consistent with the clinical experts consulted by CADTH, indicating that the most important treatment goals are achieving disease control, delaying worsening of symptoms, maintaining HRQoL, delaying disease progression, and prolonging survival. It is also important that the drug have an acceptable safety profile. The clinicians from the CGOEN also highlighted that the convenient oral route of administration of pemigatinib would contribute to improvements in QoL for patients, as fewer visits to a cancer centre and less chair time would be required compared with alternative treatment options. The clinicians from CGOEN further suggested it would be reasonable to consider pemigatinib upfront for patients deemed unsuitable for standard first-line chemotherapy. This clinician group also noted that patients with compromised hepatic function or significant hyperbilirubinemia would be least suitable for treatment with pemigatinib. The clinicians from both inputs anticipated that pemigatinib would offer clinically meaningful benefit and improved efficacy to patients, with the potential for improved QoL.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may impact their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 4.³⁶

Clinical Evidence

The clinical evidence included in the review of pemigatinib is presented in 3 sections. The first section, the systematic review, includes the pivotal study or studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those studies that were selected according to an a priori protocol. The second section includes indirect evidence from the sponsor. The third section includes an additional relevant study that was considered to address important gaps in the evidence included in the systematic review.

Systematic Review (Pivotal and Protocol Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of pemigatinib tablets (13.5 mg orally once daily on a schedule of 2 weeks on and 1 week off for each 21-day cycle) for the treatment of adult patients with previously treated, unresectable, locally advanced, or metastatic CCA with an FGFR2 fusion or other rearrangement.

Methods

Studies selected for inclusion in the systematic review included the pivotal study provided in the sponsor’s submission to CADTH and Health Canada, as well as those meeting the selection criteria presented in Table 5. Outcomes included in the CADTH review protocol reflect outcomes considered to be important to patients, clinicians, and drug plans.

Of note, the systematic review protocol presented subsequently was established before the granting of a Notice of Compliance by Health Canada.

The literature search for clinical studies was performed by an information specialist using a peer-reviewed search strategy according to the PRESS Peer Review of Electronic Search Strategies checklist.³⁷

Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946–) through Ovid and Embase (1974–) through Ovid. All Ovid searches were run simultaneously as a multi-file search. Duplicates were removed using Ovid deduplication for multi-file searches, followed by manual deduplication in Endnote. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concept was pemigatinib. Clinical trials registries were searched: the US National Institutes of Health’s clinicaltrials.gov, WHO’s International Clinical Trials Registry Platform (ICTRP) search portal, Health Canada’s Clinical Trials Database, and the European Union Clinical Trials Register.

No filters were applied to limit the retrieval by study type. Retrieval was not limited by publication date or by language. Conference abstracts were excluded from the search results. See Appendix 1 for the detailed search strategies.

The initial search was completed on July 19, 2021. Regular alerts updated the search until the meeting of the CADTH pan-Canadian Oncology Drug Review Expert Committee (pERC) on November 10, 2021.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool For Searching Health-Related Grey Literature checklist.³⁸ Included in this search were the websites of regulatory agencies (FDA and European Medicines Agency). Google was used to search for additional internet-based materials. See Appendix 1 for more information on the grey literature search strategy.

These searches were supplemented by reviewing bibliographies of key papers and through contacts with appropriate experts. In addition, the manufacturer of the drug was contacted for information regarding unpublished studies.

Two CADTH clinical reviewers independently selected studies for inclusion in the review based on titles and abstracts, according to the predetermined protocol. Full-text articles of all citations considered potentially relevant by at least 1 reviewer were acquired. Reviewers independently made the final selection of studies to be included in the review, and differences were resolved through discussion.

A focused literature search for network meta-analyses dealing with pemigatinib or CCA was run in MEDLINE All (1946–) through Ovid and Embase (1974–) through Ovid on July 16, 2021. No limits were applied, and conference abstracts were excluded from the search results.

Findings From the Literature

A total of 4 reports presenting data from 1 unique study were identified from the literature for inclusion in the systematic review (Figure 1). The included study is summarized in Table 6. A list of excluded studies is presented in Appendix 2.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Description of Studies

FIGHT-202 is an ongoing, multi-centre, open-label, single-arm phase II trial evaluating the efficacy and safety of pemigatinib in patients with advanced, metastatic, or surgically unresectable CCA with FGFR2 alterations, alterations in other FGFs or FGFRs, or no FGF or FGFR alterations, whose disease did not respond to previous therapy. Patients were assigned to 3 cohorts depending on their FGF/FGFR status (cohort A: FGFR2 fusions or rearrangements; cohort B: FGF/FGFR alterations other than FGFR2 fusions or rearrangements; or cohort C: negative for FGF/FGFR alterations). This CADTH review focuses on cohort A, since cohorts B and C were not part of the requested reimbursement criteria submitted to CADTH and not approved in the Health Canada Notice of Compliance with conditions and are therefore beyond the scope of this review. Selected results for cohorts B and C have been included in Appendix 3. The primary objective of the trial was to assess the efficacy of pemigatinib in patients with advanced, metastatic, or surgically unresectable CCA with FGFR2 alterations (patients in cohort A) that did not respond to at least 1 previous treatment. Patients in this international trial were enrolled at 67 sites across 12 countries, which are listed in Table 6. The majority of sites were in the US followed by Europe, with no sites in Canada. Enrolment started on January 12, 2017 and ended on March 22, 2019.¹⁶

A total of 146 patients were enrolled to received oral pemigatinib (13.5 mg orally once daily on a schedule of 2 weeks on, 1 week off for each 21-day cycle). Best supportive care was administered as needed and included but was not limited to palliative radiotherapy for bone lesions, stent placement, or replacement of blocked bile ducts.¹⁷ Pemigatinib was administered until documented disease progression or unacceptable toxicity.¹² The study design is depicted in Figure 2.

To be enrolled and for the initial cohort assignment, patients had to have results from a certified local laboratory; however, final cohort assignment for the statistical analyses was based on the results of next-generation sequencing (using the Foundation Medicine Clinical Trial Assay) from the central genomics laboratory.¹² The study consisted of 3 phases, the screening phase (lasting up to 28 days), the treatment phase, and the follow-up phase. During the follow-up phase, patients were followed for safety (final follow-up visit 30 to 35 days after end of treatment), for disease status every 9 weeks (for patients who discontinued pemigatinib for reasons other than disease progression), and for OS at least every 12 weeks.¹² Response assessment, according to Response Evaluation Criteria in Solid Tumours Version 1.1 (RECIST 1.1) (Eisenhauer et al.)⁴⁰ was based on radiologic imaging and performed by an IRC and occurred every 2 cycles (every 6 weeks) for the first 4 cycles and every 3 cycles (every 9 weeks) thereafter, and at the treatment discontinuation visit.¹² Tumour response was also evaluated by the investigator.¹² Safety and tolerability were evaluated by monitoring the frequency, duration, and severity of AEs.¹²

The predetermined threshold for a positive study outcome (lower limit of the 95% CI for ORR > 15%) was achieved as of the March 22, 2019, data cut-off date.¹⁶ While the timing of this data cut-off date was not pre-specified a priori in the statistical analysis plan, the sponsor’s proposed timing was agreed upon by the FDA during its review process for pemigatinib.⁵ Two additional updated analyses occurred at the August 2019 and April 2020 data cut-off dates: the former was a 4-month safety update required for the FDA NDA, the latter was performed to support the safety data summaries for another indication outside of Canada.¹⁷ The trial is still ongoing, with an estimated completion date in the first quarter of 2022.¹⁷ The FIGHT-202 trial was sponsored by Incyte Corporation.³⁹

Figure 2: Study Design

CR = complete response; D/C = discontinue; FGF = fibroblast growth factor; FGFR = fibroblast growth factor receptor; PD = progressive disease; PR = partial response; SD = stable disease; SF = screen fail; tx = therapy.

Source: Clinical Study Report.¹⁶

Populations

Inclusion and Exclusion Criteria

The key inclusion and exclusion criteria used in the FIGHT-202 trial are described in Table 6. Briefly, the trial enrolled adults, aged 18 years or older, diagnosed with advanced, metastatic, or surgically unresectable CCA, whose disease progressed after at least 1 line of prior systemic therapy (prior therapy with selective FGFR inhibitors was not permitted). Patients had to have documented FGF/FGFR gene alteration status and radiologically measurable disease according to RECIST 1.1. At screening, patients had to have an ECOG PS of 0 to 2, a life expectancy of at least 12 weeks, adequate hepatic and renal function, and no clinically significant corneal or retinal disorder, confirmed by ophthalmologic examination.¹⁶

Baseline Characteristics

The baseline characteristics of patients who comprised the safety population of the FIGHT-202 trial are summarized in Table 7. At baseline, 107 patients were identified as having FGFR2 fusions or rearrangements and were grouped into cohort A. Cohort B included 20 patients with other FGF/ FGFR alterations other than FGFR2, and cohort C included 18 patients with no identified FGF/FGFR alterations. One patient who was placed into an “undetermined” group was not assigned to any of the 3 cohorts, as the local FGF/FGFR status results could not be confirmed by the central genomics laboratory. For patients in cohort A, the mean age was 55.3 years (SD = 12.02); most patients were female (60.7%), and most were enrolled in trial sites in North America (59.8%) or Europe (29.9%). Almost all patients (89% of patients overall and 98.1% of patients in cohort A) had iCCA. The majority of patients in cohort A had metastatic disease (82.2%), with the ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Median time since diagnosis was 1.28 years (range, 0.03 to 11.1 years) in patients in cohort A. The majority of patients in cohort A had an ECOG PS of 1 (53.3%) or 0 (42.1%) and all patients had received at least 1 line of prior systemic therapy for advanced or metastatic disease (60.7%, 27.1%, and 12.1% of patients received 1, 2, and ≥ 3 prior lines, respectively). Renal and hepatic impairment grades were normal or mild for most patients in cohort A (39.3% and 43.9% had normal and mild renal impairment grades, respectively; 44.9% and 48.6% had normal and mild hepatic grades, respectively).¹⁶

All patients in cohort A had received at least 1 prior systemic cancer therapy (Table 8). The majority of patients had received platinum-based chemotherapy regimens before study enrolment, most commonly gemcitabine, reported as gemcitabine (85.0%) or gemcitabine hydrochloride (7.5%) and cisplatin (75.7%). The second and third most commonly received pyrimidine analogues were fluorouracil (29.0%) and ||||||||||||||||||||||||||||||||; the second most frequently administered platinum compound was oxaliplatin (38.3%).¹⁶ |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.³⁹

All patients in cohort A (N = 107) had centrally confirmed FGFR2 fusions or rearrangements with 56 unique rearrangement or fusion partners identified (Table 9).¹⁶ The most frequently identified FGFR2 partner was BICC1 (31 patients; 29%). There were 5 patients with FGFR rearrangements referred to as FGFR2-N/A, as no specific partner could be identified.³⁹

Interventions

Patients enrolled in the FIGHT-202 trial self-administered oral pemigatinib once daily on a schedule of 2 weeks on, 1 week off. Treatment and dose are described in Table 10.

Dose Modifications

Dose interruptions, delays, and modifications were permitted and were guided by the occurrence of toxicities (related or unrelated to the study drug). A maximum of 2 dose reductions (from a daily dose of 13.5 mg to 9 mg, and from 9 mg to 6 mg) was recommended; patients could not receive a dose below 6 mg daily. No modification of the treatment schedule was allowed. At the occurrence of any grade 3 toxicity not manageable by supportive care, or a level of aspartate transaminase and/or alanine transaminase greater than 5.0 times the upper limit of normal, treatment could be interrupted for up to 14 days until toxicity resolved to grade 1 or less. Once a dose interruption occurred, pemigatinib was either restarted at the same dose that was being taken before the dose interruption occurred, or at the next-lower dose and monitored as clinically indicated. Situations in which treatment was delayed for more than 14 days before restarting treatment had to be discussed on a case-by-case basis with the sponsor. In cases of recurrent grade 3 toxicity after 2 dose reductions or any other grade 4 toxicity, pemigatinib administration had to be discontinued; exceptions to that required the sponsor’s approval.¹⁶

Concomitant Medication

Concomitant medications were allowed to treat comorbidities or AEs as long as they did not include potent cytochrome P3A4 (CYP3A4) inhibitors and inducers or moderate CYP3A4 inducers (there was no restriction on topical ketoconazole), another selective FGFR inhibitor, an investigational study drug for any indication, or any anti-cancer medications other than the study drug.¹²

Outcomes

A list of efficacy end points identified in the CADTH review protocol that were assessed in the clinical trial included in this review is provided in Table 11. These end points are further summarized subsequently. A detailed discussion and critical appraisal of the HRQoL and symptom severity outcomes is provided in Appendix 4.

