A Study to Compare Treatments for a Type of Kidney Cancer Called TFE/Translocation Renal Cell Carcinoma (tRCC): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Compare Treatments for a Type of Kidney Cancer Called TFE/Translocation Renal Cell Carcinoma (tRCC) (NCT03595124), a randomized phase 2 study. The primary outcome is Progression Free Survival (PFS).

2. Clinical Question

The study is a randomized phase 2 trial in TFE/translocation renal cell carcinoma, a rare kidney cancer type identified in the trial title and eligibility description. Its primary endpoint focuses on whether participants remain free from progression or death during the stated follow-up period. This is a time-to-event outcome, so the analysis accounts for the timing of events and for participants whose event status may not be observed through the full follow-up period.

The registry describes a three-arm design, but the axitinib-alone arm closed to accrual. It states that prospective patients were thereafter randomly assigned to the combination or nivolumab-alone arms. The posted primary endpoint results report these two arms. The record does not provide a direct randomized comparison estimate, such as a hazard ratio or a test of the difference between arms, so the arm-specific estimates should not be treated as a conclusive estimate of comparative treatment effect.

3. Trial Design

The registry describes a randomized, open-label, parallel-group phase 2 trial sponsored by the National Cancer Institute. Its official title identifies the comparison as axitinib plus nivolumab versus nivolumab alone across all age groups. The listed enrollment is 15. The registry lists a start date of 2019-01-08, a primary completion date of 2024-09-30, and a status of active, not recruiting.

Three arms appear in the record. Arm A combines axitinib and nivolumab; Arm B uses axitinib alone; and Arm C uses nivolumab alone. The Arm B description says that it was closed to accrual as of 1/23/2020, after which prospective patients were randomly assigned to Arms A or C. The posted primary endpoint results include Arm A and Arm C, not an axitinib-alone result.

For Arm A, axitinib is taken by mouth twice daily on days 1-28. Nivolumab is given intravenously on days 1 and 15 for participants younger than 18 years, or on day 1 for participants 18 years or older. Arm C uses the same age-dependent nivolumab schedule without axitinib. Both descriptions state that treatment repeats every 28 days for up to 26 cycles (2 years), subject to disease progression or unacceptable toxicity. The Arm B description also specifies axitinib twice daily on days 1-28 and the same cycle duration, but that arm is closed to accrual.

The record identifies the primary endpoint as the percentage of participants with PFS up to 1 year from study entry. It describes a Kaplan-Meier estimate and a 70% confidence interval. The posted endpoint analysis lists 6 participants in Arm A and 5 in Arm C. These are the arm counts attached to the reported endpoint, not a basis for inferring the number enrolled, treated, or included in every other analysis. Details such as the randomization procedure and allocation ratio are not reported in the registry record.

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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The data was analyzed with revised intention-to-treat approach. The following patients were excluded: Three (3) Arm B patients (as of Amendment 2A, the arm permanently closed to accrual and patients were randomly assigned to Arms A and C), One (1) Arm A patient who is ineligible for the study. 11 total patients were analyzed.

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 Progression Free Survival (PFS), measured over Up to 1 year from study entry.

The primary endpoint is progression-free survival (PFS) up to 1 year from study entry. PFS is defined as the time from randomization to disease progression or death due to any cause. The registry reports the one-year result as a percentage of participants and says that the Kaplan-Meier method is used to estimate it.

PFS combines progression and death into a time-to-event outcome. It can describe how long participants remain alive without documented progression, rather than counting only whether progression has occurred at a particular visit. Kaplan-Meier estimation is generally used when follow-up durations vary or some participants have not had an event by the end of their observed follow-up. In general, participants who have not had an event are treated as censored at the last time their event status is known; the registry record does not give further detail about censoring rules for this trial.

The endpoint is reported up to 1 year from study entry, while its definition measures time from randomization. That wording is reproduced from the registry record; whether the two time origins were operationally identical is not reported in the registry record. The endpoint offers an arm-specific estimate at the stated time point, but does not by itself establish whether the arms differ. A comparative estimate and its uncertainty would be needed to quantify the randomized contrast. The record gives an estimate and interval for Arm A, but the Arm C interval is not reported.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome uses a time-to-event method, so the statistical question is not simply how many participants had an event by the end of follow-up. It concerns the estimated probability of remaining free from progression or death through the specified time point, taking account of the timing of events and the observed follow-up. The registry names the Kaplan-Meier method for estimating one-year PFS.

