Efficacy and Safety of Pemigatinib in Subjects With Advanced/Metastatic or Surgically Unresectable Cholangiocarcinoma Who Failed Previous Therapy - (FIGHT-202): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Efficacy and Safety of Pemigatinib in Subjects With Advanced/Metastatic or Surgically Unresectable Cholangiocarcinoma Who Failed Previous Therapy - (FIGHT-202) (NCT02924376), a non-randomized phase 2 study. The primary outcome is Objective Response Rate (ORR) in Participants With FGFR2 Rearrangements or Fusions.

2. Clinical Question

3. Trial Design

FIGHT-202 is recorded as a phase 2, open-label, non-randomized, multicenter study of pemigatinib. Its title and summary describe people with advanced or metastatic, or surgically unresectable, cholangiocarcinoma who had failed previous therapy. The study purpose was to evaluate pemigatinib in participants with FGFR2 translocation, while the listed cohorts also distinguish other FGF/FGFR alteration groups and participants negative for those alterations.

The registry lists enrollment of 147 and status as completed. The recorded study start is 2017-01-16, and the primary completion date is 2022-02-01. These dates describe the record; they do not establish when every participant enrolled, received treatment, or completed follow-up. The primary endpoint timeframe is stated as up to 1527 days.

Because allocation is non-randomized and masking is absent, the listed cohorts should not be read as randomized treatment arms. All are described as receiving pemigatinib, with cohort membership defined by FGF/FGFR alteration status. There is no like-for-like control comparison in the registry record. Differences between these groups, if reported elsewhere, would not by themselves isolate a treatment effect: the groups are defined by biological characteristics rather than random assignment.

The primary endpoint is objective response rate in participants with FGFR2 rearrangements or fusions. Secondary outcomes listed include response rates in other alteration groups and across participants with FGF/FGFR alterations, progression-free survival, duration of response, disease control rate, overall survival, and the number of participants with any treatment-emergent adverse event. The registry record does not report numerical results for these efficacy outcomes in the material available for this analysis. It also does not report details of the protocol’s sample-size justification, interim analyses, or full statistical analysis plan.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: This study enrolled participants at 68 study sites in the United States, South Korea, United Kingdom, France, Italy, Thailand, Germany, Belgium, Israel, Spain, Japan, and Taiwan.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Efficacy Evaluable Population (EEP): all participants who received ≥1 dose of pemigatinib and had a known FGF/FGFR alteration or, in the United States, were negative for FGF/FGFR alterations based on central genomics laboratory results. Two participants were assigned to a group labeled "Other" and were excluded from the EEP because the local laboratory FGF/FGFR results could not be confirmed centrally. Confidence intervals were calculated based on the exact method for binomial distribution.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Objective Response Rate (ORR) in Participants With FGFR2 Rearrangements or Fusions, measured over up to 1527 days.

The primary endpoint is objective response rate (ORR) in participants with FGFR2 rearrangements or fusions. It is defined as the percentage of participants whose best overall response at a post-baseline visit before first progressive disease was complete response (CR) or partial response (PR), using RECIST version 1.1. Response was based on scans reviewed by an independent centralized radiological review committee, and the alteration classification was based on central genomics laboratory results.

In general, an objective response endpoint reduces a set of tumor assessments to a clinically interpretable category: whether a participant met prespecified criteria for complete or partial response. Under the definition in the registry, CR requires disappearance of target and non-target lesions, no new lesions, and specified reduction in pathological lymph-node short axis. PR requires complete disappearance or at least a 30% decrease in the sum of target-lesion diameters from baseline, with no new lesions and no progression of non-target lesions. These rules make the response classification more explicit than an informal judgment of improvement.

ORR is used to summarize tumor shrinkage meeting the criteria, rather than time until progression or duration of survival. It does not by itself describe how long a response lasts or what happens after progression. The registry lists duration of response, progression-free survival, and overall survival as separate secondary outcomes. The primary endpoint’s numerical estimate and 95% confidence interval are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods determine how observations are classified, summarized, and interpreted. The registry specifies several important pieces of the endpoint definition: the response criteria, the assessment source, the population focus, and a confidence interval level. It does not report the full statistical analysis plan, so details beyond those stated cannot be assumed.