Overall Survival

OS was a secondary outcome of the FIGHT-202 trial and was defined as the time from the date of first dose until the date of death due to any cause.¹² The OS analyses were performed for patients in cohorts A, B, and C, respectively.¹²

Progression-Free Survival

PFS was a secondary outcome of the FIGHT-202 trial and was measured as the time from the date of the first dose of the study drug until the date that PD was first recorded or death, whichever occurred first. The date of PD was the time point at which progression was first documented. Disease progression for the analysis of PFS was determined according to RECIST 1.1 criteria as assessed by an independent centralized radiological review committee (see Table 29 in Appendix 3 for definitions for the evaluation of the target lesions). PFS was also analyzed based on investigator assessment. The PFS analyses were performed for patients in cohorts A, B, and C, respectively.¹²

Objective Response Rate

The ORR assessed in cohort A was the primary outcome of the FIGHT-202 trial. The ORR evaluated in cohort B, cohort A plus B, and cohort C were secondary outcomes. ORR was defined as the proportion of patients who achieved a best overall response of CR (disappearance of all target lesions) or a PR (≥ 30% decrease in the sum of the longest diameters of target lesions) at any post-baseline visit before the first instance of PD. Best overall response was defined as the best response documented post baseline before and including the first instance of PD, in the order of CR, PR, stable disease, PD, and not evaluable. Clinical response for the analysis of ORR was determined based on RECIST 1.1, assessed by an independent radiological review committee using central genomics laboratory results, and required confirmation of a CR and PR at least 4 weeks after the initial assessment (see Table 29 in Appendix 3 for definitions for the evaluation of the target lesions). The only clinical data transmitted by the site to the IRC were radiation history, prior surgeries, and investigator-documented benign radiographic abnormalities. For stable disease, the stable disease criteria had to be met at least once after the date of the first dose at a minimum interval of 39 days; in case these criteria were not met, patients were recorded as having an overall response of PD if the next available assessment indicated PD, or not evaluable if there was no additional assessment available. ORR was also analyzed based on investigator assessment; however, confirmation of a CR and PR was not required.¹²

Duration of Response

DOR was a secondary outcome of the FIGHT-202 trial and was defined as the interval from the date of a CR or PR (i.e., an overall response contributing to an objective response) to the date of death or first overall response of PD, whichever occurred first.¹² Clinical response for the analysis of DOR was determined based on RECIST 1.1, assessed by an independent radiological review committee. The DOR analyses were performed for patients in cohorts A, B, and C, respectively. DOR was also analyzed based on investigator assessment.¹²

Disease Control Rate

DCR was a secondary outcome of the FIGHT-202 trial and was defined as the proportion of patients with a best response of CR, PR, or stable disease. Clinical response for the analyses of DCR was determined based on RECIST 1.1, assessed by an independent radiological review committee using central genomics laboratory results, and required confirmation of a CR and PR at least 4 weeks after the initial assessment.¹² The DCR analyses were performed for patients in cohorts A, B, and C, respectively. DCR was also analyzed based on investigator assessment.¹²

Time to Progression

This outcome was not assessed in the FIGHT-202 trial.

Health-Related Quality of Life

The HRQoL outcomes measured in the trial included the EORTC QLQ-C30. No analysis plan, objective, or minimally important difference (MID) for the EORTC QLQ-C30 instrument were specified a priori in the statistical analysis plan; it was noted, however, that the scores for each scale were to be calculated. According to the study’s protocol scores for each scale, changes from baseline to each visit were to be measured and summarized descriptively.¹² The EORTC QLQ-C30 questionnaire was assessed at baseline and then every 3 cycles starting with cycle 3, until discontinuation of the study treatment and at the end-of-treatment visit.¹² A detailed discussion and critical appraisal of the HRQoL measures is provided in Appendix 4.

The EORTC QLQ-C30 is 1 of the most commonly used patient-reported outcome measures in oncology clinical trials. It is a multi-dimensional, cancer-specific, evaluative measure of HRQoL. It was designed specifically for the purpose of assessing changes in participants’ HRQoL in clinical trials in response to treatment. The EORTC QLQ-C30 questionnaire consists of 30 questions that are scored to create 5 multi-item functional scales, 3 multi-item symptom scales, 6 single-item symptom scales, and a 2-item QoL scale. Most questions have 4 response options (not at all, a little, quite a bit, very much), with scores on these items ranging from 1 to 4. For the 2 items that form the global QoL scale, however, the response format is a 7-point Likert-type scale, with anchors between 1 (very poor) and 7 (excellent). Version 3.0 of the questionnaire, which was used in the included trials in this report, is the most current version. Each raw scale score is converted to a standardized score that ranges from 0 to 100 using a linear transformation, with a higher score reflecting better function on the function scales, higher symptoms on the symptom scales, and better QoL (i.e., higher scores simply reflect higher levels of response on that scale). Thus, a decline in score on the symptom scale would reflect an improvement, whereas an increase in score on the function and QoL scales would reflect an improvement. It is available in 118 different languages on the EORTC Quality of Life Group website and is intended for use in adult populations only.

The reliability of the EORTC QLQ-C30 instrument was evaluated in an international study in patients with BTC.⁴¹ Internal consistency was assessed and was acceptable for all scales except for the physical function, cognitive function, and nausea/ vomiting scales, which had mixed results. Results for test–retest reliability were also mixed, with the intraclass correlation coefficient (ICC) for the scales ranging from 0.52 to 0.92. Construct validity and responsiveness were not assessed for the EORTC QLQ-C30.⁴¹ Estimates for MIDs in the literature were not found for the EORTC QLQ-C30 in patients with CCA or BTC.

Symptom Severity

Symptom severity was assessed in the trial using the EORTC QLQ-BIL21. The EORTC QLQ-BIL21 was an exploratory outcome in the FIGHT-202 trial. No analysis plan, objective, or MID for the EORTC QLQ-BIL21 instrument were specified a priori in the statistical analysis plan; it was noted, however, that the scores for each scale were to be calculated. According to the study’s protocol scores for each scale, changes from baseline to each visit were to be measured and summarized descriptively.¹² The EORTC QLQ-BIL21 questionnaire was assessed at baseline and then every 3 cycles starting with cycle 3, until discontinuation of the study treatment and at the end-of-treatment visit.¹² A detailed discussion and critical appraisal of the HRQoL measures is provided in Appendix 4.

The EORTC QLQ-BIL21 is a disease-specific module to be used in addition to the EORTC QLQ-C30 to assess HRQoL in patients with CCA and gallbladder cancer.⁴² It consists of 21 questions, with 18 of the items grouped into 5 scales: eating symptoms (4 items), jaundice symptoms (3 items), tiredness (3 items), pain symptoms (4 items), and anxiety symptoms (4 items).⁴² The remaining 3 items are single-item assessments of treatment side effects, difficulties with drainage bags or tubes, and concerns about weight loss.⁴² Patients complete the questionnaire based on a 1-week recall period by rating each item on a 4-point Likert scale (1 = not at all, 2 = a little, 3 = quite a bit, 4 = very much).⁴² The scores are then transformed linearly to a 0 to 100 scale to yield scale scores using EORTC guidelines, with higher scores indicating more severe symptoms.^41,42 The questions have been translated according to QoL group guidelines into Mandarin Chinese, Italian, German, Dutch, Spanish, and Hindi.^41,42

An international study was conducted to validate the EORTC QLQ-BIL21 in patients with BTC.⁴¹ The study included 172 adult patients with CCA and 91 patients with gallbladder cancer who had an expected minimum survival of 3 months and were undergoing treatment.⁴¹ Internal consistency was assessed and was acceptable for all multi-item scales of the EORT QLQ-BIL21 instrument. Test–retest reliability was assessed using the ICC, which showed good reproducibility. Known group validity was assessed and shown to distinguish groups based on Karnofsky performance status (KPS). There was some evidence for responsiveness to the change in the eating, jaundice, tiredness, pain, treatment side effects, and anxiety scales. The single-item assessment of difficulties with drainage bags and tubes was considered irrelevant by 29 patients.⁴¹ The study authors noted that not all patients experience drains during their treatment and that perhaps there should be a “not applicable” option for responding to that item.⁴¹ Estimates for MIDs in the literature were not found for the EORTC QLQ-BIL30 in patients with CCA or BTC.

Safety

Safety was designated as a secondary outcome in the FIGHT-202 trial and was defined as any untoward medical occurrence associated with the use of a drug in humans, whether or not the occurrence was considered drug-related, that occurred after a patient provided informed consent.¹² While the data listings included all AEs, the analysis of AEs was limited to TEAEs.¹² A TEAE included any AE reported for the first time or a worsening of a pre-existing event after the first dose of pemigatinib.¹² Abnormal laboratory values or test results observed in patients constituted AEs only if they were associated with clinical signs or symptoms, were considered clinically meaningful, required therapy (e.g., hematologic abnormality requiring transfusion), or required changes in the investigation study drug.¹² Disease progression was recorded as an AE only if there were no other identifiable AEs or serious adverse events (SAEs) associated with the disease progression at the time of reporting.¹²

AEs were organized based on the Medical Dictionary for Regularity Activities (MedDRA) preferred term and system organ class. Severity of AEs was defined according to the US National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03 (CTCAE 4.03). If a toxicity was not included in the CTCAE 4.03 criteria, it was graded on a scale of 1 to 4 (where 1 = mild, 2 = moderate, 3 = severe, and 4 = life-threatening).¹²

All AEs were documented by the investigator from the date a patient signed the informed consent form to at least 30 to 35 days after the end-of-treatment visit (or the last dose of the study drug if the end-of-treatment visit was not performed).¹²

The FIGHT-202 trial included the following parameters for the analysis of AEs:

• number of patients reporting TEAEs, SAEs, grade 3 or 4 TEAEs, fatal TEAEs, and temporarily interrupting pemigatinib or permanently discontinuing pemigatinib due to TEAEs

• summary of TEAEs and grade 3 or 4 TEAEs by system organ class, preferred term, decreasing order of frequency, and maximum severity

• summary of TEAEs leading to death, treatment-emergent SAEs, TEAEs leading to dose modifications (reductions, interruptions), and TEAEs leading to discontinuation of pemigatinib by system organ class and preferred term.¹²

The FIGHT-202 trial monitored the following parameters: patients’ physical examination, changes in vital signs, electrocardiogram findings, and changes in clinical laboratory blood and urine sample evaluations.¹²

Statistical Analysis

Sample Size Determination

For the final analysis of the primary end point (ORR in cohort A), a sample size of approximately 100 patients with FGFR2 translocation documented by the central genomics laboratory was planned.¹² A sample size of 100 patients was selected to guarantee an adequate population for robust response data and safety assessments.³⁹ With the assumption that 33% of patients treated with pemigatinib would achieve an objective response, a sample size of 100 patients (assuming 10% of patients would be lost to follow-up) was estimated to provide greater than 95% probability of having a 95% CI with a lower limit of greater than 15%.³⁹ If the lower limit of the 95% CI for ORR exceeded 15%, it was predetermined that the trial results would be considered positive.¹⁶ The minimum clinically meaningful proportion of patients with an objective response was considered to be 15%, based on ORR results reported by previous studies^43-45 of patients with CCA.³⁹ For the analyses of cohort B and cohort C, a sample size of up to 20 patients was planned for each cohort, allowing a greater than 80% chance of observing at least 4 objective responders per cohort, if the underlying ORR was 30%.³⁹

Interim Analysis

An interim analysis for futility for cohort A was planned after approximately 25 patients had enrolled and had at least 1 post-baseline tumour assessment or had permanently discontinued treatment.¹² Enrolment into cohort A could have been terminated for futility if 2 or fewer of the 25 patients in cohort A had achieved a response. Based on a sample size of 60 patients in cohort A, there was a less than 10% probability of the proportion of patients with an objective response being greater than 15% at the final analysis.¹² Initially, the trial was designed to enrol 60 patients into cohort A; however, protocol amendment 5 (October 3, 2017) was later approved to increase the sample size of cohort A to approximately 100 patients.¹² The futility analysis was conducted on October 3, 2017. Since the futility boundary was not crossed, the study proceeded as planned.¹⁷

The timing of the final analysis, at which point the predetermined threshold (i.e., a lower limit of the 95% CI for ORR > 15%) would be assessed, was not pre-specified a priori in the statistical analysis plan. According to the FDA report, the sponsor proposed a data cut-off date of March 22, 2019, during the pre-NDA meeting on August 8, 2019, during which the FDA acknowledged that the suggested data cut-off date would provide a minimum of 7 months of follow-up for all patients in the efficacy set and a minimum of 6 months of follow-up, from time of initial response, for 92% of responders. The sponsor further agreed to provide additional follow-up data for DOR (DOR only; no other efficacy outcome was analyzed at that time). This corresponded to an August 30, 2019, data cut-off date, based on data with at least 6 months of follow-up from the time of initial response for all responders, and a minimum of 12 months of follow-up for all patients in the efficacy set. The data cut-off date of August 30, 2019, also aligned with the 4-month safety update required for the NDA submitted to the FDA.¹⁷ According to the FDA, the proposed data analyses were sufficient to support the filing of an NDA under the provision of accelerated approval.⁵

An updated data analysis occurred at the April 7, 2020, data cut-off date to support safety data summaries for another indication outside of Canada.¹⁷ Since the April 7, 2020, data cut-off date included 1 additional patient in cohort A who had been enrolled after the August 30, 2019, data cut-off date, some efficacy analyses (i.e., survival and response outcomes) were performed, in addition to safety analyses, and were provided to relevant regulatory authorities.¹⁷

For the purpose of regulatory requirements outside of Canada, a data cut-off date of July 8, 2021, was set. Data for the July 8, 2021, data cut-off date has not been submitted to CADTH. Two patients were still receiving treatment at that time and, therefore, a subsequent data cut-off date will be set once the database is closed, approximately in the first quarter of 2022.¹⁷

Primary Outcome

The primary outcome in the FIGHT-202 trial was ORR in patients enrolled in cohort A. A brief overview of statistical methods used for the primary outcome is provided in Table 12.

No statistical comparisons were planned between cohorts and no formal hypothesis testing or inferential analyses were performed.³⁹ The 95% CI for ORR was estimated using the exact method for binomial distribution.¹² Patients with insufficient baseline or on-study response assessment data were considered nonresponders and were included in the denominators in the calculation of ORR.¹² One sensitivity analysis for ORR was planned in the per-protocol (PP) population.¹²

Secondary Outcomes

A brief overview of the statical methods used for the secondary outcomes is provided in Table 12.