What does Kaplan-Meier estimation do?

In general, Kaplan-Meier estimation updates the estimated event-free proportion at each observed event time. Participants who remain under observation contribute information until an event or until their follow-up ends. If a participant has no event recorded by their last observation, that participant is censored rather than treated as having experienced the event. This method can use unequal follow-up durations, provided the assumptions about censoring and observation are reasonable.

The posted record does not detail the censoring rules, how progression was assessed, how missing follow-up was handled, or whether any special rules were used for deaths or other events. Those operational details are not reported in the registry record. A time-to-event estimate should therefore be interpreted in light of the endpoint definition and the record’s stated analysis method, without assuming additional procedures.

How should the confidence interval be read?

Arm A has a one-year PFS estimate of 33% and a 70% confidence interval from 18.32% to 60.64%. In general, a confidence interval is a range generated by a statistical procedure to reflect uncertainty in an estimate. Under repeated use of the same method and assumptions, the stated proportion of such intervals would cover the underlying quantity. It is not a probability statement about the fixed underlying value for this particular trial.

The endpoint description says the 70% confidence interval was selected to match the one-sided significance level of alpha = 0.15 in the protocol analysis plan. It is therefore important not to compare its width casually with intervals at other confidence levels. A lower confidence level generally yields a narrower interval under the same method, but does not make the estimate more precise in the sense of adding information. The interval remains wide, and the arm count is 6.

For Arm C, the estimate is 0%, but the registry does not report an interval. An interval cannot be inferred from the point estimate alone. In particular, a reported zero does not establish that the underlying PFS probability is exactly zero; the number of participants and the uncertainty method matter. The record does not explain why an interval is absent or provide enough summary information to reconstruct one.

Why is a between-arm test not available?

The trial is randomized, but the posted outcome results are arm-level estimates. They do not include a comparative effect estimate, a confidence interval for the arm contrast, or a test result. The record lists zero analyses for the posted outcomes and indicates that values are missing or not estimable for a two-arm comparison of this binary-percentage endpoint. Consequently, the reported values should not be converted into a claim about a treatment difference. A valid comparison could require individual time-to-event data or a suitably reported comparative analysis, neither of which is present in the registry record.

Randomization is intended, in general, to support fair comparisons by balancing prognostic factors on average. It does not guarantee balance in a small trial, nor does it eliminate uncertainty. The endpoint analysis counts are small, and the registry does not report baseline balance or adjusted results. No adjustment for prognostic factors should be assumed.

What do the planning illustrations represent?

Any sample-size calculations associated with these results are planning illustrations based on a fraction of the observed effect, not observed power. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. For time-to-event outcomes, the stated calculation approach concerns the total number of events; other outcomes use participants per arm. These illustrations do not describe the trial’s original design justification, and that justification is not reported in the registry record.

In general, sample-size planning asks how much information a future study might need to detect a chosen effect under specified assumptions. It cannot change the information collected in this trial, and it does not show the probability that this trial detected an effect. The record does not provide planning illustration values in the endpoint results, so no sample-size target or required enrollment can be inferred from the assumptions alone.

What does over-dispersion mean in general?

Over-dispersion is a general issue for count data when variation exceeds what a model assumes. It can make standard errors too small if ignored. The primary endpoint is a time-to-event outcome reported as a percentage, and the registry does not describe an over-dispersion model or a count-data analysis for PFS. Over-dispersion should not be assumed to be part of this endpoint analysis.

