An open-label study of pemigatinib in cholangiocarcinoma: final results from FIGHT-202

Background Fibroblast growth factor receptor 2 (FGFR2) fusions and rearrangements are clinically actionable genomic alterations in cholangiocarcinoma (CCA). Pemigatinib is a selective, potent, oral inhibitor of FGFR1-3 and demonstrated efficacy in patients with previously treated, advanced/metastatic CCA with FGFR2 alterations in FIGHT-202 (NCT02924376). We report final outcomes from the extended follow-up period. Patients and methods The multicenter, open-label, single-arm, phase II FIGHT-202 study enrolled patients ≥18 years old with previously treated advanced/metastatic CCA with FGFR2 fusions or rearrangements (cohort A), other FGF/FGFR alterations (cohort B), or no FGF/FGFR alterations (cohort C). Patients received once-daily oral pemigatinib 13.5 mg in 21-day cycles (2 weeks on, 1 week off) until disease progression or unacceptable toxicity. The primary endpoint was objective response rate (ORR) in cohort A assessed as per RECIST v1.1 by an independent review committee; secondary endpoints included duration of response (DOR), progression-free survival (PFS), overall survival (OS), and safety. Results FIGHT-202 enrolled 147 patients (cohort A, 108; cohort B, 20; cohort C, 17; unconfirmed FGF/FGFR alterations, 2). By final analysis, 145 (98.6%) had discontinued treatment due to progressive disease (71.4%), withdrawal by patient (8.2%), or adverse events (AEs; 6.8%). Median follow-up was 45.4 months. The ORR in cohort A was 37.0% (95% confidence interval 27.9% to 46.9%); complete and partial responses were observed in 3 and 37 patients, respectively. Median DOR was 9.1 (6.0-14.5) months; median PFS and OS were 7.0 (6.1-10.5) months and 17.5 (14.4-22.9) months, respectively. The most common treatment-emergent AEs (TEAEs) were hyperphosphatemia (58.5%), alopecia (49.7%), and diarrhea (47.6%). Overall, 15 (10.2%) patients experienced TEAEs leading to pemigatinib discontinuation; intestinal obstruction and acute kidney injury (n = 2 each) occurred most frequently. Conclusions Pemigatinib demonstrated durable response and prolonged OS with manageable AEs in patients with previously treated, advanced/metastatic CCA with FGFR2 alterations in the extended follow-up period of FIGHT-202.


INTRODUCTION
Cholangiocarcinoma (CCA) accounts for w10%-25% of primary hepatic cancers and 3% of gastrointestinal tumors. 1,2n the United States, CCA incidence is increasing. 3Older patients, men, and people identifying as Asian/Pacific Islander generally have a higher CCA incidence. 3CCA is classified as intrahepatic (iCCA) or extrahepatic (perihilar or distal) based on location.Among CCA tumors, w10%-56% are iCCA. 1,4,5iCCA has high genomic heterogeneity, with 40%-50% of patients with CCA harboring one or more clinically actionable genomic alteration. 60][11] Compared to CCA without FGFR2 alterations, FGFR2 fusions are associated with longer overall survival (OS) from diagnosis. 12GFRs regulate several cellular processes, including cell proliferation, survival, migration, and angiogenesis; dysregulation of these pathways drives tumorigenesis. 13Therefore, FGFR inhibitors are a rational targeted therapy to disrupt pathogenic FGFR signaling in CCA. 14 Because of the asymptomatic nature of early-stage disease and nonspecific symptoms in later stages, CCA is often diagnosed in advanced stages when patients are ineligible for curative surgery. 1,15Approximately 65% of patients have unresectable disease, and up to half of them have lymph node metastases at time of diagnosis. 16,17Until recently, gemcitabine plus cisplatin chemotherapy was the first-line standard of care for treatment of unresectable or metastatic CCA. 15,180][21][22][23] Modest response rates [w20% objective response rate (ORR)] and a median survival of w11 months are typical with first-line chemotherapy. 24,25he addition of durvalumab to chemotherapy improves ORR to w27% and extended median OS to nearly 13 months. 26Improvement in OS has also been observed with the addition of pembrolizumab to gemcitabine and cisplatin, resulting in an OS of nearly 13 months. 27espite this recent advance in therapy for unresectable or metastatic CCA, 15 treatment options that exploit clinically actionable genomic alterations, including FGFR2 rearrangements, are needed.Pemigatinib is an oral, potent, selective FGFR1-3 inhibitor for treatment of adults with previously treated, unresectable, locally advanced or metastatic CCA with FGFR2 fusions or other rearrangements. 28n the primary analysis of FIGHT-202, a phase II study evaluating the safety and efficacy of pemigatinib in previously treated locally advanced or metastatic CCA, patients with FGFR2 fusions or rearrangements had an ORR of 35.5% at a median follow-up of 15.4 months. 29Here we report final efficacy and safety analyses from the extended followup period of the FIGHT-202 study (NCT02924376; EudraCT 2016-002422-36).