Three secondary outcomes involved ORR in the FIGHT-202 trial: ORR in patients in cohort B, ORR in patients in cohort A plus cohort B, and ORR in patients in cohort C. The 3 secondary ORR outcomes were planned to be analyzed in the same manner as the primary ORR outcome, and the 95% CI for ORR was estimated using the exact method for binomial distribution. No sensitivity analyses were planned for the 3 secondary ORR outcomes.¹²

PFS was a secondary outcome in the FIGHT-202 trial and was assessed for cohorts A, B, and C, respectively; no statistical comparisons were planned between cohorts and no formal hypothesis testing or inferential analyses were performed. The number of patients who progressed, died, and were censored were summarized. A Kaplan–Meier (KM) plot of PFS was presented with its 95% CI; 95% CI was estimated using the Brookmeyer and Crowley method.^46,12 Censoring was based on the FDA Guidance for Industry: Clinical Trial End points for the Approval of Cancer Drugs and Biologics (FDA, 2015⁴⁷; FDA, 2018⁴⁸).¹² Reasons for censoring included no baseline tumour assessment, no adequate post-baseline response assessment, no progression, study discontinuation for undocumented progression or for toxicity or other reason, starting of new anti-cancer treatment (date of censoring was the last adequate response assessment before the starting of the new anti-cancer treatment), and death or progression after more than 1 missed assessment. Outcomes of progression included progression documented between scheduled response assessments (date of progression was the date of the first overall response assessment showing PD), death before first assessment of PD (date of progression was date of death), and death between adequate assessment visits (date of progression was date of death). No sensitivity analysis was planned for PFS.¹²

DOR was a secondary outcome in the FIGHT-202 trial and was assessed for cohorts A, B, and C, respectively. The number of patients who responded, who progressed or died, and who were censored were summarized. A KM plot of DOR was presented with its 95% CI; 95% CI was estimated using the Brookmeyer and Crowley method.^46,12 Censoring of DOR was done in the same manner as the censoring of PFS (see details provided earlier for PFS). No sensitivity analysis was planned for DOR.¹²

DCR was a secondary outcome in the FIGHT-202 trial and was assessed for cohorts A, B, and C, respectively. The 95% CI for DCR was estimated using the exact method for binomial distribution.¹² Patients with insufficient baseline or on-study response assessment data were considered nonresponders and were included in the denominators in the calculation of DCR.¹² No sensitivity analysis for DCR was planned.¹²

OS was a secondary outcome in the FIGHT-202 trial and was assessed for cohorts A, B, and C, respectively. The number of patients who died and those who were censored were summarized. The KM plot of OS was presented with its 95% CI; 95% CI was estimated using the Brookmeyer and Crowley method.^46,12 Reasons for censoring included lost to follow-up or still alive at the time of analysis (censoring occurred at the earlier of the date the patient was last known alive and the clinical data cut-off date for the analysis). The “last known alive” date was specified as the date of the last study visit or the date the patient was last known alive from the survival follow-up, whichever was later.¹² No sensitivity analysis for OS was planned.¹²

All analyses performed on the HRQoL outcome (EORTC QLQ-C30) and symptom severity outcome (EORTC QLQ-BIL21) were done in the efficacy evaluable population (defined in the section that follows) for cohorts A, B, and C, respectively. Analyses were considered descriptive (i.e., non-inferential) in nature. No statistical comparisons were planned between cohorts. The standardized scores for each scale of the EORT QLQ-C30 and the EORTC QLQ-BIL21 were calculated as per the scoring guidance of the respective measure. Changes from baseline to each visit were presented for the March 22, 2019¹⁶ and April 7, 2020¹⁷ data cut-off dates. A raw score was considered as missing if the number of missing item values totalled 50% or more of the items that contributed to a scale.¹² No MID was defined.¹²

In addition, the results of a post-hoc analysis were published in abstract presented at the 2021 Gastrointestinal Cancers Symposium.⁴⁹ Additional information about this post-hoc analysis and its results was provided by the sponsor in its clinical summary document.¹² Valle et al. (2021) performed subgroup analyses in cohort A whereby the changes from baseline to week 16 (cycle 6) for the EORTC QLQ-C30 and EORTC QLQ-BIL21 were summarized by the following subgroups: patients with a CR or PR, stable disease, and PD. Valle et al. (2021) noted that treatment-related changes in HRQoL would be expected to be apparent by week 16. Upon request, the sponsor explained that due to a rapid decline in the number of patients available to complete the questionnaires, a robust assessment of the data could only be conducted between baseline and cycle 6.¹⁷ Three key independent physicians consulted by the sponsor on how the drop-off in patients after cycle 6 might impact the interpretation of the data noted that changes in HRQoL would be apparent within 4 to 6 cycles after starting treatment.¹⁷ Post-hoc analyses were performed on the data from the March 22, 2019, data cut-off date and were considered descriptive (i.e., non-inferential) in nature. It was reported by the sponsor that nominal statistical significance was determined by a lack of overlap of the associated standard error bars around the point estimate.¹² Graphs of the observed mean changes from baseline within each subgroup were generated and presented in the Clinical Summary Report.¹²

Subgroup Analyses

Subgroup analyses were planned a priori in the statistical analysis plan for the groups of patients listed subsequently. For each subgroup, a forest plot and the respective outcome’s 95% CI was provided. Subgroup analyses were conducted for the primary outcome, the ORR in cohort A (subgroup analyses were conducted for all patient groups listed subsequently), the PFS in cohort A (subgroup analyses were performed for all groups listed except for renal or hepatic impairment), and the DOR for cohort A (subgroup analyses were done only in patients with renal or hepatic impairment):

• age category (< 65 years versus 65 to < 75 years versus ≥ 75 years)

• sex (female versus male)

• region (North America versus Western Europe versus rest of world)

• baseline ECOG PS (0 versus 1 or 2)

• metastatic disease present (yes versus no)

• lines of prior therapy (1 line versus 2 lines versus ≥ 3 lines)

• received previous platinum treatment (yes versus no)

• renal impairment grade (normal versus mild versus moderate versus severe)

• hepatic impairment grade (normal versus mild versus moderate versus severe).

Only the subgroup, baseline ECOG PS, which was identified in the CADTH review protocol, is reported in the efficacy section that follows.

Amendments¹²

The study protocol of the FIGHT-202 trial was amended 7 times (amendment 1: September 14, 2016, amendment 2: December 5, 2016, amendment 3: January 18, 2017, amendment 4: March 21, 2017, amendment 5: October 3, 2017, amendment 6: February 15, 2018, amendment 7: April 2, 2020).

Amendment 1 included changes to the patient eligibility criteria for completing the EORTC QLQ-BIL21 questionnaire. Only patients in the US, UK, Italy, Germany, and Korea were administered the QLQ-BIL21 instrument, given that this questionnaire was only translated and validated in the primary languages of those countries.

Amendment 2 included clarifications to the inclusion and exclusion criteria, guidance for dose reductions, the addition of updated clinical experience data, and minor administrative changes.

Amendment 3 included clarifications to requirements for HIV screening and enrolment parameters for cohort C (i.e., only patients from the US were allowed to enrol in cohort C).

Amendment 4 included revisions to permit patients to enrol based on local genomics testing results, with the final results to be determined by the central genomics laboratory. Statistical analyses of the primary and secondary outcomes and final cohort assignment were to be done based on the central genomics testing results. This amendment also included changes to the exclusion and screening criteria, the addition of updated clinical experience data, and minor administrative changes.

Amendment 5 implemented changes to increase the number of patients to be enrolled into the FIGHT-202 trial. The overall number of patients, including all 3 cohorts, was increased from 100 to 140; the number of patients to be enrolled into cohort A was increased from 60 to 100. A sample size of 100 patients was estimated to provide a greater than 95% probability (previously 80% probability with a sample size of 60) of having a 95% CI with a lower limit of greater than 15%. This change was done to ensure the most robust efficacy data to inform any future development decision. Following the European Medicines Agency’s Scientific Advice Working Party oral explanation, a sensitivity analysis was done at the time of the primary analysis showing that the proportion of the first 60 patients in cohort A (sample size as per original protocol) who achieved an objective response was consistent with the proportion of patients with an objective response among the 47 patients enrolled in cohort A after amendment 5.^17,39 The European Medicines Agency assessment report noted that the increase in sample size could be considered acceptable.⁹ This amendment also included clarifications to study eligibility (a list of possible FGF and FGFR alterations that can be considered eligible for the study was added) and inclusion criteria, criteria for the futility analysis, and minor administrative changes.

Amendment 6 included clarifications to inclusion criteria, guidelines for dose reductions, ophthalmologic testing, and hyperphosphatemia grading. The comprehensive eye examination was revised to add fundoscopy with digital imaging.

Amendment 7 included guidelines for the management of serous retinal detachment and retinal pigmented epithelium detachment, inclusion of optical coherence tomography at the time of regular scheduled eye examinations and updates to the restricted medications (removal of CYP3A4 inducers and proton pump inhibitors and addition of organic cation transporter 2 substrates) and the prohibited medications (inclusion of moderate CYP3A4 inducers). This amendment also included other minor administrative changes.¹²

Analysis Populations

All efficacy data were analyzed using the efficacy evaluable population, as defined in Table 13. In sensitivity analyses, ORR in cohort A was analyzed using the PP population as defined in Table 13. Analyses of safety were performed using the safety population (Table 13).

Results

Patient Disposition

Details of the patient disposition in cohort A of the FIGHT-202 trial are summarized in Table 14. A total of 171 patients were screened and, of those, 146 patients were enrolled; 107 patients were enrolled into cohort A, 20 patients into cohort B, and 18 patients into cohort C. Patients (N = 25) who were not enrolled failed to meet the trial eligibility criteria. All patients enrolled into cohort A received pemigatinib. As of the primary (March 22, 2019) and updated (April 7, 2020) data cut-off dates, 76 (71.0%) and 98 (90.7%) patients in cohort A had discontinued treatment, respectively. The most common reason for treatment discontinuation was PD.¹⁶

As of the primary (March 22, 2019) and updated (April 7, 2020) data cut-off dates, a total of |||||||||||| and |||||||||||| patients in cohort A, respectively, had terminated the study. The main reason for study termination was death, and the percentage of patients who had died was |||||||||||| at the March 22, 2019, data cut-off date, and |||||||||||| at the April 7, 2020, data cut-off date. At the March 22, 2019, and April 7, 2020, data cut-off dates, respectively, 6 (5.6%) and |||||||||||| patients discontinued the study due to withdrawal of consent and 2 (1.9%) and |||||||||||| patients discontinued the study due to being lost to follow-up. There were 31 |||||||||| and ||||| patients (9.3%) participating in ongoing on-study treatment at the March 22, 2019, and April 7, 2020, data cut-off dates, respectively. A total of |||||||||||| and |||||||||||| patients remained in the study at the primary and updated data cut-off dates, respectively.¹⁶

Protocol Violations

Protocol deviations in cohort A of the FIGHT-202 trial are summarized in Table 15. Missed assessments (most of which were missed laboratory assessments) were the most commonly reported protocol deviations, occurring in 95 patients (88.8%). The next most commonly reported deviations were related to non-compliance with study treatment, none of which were recorded as substantial deviations.¹⁶ Three patients in cohort A had study entry–criteria deviations: for 1 patient, the electrocardiogram displayed incomplete information at study entry (this patient achieved a PR with pemigatinib); 1 patient received treatment with an FGFR inhibitor for 11 days that ended 30 days before the first dose with pemigatinib (this patient did not achieve a response on pemigatinib); and 1 patient had missing serology results at screening (this patient did not achieve a response with pemigatinib).⁵ It was noted in the sponsor’s submission that none of the protocol deviations significantly affected the internal validity of the study data.¹⁶ This statement was echoed in the FDA report, which stated that none of the deviations were likely to impact the efficacy results in favour of pemigatinib.⁵

Exposure to Study Treatments

Exposure to pemigatinib in cohort A at the March 22, 2019, and April 7, 2020, data cut-off dates is summarized in Table 16. The median duration of treatment with pemigatinib was 219 days (range, 7 to 730) and |||||||||||||||||||||||||||||||||||||| at the primary (March 22, 2019) and updated (April 7, 2020) data cut-off dates, respectively. The median overall compliance rates were high at both data cut-off dates (100%; mean = 100.37; SD = 3.296) and |||||||||||||||||||||||||||||||||||||||||||||||||||||| for the primary and updated data cut-off dates, respectively), indicating high treatment adherence in cohort A. Treatment compliance was evaluated by pill counts completed at the site.¹⁶

Dose Modifications (Interruption, Reduction)

At the primary (March 22, 2019) and updated (April 7, 2020) data cut-off dates, respectively, 47 patients (43.9%) and |||||||||||||||||||||| of patients in cohort A (safety population) required a study treatment interruption due to a TEAE. The most common TEAE leading to dose interruption was stomatitis, occurring in 10 patients (9.3%) at the primary and updated data cut-off dates. Other commonly reported TEAEs leading to pemigatinib interruption included palmar-plantar erythrodysesthesia syndrome, arthralgia, and fatigue, occurring in 8 (7.5%), 5 (4.7%), and 4 (3.7%) patients at both data cut-off dates. ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The most commonly reported TEAEs leading to dose reductions were ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| of patients, respectively.¹⁶

