Nivolumab in Treating Patients With Primary Myelofibrosis, Post-Essential Thrombocythemia Myelofibrosis, or Post-Polycythemia Vera Myelofibrosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Nivolumab in Treating Patients With Primary Myelofibrosis, Post-Essential Thrombocythemia Myelofibrosis, or Post-Polycythemia Vera Myelofibrosis (NCT02421354), a na phase 2 study. The primary outcome is Objective Response Rate, Defined as Complete Remission + Partial Remission + Clinical Improvement.

2. Clinical Question

This question is descriptive as well as clinical: the primary endpoint counts participants who meet one of the specified response categories. It does not, by itself, establish how response compares with what would have happened without nivolumab. A single-group study can document outcomes among treated participants, but without a concurrent comparator it cannot separate treatment-associated changes from other influences on outcomes.

The registry record names laboratory biomarker analysis and quality-of-life assessment as additional interventions. It does not report results for those assessments in the posted outcomes. The primary response result is also not reported in the registry record, so the observed response count and any uncertainty around it cannot be described from the recorded results.

3. Trial Design

This is a phase 2, single-group study of nivolumab in people with primary myelofibrosis, post-essential thrombocythemia myelofibrosis, or post-polycythemia vera myelofibrosis. M.D. Anderson Cancer Center is listed as the sponsor. Enrollment is recorded as 8. The study began on 2015-05-14, and primary completion is recorded as 2018-04-13. Its status is terminated; the reason for termination is not reported in the registry record.

The design has no separate control arm: allocation is listed as not applicable, and masking is listed as none. The treatment description says that participants received nivolumab intravenously over 60 minutes once every 2 weeks for 8 doses, then once every 12 weeks. Treatment could continue for up to 4 years if disease progression or unacceptable toxicity did not occur. The registry record lists laboratory biomarker analysis and quality-of-life assessment in addition to nivolumab.

The primary outcome is an objective response count defined using complete remission, partial remission, and clinical improvement. The assessment time is 14 weeks after 8 doses of therapy. Responses were to be categorized according to the revised International Working Group-Myeloproliferative Neoplasms Research and Treatment and European LeukemiaNet consensus criteria for myelofibrosis.

Because the study has a single group and no masking, its design does not support a randomized, masked comparison. The registry record does not report a statistical analysis plan, a sample-size justification, interim analyses, or a reason for termination. It also does not report the primary response result. These absences limit the conclusions that can be drawn from the posted record; they do not establish why the study ended or what investigators observed.

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 Objective Response Rate, Defined as Complete Remission + Partial Remission + Clinical Improvement, measured over 14 weeks (after 8 doses of therapy).

The primary endpoint is the number of participants with an objective response, where response includes complete remission, partial remission, or clinical improvement. The assessment was scheduled for 14 weeks, after 8 doses of therapy. Categorization was to follow the revised consensus criteria named in the registry record for myelofibrosis.

A response endpoint of this kind turns clinical status into categories that can be counted. Combining complete remission, partial remission, and clinical improvement allows the endpoint to capture more than the most stringent response category. The count can describe how many participants met at least one of the listed criteria at the specified assessment time. It does not indicate, on its own, how durable a response was, whether participants improved in quality of life, or whether the treatment caused the outcome.

The endpoint is reported as a participant count, not as a time-to-event measure. Its interpretation depends on the criteria used, the assessment timing, and which participants had an evaluable assessment. Although the registry record specifies the criteria and timeframe, it does not report the response count, category-specific counts, missing assessment information, or an analysis population for this endpoint. Consequently, the actual observed response result and its precision are not reported in the registry record.

Such endpoints are used to describe disease response in a defined study population. For comparative claims, the endpoint would need to be interpreted alongside an appropriate comparison and a prespecified analysis. This record describes a single group and provides no primary endpoint analysis result, so it does not permit that comparative assessment.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods determine how study observations are translated into estimates and how uncertainty is assessed. In this record, the primary endpoint is a count of participants who meet a response definition. The registry record gives the endpoint definition, assessment timeframe, and criteria, but does not report an analysis plan or a result for that endpoint. The explanation below is general; it does not describe unreported methods as if they were used in this study.