Study design
FIGHT-202 was an open-label, single-arm, multicenter, phase II study conducted at 146 sites in the United States, Republic of Korea, UK, France, Italy, Thailand, Germany, Belgium, Israel, Spain, Japan, and Taiwan.The data cut-off date was 8 July 2021.FIGHT-202 consisted of three cohorts based on tumor FGF/FGFR alteration status: (A) FGFR2 rearrangements or fusions, (B) other FGF/FGFR alterations, or (C) no FGF/FGFR alterations (United States only).Enrollment and initial cohort assignment were permitted based on genomic testing results from a local laboratory.Final cohort assignment for statistical analyses was based on centrally confirmed next-generation sequencing results using the Foundation Medicine clinical trial assay (Founda-tionOne, Foundation Medicine, Cambridge, MA).FIGHT-202 was carried out as per the International Council for Harmonisation Guideline for Good Clinical Practice, Declaration of Helsinki, and local regulatory requirements.The study protocol was approved by the institutional review board of each site before patient enrollment.All patients provided written informed consent.

Patients
Eligibility requirements have been published previously. 29riefly, eligible patients were !18 years old, had advanced/ metastatic or surgically unresectable CCA with radiographically measurable disease as per RECIST v1.1, disease progression after one or more line of prior systemic therapy, documented FGF/FGFR gene alteration, life expectancy !12 weeks, and Eastern Cooperative Oncology Group (ECOG) performance status 2. Patients with inadequate hepatic or renal function, history or current evidence of ectopic mineralization or calcification, or current evidence of clinically significant corneal or retinal disorder were ineligible.

Treatment
All patients self-administered pemigatinib over 21-day cycles (2 weeks on/1 week off) at a starting oral dose of 13.5 mg once daily until documented radiologic disease progression, unacceptable toxicity, consent withdrawal, or physician decision.

Endpoints and assessments
The primary endpoint was ORR in patients with FGFR2 fusions or rearrangements (cohort A) as determined by an independent review committee (IRC).ORR was defined as the percentage of patients with complete (CR) or partial responses (PR) as per RECIST v1.1.Disease was assessed by computed tomography or magnetic resonance imaging every 6 weeks through week 12, and every 9 weeks thereafter for all cohorts; patients who discontinued study treatment for reasons other than disease progression were assessed every 9 weeks during follow-up.
Secondary endpoints were ORR in patients with FGF/ FGFR alterations other than FGFR2 fusions or rearrangements (cohort B) and ORR in patients without FGF/FGFR alterations (cohort C).Additional secondary endpoints assessed in all cohorts were progression-free survival [PFS; time from first dose to progressive disease (PD) or death], duration of response (DOR; time from the date of CR or PR until PD), disease control rate (DCR; CR þ PR þ stable disease), and OS (time from first dose to death due to any cause) for all cohorts.
Safety and tolerability were based on the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.03 and were assessed at screening, during treatment, at the end of treatment, and during follow-up.
Genomic analysis of baseline tumor samples was carried out as previously described. 6

Statistical analyses
The efficacy-assessable population included all patients with centrally confirmed FGF/FGFR alteration status who received one or more dose of pemigatinib.The primary analysis of ORR was carried out in cohort A based on IRCassessed tumor responses.The ClopperePearson method was used to estimate the 95% confidence interval (CI) for ORR.Analyses of ORR and 95% CI estimation in cohorts A and B combined, cohort B, and cohort C, as well as DCR analyses, were carried out in the same way as the analysis of ORR for cohort A. The KaplaneMeier method was used to assess PFS, DOR, and OS.Exploratory analysis of ORR, PFS, and OS in subgroups based on demographic and baseline clinical characteristics was carried out for cohort A. The safety-assessable population included all enrolled patients who received one or more dose of pemigatinib; safety data were summarized descriptively.Statistical analysis of the effect of co-alterations on OS was carried out using the log-likelihood ratio test and KaplaneMeier method as described previously. 6

Response to treatment
Overall, median (range) follow-up for the efficacy-assessable population was 45.4 (19.9-53.7)months.