Concomitant Medication

Concomitant medications were generally administered similarly across both data cut-off dates and reported for almost all patients in cohort A (safety population) (98.6% and |||||||||||||| of patients at the March 22, 2019, and April 7, 2020, data cut-off dates, respectively). As of the March 22, 2019, data cut-off date, the medications most commonly used (≥ 20%) included paracetamol (37.0%), ondansetron (|||||||), lorazepam (||||||||), sodium chloride (|||||||||), and omeprazole (21.9%). As of the April 7, 2020, data cut-off date, the medication most commonly used (≥ 20%) |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||. Concomitant phosphate binders were used by |||||||||||||| and ||||||||||| of patients at the March 22, 2019, and April 7, 2020, data cut-off dates, respectively; the most commonly used phosphate binder was sevelamer (March 22, 2019, data cut-off date: sevelamer = 8.2%, sevelamer carbonate = 4.1%, and sevelamer hydrochloride = 0.7%; April 7, 2020, data cut-off date: |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||).¹⁶

Subsequent Treatments

Information on any subsequent treatments received is available for only some patients.¹⁷ According to the sponsor, the study sites of the FIGHT-202 trial were not required (but were requested) to provide information on post-treatment therapies once pemigatinib was discontinued.¹⁷ For cohort A, information on subsequent treatments was available for ||||||||||||||, the majority of whom received FOLFIRI (|||||||||||||) as the first subsequent treatment.¹⁷ The next most commonly received subsequent treatments following pemigatinib included immunotherapies (||||||||||||||| received nivolumab and ||||||||||||||| received pembrolizumab) and targeted therapies (i.e., TAS-||||||||||||||||| and -||||||||||||||||||||||||], capecitabine [||||||||||||], and gemcitabine plus cisplatin [||||||||||||]). A range of chemotherapies was received by only | ||||||||||||||||||||||||||||||, including fluorouracil plus gemcitabine, irinotecan plus leucovorin, epirubicin plus cisplatin, and gemcitabine plus oxaliplatin).¹⁷

Efficacy

Only those efficacy outcomes and analyses of subgroups identified in the review protocol are reported subsequently. This CADTH review focuses on cohort A, as cohorts B and C were not part of the requested reimbursement criteria submitted to CADTH and not approved in the Health Canada Notification of Compliance with conditions; therefore, cohorts B and C are beyond the scope of this review. Selected results for cohorts B and C have been included in Appendix 3.

Overall Survival

The OS results of the FIGHT-202 trial for cohort A at the March 22, 2019, and April 7, 2020, data cut-off dates are summarized in Table 17. With an additional 11 months of follow-up, OS at the April 7, 2020, data cut-off date was generally consistent with the OS results seen at the March 22, 2019, data cut-off date. As of the primary analysis (March 22, 2019), with a median follow-up time of 15.44 months, |||||||||||||||||||||| death events occurred in cohort A. There were 67 (62.6%) patients censored. Median OS was 21.06 months (95% CI, 14.82 to not evaluable). The KM curve is depicted in Figure 3 (panel A). The survival probabilities of patients surviving to 6 and 12 months were 88.6% (95% CI, 80.8 to 93.4) and 67.5% (95% CI, 56.4 to 76.3), respectively.¹⁶

As of the updated analysis (April 7, 2020), with a median follow-up time of 27.9 months,12 |||||||||||||||||||| death events occurred in cohort A. Forty-five patients (41.7%) were censored. Median OS was 17.48 months (95% CI, 14.42 to 22.93). The KM curve is depicted in Figure 3 (panel B). The survival probabilities of patients surviving to 6 and 12 months were ||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||) and 67.3% (95% CI, 57.4 to 75.4), respectively.¹⁶

Figure 3: Kaplan–Meier Estimates of Overall Survival — Cohort A, Efficacy Evaluable Population

CI = confidence interval; NE = not evaluable.

Source: Clinical Study Report¹⁶ and sponsor’s response.¹⁷

Progression-Free Survival

The PFS results (based on IRC assessment) of the FIGHT-202 trial for cohort A at the March 22, 2019, and April 7, 2020, data cut-off dates are summarized in Table 17. With an additional 11 months of follow-up, PFS at the April 7, 2020, data cut-off date was generally consistent with the PFS results seen at the March 22, 2019, data cut-off date. As of the analysis of the data from the March 22, 2019, data cut-off date, with a median follow-up time of 15.44 months, 63 patients (58.9%) had experienced disease progression, 8 patients (7.5%) had died, and 36 patients (33.6%) were censored. Median PFS was 6.93 months (95% CI, 6.18 to 9.59). The KM curve is depicted in Figure 4 (panel A). The PFS probabilities at 6 and 12 months were 61.7 months (95% CI, 51.5 to 70.4) and 29.2 months (95% CI, 18.9 to 40.2), respectively. Among the censored patients (33.6%), ||||| patients had an ongoing response or stable disease (3 patients had a CR with a PFS duration of 6.24 to 22.57 months, 12 patients had a PR with a PFS duration of 6.87 to 19.32 months, and 13 patients had stable disease with a PFS duration of 2.73 to 19.32 months). The PFS results based on investigator assessment showed results consistent with those based on IRC assessment; median PFS was 7.03 months (95% CI, 6.87 to 9.63).¹⁶

As of the updated analysis (April 7, 2020), with a median follow-up time of 27.9 months, |||||||||||||||||| patients had experienced disease progression, 9 patients (8.3%) had died, and 27 patients (25.0%) were censored. Median PFS was 7.03 months (95% CI, 6.08 to 10.48). The KM curve is depicted in Figure 4 (panel B). The PFS probabilities at 6 and 12 months were |||||||||||||||||||||||||||||||||||||||| and ||||||||||||||||||||||||||||||||||||||||||, respectively. Among the censored patients, |||||||||||||||||||| patients had ongoing response or stable disease (||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||). Results for PFS based on investigator assessment were not reported for the April 7, 2020, data cut-off date in the sponsor’s submission.¹⁶

Figure 4: Kaplan–Meier Estimates of Progression-Free Survival — Cohort A, Efficacy Evaluable Population

CI = confidence interval.

Source: Clinical Study Reports.¹⁶

The PFS results by subgroup of interest, as specified a priori in the protocol for this CADTH review, are summarized in Table 18. The treatment effect on PFS was consistent with the PFS analysis for all patients in cohort A across patients with ECOG PS of 0 and patients with ECOG PS of 1 or 2. Of note, the sample sizes of these subgroups were small (45 patients with ECOG PS of 0 and 62 patients with ECOG PS of 1 or 2) and relatively wide CIs in the subgroups reflected uncertainty in the effect estimates.

Objective Response Rate

The ORR results (based on IRC assessment) for the FIGHT-202 trial for cohort A at the March 22, 2019, and April 7, 2020, data cut-off dates are summarized in Table 17. The ORR results at the April 7, 2020, data cut-off date were generally consistent with the ORR results seen at the March 22, 2019, data cut-off date.

As of the analysis at the March 22, 2019, data cut-off date, the proportion of patients who achieved an objective response was 35.5% (N = 38) (95% CI, 26.50 to 45.35), including 3 patients (2.8%) with a CR and 35 patients (32.7%) with a PR. It was reported in the sponsor’s submission that since the lower limit of the 95% CI for ORR exceeded 15% (the predetermined threshold), the trial results would be considered positive. The ORR sensitivity analysis using the PP population showed results consistent with the ORR results for the efficacy evaluable population. The ORR results based on investigator assessment showed results that were generally consistent with those based on IRC; the proportion of patients with an objective response was 32.7% (N = 35) (95% CI, 23.95 to 42.45), including 4 patients (3.7%) with CRs and 31 patients (29.0%) with PRs.¹⁶

As of the updated analysis (April 7, 2020), the proportion of patients who achieved an objective response was 37.0% (N = 40) (95% CI, 27.94 to 46.86), including 4 patients (3.7%) with a CR and 36 patients (33.3%) with PRs. ||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||Results for ORR based on investigator assessment for the PP population were not reported for the April 7, 2020, data cut-off date in the sponsor’s submission.¹⁶

The ORR results by subgroup of interest, as specified a priori in the protocol for this CADTH review, are summarized in Table 19. The ORR results for the subgroup of interest suggest that the treatment effects on ORR for the subgroups of patients with an ECOG PS of 0 and 1 plus 2 were generally consistent with the overall population in cohort A. Of note, the sample sizes of these subgroups were small (45 patients with an ECOG PS of 0 and 62 patients with an ECOG PS of 1 or 2) and the relatively wide CIs in the subgroups reflected uncertainty in the effect estimates.

Duration of Response

The DOR results (based on IRC assessment) for the FIGHT-202 trial for cohort A at the March 22, 2019, and April 7, 2020, data cut-off dates are summarized in Table 17. In addition, results were presented for DOR at the August 12, 2019, data cut-off date. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| The results for each data cut-off date are described in text that follows.

At the March 22, 2019, data cut-off date, among the 38 patients who achieved an objective response, median DOR was 7.49 months (95% CI, 5.65 to 14.49). The KM curve is depicted in Figure 5 (panel A). Of the 38 patients with an objective response, 35 patients (92%) had at least 6 months of follow-up from the time of initial response; the other 3 patients with an objective response had 5.2, 5.7, and 5.85 months of follow-up from the time of initial response. Median time-to-first response was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||.³⁹ There were 17 patients (44.7%) censored, of which 12 patients had an ongoing PR with a DOR ranging from 4.17 to 14.55 months, and 3 patients had an ongoing CR with a DOR of 4.83, 6.34, and 19.52 months, respectively. The probabilities of maintaining a response for at least 6 and 12 months were |||||||||||||||||||||||||||||||||||||||| and 37.4 (95% CI, 18.6 to 56.2), respectively. The DOR results that were based on investigator assessment were generally consistent with those based on IRC results; median DOR was ||||||||||||||||||||||||||||||||||||||||||||||||).¹⁶

As of the updated analysis at the August 12, 2019, data cut-off date, median DOR was ||||||||||||||||||||||||||||||||||||||||| months among the 38 respondents.¹⁶ All 38 patients with an objective response had at least ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. There were ||||||||||||||| patients censored. Among the 38 patients with an objective response, |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||, respectively.

As of the updated analysis (April 7, 2020), among the 40 patients who achieved an objective response, median DOR was 8.08 months (95% CI, 5.65 to 13.14). The KM curve is depicted in Figure 5 (panel B). Among the 40 patients with an objective response, 23 patients (57.5%) had a DOR of at least 6 months, 15 patients (37.5%) had a DOR of at least 9 months, and 10 patients (25.0%) had a DOR of at least 12 months. There were ||||||||||||||| patients censored. The probabilities of maintaining a response for at least 6 and 12 months were ||||||||||||||||||||||||||||||||| and |||||||||||||||||||||||||||||||||||||||||||||, respectively. Results for DOR, based on investigator assessment, were not reported for the April 7, 2020, data cut-off date in the sponsor’s submission.¹⁶

Figure 5: Kaplan–Meier Estimate of Duration of Response Based on IRC Assessment, Cohort A, Efficacy Evaluable Population

CI = confidence interval; IRC = independent review committee RECIST 1.1 = Response Evaluation Criteria in Solid Tumours Version 1.1.

Source: Clinical Study Reports.¹⁶

Disease Control Rate

The DCR results (based on IRC assessment) for the FIGHT-202 trial for cohort A at the March 22, 2019, data cut-off date are summarized in Table 17. As of that date, the proportion of patients with a best response of CR, PR, or stable disease was 82.2% (n = 88) (95% CI, 73.7 to 89.0), including 3 patients (2.8%) with a CR, 35 patients (32.7%) with a PR, and 50 patients (46.7%) with stable disease for 39 or more days since the first pemigatinib dose. The DCR results based on investigator assessment (DCR = 86.9%; 95% CI, 79.0 to 92.7) were generally consistent with those based on IRC assessment.

