Nivolumab for Relapsed, Refractory, or Detectable Disease Post Chimeric Antigen Receptor T-cell Treatment in Patients With Hematologic Malignancies: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Nivolumab for Relapsed, Refractory, or Detectable Disease Post Chimeric Antigen Receptor T-cell Treatment in Patients With Hematologic Malignancies (NCT04205409), a na phase 2 study. The primary outcome is Best Overall Response Rate (ORR).

2. Clinical Question

3. Trial Design

This phase II study evaluated nivolumab in a single group of participants with relapsed, refractory, or detectable hematologic malignancy after CAR T-cell treatment. The University of Washington is listed as the sponsor. The registry lists enrollment of 20 and the study status as completed. The recorded start date is 2020-06-05, and the primary completion date is 2024-09-20.

Participants received nivolumab intravenously over 30 minutes on day 1, with cycles repeating every 28 days in the absence of disease progression or unacceptable toxicity. The registry lists no masking. Because this was a single-group study, there is no concurrent control group against which to compare response, survival, or safety outcomes. The listed allocation value is not applicable; it does not indicate random assignment.

The eligible disease categories span several hematologic malignancies, including chronic lymphocytic leukemia, several lymphoma types, and plasma cell myeloma. The primary response assessment was to use disease-specific criteria: International Myeloma Working Group response criteria for multiple myeloma, Lugano Criteria for non-Hodgkin lymphoma, and International Workshop on Chronic Lymphocytic Leukemia Criteria for chronic lymphocytic leukemia. The record does not report disease-specific enrollment counts or further details about how results were to be combined across these groups.

The primary endpoint was best overall response rate, with a stated assessment timeframe of up to 1 year and 8 months. Secondary outcomes included overall survival, progression-free survival, duration of response, and incidence of adverse events. The registry record does not report a protocol rationale for the enrollment target, a detailed statistical analysis plan, or the methods used to analyze each secondary outcome. The absence of a comparator means the study can describe outcomes among treated participants, but cannot by itself estimate a treatment effect relative to another approach.

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

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 Best Overall Response Rate (ORR), measured over Up to 1 year and 8 months.

Best overall response rate (ORR) is intended to describe the proportion of participants who achieve a qualifying response during the assessment period. The registry describes the primary measure as a count of participants and states that response was to be assessed using disease-specific criteria: International Myeloma Working Group response criteria for multiple myeloma, Lugano Criteria for non-Hodgkin lymphoma, and International Workshop on Chronic Lymphocytic Leukemia Criteria for chronic lymphocytic leukemia. Applying disease-specific criteria reflects that response is evaluated differently across these conditions.

The registry states that ORR would be estimated with a 95% exact binomial confidence interval. In general, a confidence interval conveys the statistical uncertainty around an estimated response rate; it does not guarantee that the true response rate lies within its limits. An exact binomial interval is a method for estimating uncertainty around a count-based response result, particularly when the number of participants is limited. The registry record does not report the response count, the estimated ORR, or its confidence interval, so the primary endpoint result is not reported in the registry record.

The stated timeframe is up to 1 year and 8 months. The record does not give further detail about response confirmation, assessment schedules, or how participants without an evaluable response were handled. Those details matter when interpreting a response endpoint because they define which observations contribute to the estimate and when a response qualifies.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The methods described in the registry record concern a single-group phase II study and a count-based response endpoint. The primary outcome is best overall response rate (ORR), described as a participant count, with response determined using criteria specific to the underlying disease. The registry says ORR would be estimated and accompanied by a 95% exact binomial confidence interval. It does not report an ORR estimate or interval, and the record lists zero analyses.

What does a response-rate estimate represent?

In general, a response rate summarizes the number of participants meeting a response definition in relation to the participants evaluated under the analysis rules. The definition of response is substantive as well as statistical: it depends on the disease-specific criteria stated in the registry record. A response count without the corresponding analysis population and handling rules may not be enough to reproduce an estimate. Those details, including rules for participants with missing or unevaluable assessments, are not reported in the registry record.

The record describes the measure as a count of participants and says ORR will be estimated. A count and a rate are related but not identical summaries: the count is the number meeting the criteria, whereas a rate also depends on the denominator used. Since neither a response count nor an estimated rate is posted, no numerical response interpretation is available. Enrollment of 20 is the listed study enrollment; it should not automatically be treated as the denominator for the response endpoint because the endpoint analysis population and evaluability rules are not specified.

What does the exact binomial confidence interval add?

In general, a confidence interval communicates uncertainty around an estimate. For a binary outcome such as whether a participant meets a response definition, an exact binomial method uses the observed count and denominator to estimate a range of values compatible with the data under the method's assumptions. With limited observations, the interval can be wide, so an estimate should be read together with its interval rather than as a precise fixed value. The registry specifies a 95% exact binomial interval, but gives neither endpoint of the interval nor the data needed to reconstruct it.

A confidence interval is not a guarantee that the interval contains the underlying response rate, nor does it address every source of uncertainty. It does not correct for a nonrandomized design, differences among disease groups, missing assessments, or limitations in outcome ascertainment. It also does not substitute for a comparator when the aim is to estimate a difference between treatment strategies.