What does a participant count describe?

A count states how many participants met a specified outcome definition. To interpret it, readers also need to know the relevant analysis population and how participants without an assessable outcome were handled. A rate or proportion can be informative when its denominator is clearly specified, but it still describes the study group rather than automatically establishing a treatment effect. The registry record gives the planned response endpoint as a participant count, but does not report the observed response count, denominator for that analysis, or analysis population.

The response definition combines complete remission, partial remission, and clinical improvement. Composite endpoints can gather multiple clinically meaningful categories into a single outcome. This can make the overall endpoint broader than any component alone, while making the combined result less specific about which kind of response occurred. Component results are useful for understanding that detail. The registry record does not report component counts or a primary response analysis result.

Why does the study design matter?

Randomization, masking, and a comparator address different potential sources of bias. Randomization aims to make comparison groups similar, on average, with respect to factors that could affect outcomes. Masking can reduce the influence of knowledge about treatment assignment on care, assessment, or reporting. A concurrent comparator provides a reference outcome under a different study condition. This study is recorded as single-group, with allocation not applicable and masking none; no separate comparator is listed.

As a general point, an outcome observed after treatment in a single group cannot by itself show what would have happened to the same participants under another condition. The endpoint may still describe outcomes among those studied, but causal interpretation is more limited. The record does not report a comparative analysis, so a treatment contrast cannot be estimated from the posted results.

What would an uncertainty interval add?

A confidence interval is a general statistical tool for expressing the precision of an estimated quantity under a specified model and sampling process. For a response proportion, an interval would help show the range of values compatible with the observed data under the method used. A wide interval signals limited precision; a narrow interval signals greater precision, conditional on the assumptions. An interval is not a guarantee that a particular value is correct, and it does not remove design limitations such as the absence of a comparator.

No confidence interval for the primary endpoint is reported in the registry record. Since the observed response result is not reported either, neither the magnitude nor the precision of that outcome can be assessed from the record. A calculated interval would require an outcome count and denominator, along with a stated method; inventing those inputs would not be appropriate.

What can a significance test establish?

In general, a significance test evaluates how compatible observed data are with a specified null hypothesis, given a statistical model. A p-value is not the probability that a treatment works, nor does it measure the size or clinical importance of an effect. Its meaning depends on the hypothesis, the analysis, and the assumptions. A test result should be interpreted alongside an effect estimate and its uncertainty, not in isolation.

No primary endpoint test or p-value is reported in the registry record, and the registry record does not report a statistical analysis plan. It is therefore not possible to identify the planned null hypothesis, whether the analysis was adjusted, or what inferential procedure investigators intended to use.

How should planning calculations be read?

Sample-size calculations are planning tools, not observed findings. The registry record describes planning illustrations based on a stated fraction of an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It also distinguishes calculations for time-to-event outcomes from those for other outcomes. These are assumptions for illustrations, not evidence that the trial had that design or that the calculations were its original sample-size justification.

The actual study is recorded as single-group, not as an equal-allocation, two-arm comparison. The registry record does not report a sample-size justification for the trial. Any planning illustration should therefore be understood as a hypothetical design calculation, not as observed power or proof that the study could detect a particular effect. The primary outcome result is not reported, so a planning calculation based on an observed effect cannot be independently interpreted from the posted results.

Why do analysis choices matter?

Analysis populations and missing-outcome rules can affect the interpretation of a clinical endpoint. For example, an analysis restricted to participants with an assessed outcome may differ from one that accounts for every enrolled participant. A general analysis plan should specify how incomplete assessments, withdrawals, and protocol deviations are handled before results are examined. The registry record does not report those choices or the number of participants assessed for the primary endpoint.