The DCR outcome was not generated for the April 7, 2020, data cut-off date since the primary focus of the analyses at that date was on an integrated safety summary for a new regulatory submission.¹⁷

Patient-Reported Outcomes

Health-Related Quality of Life

EORTC Quality of Life Questionnaire Core 30

Completion rates for the EORT QLQ-C30 instrument declined over time. After week 16 (cycle 6) there were approximately ||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁷ The descriptive summary statistics of observed mean scores and mean changes from baseline at each assessment point for the EORT QLQ-C30 questionnaire (global health status [QoL] scale) at the April 2, 2020, data cut-off date are summarized in Figure 6 and Figure 7, respectively. ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Overall observed scores from baseline to cycle 33 (March 22, 2019, data cut-off date) or to cycle 42 (April 7, 2020, data cut-off date) |||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

Figure 6: Redacted

Note: This figure has been redacted at the sponsor’s request.

||||||||||||||||||||||||||||||||||||.¹⁷

Figure 7: Redacted

Note: This figure has been redacted at the sponsor’s request.

|||||||||||||||||||||||||||||||||||||||||||||.¹⁷

Post-Hoc Analysis on HRQoL

Descriptive statistics of observed mean changes from baseline to week 16 (cycle 6) by subgroups of patients (i.e., patients with a CR or PR, stable disease, or PD) are summarized in Table 20 (baseline to cycle 6) and Figure 8 (baseline to cycle 39). The analysis population included 100 evaluable patients of the total 107 patients in cohort A. A definition of the evaluable population was not provided.⁴⁹ The 3 subgroups included 36, 48, and 15 patients with a CR or PR, stable disease, and PD, respectively. The results suggested that the change in overall mean score for the overall health status scale appeared to be maintained in patients with a CR or PR (−0.3; SD = 1.3) and stable disease (−0.3; SD = 1.1), and declined in patients with PD (−1.2; SD = 0.8) (Table 20 and Figure 8).⁴⁹ While similar results were observed for the emotional function scale, all patients appeared to show decline for the role and social functioning scales.⁴⁹ The sponsor reported that the difference in mean change from baseline between patients with PD and those with either a CR or PR, or stable disease were driven by a reported increase in feelings of worry and tension in patients with PD at cycle 6.¹² The sponsor additionally reported that differences in mean change from baseline were seen for the constipation scale between patients with PD and those with either a CR or PR, or stable disease; however, results were not reported and no definition of what constitutes a meaningful change was provided.¹² Overall, Valle et al. (2021) concluded that changes in HRQoL appeared directionally more favourable in patients with a CR or PR, or stable disease than in patients with PD.⁴⁹

Figure 8: Redacted

Note: This figure has been redacted at the sponsor’s request.

||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹²

Symptom Severity

EORTC Quality of Life Questionnaire Cholangiocarcinomas and Gallbladder Cancer Module 21

Completion rates for the EORT QLQ-BIL21 instrument declined over time. After week 16 (cycle 6) there were approximately ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||.¹⁷ The descriptive summary statistics of observed mean scores and mean changes from baseline at each assessment point for the EORT QLQ-BIL21 questionnaire (eating and pain scales; other scales are found in Appendix 3, Figures 19 to 30) at the April 2, 2020, data cut-off date are summarized in Figure 9 (eating scale), Figure 10 (eating scale), Figure 11 (pain scale), and Figure 12 (pain scale), respectively. A definition for what constituted a clinically meaningful change from baseline in the present target population was not provided. Overall observed scores from baseline to cycle 33 (March 22, 2019, data cut-off date) or to cycle 42 (April 7, 2020, data cut-off date) | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

Figure 9: Redacted

Note: This figure has been redacted at the sponsor’s request.

|||||||||||||||||||||||||||||||||||||||||||.¹⁷

Figure 10: Redacted

Note: This figure has been redacted at the sponsor’s request.

||||||||||||||||||||||||||||||||||||||||||||||||.¹⁷

Figure 11: Redacted

Note: This figure has been redacted at the sponsor’s request.

||||||||||||||||||||||||||||||||||||||||||||||||.¹⁷

Figure 12: Redacted

Note: This figure has been redacted at the sponsor’s request.

|||||||||||||||||||||||||||||||||||||||||||.¹⁷

Post-Hoc Analysis on HRQoL

Descriptive statistics of observed mean changes from baseline to week 16 (cycle 6) by subgroups of patients (i.e., patients with a CR or PR, stable disease, or PD) are summarized in Table 20 (baseline to cycle 6) and Figure 13 (baseline to cycle 39). The analysis population included 100 evaluable patients out of the 107 patients in cohort A; a definition of the evaluable population was not provided.⁴⁹ The 3 subgroups included 36, 48, and 15 patients for patients with a CR or PR, stable disease, and PD, respectively.⁴⁹ The results suggested that the overall mean score change for the pain scales appeared to decline in patients with a CR or PR (−5.7; SD = 20.0), and stable disease (−3.5; SD = 11.4) and increase in patients with PD (8.3; SD = 5.9) (Table 20 and Figure 13).⁴⁹ The sponsor reported that the difference in mean change from baseline between patients with PD and those with either a CR or PR, or stable disease were driven by a reported increase in pain at night in patients with PD at cycle 6.¹² Similar results were observed for the anxiety scale (Table 20), and the sponsor reported that the difference in mean change from baseline between patients with PD and those with either a CR or PR, or stable disease were driven by a reported increase in worry and decreased ability to enjoy oneself in patients with PD at cycle 6.^12,49 All patients appeared to experience an increase in treatment side effects (Table 20).⁴⁹ Overall, Valle et al. (2021) concluded that changes in HRQoL appeared directionally more favourable in patients with a CR or PD, or stable disease than in patients with PD.⁴⁹

Figure 13: Redacted

Note: This figure has been redacted at the sponsor’s request.

||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹²

Harms

Only those harms identified in the review protocol are reported subsequently. Refer to Table 21 for detailed harms data in cohort A at the March 22, 2019, and April 7, 2020, data cut-off dates. Results are similar at both data cut-off dates with no new safety concerns identified as of the latest data cut-off date; therefore, the safety results of the April 7, 2020, analyses are described in the text that follows.

Adverse Events

All patients in cohort A experienced at least 1 TEAE (100.0%). The most commonly reported TEAEs were alopecia (|||||||||), hyperphosphatemia (|||||||||), diarrhea (|||||||||), dysgeusia (||||||||||), fatigue (||||||||||), and nausea (||||||||||).¹⁶

Grade 3 or higher TEAEs occurred in ||||||||| of patients in cohort A (Table 21). The most commonly reported grade 3 or higher TEAE was hypophosphatemia. The percentage of patients experiencing |||||||||||||||||||||||||||||||||||| was ||||||||||. Other grade 3 or higher TEAEs included ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||.¹⁶

Serious Adverse Events

The percentage of patients experiencing serious TEAEs was ||||||||||| in cohort A. The most common serious TEAEs were pyrexia and cholangitis, each occurring in |||||||| of patients; abdominal pain, occurring in |||||||| of patients; and infective cholangitis, occurring in |||||||| of patients.¹⁶

Withdrawals Due to Adverse Events

|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. AEs led to discontinuation of study treatment in ||||||||||||||||||||||||||||||| in cohort A. ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

Mortality

TEAEs leading to death occurred relatively rarely in |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

Notable Harms

Notable harms as specified in the protocol included nail toxicity, serous retinal detachment, hyperphosphatemia, and hypophosphatemia.¹⁶

Nail Toxicity

The percentage of patients experiencing nail toxicity TEAEs was ||||||||||| in patients in cohort A. The most commonly reported ||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The percentage of patients experiencing nail toxicity TEAEs of grade 3 or higher was |||||||, including nail discoloration, nail disorder, onychoclasis, and paronychia. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

Serous Retinal Detachment

The percentage of patients experiencing serous retinal detachment TEAEs in cohort A was ||||||||. The most commonly reported serous retinal detachment was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

Hyperphosphatemia

The percentage of patients experiencing hyperphosphatemia TEAEs in cohort A was |||||||||. The most commonly reported hyperphosphatemia events were ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.¹⁶

Hypophosphatemia

The percentage of patients experiencing hypophosphatemia TEAEs in cohort A was |||||||||. The most commonly reported hypophosphatemia events were ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Hypophosphatemia TEAEs of grade 3 or higher occurred in |||||||| of patients, including only hypophosphatemia events. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||.¹⁶

Critical Appraisal

Internal Validity

• Primary objective of phase II design: The primary objective of phase II (randomized or non-randomized) trials is to document the safety outcomes and investigate whether the estimate of the effect for a new drug is large enough to use it in confirmatory phase III trials. Phase II trials may not accurately predict harm and/or effectiveness of treatments. There are numerous examples of phase III trials whose results did not support the phase II trial results.⁵⁰ It is uncertain whether the results observed in this phase II trial would translate into positive phase III trials or into real-world clinical practice. There are currently no randomized phase III trials under way for this review’s target population. The clinical experts consulted by CADTH noted that, despite the high unmet need, conducting an RCT in this setting with a targeted therapy, such as pemigatinib, compared with the current therapies available in the second line in Canadian clinical practice, would likely not be feasible. According to the clinical experts, developing phase III RCTs is hindered by the overall low number of patients who meet the current indication and equipoise between pemigatinib and other chemotherapy drugs does not exist.

• Limited interpretation of time-to-event end points: Interpretation of time‐to‐event end points such as OS or PFS is limited in single‐arm studies. While ORR may be directly attributable to the drug effect, the non-randomized design makes interpreting PFS and OS events attributable to pemigatinib challenging, since all patients in cohort A received the same treatment. The extent to which the observed survival is due to the natural history of the tumour or the intervention remains unclear.⁴⁸ The FDA’s multi-discipline review for the assessment of pemigatinib in the present indication included the following reviewer comment: “In a single arm trial, FDA considers time-to-event end points to be uninterpretable, and the results will not be described in this review.”⁵ Consequently, the efficacy outcomes contributing to the FDA’s accelerated approval of pemigatinib in the current setting included response outcomes and no survival end points.⁵

• Prognostic value of FGFR2: While there is strong genetic and functional evidence that FGFR genetic alterations can drive the formation of tumours,⁷ it is currently not known whether patients with an FGFR2 alteration represent a distinct prognostic subgroup.¹¹ Retrospective studies^19,21-23 in patients with CCA suggest that patients with FGFR alterations appear to have better prognoses compared with an unselected CCA population.^9,19 Given the single‐arm design of the FIGHT-202 trial, it could be challenging to disentangle to what extent the observed results are due to a potential prognostic effect of FGFR2 fusions or rearrangements, or due to drug-associated effects. The clinical experts consulted on this issue by CADTH agreed that, although there is exploratory, preliminary data that FGFR2 genetic alterations may be associated with a more indolent disease progression,¹⁹ there is currently insufficient evidence to determine whether patients with FGFR2 fusion–positive CCA represent a distinct prognostic subgroup. The clinical experts were of the opinion that FGFR2 mutation status was likely not an important prognostic factor in the indicated patient population. The clinical experts considered that the median time from initial diagnosis to enrolment into cohort A of the FIGHT-202 trial was approximately 1 year, which is reflective of median OS with standard first-line therapy in this setting. The duration of first-line therapy for patients in cohort A of the FIGHT-202 trial was also reflective of clinical practice, according to the clinical experts (see Table 34 in Appendix 3). The clinical experts further agreed that progression on prior systematic therapy is a key prognostic factor in these patients, and they did not anticipate that patients would derive any substantial benefit from their underlying disease biology at the time they enrolled into the FIGHT-202 trial.

• Open-label design: The FIGHT-202 trial had an open-label design whereby the investigator and the study participants are aware of their treatment status, which increases the risk of detection bias and performance bias. This has the potential to bias results and outcomes in favour of pemigatinib if the assessor (investigator or patient) believes the study drug is likely to provide a benefit. However, to mitigate the impact of this bias, the investigators used an IRC to evaluate responses using standardized criteria (i.e., based on RECIST 1.1 and assessed by an independent radiological review committee using central genomics laboratory results,¹² with confirmation of a CR and PR at least 4 weeks after the initial assessment).¹² Results for the investigators’ and IRC’s response assessments were generally consistent, suggesting that the potential of confounding effects on response outcomes due to the open-label design is likely not substantial. Furthermore, subjective outcomes (i.e., adverse outcomes and patient-reported outcomes) may be biased due to the open-label design. For example, if study personnel and patients knew that the treatment was pemigatinib (which is known to cause nail toxicity, hyperphosphatemia, and other AEs), this could have influenced the reporting of harms. Overall, the magnitude and direction of this bias remains unclear.

• Statistical analyses: No formal statistical significance and hypotheses testing were performed and, thus, no P values were reported. Point estimates with 95% CIs were reported to estimate the magnitude of treatment effect. A greater than 95% probability of having a 95% CI for ORR in cohort A with a lower limit larger than 15% was the basis for the sample size determination and was regarded as the threshold for a positive study outcome. Results for ORR appeared consistent with the sample size assumptions, and the study recruited the intended number of patients.

• Subgroup analysis: Methodological issues limited the ability to interpret the results from the subgroup analyses. The subgroup analyses were non-inferential, wide CIs reflected uncertainty in the effect estimates, and small sample sizes limited the generalizability to a broader population.

• Small sample size: A limited number of patients were included in the efficacy evaluable dataset (n = 107) of cohort A. The magnitude of the treatment effect estimates observed in a small study sample may not be replicable in a larger study sample or generalizable to the target population in real-world clinical practice.

• HRQoL and symptom severity assessments: The interpretation of results for the EORTC QLQ-C30 and EORTC QLQ-BIL21 questionnaire (i.e., the ability to assess trends over time) at later cycles is limited by the ||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. In addition, selection bias over time should be considered when interpreting results, as the long-term survivors tend to be the healthier patients. Furthermore, given the lack of pre-specified statistical analyses a priori in the statistical analysis plan for patient-reported outcomes, results from post-hoc analyses (i.e., changes from baseline to week 16 summarized by disease response subgroups) are considered exploratory in nature. As well, patient-reported outcomes in the post-hoc analysis were measured up to week 16, which may not represent an accurate picture of patients’ experiences with pemigatinib for a prolonged period of time. Given that the trial was non-randomized, the impact of pemigatinib on patient-reported outcomes in relation to other therapies is unknown.

The reliability of the EORTC QLQ-C30 instrument was evaluated in an international study in patients with BTC⁴¹ and showed that internal consistency was acceptable for most scales and that results for test–retest reliability were mixed.⁴¹ Estimates for MIDs in the literature were not found for the EORTC QLQ-BIL30 in patients with CCA or BTC.

||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Therefore, it is unclear if the changes from baseline experienced by patients in the FIGHT-202 trial are reflective of a clinically meaningful change in patients with unresectable, locally advanced, or metastatic CCA with FGFR2 alterations.

The EORTC QLQ-BIL21 instrument was validated in an international study in patients with BTC that showed acceptable internal consistency for all multi-item scales and good test–retest reliability. Known group validity also showed an ability to distinguish between subgroups and there was some evidence of responsiveness. Estimates for MIDs in the literature were not found for the EORTC QLQ-BIL30 in patients with CCA or BTC| |||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Therefore, it is unclear if the changes from baseline experienced by patients in the FIGHT-202 trial are reflective of a clinically meaningful change in patients with unresectable, locally advanced, or metastatic CCA with FGFR2 alterations.