Why does the single-group design matter?

In general, a single-group study estimates outcomes within the enrolled group. Without a concurrent comparison group, differences between observed outcomes and what might have happened under another treatment cannot be isolated. Randomization, when used in a comparative trial, helps balance both measured and unmeasured prognostic factors across groups. This study is recorded as single group, with allocation not applicable and no masking; there is no like-for-like control comparison in the registry record. The design is therefore suited to describing the treated cohort, not to directly estimating a randomized treatment contrast.

Masking is recorded as none. In general, masking can reduce the influence of expectations on assessments or reporting, although its relevance differs by outcome and study design. For a disease-specific response assessment, knowing treatment assignment could potentially matter to subjective components of assessment. The registry record does not describe safeguards or independent review procedures, so no further claim about response assessment bias is warranted.

How should planning illustrations be understood?

The sample-size planning assumptions described for the statistical illustrations are not the trial's own design justification. They specify a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect to detect expressed as a fraction of the observed effect. For time-to-event illustrations, the stated quantity is total events using the Schoenfeld approach; other illustrations are expressed as participants per arm. These are planning assumptions, not observed outcomes or evidence of the study's actual design.

That distinction is especially important because the trial itself is a single-group study, while the stated illustrations assume allocation between two arms. Such an illustration is not a description of participants actually randomized in this study. It should not be interpreted as observed power, and it cannot establish that the completed study was adequately sized for its primary endpoint. The registry record does not report the trial's own sample-size justification. In general, planning calculations depend on the assumed effect, variability or event pattern, significance level, allocation, and analysis method; changing those assumptions changes the resulting planning requirement. No additional planning result or effect estimate is reported in the registry record.

What cannot be inferred from the posted statistical information?

The registry lists overall survival, progression-free survival, duration of response, and incidence of adverse events as secondary outcomes, but does not post their numerical analyses in the record. It does not report estimates, confidence intervals, hypothesis tests, or detailed analysis methods for those outcomes. The record also does not report interim analyses, adjustment for multiple outcomes, missing-data procedures, or the analysis populations. Those absences limit assessment of how the planned statistical methods were implemented and whether outcome estimates could be reproduced.

General statistical interpretation therefore needs to remain proportionate to the information recorded: distinguish planned endpoint methods from posted results, distinguish descriptive counts from comparative effects, and consider uncertainty and design limitations together. Here, the response method is described, but the response result itself is not reported in the registry record.

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("Treatment (Nivolumab)")
n      <- c(20)
events <- c(3)
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  Treatment (Nivolumab)  20  3  15.0  3.2  37.9  5.2  36.0  5.4  36.3

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome, best overall response rate, and several secondary outcomes: overall survival, progression-free survival, duration of response, and incidence of adverse events. It does not report a procedure for controlling multiplicity across outcomes, analyses, or disease groups. In general, testing or estimating many outcomes can increase the chance of observing apparently notable findings by chance; a prespecified hierarchy or adjustment can help define how such results should be interpreted. No such approach is reported in the registry record, so it is not possible to determine whether multiplicity was addressed.

The record does not report missing-data rules, the handling of participants without an evaluable response, or the analysis populations for the primary or secondary outcomes. In general, missing outcome information can affect both the numerator and denominator of a response-rate estimate, and different handling assumptions can lead to different interpretations. The stated enrollment of 20 does not establish that all participants contributed to every endpoint.

The registry record also does not report interim analyses, stopping rules, or a detailed statistical analysis plan. No specific conclusions about these procedures should be inferred from the phase designation or from the listed endpoint methods. The record lists zero analyses, and it does not post numerical efficacy results. These gaps mean that the reported endpoint description cannot be used to assess implementation of a multiplicity strategy, missing-data plan, or analysis-population definition.

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: sepsis (15.0%), anemia (10.0%), febrile neutropenia (10.0%), atrial fibrillation with rvr (5.0%), cardiac arrest (5.0%).

The treatment group lists 20 participants at risk for serious adverse events and 10 affected, alongside 20 at risk for other adverse events and 18 affected. The record also lists 13 deaths. These are descriptive counts for the nivolumab group, not comparative risk estimates. The record does not state that each event was caused by nivolumab, and the death count alone does not identify causes or timing.

The most frequently listed serious adverse event terms include sepsis (15.0%), anemia (10.0%), febrile neutropenia (10.0%), atrial fibrillation with RVR (5.0%), cardiac arrest (5.0%), and chest pain (5.0%). These percentages describe the posted event terms; they should not be added together or treated as mutually exclusive categories. A participant may have experienced more than one event, and the record does not provide enough detail to determine overlap, duration, outcome, or attribution.

In general, safety counts are best interpreted with information about event definitions, seriousness, timing, exposure, and investigators' assessment of relatedness. The registry record does not report those details for the listed counts. With no control group, the figures cannot establish how event frequency compares with another treatment or with the expected background in a similar population.