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("Arm A (Axitinib, Nivolumab)", "Arm C (Nivolumab)")
n      <- c(6, 5)
events <- c(2, 0)
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  Arm A (Axitinib, Nivolumab)  6  2  33.3  4.3  77.7  9.7  70.0  9.9  71.0
ARM  Arm C (Nivolumab)  5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9

11. Multiplicity, Missing Data and Analysis Populations

The registry lists secondary outcomes of overall survival (OS), overall response rate (ORR), and the percentage of patients experiencing adverse events. It also lists 7 outcome measures, while 3 secondary outcomes are posted. The record does not describe how testing across multiple outcomes was controlled or prioritized. A multiplicity adjustment or a prespecified hierarchy of tests is not reported in the registry record. Without that information, nominal findings across several outcomes should not automatically be treated as confirmatory evidence.

The endpoint description identifies a one-sided significance level of alpha = 0.15 in the protocol analysis plan and says the 70% confidence interval was chosen to match it. The record does not provide details of any other inferential testing plan or explain how that endpoint-level choice relates to analyses of secondary outcomes. The posted results contain no analyses for the primary endpoint comparison.

For the primary endpoint population, the record states that a revised intention-to-treat approach was used. It describes exclusion of three Arm B patients because the arm permanently closed to accrual after Amendment 2A and patients were randomly assigned to Arms A and C, and exclusion of one Arm A patient who was ineligible. The population description is truncated after “11 to”, so the remaining explanation is not reported in the registry record. The precise analysis-population definition and its relation to all randomized participants are therefore not fully clear.

Missing-data procedures, censoring details, and sensitivity analyses are not reported in the registry record. For PFS, incomplete follow-up is relevant because the method depends on participants’ observed event times and censoring. The posted summary does not show how many observations were censored, how missing progression assessments were handled, or whether alternative assumptions were tested. These omissions limit assessment of robustness.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: disease progression (100.0%), death nos (60.0%), anorexia (20.0%), dehydration (20.0%), lung infection (20.0%).

The registry reports serious adverse event figures separately from other adverse events. For serious adverse events, the listed affected counts are 4 of 6 in Arm A, 2 of 2 in Arm B, and 5 of 5 in Arm C. For other adverse events, the corresponding counts are 2 of 6, 0 of 2, and 1 of 5. These figures are descriptive counts for the groups shown in the safety record; they are not a randomized comparison and should not be used alone to rank treatment safety.

The record also lists deaths as 1 in Arm A, 2 in Arm B, and 4 in Arm C. It does not specify in this summary whether these deaths were treatment-related or provide their causes, so the figures should not be interpreted as treatment-attributed deaths. The listed serious adverse event terms include Disease progression at 100.0%, Death NOS at 60.0%, and Anorexia, Dehydration, Lung infection, and Nausea at 20.0% each. These are the percentages displayed for the listed terms; the registry summary does not give enough context to interpret them as comparative risks.

Safety counts can be affected by the number of participants at risk, follow-up duration, and the rules for recording events. The arm sizes differ, and Arm B was closed to accrual. The registry record does not provide enough detail in this summary to determine whether participants had comparable exposure time or how recurrent events were counted. These figures describe recorded events and affected participants, not causal effects.

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

The posted results establish that the registry reports an arm-specific one-year PFS estimate of 33% for axitinib plus nivolumab, with a 70% confidence interval from 18.32% to 60.64%. It reports 0% for nivolumab alone, without a confidence interval. These are descriptive estimates for the participants included in the endpoint analysis, whose listed arm counts are 6 and 5, respectively.

The results do not establish the size of a treatment difference. There is no posted comparative estimate, confidence interval for a contrast, or test result for the two arms. A difference between point estimates is not a substitute for a randomized comparison that accounts for time-to-event follow-up and uncertainty. The missing interval for Arm C also means that the precision of that estimate cannot be read from the registry summary.

The results also do not establish how well either treatment performs in all people with this cancer. The analyzed groups are small, the analysis population excludes some participants as described in the record, and the registry does not provide a full account of missing follow-up or censoring. Generalizing from these estimates requires caution. The figures should be understood as posted trial results, not as a definitive statement about outcomes for an individual.

The trial record lists secondary outcomes and safety figures, but the summary does not provide comparative analyses for them. Adverse event counts and death counts are descriptive and do not by themselves establish that a treatment caused an event or death. The record also does not establish that the observed estimates are reproducible in another study.