Overall, these methodological issues render results from the EORT QLQ-C30 and EORTC QLQ-BIL21 instruments inconclusive.

External Validity

• Overall, the clinical experts consulted by CADTH agreed that the baseline patient characteristics of cohort A of the FIGHT-202 trial were reflective of patients they see in Canadian clinical practice for the present indication. Although the majority of patients in cohort A were enrolled in trial sites in the US and Europe, according to the clinical experts consulted by CADTH, the population enrolled in the trial was consistent with the population expected to be treated in Canadian clinical practice; furthermore, no different treatment effect would be expected based on different disease management practices across countries. According to the clinical expects’ opinion, as long as patients have the FGFR2 alteration, pemigatinib would be appropriate to administer after any of the prior therapies received by patients in the trial. However, the clinical experts agreed that patients should not have previously been treated with an FGFR2-targeted therapy. The majority of patients (61%) in cohort A of the FIGHT-202 trial had received 1 prior line of systemic therapy before trial enrolment, and 27% and 12% of patients had received 2, or 3 lines or more of prior systemic therapy, respectively. The clinical experts anticipated seeing the benefit of treatment with pemigatinib regardless of the number of previous lines of systemic therapy, as long as patients have the FGFR2 alteration. Furthermore, FGFR2 alterations occur rarely in eCCA and there was 1 patient in cohort A of the FIGHT-202 trial with FGFR2-positive eCCA, whereas all other patients in cohort A had iCCA. The clinical experts noted that patients with iCCA and eCCA are managed in a similar way in clinical practice and that the results observed in cohort A are generalizable to patients with FGFR2-positive eCCA based on the fact that FGFR2 is the target of the mechanism of action of pemigatinib and there is no biologic rationale to assume that pemigatinib’s safety profile would be different in patients with eCCA.

• Concomitant medications received by patients in the trial appeared reflective of the medications patients would receive in Canadian clinical practice, according to the clinical experts consulted by CADTH.

• Non-comparative design: The non-comparative design of the FIGHT-202 trial precludes the ability to assess the relative therapeutic benefit or safety of pemigatinib against currently available therapies in Canadian clinical practice. As noted previously, the clinical experts consulted by CADTH agreed that direct randomized comparisons between pemigatinib and currently used therapies are unlikely to take place in the setting of previously treated, unresectable, locally advanced, or metastatic CCA with an FGFR2 fusion or other rearrangement. In the absence of a direct comparison of pemigatinib with relevant treatment options, the sponsor submitted an ITC⁵¹ in the form of an unanchored MAIC comparing the efficacy of pemigatinib (cohort A of the FIGHT-202 trial) with each of the 2 treatment groups in the ABC-06 study. Please refer to the following section on indirect evidence for a detailed summary and critical appraisal of the sponsor’s submitted ITC. The results of the ITC favoured pemigatinib for PFS and OS in comparison with mFOLFOX plus ASC as well as with ASC alone. The clinical experts agreed that in the absence of robust comparative data on PFS and OS, no firm conclusions could be drawn on how pemigatinib compared with other relevant treatment options. The clinical experts consulted by CADTH anticipated, however, that based on the FIGHT-202 results and on poor results with existing treatment options in clinical practice, pemigatinib appeared to offer at least similar or improved clinical benefits compared with current therapies, with better tolerability.

• Relevance of trial efficacy outcomes: The primary outcome in the FIGHT-202 trial was ORR in cohort A and secondary outcomes included DOR, DCR, PFS, and OS. According to the clinical experts consulted by CADTH, ORR, DOR, and DCR are clinically meaningful end points for patients with unresectable, locally advanced, or metastatic CCA that has progressed on prior therapy. Responses in this patient population are important because of accompanying delay in the worsening of symptoms and a slower decline in ECOG PS. According to the clinical experts, the majority of patients will have stable disease, followed by a PR and, rarely, a CR in response to first-line treatment; therefore, the clinical experts were not concerned about the low number of patients who achieved CR in the FIGHT-202 trial, but emphasized the clinical relevance and importance of maintaining stable disease in preventing an otherwise fast decline in patients in this setting. While the clinical experts agreed that, based on the available evidence, it was not possible to conclude whether the antitumour activity expressed as responses would translate into clinical benefits in terms of PFS and OS, they felt that durable responses could potentially delay tumour progression and result in prolonged survival benefit in this patient population.

• Excluded patient subgroups: The FIGHT-202 trial excluded patients who would be intolerant to standard first-line therapy without experiencing PD. The clinical experts consulted by CADTH felt it would be reasonable to generalize the results from cohort A to patients with FGFR2 alterations who are intolerant to first-line therapy, given the favourable safety profile of oral pemigatinib. The clinical experts noted that as patients with an ECOG PS greater than 2 were excluded from the FIGHT-202 trial, there are no data to support the generalizability of treatment benefit in this patient population. The clinical experts were of the opinion that it would be reasonable to leave it up to the discretion of the treating physician to apply some flexibility in terms of using pemigatinib with slightly lower laboratory parameters than those outlined in the trial. It was agreed by the clinical experts that since patients with brain or central nervous system metastases that are untreated or have progressed were excluded from the trial, there are no data to support the generalizability of treatment benefit in this patient population.

Indirect Evidence

Objectives and Methods for the Summary of Indirect Evidence

The FIGHT-202 study was a non-comparative study and the systematic review therefore does not provide any direct evidence for the relative efficacy of pemigatinib versus a relevant comparator. A focused literature search for network meta-analyses dealing with pemigatinib or CCA was run in MEDLINE All (1946–) through Ovid and Embase (1974–) through Ovid on July 16, 2021. No limits were applied, and conference abstracts were excluded from the search results. No relevant studies were identified comparing pemigatinib versus mFOLFOX, FOLFIRI, fluorouracil alone or in combination with cisplatin or oxaliplatin, capecitabine alone or in combination with cisplatin or oxaliplatin, or best supportive care in adults with previously treated, unresectable, locally advanced, or metastatic CCA with FGFR2 fusions or other rearrangements.

The sponsor submitted an ITC⁵¹ in the form of a MAIC between cohort A of the FIGHT-202 study and each of the 2 treatment groups in the ABC-06 study. The ABC-06 study compared an mFOLFOX regimen plus ASC versus ASC alone in patients with BTC. The results of the MAIC were used to inform the sponsor’s pharmacoeconomic submission.

Methods of Sponsor-Submitted MAIC

Objectives

Given the lack of an RCT comparing pemigatinib with a standard-of-care regimen, an ITC was conducted to provide evidence for the relative efficacy of pemigatinib versus relevant comparators.

Study Selection Methods

There were 2 literature searches, 1 performed on November 9, 2018 (original search) and a second performed on April 21, 2020, (updated search), to identify available clinical efficacy, safety, and tolerability evidence related to second-line treatment of advanced and/or metastatic CCA with FGFR2 fusions or rearrangements. Both searches used the same English-language search strategy in multiple databases (MEDLINE In-Process, Embase, MEDLINE, and the Cochrane Library) and the date ranges were from database inception to November 9, 2018 for the original search, and from October 1, 2018 to April 21, 2020, for the updated search. The following sources were also searched: conference proceedings from 6 oncology conferences starting from the year 2016; reference lists of relevant studies, systematic reviews within the previous 2 years, and meta-analyses; and reference lists from relevant articles from 7 health technology assessment agencies. For the original search, 2 independent reviewers screened abstracts, with a third reviewer assessing abstracts where there was disagreement or uncertainty. In those cases, the consensus of the majority was used to make a final decision. Full-text screening of relevant abstracts was performed in the same manner as for abstract screening. For the updated search, 2 independent reviewers screened abstracts and full-text articles, with a third reviewer independently resolving uncertainties regarding study inclusion at each screening stage.

Studies were eligible if they included adults with advanced and/or metastatic or surgically unresectable CCA with FGFR2 fusions or rearrangements that had failed to respond to at least 1 treatment. Studies were excluded if they reported on pediatric patients, patients that did not have metastatic or advanced cancer or FGFR2 fusions or rearrangements, patients who were treatment-naive, or patients with resectable CCA. Articles with unclear disease stage, FGFR2 status, or treatment line were included during abstract screening. Studies either had to be single-arm with a pharmacological intervention or had to compare a pharmacological intervention with placebo, best supportive care (as defined by the study author), or any other pharmacological intervention. Eligible study designs were RCTs, single-arm studies, observational studies, and systematic reviews, with the latter used only for bibliography searches. Preclinical studies, case reports, case series, pharmacokinetic studies, and economic studies were excluded. A list of relevant outcomes was provided, though it was unclear whether articles were screened based on available outcomes. These outcomes were: response rate; OS; PFS; time to treatment discontinuation; DOR; mortality; HRQoL; incidence of AEs; study or treatment discontinuation; and relationship between intermediate outcome (PFS and response rate) and OS, DCR, stable disease, time on treatment, time to response, overall response rate, and patient-reported outcomes.

In the original search, a total of 35 relevant results were identified reporting on 8 non-comparative studies, 1 retrospective observational study, and 11 ongoing studies with no results available. In addition, 111 articles were flagged: 32 were flagged as having a patient population with BTC and 79 were flagged as having a patient population with unclear FGFR2 mutation status. From the updated search, 829 new articles were included at the full-text screening stage; in addition, the 111 flagged articles from the original search were also included in the full-text screening. One of the most common reasons given for excluding full-text publications was “biliary tract cancer” (n = 117). In response⁵² to a request for clarification by the CADTH review team, the sponsor indicated that the following types of studies were excluded: studies that did not report a subgroup of patients with CCA, studies in which patients with CCA made up less than 80% of the study patients, and studies in which the percentages of patients with each type of BTC were not reported.

After full-text screening in the updated search, a total of 209 relevant publications were identified (of which 23 were identified from conference proceedings and 35 were the relevant results from the original search), reporting on 108 studies. Study quality for the 108 studies was assessed using the Downs and Black checklist, though it was unclear how or if these assessments were used.

For potential inclusion in the MAIC, additional criteria were applied to the 108 studies identified in the systematic literature searches. It was not clear at what point these criteria were established, and rationales were not provided for all of the criteria. KM plots of both OS and PFS were required so that pseudo patient-level data for these outcomes could be derived. The study had to include a treatment (if not pemigatinib) that was representative of standard of care, which seemed to be defined as chemotherapy. A minimum sample size of 20 patients was established, though a justification for this cut-off was not provided. After applying these additional criteria and eliminating a study in patients receiving a fourth-line or later-line therapy, 8 studies remained. Although the proportions of patients with an ECOG PS of 0 or 1 and patients with iCCA had to be high to match the patient population in the FIGHT-202 study, these criteria did not appear to be applied at this stage of the study selection process.

Of the 8 remaining studies, 2 were single-arm trials (including the FIGHT-202 study), 2 were RCTs, and 4 were retrospective studies. In the studies other than the FIGHT-202 study, publication dates ranged from 2012 to 2020, sample sizes for each treatment group ranged from 30 to 255, treatments included chemotherapy and ASC, median age ranged from 54 years to 65 years where reported, the percentage of male patients ranged from 43% to 66.7%, the percentage of patients with iCCA ranged from 16.7% to 94.6%, and the percentage of patients with an ECOG PS of 0 or 1 ranged from 64% to 100% where reported. FGFR2 mutation status was not reported for any of the studies, aside from the FIGHT-202 study.

Despite the sponsor’s explanation that studies were excluded during full-text screening if they did not report on the percentage of patients with CCA or if the percentage of patients with CCA was less than 80%, CADTH reviewers noted examples within the 8 considered studies that appeared to contradict this. The types of biliary cancers included in 1 study were not reported in the cited conference abstract⁵³ and the percentage of patients with CCA was less than 80% in 3 studies.^13,54,55

The ITC authors focused on 2 studies, Kim et al.⁵⁵ and Lemarca et al.,¹³ based on sample size and recent date of publication (thought to more accurately reflect the current standard of care). The ITC was originally developed for the pemigatinib submission to the National Institute for Health and Care Excellence (NICE) and the study by Lemarca et al.,¹³ also known as the ABC-06 study, was chosen because it recruited patients in the UK.

MAIC Analysis Methods

The MAIC approach was selected due to the non-comparative nature of the FIGHT-202 study. The choice of the ABC-06 study for comparison with the FIGHT-202 study did not appear to be based on its specific interventions; however, the mFOLFOX regimen was considered by the clinical experts consulted by the ITC authors to be a relevant comparator in the second-line setting. Both ASC alone and mFOLFOX plus ASC from the ABC-06 study were compared with pemigatinib from the FIGHT-202 study.

According to the ITC authors, the following baseline characteristics were available for cohort A of the FIGHT-202 study and the patients in the ABC-06 study: median age, percentage of male patients, percentage of patients with iCCA, percentage of patients with an ECOG PS of 0 or 1, and percentage of patients with a serum albumin concentration of less than 35 g/L. Of these, the following covariates were chosen for adjustment: |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The selection of covariates for adjustment did not appear to be preplanned and no rationale was provided for the covariates selected. Additionally, the following baseline characteristics were available for both studies and did not appear to be considered: disease stage, percentage of patients with prior surgery for cancer, and number of lines of prior systemic therapy for advanced or metastatic cancer. In the FIGHT-202 study, 9 patients from cohort A were excluded from the MAIC due to missing serum albumin values.