Likewise, a prespecified analysis reduces the risk that analytic choices are selected only after results are known. The registry record does not report whether the endpoint analysis was prespecified in a statistical plan, how any interim looks were handled, or whether any adjustment for multiple analyses was planned. These methods cannot be inferred from the endpoint description alone.

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(8)
events <- c(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  Treatment (Nivolumab)  8  0  0.0  0.0  36.9  0.0  32.4  0.3  33.6

11. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates several outcomes, response categories, timepoints, or statistical hypotheses. In general, examining many hypotheses can increase the chance of at least one apparently notable result if no adjustment or interpretive strategy is used. Prespecifying a primary outcome and defining how other results will be treated are common ways to make the inferential framework clearer. The registry record identifies one primary outcome measure, but it does not report a statistical analysis plan or a multiplicity strategy.

The primary outcome combines complete remission, partial remission, and clinical improvement. The record does not report separate results for those categories. It also lists laboratory biomarker analysis and quality-of-life assessment, but outcomes for those assessments are not reported in the posted results. The number of outcome measures recorded does not establish how many hypotheses were tested or whether any multiplicity adjustment was made.

Missing data require explicit handling because participants without an outcome assessment can affect both the analysis population and the estimated result. The registry record does not report missing primary endpoint assessments, reasons for missingness, a missing-data method, or how participants who discontinued treatment were treated in the analysis. It does not report the primary endpoint analysis population either. These details are necessary to assess whether the reported analysis, if available, would represent all enrolled participants or a subset.

No analysis results are reported in the registry record. Thus there are no reported primary endpoint estimates, tests, or analysis populations to evaluate for multiplicity or missing-data handling. The statistical analysis plan, any interim analyses, and the approach to analysis populations are not reported in the registry record.

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: anemia (25.0%), gastric hemorrhage (25.0%), abdominal pain (12.5%), acute kidney injury (12.5%), back pain (12.5%).

The registry lists safety figures for the nivolumab treatment group. Serious adverse events are recorded for 6 of 8 participants at risk, and other adverse events are recorded for 7 of 8 participants at risk. The record also lists 3 deaths. These figures describe what the registry reports for this group; they are not a comparison of event rates between treatment groups.

The serious adverse-event terms listed are anemia and gastric hemorrhage, each at 25.0%, and abdominal pain, acute kidney injury, back pain, and chest pain, each at 12.5%. The registry record does not provide additional context in these figures about event timing, duration, severity beyond the serious classification, or whether the events were judged related to nivolumab. The listed terms should therefore be read as recorded safety information, not as proof that treatment caused each event.

Adverse-event counts and percentages are descriptive. In general, a safety comparison requires attention to how events were defined and collected, the time participants were observed, and the characteristics of any comparison group. No separate control group is recorded for this study. The registry record also does not explain the reason for termination, so the status and the safety figures should not be treated as evidence that safety caused the trial to end.

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

The posted record establishes that a phase 2, single-group study of nivolumab was conducted in people with primary myelofibrosis or myelofibrosis following essential thrombocythemia or polycythemia vera. It identifies the intended primary endpoint as a count of participants with complete remission, partial remission, or clinical improvement at 14 weeks after 8 doses. It also reports safety figures for the treatment group. These registry entries describe the study and its recorded outcomes; they do not supply the missing primary response result.

Without the response count, the record cannot show how many participants met the endpoint, which response categories were represented, or how precisely an estimate could be made. There is no reported primary endpoint analysis or confidence interval. The outcome therefore cannot be characterized as frequent or uncommon on the basis of this record, and no conclusion about the magnitude of a response can be drawn.

The single-group design also matters. In general, observing an outcome after treatment is not the same as estimating what treatment changed compared with an alternative. The absence of a concurrent comparator means the record does not provide a treatment-versus-control effect estimate. No claim of causal benefit or lack of benefit follows from the design description alone.

The safety figures show that serious and other adverse events, as well as deaths, were recorded in the treatment group. They do not establish comparative safety or determine whether events were caused by treatment. A small group and the absence of a control group limit what can be inferred about how the recorded events compare with another treatment condition. The registry does not report a reason for termination, so the trial status does not explain the safety findings or the missing efficacy result.