DISCUSSION
In the final analysis of FIGHT-202, continued benefit of pemigatinib in patients with previously treated advanced or metastatic CCA with FGFR2 rearrangements or fusions was observed over an extended follow-up period, including a 37% ORR, a median DOR of 9.1 months, and a median PFS and OS of 7.0 and 17.5 months, respectively.No patients with other or without FGF/FGFR alterations responded to pemigatinib.No new safety concerns were identified; the most common treatment-related TEAE was hyperphosphatemia (54%; all cases were grade 1 or 2).
CCA is typically unresectable at diagnosis, 16 and mortality rates, primarily driven by iCCA, are increasing. 30,31tandard-of-care first-line treatment for unresectable or metastatic CCA is gemcitabine and cisplatin plus durvalumab in many countries, whereas gemcitabine plus cisplatin chemotherapy remains the standard of care where durvalumab has not yet been approved. 15,18,32However, many The dashed line indicates criterion for PR (!30% decrease in sum of target lesion diameters).CR, complete response; FGFR, fibroblast growth factor receptor; IRC, independent review committee; PD, progressive disease; PR, partial response; SD, stable disease.patients do not respond to treatment, and second-line therapies provide only limited benefit. 33A meta-analysis of retrospective and phase II studies reporting second-line chemotherapy for bile duct cancers showed a mean response rate of 7.7% and a mean OS of only 7.2 months. 34 post hoc analysis of FIGHT-202 assessing PFS in patients by prior systemic therapy showed that median PFS in patients with FGFR2 fusions or rearrangements treated with second-line pemigatinib was 7.0 months. 35Patients with prior second-line therapy (chemotherapy, 93%) had a median PFS of 4.2 months, possibly suggesting that second-line targeted therapy may improve outcomes in patients with FGFR2 fusions or rearrangements over chemotherapy. 35In contrast with the historically poor responses to second-line treatments for CCA, we demonstrated the continued clinical benefit of pemigatinib in previously treated CCA with FGFR2 fusions or rearrangements over an extended follow-up period.

Time to event, months
The success of the FGFR inhibitors pemigatinib, erdafitinib, futibatinib, and RLY-4008 in patients with previously treated CCA tumors with FGFR2 fusions or rearrangements 29,[36][37][38][39][40] represents a paradigm shift toward personalized medicine. 41The European Society for Medical Oncology (ESMO) and United States National Comprehensive Cancer Network (NCCN) treatment guidelines currently suggest biomarker-guided treatments based not only on FGFR2 alterations but also microsatellite instability-high/mismatch repair-deficient, ERBB2-positive/mutated, NTRK fusion-positive, and BRAF-and IDH1-mutated tumors. 15,18NCCN also suggests biomarker-guided treatments for tumor mutation burden-high and RET fusion-positive tumors, 15 and ESMO suggests treatment for patients with BRCA1/2 or PALB2 mutations. 18Targeted therapies are currently recommended as second-line treatments; however, extending these findings into the first-line setting may improve clinical outcomes in patients with CCA with specific genomic alterations.For example, the ongoing randomized phase III FIGHT-302 clinical study (NCT03656536) 42 is evaluating first-line pemigatinib versus chemotherapy in patients with CCA with FGFR2 rearrangements.
The clinical utility of FGFR inhibition in CCA is hampered by primary and secondary resistance to FGFR inhibitors.Understanding mechanisms of tumor resistance is therefore critical.In FIGHT-202, patients with cooccurring alterations in one or more tumor suppressor gene had significantly shorter PFS than those without tumor suppressor gene alterations; ORR was not significantly different. 6Consistent with previous PFS and ORR findings, TP53 and PBRM1 co-alterations were also associated with significantly shorter OS in this final analysis of FIGHT-202. 6Patients with co-alterations in BAP1 also tended to have numerically shorter OS, consistent with the worse PFS previously reported in patients with this co-alteration. 6In FIGHT-202, all patients with reductions in tumor size followed by PD (n ¼ 8) had developed one or more mutation in the kinase domain of FGFR2 predicted to promote kinase activation or impair pemigatinib binding. 6Larger-scale molecular profiling may enhance understanding of alterations that prevent or reverse FGFR inhibition in CCA.
FGFR inhibitors that irreversibly bind to FGFR2 (e.g.futibatinib, RLY-4008) may have greater antitumor activity in CCA with FGFR2 resistance mutations compared with ATP-competitive inhibitors (e.g.pemigatinib). 14However, whether the covalent binding mechanism translates to clinically meaningful longer PFS is unknown.Ongoing phase I/II studies of RLY-4008 will provide more data to address this question. 40Preliminary work suggests that sequential treatment of patients with acquired 'on target' resistance to ATP-competitive inhibitors with irreversible FGFR2 inhibitors is possible in a subgroup of patients; this concept should be evaluated in future studies.
Although targeting the tumor microenvironment (TME) with immune checkpoint inhibitors has proven to be an effective therapeutic strategy in many solid tumors, these drugs, with the exception of durvalumab 26 and pembrolizumab, 27 have been used with limited success in CCA. 1 Recent TME-based transcriptomic analyses demonstrated that approximately two-thirds of iCCA tumors exhibit an immunologically cold 'non-inflamed' TME. 43The largest subtype of non-inflamed tumors was significantly enriched in FGFR2 fusions as well as BAP1 and IDH1/2 mutations. 43A TOPAZ-1 post hoc analysis found that patients with biliary tract cancer with clinically actionable genomic alterations, including FGFR2 rearrangements, had an OS benefit from first-line durvalumab plus gemcitabine and cisplatin versus chemotherapy alone. 44The number of patients with FGFR2 rearrangements was very small; therefore, whether adding durvalumab to standard chemotherapy benefits patients with FGFR2 alterations to the same extent as patients without FGFR2 alterations remains an open question.FGFR inhibition coupled with TME-targeted therapies, such as those that deplete cells contributing to immunosuppressive TME phenotypes, may improve clinical outcomes in specific patient populations. 43Further characterization of the TME in FGFR2-altered CCA may be warranted to understand the impact of dysregulated FGFR2 signaling on the TME and to identify patients who might benefit most from combination therapies.
Limitations of FIGHT-202 have been discussed previously. 29The study design included no active comparator treatment arm.Small numbers in patient subgroups also limit interpretation of efficacy based on demographic and disease characteristic factors.