In general, a randomized trial can support causal comparisons when the design, follow-up, analysis, and reporting allow the treatment groups to be compared appropriately. Here, the record documents randomization and reports arm-specific estimates, but the comparative analysis needed to interpret the contrast is not posted. The main conclusion is therefore limited to describing those estimates and their uncertainty where reported.

14. Why This Trial Matters Statistically

This trial is statistically notable because it studies a rare kidney cancer subtype and includes participants across all age groups, while the posted endpoint analysis contains small arm counts. In small studies, each participant’s event status can have a strong influence on an estimate. That makes uncertainty, transparent reporting, and careful separation of descriptive results from comparative inference especially important.

The registry’s chosen primary endpoint is time-to-event PFS, analyzed with Kaplan-Meier estimation. This approach is designed to use the timing of progression or death and accommodate follow-up that may end before an event is observed. The record also specifies a 70% confidence interval, linked to a one-sided significance level in the protocol analysis plan. Readers need to understand that interval level when interpreting the displayed range; it is not directly interchangeable with a confidence interval at another level.

The treatment arms changed during the trial: the axitinib-alone arm closed to accrual, and prospective patients were thereafter assigned to the combination or nivolumab alone. This makes clear reporting of analysis populations and comparisons important. The record describes a revised intention-to-treat approach and exclusions, but its population description is incomplete. It also lacks a posted comparative analysis for the two reported arms.

These features make the record useful for understanding how a trial can yield arm-specific estimates without enough posted information to draw a clear comparative conclusion. The distinction between a point estimate, its uncertainty, and an estimated treatment contrast is central to interpreting clinical research. Here, that distinction prevents the arm-level figures from being overstated.

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.

The endpoint analysis is based on small reported arm counts: 6 for Arm A and 5 for Arm C. Estimates from small groups can be unstable, and the Arm A confidence interval is wide. The registry gives no interval for the 0% estimate in Arm C, so its precision cannot be assessed from the posted results.

The registry does not post a comparative estimate or inferential test for PFS. The displayed arm-specific results therefore do not quantify the randomized contrast. The axitinib-alone arm was closed to accrual, and its endpoint estimate is not included in the posted primary results. The record describes subsequent assignment to Arms A and C but does not provide a full account of how this design change affected the analysis.

The revised intention-to-treat population description is incomplete, ending with “11 to”. It identifies exclusions but does not fully clarify the final population definition. Details of missing-data handling, censoring, progression assessment, follow-up duration by arm, and sensitivity analyses are not reported in the registry record. These details matter for evaluating a Kaplan-Meier estimate.

The open-label design may affect assessment or reporting of outcomes, although the registry summary does not describe whether outcome assessment was blinded or how progression was determined. Baseline characteristics and adjusted analyses are not included in the registry entries. Safety counts are descriptive, and the record does not provide enough detail about exposure time, recurrent events, or attribution to treatment. Together, these limitations call for a cautious, descriptive reading rather than a claim of comparative benefit or harm.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study to Compare Treatments for a Type of Kidney Cancer Called TFE/Translocation Renal Cell Carcinoma (tRCC) — NCT03595124.
• Registry record last updated 2026-10-22; status last verified 2026-05.
• Results first posted 2025-10-10.
• 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

NCT03595124 is a randomized, open-label, parallel-group phase 2 trial sponsored by the National Cancer Institute. It studies axitinib plus nivolumab and nivolumab alone for TFE/translocation renal cell carcinoma. Axitinib alone was also listed as an arm but closed to accrual as of 1/23/2020. The registry lists enrollment of 15 and a status of active, not recruiting.

The primary endpoint is PFS up to 1 year, defined as time from randomization to progression or death due to any cause and estimated using the Kaplan-Meier method. The combination arm has a posted estimate of 33% (70% confidence interval 18.32% to 60.64%); nivolumab alone has a posted estimate of 0% without a reported interval. The registry does not post a comparative analysis for these arms.

The record lists OS, ORR, and adverse events as secondary outcomes. Safety figures include serious and other adverse events and deaths by arm. The available summaries do not provide full details of missing-data handling, the final analysis-population definition, or comparative safety and efficacy analyses.