What does the response endpoint count?

ORR is a binary participant-level outcome: a participant is classified as having a response or not having a response under the stated criteria. In this study, a response means a best overall response of CR or PR at a post-baseline assessment before first progressive disease. The endpoint is expressed as a percentage of participants. The registry defines CR and PR using RECIST version 1.1, including rules for disappearance or reduction of lesions and the absence of new lesions or non-target progression.

In general, binary outcomes simplify a complex sequence of clinical assessments into a clear category. That simplicity depends on consistent definitions and complete enough assessment to classify participants. The registry specifies that scans were reviewed by an independent centralized radiological review committee. Central review can apply a common assessment process across participants; it does not eliminate uncertainty in the estimate or answer questions about participants whose outcome data are missing.

What does a 95% confidence interval communicate?

The endpoint is described with a 95% confidence interval. In general, a confidence interval expresses the statistical uncertainty associated with an estimate under the method used. It should be read together with the response percentage: the estimate describes the result in the analyzed population, while the interval indicates its precision. It is not a guarantee that a particular future population will have a result inside those bounds. Neither the estimate nor the interval bounds are reported in the registry record, so the precision of the primary response result cannot be judged.

Why does the study design matter?

FIGHT-202 is listed as non-randomized, open-label, and single-arm. General statistical interpretation of a single-arm study differs from interpretation of a randomized comparison. In a randomized study, allocation is intended to support a comparison between groups by balancing prognostic factors on average. Here, the registry describes treatment with pemigatinib and cohorts defined by alteration status, not random assignment to pemigatinib or a comparator. Thus, a response proportion can describe the treated group, but it cannot by itself estimate a treatment difference against a concurrent control.

Open-label status means masking was not used. The primary scan review was nevertheless described as independent and centralized. Those design features address different potential influences: masking concerns knowledge of treatment assignment, while centralized review concerns the assessment process. The record does not report further details that would establish how other potential sources of bias were managed.

How should cohort results be compared?

The registry lists Cohort A for FGFR2 rearrangements or fusions, Cohort B for other FGF/FGFR alterations, Cohort C for participants negative for FGF/FGFR alterations, and an “Other” group. These categories describe biological subgroups among participants receiving pemigatinib. In general, subgroup estimates may be less precise when based on fewer participants, and apparent differences can reflect variation in participant characteristics as well as chance. The record’s counts include 108 participants at risk in Cohort A, 20 in Cohort B, 17 in Cohort C, and 2 in “Other.” Those counts do not provide the missing primary response numerator or confidence interval.

What do the planning illustrations represent?

The registry describes a sample-size planning illustration based on the observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is stated as a fraction of the observed effect. For time-to-event outcomes, the illustration uses a total number of events, while other outcomes are expressed as participants per arm. These are planning assumptions, not observed power and not evidence that the original trial was designed with those assumptions. No numerical planning estimates or actual sample-size justification are reported in the registry record.

In general, a planning calculation asks how much information a hypothetical design might need to detect a chosen effect under specified assumptions. It does not show the probability that a completed study’s result is true, and it does not repair the absence of a randomized comparison. Because this study is single-arm, a two-arm planning scenario should be treated as a separate illustration rather than a description of the trial’s allocation.

Why can event counts and time-to-event outcomes differ?

ORR counts whether a qualifying response occurred. Progression-free survival, duration of response, and overall survival are time-to-event outcomes, which concern the timing of an event and can involve participants who have not yet had that event when follow-up ends. In general, time-to-event methods account for differing follow-up and censoring, but the exact methods used for those secondary endpoints are not reported in the registry record. Their names alone do not establish how censoring, missing assessments, or analysis populations were handled.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Cohort A: FGFR2 Rearrangements or Fusions")
n      <- c(108)
events <- c(40)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Cohort A: FGFR2 Rearrangements or Fusions  108  40  37.0  27.9  46.9  28.5  46.4  28.5  46.5

11. Multiplicity, Missing Data and Analysis Populations

The registry lists several secondary outcomes in addition to the primary ORR endpoint. These include response rates for alteration groups, progression-free survival, duration of response, disease control rate, overall survival, and the number of participants with any treatment-emergent adverse event. When a study evaluates multiple outcomes or subgroups, general statistical practice recognizes a multiplicity issue: the more statistical questions asked, the greater the chance that some apparently notable findings occur by chance. Whether and how multiplicity was controlled for this study is not reported in the registry record.