For each arm in the ASC-06 study, a logistic regression model was estimated using the method of moments based on individual patient data from the FIGHT-202 study and summary data from the relevant arm in the ABC-06 study. The model was used to approximate propensity scores for patients in the FIGHT-202 study, which reflected the odds for each patient being included in the ABC-06 study versus the FIGHT-202 study based on the distributions of the covariates included in the model. The baseline characteristics used to reweight the pemigatinib group, along with effective sample size, were presented for the pemigatinib group before weighting and after weighting to the mFOLFOX plus ASC and ASC only groups. The outcomes of PFS and OS were the only disease-related outcomes available for both studies and were generated on a pseudo patient level for the ABC-06 study by estimating times and survival probabilities from the published KM PFS and OS curves using a software tool. HRs for PFS and OS were determined using Cox proportional hazard models with bootstrapping used to estimate standard errors and CIs.

Results of Sponsor-Submitted MAIC

Summary of Included Studies

The ABC-06 study, the study selected for the MAIC with cohort A of the FIGHT-202 study (the pemigatinib group), was a phase III, open-label, multi-centre RCT conducted in the UK with patients enrolled from 2014 to 2018. Adult patients with locally advanced or metastatic BTC with disease progression in the previous 6 weeks on first-line cisplatin and gemcitabine chemotherapy were randomized (1:1) to receive ASC alone (ASC group) or mFOLFOX plus ASC (mFOLFOX plus ASC group). Randomization was stratified by platinum sensitivity (sensitive or refractory/resistant), serum albumin concentration (< 35 g/L or ≥ 35 g/L), and disease stage (locally advanced or metastatic).

Patients in the ABC-06 study had to have an ECOG PS of 0 to 1; a life expectancy of longer than 3 months; adequate hematological, renal, and hepatic function; and no evidence of ongoing infection, inadequate biliary drainage, metastatic disease to the brain, or clinically significant cardiovascular disease.

All patients received ASC, which consisted of early identification and treatment of biliary-related complications and cancer-related symptom management. Interventions included biliary drainage, antibiotics, analgesia, steroids, antiemetics, other palliative treatment for symptom control, palliative radiotherapy (e.g., for painful bone metastases), and transfusion of blood products. Patients had study visits every 4 weeks for ASC, which included physical examination, assessment of ECOG PS, symptom monitoring, review of concomitant medications, and assessment of liver and renal function with full blood count. Patients in the mFOLFOX plus ASC group also received chemotherapy every 2 weeks for a maximum of 12 cycles. At each cycle, patients received oxaliplatin 85 mg/m² and L-folinic acid 175 mg (or folinic acid 350 mg) through IV infusion over 2 hours, and fluorouracil 400 mg/m² through a 5- to 10-minute bolus on day 1. Fluorouracil 2,400 mg/m² was started as a continuous IV infusion on day 1 and was finished on day 2. A maximum of 2 dose reductions for each drug was allowed, representing a 20% and 50% reduction from the initial dose. If treatment was delayed for more than 28 days due to toxicity, the patient permanently discontinued treatment. If oxaliplatin was discontinued due to toxicity, treatment with the other components of the regimen could continue with an increase in fluorouracil dose, according to local practice. Patients in the mFOLFOX plus ASC group with disease progression were subsequently treated at the clinician’s discretion. Following disease progression, patients in both the ASC alone group and mFOLFOX plus ASC group could receive treatment with experimental therapies in phase I trials.

Radiological assessment with CT (and, optionally, MRI) was performed in the mFOLFOX plus ASC group every 12 weeks until disease progression, and images were evaluated by investigators according to RECIST 1.1 criteria. Patients in the ASC alone group had radiological assessments only when clinically indicated. The primary end point of the ABC-06 study was OS in the intention-to-treat population. Secondary end points included PFS, radiological response, and QoL.

The key baseline characteristics of patients in cohort A of the FIGHT-202 study, and the ASC alone group and mFOLFOX plus ASC group in the ABC-06 study, are presented in Table 22. Compared with both treatment groups in the ABC-06 study, patients in cohort A of the FIGHT-202 study were younger (median age of 56 years versus 65 years) and less likely to be male (39% male versus 53% and 46% male). Almost all patients in cohort A of the FIGHT-202 study had iCCA, while 42% and 47% of patients in the mFOLFOX plus ASC and ASC alone groups, respectively, had iCCA. Almost all patients in both studies had an ECOG PS of 0 or 1, with a greater proportion of patients in cohort A of the FIGHT-202 study having an ECOG PS of 0 compared with the mFOLFOX plus ASC and ASC alone groups. FGFR2 mutation status was not reported in the ABC-06 study. Almost 40% of patients in cohort A of the FIGHT-202 study had more than 1 line of prior systemic therapy compared with none in the ABC-06 study.

Results

The regression model covariates are presented in Table 23 for the pemigatinib group before weighting and after weighting to match the mFOLFOX plus ASC group and the ASC alone group. ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| The effective sample size of the pemigatinib group was reduced to ||||||||||||||||||||||||| after weighting to match the mFOLFOX plus ASC and ASC alone groups, respectively. Other baseline characteristics after weighting were not presented in the sponsor-submitted ITC report.

Median follow-up duration in the ABC-06 study at data cut-off was 21.7 months (interquartile range, 17.2 months to 30.8 months). There were 2 patients and 1 patient lost to follow-up in the ASC alone and mFOLFOX plus ASC groups, respectively.

Overall Survival

Results for OS for pemigatinib versus mFOLFOX plus ASC are presented in Table 24. Following weighting of the pemigatinib group to match the mFOLFOX plus ASC group, median OS was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| for the pemigatinib group versus |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| for the mFOLFOX plus ASC group, based on the March 22, 2019, data cut-off for the FIGHT-202 study. The corresponding HR was |||||||||||||||||||||||||||||||||||||||||||||||||||||||) and the HR using the results from the April 7, 2020, data cut-off was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||. KM curves for OS are shown in Figure 14 for the unweighted pemigatinib group, weighted pemigatinib group, and mFOLFOX plus ASC group.

Figure 14: Redacted

Note: This figure has been redacted at the sponsor’s request.

|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.

||||||||||||||||||||||||||.⁵¹

Results for OS for pemigatinib versus ASC are presented in Table 25. Following weighting of the pemigatinib group to match the ASC alone group, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||, based on the March 22, 2019, data cut-off for the FIGHT-202 study. The corresponding HR was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. KM curves for OS are shown in Table 15 for the unweighted pemigatinib group, weighted pemigatinib group, and ASC alone group.

Figure 15: Kaplan–Meier Overall Survival Curves, Pemigatinib Versus ASC (Redacted)

Note: This figure has been redacted at the sponsor’s request.

||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.⁵¹

Progression-Free Survival

Results for PFS for pemigatinib versus mFOLFOX plus ASC are presented in Table 26. Following weighting of the pemigatinib group to match the mFOLFOX plus ASC group, median PFS was ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| for the pemigatinib versus mFOLFOX plus ASC groups, based on the March 22, 2019, data cut-off for the FIGHT-202 study. The corresponding HR was |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. KM curves for PFS are shown in Figure 16 for the unweighted pemigatinib group, the weighted pemigatinib group, and the mFOLFOX plus ASC group.

Figure 16: Kaplan–Meier Progression-Free Survival Curves, Pemigatinib vs. mFOLFOX Plus ASC (Redacted)

Note: This figure has been redacted at the sponsor’s request.

||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.⁵¹

Critical Appraisal of Sponsor-Submitted MAIC

Due to the non-comparative design of the FIGHT-202 study, the approach of using a MAIC to compare pemigatinib with a relevant comparator was appropriate. The following limitations of the systematic literature search methods were identified: the final stage of study selection for the MAIC did not appear to use preplanned criteria and the criteria for excluding publications during full-text screening for the reason “biliary tract cancer” may not have been consistently applied. However, the clinical experts consulted by CADTH for this review agreed that the ABC-06 study was likely the most relevant trial in this setting, given that it represents the only RCT evidence for treating CCA beyond the first-line setting. The selection of an RCT rather than a retrospective study was an appropriate choice, since patients who enrol in a clinical trial likely differ overall compared with those included in retrospective studies.

Although the statistical methods used for reweighting the pemigatinib group (cohort A of the FIGHT-202 study) and for estimating the 95% CIs were appropriate, there were potentially important underlying differences between the FIGHT-202 and ABC-06 studies. In particular, all patients in cohort A of the FIGHT-202 study had FGFR2 fusions or rearrangements, while patients in the ABC-06 study were not selected based on FGFR2 mutation status, and FGFR2 mutation status was not reported. Given that FGFR2 fusions and rearrangements occur almost exclusively in iCCA and that the prevalence of FGFR2 fusions and rearrangements is less than 20%⁵⁶ in patients with iCCA, there is likely a large disparity in FGFR2 mutation status between the study populations. While the FIGHT-202 study included only patients with CCA, the ABC-06 study included patients with BTC, which encompasses gallbladder cancer and ampullary cancer in addition to CCA. Due to this difference and the distribution of FGFR2 fusions and rearrangements between the types of CCA, 98.1% of patients in cohort A of the FIGHT-202 study had iCCA compared with 42% and 47% in the mFOLFOX plus ASC and ASC alone groups, respectively. Since disease type and FGFR2 status were more restricted in the FIGHT-202 study, these differences could not be addressed through the weighting of patients in the pemigatinib group. The natural history among bile duct cancer subtypes appears variable with different prognosis.⁵⁷

The covariates chosen for adjustment were based on |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Although the ITC authors claimed that this selection was based on the availability of this information for both the FIGHT-202 and ABC-06 studies, the following baseline characteristics were also available for both studies and did not appear to be considered: disease stage, percentage of patients with prior surgery for cancer, and number of lines of prior systemic therapy for advanced or metastatic cancer. It was not clear if the selection of covariates for adjustment was based on the need to adjust for predetermined prognostic factors and effect modifiers. Baseline characteristics other than the chosen covariates were not reported for the pemigatinib group after weighting. The clinical experts consulted by CADTH for this review were of the opinion that the number of lines of previous therapy was of key importance in terms of prognosis. The clinical experts were not aware of any additional prognostic factors and/or effect modifiers that were not reported in both studies and that should have been considered.

Another difference was noted related to the median duration of follow-up; which was 21.7 months (interquartile range = 17.2 to 30.8) in the ABC-06 trial and 15.44 months (minimum = 7.0; maximum = 24.7) and 27.9 months (||||||||||||||||||||||||||||||||||), respectively, at the March 22, 2019, and April 7, 2020, data cut-off dates in cohort A of the FIGHT-202 trial. The differing lengths of follow-up between the trials were not adjusted in the MAIC and may also contribute to heterogeneity, especially for the survival analyses.

While there are retrospective studies suggesting that the presence of FGFR2 mutations in CCA may be associated with a better prognosis,^18,19 the clinical experts consulted by CADTH were of the opinion that FGFR2 mutation status was not an important prognostic factor in the indicated patient population. The clinical experts considered the fact that patients in both the FIGHT-202 and ABC-06 trials had progressed on prior systemic therapy to be of greater importance in terms of prognosis. The clinical experts expected patients in the FIGHT-202 study to have more advanced disease than patients in the ABC-06 study because the FIGHT-202 study population was more heavily pre-treated overall. It is unclear whether the pemigatinib group was more or less similar to the ABC-06 groups in this respect following weighting, as the weighting process did not take the number of prior lines of systemic therapy into account. If substantial differences remained, these differences could have led to bias against pemigatinib in all of the comparisons.

The limitations with regard to the external validity identified for cohort A in the FIGHT-202 study also apply to the MAIC results. In addition, the effective sample size of the pemigatinib group was reduced by approximately ||||||| after weighting to the mFOLFOX plus ASC and ASC alone groups, and it is unclear how representative the post-weighting pemigatinib groups are of cohort A of the FIGHT-202 study. It is apparent that patients from cohort A of the FIGHT-202 study with an ECOG PS of 2 were effectively excluded from the pemigatinib group during the weighting process. Other changes noted in the pemigatinib group following weighting were that mean age increased by approximately ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||| for comparison with mFOLFOX plus ASC and ASC alone, respectively. While both male and female patients would have been adequately represented in the post-weighting pemigatinib groups, it is not clear how the age distribution was affected.

After adjustment based on the 4 chosen covariates, there was substantial loss in precision for the efficacy estimates, as the effective sample size was reduced by approximately |||||||. Adding more covariates for adjustment may have further reduced the available precision.

Comparisons of pemigatinib with other relevant comparators (FOLFIRI, fluorouracil alone or in combination with cisplatin or oxaliplatin, and capecitabine alone or in combination with cisplatin or oxaliplatin) were not available. Given that mFOLFOX plus ASC is the only therapy beyond the first-line setting with RCT evidence of an OS benefit, the clinical experts consulted by CADTH expected that mFOLFOX plus ASC would have the greatest efficacy out of all the relevant comparators.

Summary

For the unanchored MAIC to produce unbiased treatment effect estimates, both effect modifiers and prognostic variables need to be adjusted for in the analysis. Residual confounding remains the major limitation of the MAIC despite adjusting for age, sex, ECOG PS, and serum albumin in the comparisons of pemigatinib with mFOLFOX plus ASC and ASC alone. While any bias introduced by the differences between the FIGHT-202 and ABC-06 studies in the number of prior lines of systemic therapy may have been against pemigatinib, the substantial differences in FGFR2 mutation status and tumour site between trials introduce a high degree of uncertainty in the OS and PFS results. Furthermore, MAICs cannot account for unknown cross-trial differences; thus, the MAIC estimates are susceptible to bias from unknown confounding. An evaluation of potential bias from residual confounding was not reported; therefore, the magnitude of this bias in the relative treatment effect estimates is unclear. Overall, it remains uncertain whether pemigatinib provides additional OS or PFS benefit versus mFOLFOX plus ASC or ASC alone.