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

The registry record establishes that a completed phase II, single-group study evaluated nivolumab in people with relapsed, refractory, or detectable hematologic malignancy after CAR T-cell treatment. It describes the dosing schedule, the disease-specific response criteria, the primary endpoint, and selected adverse-event counts. It does not report the primary response result. As a result, the posted information does not show the number of participants with a best overall response, the ORR estimate, or its 95% exact binomial confidence interval.

Even if a response estimate were available, a single-group result would describe outcomes among participants in the study rather than identify the effect of nivolumab compared with another treatment. In general, outcomes can reflect factors other than the intervention, including participant characteristics, disease type, prior treatment, subsequent care, and the course of the disease. Without a concurrent comparator, those influences cannot be separated through a randomized contrast.

The range of listed malignancies also matters. Disease-specific response guidelines were planned, but the registry record does not provide condition-specific participant counts or outcomes. A pooled summary, if one were available, would need to be interpreted in light of that mixture and the distinct assessment criteria. The record does not provide enough information to characterize response separately for any listed diagnosis.

Safety counts offer descriptive information, but do not establish causation. The reported serious adverse events, other adverse events, and deaths should not be interpreted as treatment-attributable effects without the missing context on event timing, causes, relatedness, and an appropriate comparison. The listed secondary endpoints are not accompanied by posted numerical analyses.

Accordingly, the trial description and safety counts can inform an understanding of what was studied and what events were recorded, but the registry record does not permit a conclusion about the observed primary response outcome or a comparative treatment effect. This is a limit of the available record, not evidence that an outcome was absent.

14. Why This Trial Matters Statistically

Statistically, this study illustrates the difference between a clearly named endpoint and a reported endpoint result. The registry specifies best overall response rate, disease-specific response criteria, and a 95% exact binomial confidence interval. Yet the response count, estimate, and interval are not reported in the registry record. Without those values, readers cannot assess the magnitude or precision of the primary outcome, regardless of the study's enrollment.

The single-group design also makes the estimand important. A response rate among treated participants is a descriptive quantity; it is not the same as a comparative effect against another treatment. The distinction is consequential in a population spanning multiple hematologic malignancies and with disease-specific response assessments. A pooled estimate may not describe every disease group in the same way, and the record does not report the disease-specific enrollment or results needed to evaluate that issue.

The registry's safety figures demonstrate another basic interpretive point: event counts and percentages are not causal comparisons. They describe recorded events within the treatment group. Without a control group and context on attribution and timing, the figures cannot determine how much risk is associated with treatment relative to another course of care.

Finally, the difference between planning assumptions and trial design matters. The statistical planning illustrations described assume two-arm allocation, while the trial is recorded as single group. Those planning assumptions cannot be mistaken for actual randomization or the study's own sample-size justification. Together, these features make the record useful for understanding the study's structure and planned endpoint, while underscoring the limits created by absent efficacy analyses and lack of a comparator.

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 main trial-specific limitation is the absence of a posted primary response result. The registry record does not report the response count, ORR estimate, or 95% exact binomial confidence interval. It lists zero analyses, and the secondary outcomes are not accompanied by numerical analyses. The description of an intended analysis is therefore not evidence of the outcome observed.

The single-group design has no randomized control arm, so observed outcomes cannot be interpreted as a comparative treatment effect. Masking is listed as none. The study includes several hematologic malignancies, but disease-specific participant numbers and results are not reported. Although disease-specific response criteria are named, the record does not describe how results across diagnostic groups would be combined or how differences in assessment might affect an overall summary.

The registry lists enrollment of 20 but does not report the primary analysis population, the number evaluable for response, or missing-data rules. The stated primary endpoint timeframe is up to 1 year and 8 months, but the record does not give further details on assessment timing, response confirmation, or censoring. It also does not report the trial's sample-size justification, multiplicity procedures, interim analyses, or a detailed analysis plan.

The safety information is limited to group counts and selected serious adverse event terms and percentages. The record does not provide context on timing, event overlap, duration, seriousness criteria, attribution, or cause of death. With no comparison group, safety counts cannot establish relative risk. These limitations concern what can be interpreted from the registry record and do not imply that unreported methods or outcomes were not collected.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Nivolumab for Relapsed, Refractory, or Detectable Disease Post Chimeric Antigen Receptor T-cell Treatment in Patients With Hematologic Malignancies — NCT04205409.
• Registry record last updated 2026-06-26; status last verified 2026-06.
• Results first posted 2025-11-03.
• 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

NCT04205409 is a completed phase II, single-group study sponsored by the University of Washington. It enrolled 20 participants with relapsed, refractory, or detectable hematologic malignancy after CAR T-cell treatment. Participants received nivolumab intravenously on day 1, with cycles repeating every 28 days in the absence of progression or unacceptable toxicity.

The primary outcome was best overall response rate, assessed by disease-specific criteria, with a planned 95% exact binomial confidence interval. The record does not report the response count or estimate. Secondary outcomes included overall survival, progression-free survival, duration of response, and incidence of adverse events. The treatment group lists 10 participants affected by serious adverse events, 18 affected by other adverse events, and 13 deaths; these are descriptive counts, not comparative or necessarily treatment-attributed effects.

The study has no control group, and the registry record does not report detailed analysis populations, missing-data procedures, multiplicity handling, or numerical efficacy analyses.