The record does not report a complete analysis plan, a multiplicity adjustment, or interim analyses. No assumptions about endpoint hierarchy, confirmatory status of secondary outcomes, or stopping rules can therefore be made. The primary endpoint is identified, but the registry information does not explain how the other listed outcomes were to be interpreted in relation to it.

The stated efficacy evaluable population includes participants who received at least 1 dose of pemigatinib and had a known FGF/FGFR alteration or, in the United States, were negative for FGF/FGFR alterations based on central genomics laboratory results. The record also says two participants were assigned to “Other.” It does not provide the primary endpoint denominator, the number with evaluable scans, or a detailed account of exclusions and missing outcome data. How participants with missing or unevaluable assessments were handled is not reported in the registry record.

In general, missing outcomes can affect a percentage estimate if the participants without an assessment differ from those with an assessment. A transparent interpretation therefore needs the analysis population, denominator, missing-data rules, and sensitivity analyses where applicable. Those details are not reported in the registry record for the primary response estimate. The count enrolled should not be treated as the endpoint denominator without that information.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: abdominal pain (11.8%), back pain (11.8%), hyponatraemia (10.0%), pleural effusion (10.0%), acute kidney injury (5.9%).

The serious adverse event figures are counts of participants listed as affected among those at risk in each group. Cohort A lists 46 affected among 108 at risk; Cohort B lists 10 among 20; Cohort C lists 12 among 17; and “Other” lists 0 among 2. The total is 68 affected among 147 at risk. These are participant counts, not counts of individual event episodes.

The record also lists the most frequent serious-event terms with percentages: abdominal pain, 11.8%; back pain, 11.8%; hyponatraemia, 10.0%; pleural effusion, 10.0%; acute kidney injury, 5.9%; and anaemia, 5.9%. Those percentages describe the listed terms; they do not establish when events occurred, how long they lasted, or whether they were considered related to treatment. The denominator for each term is not stated in the information.

Deaths are listed as 76 in Cohort A, 18 in Cohort B, 15 in Cohort C, 2 in “Other,” and 111 overall. A death count is not, by itself, a count of deaths attributed to treatment. The registry information does not report causal attribution in these figures. In general, safety counts are best interpreted alongside exposure duration, event definitions, severity, timing, and attribution; those details are not reported in the registry record.

13. What the Result Does, and Does Not, Mean

The central result this study was designed to assess was the objective response rate among participants with FGFR2 rearrangements or fusions who received pemigatinib. The endpoint definition is explicit: a response required a best overall response of complete or partial response under RECIST version 1.1, with scans reviewed by an independent centralized radiological review committee. However, the numerical ORR and its 95% confidence interval are not reported in the registry record. The response result’s size and precision therefore cannot be described from the record information stated.

The study design permits descriptive statements about outcomes among enrolled participants and the recorded safety experience. It does not offer a randomized comparison against another treatment. A one-group response estimate cannot establish the outcome that would have occurred in the same population under a different treatment. Nor does the absence of a comparator show that there is no difference; it means the registry record does not contain the design needed for a direct randomized comparison.

The listed cohort categories help identify groups by FGF/FGFR alteration status, but they should not be treated as exchangeable arms. In general, biology-defined groups may differ in ways that affect outcomes. The registry’s cohort counts and safety summaries do not provide the efficacy estimates needed to compare these groups, and any group contrast would require careful interpretation even if outcome estimates were available.

ORR also answers a narrower question than time-to-event endpoints. A qualifying response does not establish its duration, delay in progression, or survival. Those outcomes appear in the list of secondary outcomes, but their numerical results are not reported in the registry record. Likewise, the serious adverse event and death counts describe recorded events, not causal effects of treatment.