In short, the record is useful for identifying the study population, treatment schedule, planned response assessment, and recorded safety events. It does not report the primary response outcome, a comparative analysis, or details needed to assess missing assessments and analysis populations. These limits should remain central when interpreting what the trial does and does not establish.

14. Why This Trial Matters Statistically

This study is statistically informative as an example of the distinction between a defined endpoint and a reported result. The registry record specifies an objective response endpoint, a response framework, and an assessment time. Those details help readers understand what investigators intended to count. But because the primary endpoint result is not reported, the planned measure cannot be converted into an estimate of observed response or its precision.

The single-group design is also central to interpretation. A single-group study can describe outcomes in the participants who received treatment. It cannot, without an appropriate comparison, directly quantify how outcomes differ from a separate treatment condition. This distinction is especially important when reading clinical response outcomes: a defined response category is not itself a comparative effect estimate.

The record also illustrates why safety results must be kept separate from efficacy conclusions. Serious adverse events, other adverse events, and deaths are recorded, but these descriptive figures do not supply a control-group comparison or establish event causality. The status of terminated does not fill that gap: the reason for termination is not reported in the registry record.

Finally, transparency about analysis populations, missing assessments, planned statistical methods, and multiplicity can shape how confidently readers interpret results. Those details are not reported in the registry record. Their absence does not establish that a particular method was or was not used; it means the posted information does not allow that assessment. For researchers and biostatisticians, this is a reminder to distinguish a trial's intended questions from the information available to answer them, and to keep design limitations separate from unreported results.

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 most consequential limitation of the posted record is that it does not report the primary endpoint result. Although the endpoint is defined as a participant count and its assessment time is specified, the response count, category-specific results, and uncertainty estimate are not reported. The central efficacy question therefore remains unanswered by the registry record.

The design is single-group, with no separate comparator and no masking. As a result, the posted information does not provide a randomized treatment contrast or a basis for attributing observed outcomes to nivolumab in comparison with another condition. The recorded enrollment is 8, which also means the study description concerns a limited group; the record does not provide information that would establish how representative that group is of a wider population.

Several details needed for a fuller statistical appraisal are not reported in the registry record: the reason for termination, the statistical analysis plan, a sample-size justification, interim analyses, the primary endpoint analysis population, the number of assessable endpoint observations, and methods for handling missing data. Results for laboratory biomarker analysis and quality-of-life assessment are not reported. These omissions prevent assessment of how the outcome analysis was conducted and whether other planned measures were completed.

The safety figures are descriptive and have no listed comparison group. The registry record does not give the context needed to establish event timing, treatment attribution, or the reason for the recorded deaths. The adverse-event information therefore cannot support a comparative safety conclusion. In combination, the missing primary outcome result, single-group design, and unreported analysis details sharply limit the conclusions available from this record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Nivolumab in Treating Patients With Primary Myelofibrosis, Post-Essential Thrombocythemia Myelofibrosis, or Post-Polycythemia Vera Myelofibrosis — NCT02421354.
• Registry record last updated 2019-05-15; status last verified 2019-04.
• Results first posted 2019-05-15.
• 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

The registry record describes a phase 2, single-group study of nivolumab sponsored by M.D. Anderson Cancer Center. It enrolled 8 participants with primary myelofibrosis, post-essential thrombocythemia myelofibrosis, or post-polycythemia vera myelofibrosis. The study began on 2015-05-14, primary completion is recorded as 2018-04-13, and the status is terminated.

The primary endpoint was the count of participants with complete remission, partial remission, or clinical improvement at 14 weeks after 8 doses, using the stated consensus criteria. The primary endpoint result and analysis results are not reported in the registry record. The record lists 6 of 8 participants affected by serious adverse events, 7 of 8 affected by other adverse events, and 3 deaths. No separate comparator is listed, and the reason for termination is not reported.