Conclusions
This final analysis of FIGHT-202 demonstrated continued durable response, prolonged OS, and manageable AEs in patients with previously treated advanced or metastatic CCA with FGFR2 fusions or rearrangements, further supporting regulatory approvals of pemigatinib based on this single-arm, phase II study. 28,45These results highlight the need for early molecular testing in CCA.The phase III FIGHT-302 study will further elucidate the role of FGFR inhibitors in biomarker-selected CCA.Routine comprehensive genomic profiling is needed to discover novel actionable FGFR2 alterations and identify patients who might benefit from FGFR inhibition.

Figure 1 .
Figure1.Best percentage change from baseline in target lesion size based on IRC assessment among efficacy-assessable patients in cohort A (FGFR2 rearrangements or fusions).The dashed line indicates criterion for PR (!30% decrease in sum of target lesion diameters).CR, complete response; FGFR, fibroblast growth factor receptor; IRC, independent review committee; PD, progressive disease; PR, partial response; SD, stable disease.

Table 1 .
51.1) months.No objective responses were observed (Table 2).Median (range) best percentage change from baseline in sum of Median (range) follow-up for efficacy-assessable patients in cohort C was 51.9 (49.5-53.7)months.No objective responses were observed (Table 2).Median (range) best percentage change from baseline in sum of target lesion diameters was 6.2% (À33% to þ74%; Supplementary Figure S4B, available at https://doi.org/10.1016/j.esmoop.2024.103488).Patient demographics and baseline clinical characteristics (safety-assessable population) CCA, cholangiocarcinoma; ECOG, Eastern Cooperative Oncology Group; FGF, fibroblast growth factor; FGFR, FGF receptor.aTotal number includes two patients who did not have confirmed FGF/FGFR status by central laboratory testing and were not assigned to any cohort.bRest of the world includes Israel, Japan, Republic of Korea, Taiwan, and Thailand.cPatients with nonmetastatic disease have no evidence of extrahepatic metastasis.Patients with metastatic disease may have had intrahepatic and extrahepatic metastases.dSpecific sites reported in >5% of patients overall are shown.medianPFS (2.1 and 1.5 months, respectively) and Kaplane Meier estimates of PFS at evaluable time points were significantly lower than in cohort A.At data cut-off, 32 patients (29.6%) in cohort A were alive and censored for survival (

Table 2 .
Efficacy outcomes (efficacy-assessable population) CR, complete response; DCR, disease control rate; DOR, duration of response; FGF, fibroblast growth factor; FGFR, FGF receptor; NE, not evaluable; ORR, objective response rate; OS, overall survival; PFS, progression-free survival; PR, partial response; SD, stable disease.a Calculated as the percentage of patients with DOR !12 months among all patients with CR or PR (n ¼ 40).

Table 3 .
Treatment-related treatment-emergent adverse events (safety-assessable population) FGF, fibroblast growth factor; FGFR, FGF receptor; TEAE, treatment-emergent adverse event.aTotal number includes two patients who did not have confirmed FGF/FGFR status by central laboratory testing and were not assigned to any cohort.b All any-grade TEAEs occurring in !10% and grade !3 TEAEs occurring in !2% of the total population are shown.