In general, the meaning of a study result depends on its endpoint, analysis population, precision, missing-data handling, and design. Here, the endpoint rules and broad design are described, but the primary estimate, its interval, and several analysis details are not reported in the registry record. Conclusions should remain within that boundary.

14. Why This Trial Matters Statistically

FIGHT-202 illustrates why a trial’s design and endpoint definition must be read alongside any reported estimate. ORR is a concrete response measure: the registry specifies RECIST version 1.1 criteria and an independent centralized review of scans. Those details clarify what a counted response means and reduce reliance on an undefined impression of tumor change.

At the same time, the non-randomized, open-label, single-arm design places a limit on comparative inference. The study can describe outcomes among participants receiving pemigatinib, but it does not include a randomized control group. Cohorts based on alteration status are useful descriptive categories, yet they are not randomized comparisons. Distinguishing descriptive evidence from comparative evidence is a basic statistical discipline, particularly when interpreting subgroup outcomes.

The registry also demonstrates the importance of precision and denominator information. A response percentage without its confidence interval would not fully communicate uncertainty, and the primary response percentage itself is not reported in the registry record. Enrollment of 147 provides context but does not establish the denominator for the primary efficacy analysis. The efficacy evaluable population is defined in general terms, while the endpoint-specific count and missing-data handling are not reported.

Finally, safety summaries need careful wording. Serious adverse event counts, selected event-term percentages, and deaths provide information about events recorded during the study, but do not on their own establish causation. These statistical distinctions help readers understand what the registry can support: a description of the study structure and recorded safety counts, but not a numerical assessment of the primary response result or a controlled estimate of treatment benefit.

15. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

A central limitation of the available registry information is that it does not report the primary ORR estimate, the number of participants with a complete or partial response, or the 95% confidence interval. The endpoint definition and timeframe are given, but without those results its magnitude and precision cannot be assessed. The primary population’s endpoint denominator is also not reported.

The study is non-randomized, open-label, and single-arm. There is no concurrent randomized comparator, so the registry record does not support a controlled estimate of treatment effect. Cohorts are defined by FGF/FGFR alteration status and should not be interpreted as randomized arms. The registry lists 147 enrolled participants, but does not provide enough endpoint-specific information to establish how many contributed to the primary analysis.

The record does not report detailed handling of missing or unevaluable scans, the full statistical analysis plan, multiplicity procedures, interim analyses, or the original sample-size justification. These omissions limit evaluation of how analyses were selected, how uncertainty was handled, and how secondary outcomes should be interpreted. While a planning illustration is described, it is not the original trial design justification and no numerical planning outputs are stated.

Safety information is also limited in scope. The record lists participants affected by serious adverse events, selected serious-event terms, and deaths, but the figures do not establish causal attribution. Exposure duration, timing and duration of events, and further details needed for a full safety assessment are not reported in the registry record. Numerical results for the listed secondary efficacy outcomes are also not reported.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Efficacy and Safety of Pemigatinib in Subjects With Advanced/Metastatic or Surgically Unresectable Cholangiocarcinoma Who Failed Previous Therapy - (FIGHT-202) — NCT02924376.
• Registry record last updated 2025-08-14; status last verified 2025-08.
• Results first posted 2023-02-23.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

FIGHT-202 (NCT02924376) is recorded as a completed phase 2, open-label, non-randomized, multicenter study of pemigatinib for previously treated advanced or metastatic, or surgically unresectable, cholangiocarcinoma. Enrollment was 147. The study lists cohorts based on FGFR2 rearrangements or fusions, other FGF/FGFR alterations, and negative FGF/FGFR alteration status, plus an “Other” group.

The primary endpoint was ORR in participants with FGFR2 rearrangements or fusions, assessed under RECIST version 1.1 with independent centralized scan review. The numerical primary response result and its 95% confidence interval are not reported in the registry record. Serious adverse events affected 68 of 147 participants overall, and 111 deaths are listed; these counts do not establish treatment causation. The registry does not report the full statistical analysis plan, missing-data handling, or numerical secondary efficacy results.