Other Relevant Evidence

This section includes 1 additional relevant report included in the sponsor’s submission to CADTH that was suggested to be a study for the indication for this CADTH review. FIGHT-101 is an ongoing, open-label phase I/II dose-escalation and expansion study of pemigatinib among participants with previously treated advanced malignancies with and without an FGF/FGFR alteration. As of February 2019, FIGHT-101 enrolled 160 participants from 14 study sites in the US and Denmark, 116 of which received at least 1 dose of pemigatinib monotherapy. Sixteen participants who were treated with pemigatinib monotherapy had CCA, ||||| of whom had FGFR2 rearrangements or fusions and received pemigatinib 13.5 mg orally once daily on a schedule of 2 weeks on, 1 week off, for each 21-day cycle.

|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The best overall response of a PR was observed in ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. In terms of safety outcomes, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. However, due to the open-label design and the limited data on the efficacy of pemigatinib on CCA within the report, the ability to interpret these results is considerably limited.

Discussion

Summary of Available Evidence

The CADTH systematic review included 1 phase II trial of pemigatinib that evaluated the efficacy and safety of pemigatinib in patients with advanced, metastatic, or surgically unresectable CCA with FGFR2 alterations, other FGF or FGFR alterations, or no FGF or FGFR alterations, whose disease did not respond to previous therapy. The FIGHT-202 trial (N = 146) is an ||||||||||||||||, multi-centre, open-label, single-arm phase II trial that assigned patients to 3 cohorts, depending on the patient’s FGF/FGFR status (cohort A: FGFR2 fusions or rearrangements; cohort B: FGF/FGFR alterations other than FGFR2 fusions or rearrangements; or cohort C: negative for FGF/FGFR alterations). This CADTH review focuses on cohort A, as cohorts B and C were not part of the requested reimbursement criteria submitted to CADTH and not approved in the Health Canada Notice of Compliance with conditions. All enrolled participants received oral pemigatinib (13.5 mg orally once daily on a schedule of 2 weeks on, 1 week off for each 21-day cycle). The primary outcome was ORR in cohort A, and secondary outcomes included ORR in cohort B, cohort A plus B, and cohort C; PFS, DOR, DCR, OS, and safety were assessed in all 3 cohorts. Exploratory end points included HRQoL and symptom severity.

Adults diagnosed with advanced, metastatic, or surgically unresectable CCA with FGFR2-positive disease who had documented disease progression after at least 1 line of prior systemic therapy, were enrolled into cohort A of the Fight-202 trial. The majority of patients had iCCA, an ECOG PS of 0 or 1, had received 1 or 2 previous lines of systemic therapies for advanced or metastatic disease, and were 56 years old (range, 26 to 77 years).

In addition to the systematic review, 1 sponsor-submitted ITC was summarized and appraised for this review. The other relevant evidence section included the summary of an open-label phase I/II trial (FIGHT-101) of pemigatinib in patients with advanced malignancies.

Interpretation of Results

Efficacy

The FIGHT-202 trial achieved the predetermined threshold for a positive outcome (lower limit of the 95% CI for ORR > 15%) in cohort A. As of the March 22, 2019, data cut-off date, after a median follow-up time of 15.44 months, the proportion of patients with an objective response, (the primary end point in cohort A) was 35.5% (95% CI, 26.50 to 45.35). A total of 3 patients had achieved a CR, 35 patients had a PR, and 50 patients had stable disease as the best response; the DCR was 82.2%. The results at the April 7, 2020, data cut-off date, with 27.9 months of follow-up, were consistent with the previous data cut-off date, with an ORR of 37% (95% CI, 27.94 to 46.86) and 4, 36, and 49 patients achieving a CR, PR, and stable disease as best response, respectively; DCR was not reported for the April 7, 2020, data cut-off date. Median DOR was 7.49 months (95% CI, 5.65 to 14.49) and 8.08 months (95% CI, 5.65 to 13.14) at the March 22, 2019, and April 7, 2020, data cut-off dates, respectively. The FIGHT-202 trial included no formal statistical significance and hypotheses testing, and point estimates with 95% CIs were reported to estimate the magnitude of treatment effect. A greater than 95% probability to have a 95% CI for ORR in cohort A with a lower limit larger than 15% was the basis for the sample size determination and was regarded as the threshold for a positive study outcome. Results for the subgroup of interest, as pre-specified in the protocol for this CADTH systemic literature review, |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. However, given that the trial was not designed to detect differences in treatment effects across subgroups, no conclusions can be made on the basis of subgroup results. OS and PFS were assessed as secondary outcomes in the FIGHT-202 trial; median OS and PFS, respectively, were 21.06 months (95% CI, 14.82 to not evaluable) and 6.93 months (95% CI, 6.18 to 9.59) at the March 22, 2019, data cut-off date. Consistent results were observed at the April 7, 2020, data cut-off date, with median OS and PFS being 17.48 months (95% CI, 14.42 to 22.93) and 7.03 months (95% CI, 6.08 to 10.48), respectively. Interpretation of time‐to‐event end points such as OS or PFS is limited in single‐arm studies; since all patients in cohort A received the same treatment, the extent to which the observed survival is due to the natural history of the tumour or the intervention remains unclear.⁴⁸ The primary objective of phase II (randomized or non-randomized) trials is to document the safety outcomes and investigate whether the estimate of effect for a new drug is large enough to use it in confirmatory phase III trials. A phase II trial may not accurately predict harm and/or effectiveness of treatments. The clinical experts consulted by CADTH noted that, despite the high unmet need, conducting an RCT in this setting with a targeted therapy, such as pemigatinib, compared with the available therapies currently used in the second line in Canadian clinical practice would likely not be feasible. According to the clinical experts, developing phase III RCTs is hindered by the overall low number of patients who meet the current indication and equipoise between pemigatinib and other chemotherapy drugs does not exist.

According to the clinical experts consulted by CADTH and the registered clinician groups providing input for this submission, the responses achieved with pemigatinib were clinically relevant, important to patients in this setting, and clearly higher in comparison with what could be observed with the therapies currently used in this setting. The clinically experts consulted by CADTH noted that durable responses in this patient population are important because of accompanying delay in the worsening of symptoms and a slower decline in ECOG PS. The clinical experts emphasized the clinical relevance and importance of even maintaining stable disease in preventing an otherwise fast decline in patients in this oftentimes last line of treatment. This view was echoed by the input provided by the patient advocacy group, which highlighted tumour response, maintenance of response, delay in disease progression, and QoL as important treatment goals for patients. While the clinical experts agreed that, based on the available evidence, it was not possible to conclude whether the antitumour activity expressed as responses would translate into clinical benefits in terms of PFS and OS, they felt that the preliminary survival results from the trial were very encouraging and that durable responses could potentially delay tumour progression and result in prolonged survival benefit in this patient population.

In the absence of a direct comparison of pemigatinib with relevant treatment options, the sponsor submitted an ITC⁵¹ in the form of an unanchored MAIC comparing the efficacy of pemigatinib (cohort A of the FIGHT-202 trial) with each of the 2 treatment groups in the ABC-06 study. The results of the ITC favoured pemigatinib for PFS and OS in comparison with mFOLFOX plus ASC as well as compared with ASC alone. The CADTH critical assessment identified several limitations with the sponsor’s submitted MAIC, including heterogeneity across study designs and populations and the inability to adjust for all potential confounders and prognostic variables in the MAIC. The clinical experts agreed with the CADTH clinical review team that, given the absence of robust comparative data on PFS and OS, the ability to interpret the relative treatment effects observed between pemigatinib and FOLFOX plus ASC and ASC alone was limited, and no firm conclusions could be drawn on how pemigatinib compared with other relevant treatment options. The clinical experts consulted by CADTH anticipated, however, that based on the FIGHT-202 results and on poor results with existing treatment options in clinical practice, pemigatinib appeared to offer at least similar or improved clinical benefits compared with current therapies, with better tolerability. The clinical experts consulted by CADTH noted there is currently insufficient evidence to determine whether patients whose cancer is positive for an FGFR2 fusion represent a distinct prognostic subgroup. The clinical experts agreed that progression on prior systematic therapy is a key prognostic factor in these patients, and they did not anticipate that patients would derive any substantial benefit from their underlying disease biology at the time they enrolled into the FIGHT-202 trial.

Input received by the patient advocacy group, the registered clinicians, and the clinical experts consulted by CADTH highlighted HRQoL as an important outcome and treatment goal for patients. The overall observed scores from baseline to cycle 42 were |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. However, given the non-comparative, open-label design of the trial, the lack of a pre-specified analysis of the patient-reported outcomes data, |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The clinical experts consulted by CADTH noted that HRQoL in this setting is low and unstable. They anticipated that, given the observed responses in the FIGHT-202 trial, pemigatinib would likely improve or at least maintain patients’ HRQoL.

While patients recruited in the FIGHT-202 trial were considered representative of patients in Canadian clinical practice, the clinical experts consulted by CADTH noted that it would be reasonable to generalize the results of cohort A of the FIGHT-202 trial to patients who are intolerant to first-line therapy, which was a group of patients excluded from the trial. As well, given the acceptable safety profile of pemigatinib, the clinical experts felt it would be reasonable to leave it up to the discretion of the treating physician to apply some flexibility in terms of using pemigatinib in patients with slightly lower laboratory parameters compared with those outlined in the trial. The clinical experts anticipated seeing the benefit of treatment with pemigatinib regardless of the number of previous lines of systemic therapy, as long as patients have the FGFR2 alteration. However, the clinical experts agreed that patients should not have been treated previously with an FGFR2-targeted therapy. Furthermore, the clinical experts noted that patients with iCCA and eCCA are managed in a similar way in clinical practice and that results observed in cohort A are generalizable to patients with FGFR2-positive eCCA, based on the fact that FGFR2 is the target of the mechanism of action of pemigatinib and there is no biologic rationale to assume that pemigatinib’s safety profile would be different in patients with eCCA.

Harms

The single-arm, non-randomized design of the FIGHT-202 trial made interpreting the safety events attributable to pemigatinib challenging, since all patients in cohort A received the same treatment. This should be considered when reviewing the incidence of TEAEs. All patients in cohort A experienced at least 1 TEAE. The most commonly reported TEAEs included alopecia, hyperphosphatemia, diarrhea, and dysgeusia. The most commonly reported grade 3 or higher and serious TEAEs, respectively, were hypophosphatemia and pyrexia. The clinical experts consulted by CADTH noted that most TEAEs associated with pemigatinib could be managed with dose modifications, and that treatment discontinuation due to TEAEs was relatively rare. From the review of notable harms, it appeared that toxicities from pemigatinib were mostly seen as hyperphosphatemia and nail toxicities. Overall, deaths were few and no TEAE leading to death was considered treatment-related. Overall, the clinical experts consulted by CADTH agreed with the registered clinicians who provided input to this submission that the TEAEs observed with pemigatinib were generally acceptable and could be adequately managed in clinical practice. This was reflective of the patient experience with pemigatinib reported in the patient input received, which stated that, overall, patients had little challenge dealing with the side effects of pemigatinib. Furthermore, it was emphasized by the clinical experts consulted by CADTH that the toxicity with pemigatinib appeared favourable compared with currently available chemotherapy options. Examples of side effects from chemotherapy that may be avoided with pemigatinib were neuropathy and neutropenia, according to the clinical experts.

Conclusions

One phase II, singe-arm, open-label trial (FIGHT-202) provided evidence regarding the efficacy and safety of pemigatinib in patients with advanced, metastatic, or surgically unresectable CCA with FGFR2 alterations (cohort A) whose disease did not respond to a previous therapy. The FIGHT-202 trial achieved the predetermined threshold for a positive outcome (lower limit of the 95% CI for ORR > 15%) in cohort A. The clinical experts consulted by CADTH felt that the achieved ORR of 37% (April, 7, 2020, data cut-off date) was clinically meaningful for the target population and durable (median of 8.08 months; range, 5.65 to 13.14). In the opinion of the clinical experts, the observed responses appeared higher than what is seen with the therapies currently used in the second line in this setting. There was uncertainty around the magnitude of the clinical benefit, given the limitations in the evidence from this non-comparative phase II clinical trial. The clinical experts consulted by CADTH noted that, despite the high unmet need, conducting an RCT with a targeted therapy, such as pemigatinib, compared with the currently available therapies used in the second line in Canadian clinical practice would likely not be feasible. While the secondary efficacy outcomes, OS and PFS, appeared supportive of the observed ORR achievements, the non-randomized design of the FIGHT-202 trial made interpreting the PFS and OS events attributable to pemigatinib challenging. In the absence of a direct comparison of pemigatinib with relevant treatment options, the sponsor submitted an ITC. However, the CADTH critical assessment identified limitations with the sponsor’s submitted unanchored MAIC (including heterogeneity across study designs and populations and the inability to adjust for all potential confounders and prognostic variables), which limited the ability to interpret the relative treatment effects observed between pemigatinib and other treatments. The results for the HRQoL and symptom severity exploratory outcomes remained inconclusive due to a number of important limitations. The toxicity profile of pemigatinib was considered manageable by the clinical experts consulted by CADTH and appeared favourable compared with currently available chemotherapy options. However, the non-comparative design of the FIGHT-202 trial made interpreting the safety events attributable to pemigatinib challenging, since all patients in cohort A received the same treatment.

References

1.Zhou X, Wang J, Tang M, et al. Hepatocellular carcinoma with hilar bile duct tumor thrombus versus hilar Cholangiocarcinoma on enhanced computed tomography: a diagnostic challenge. BMC Cancer. 2020;20(1):54. PubMed