Testing the Combination of Two Experimental Drugs MK-3475 (Pembrolizumab) and Interferon-gamma for the Treatment of Mycosis Fungoides and Sézary Syndrome and Advanced Synovial Sarcoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Testing the Combination of Two Experimental Drugs MK-3475 (Pembrolizumab) and Interferon-gamma for the Treatment of Mycosis Fungoides and Sézary Syndrome and Advanced Synovial Sarcoma (NCT03063632), a non-randomized phase 2 study. The primary outcome is Overall Response Rate (ORR).

2. Clinical Question

3. Trial Design

This was a phase II, non-randomized, parallel-group trial with no masking. The registry lists an enrollment of 28 and the status as completed. The trial began on 2017-12-14, and primary completion was recorded as 2021-04-08. The sponsor was the National Cancer Institute.

The two groups addressed different clinical populations and used different interferon gamma-1b schedules. Group I was for previously treated mycosis fungoides or Sézary syndrome; Group II was for advanced synovial sarcoma. Pembrolizumab was administered in both groups. As participants were not randomly allocated between treatment strategies, and the groups differed in disease and treatment schedule, comparisons between their outcomes would not isolate the effect of a treatment difference.

The Group I description specifies pembrolizumab intravenously on day 1, with cycles repeated every 3 weeks for up to 2 years in the absence of progression or unexpected toxicity. Interferon gamma-1b was given subcutaneously 3 times per week for 12 weeks, followed by a 3 weeks on and 3 weeks off schedule; further details are not reported in the registry record. Group II received pembrolizumab on day 1 and interferon gamma-1b once a week, with cycles repeated every 3 weeks for up to 2 years in the absence of progression or unexpected toxicity.

The registry states that participants in both groups received at least one dose of pembrolizumab and that 27 participants received at least one dose of interferon gamma. It also notes that not all participants reached the first disease assessment timepoint specified in the protocol. Further detail on the analysis populations and the consequences of missing that assessment is 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: Both Treatment Group 1 and Treatment Group 2 study participants received at least one dose of IV pembrolizumab on study, and 27 participants received at least one dose of Interferon gamma on study. While not all participants reached the first disease assessment timepoint specified in the protocol, all had documented disease status at the time of treatment or study discontinuation. Therefore, no study participants were excluded from ORR analysis due to missing data.

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.

Population: Patients do not have to complete the study in order to be included in the analysis population for the primary outcome measure.

6. Primary Endpoint

The registry lists the primary outcome as Overall Response Rate (ORR), measured over Up to 2 years.

The primary endpoint was overall response rate, defined as the number or proportion of participants whose best response was complete response or partial response. The registry says it was assessed using a binomial proportion, with best response at any timepoint used to determine response. The timeframe was up to 2 years.

Response criteria differed by disease group. In mycosis fungoides and Sézary syndrome, the Global Response Score evaluates skin, lymph nodes, internal organs and blood specimens. Complete response means disappearance of all clinical evidence of disease; partial response means regression of measurable disease. For synovial sarcoma, RECIST 1.1 was used. Under the registry description, complete response means disappearance of all lesions and no new lesion, while partial response means 30% or greater reduction in tumor size and no new lesions.

Response rate is a useful way to summarize whether measurable disease met specified criteria for improvement during assessment. It is a binary outcome at the participant level: a participant either meets the definition of response or does not. It does not, by itself, describe how long a response lasts, how quickly it occurs, or what happens to participants who do not respond. Those are distinct questions reflected among the listed secondary outcomes, including time to response, duration of response and survival outcomes.

The response counts and resulting rate are not reported in the registry record. Consequently, no estimate of the primary endpoint or its precision can be interpreted from the available record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary endpoint is a binary participant-level outcome. A participant’s best recorded response is classified according to the disease-specific criteria, and the overall response rate combines complete and partial responses. A binomial proportion is an appropriate basic summary for this type of outcome: it expresses the number meeting the response definition relative to the participants in the response analysis. The registry record describes this method but does not report the counts or the calculated proportion.

What does a confidence interval communicate?

In general, a confidence interval describes the precision of an estimate under the statistical method used. A narrower interval indicates greater precision, while a wider interval indicates more uncertainty about the underlying value. The interval should be interpreted alongside the number of participants assessed and the analysis approach. For a response proportion, the observed estimate alone does not convey its precision. No confidence interval for the primary endpoint is reported in the registry record, so the precision of the response estimate cannot be assessed from the available information.

Why do the response criteria matter?

The trial did not apply one common response definition to both groups. The Global Response Score for mycosis fungoides and Sézary syndrome uses assessments of skin, lymph nodes, internal organs and blood. RECIST 1.1 was used for synovial sarcoma, with the complete and partial response definitions described in the registry. These criteria give the endpoint a disease-specific meaning. A response rate calculated under one set of criteria should not automatically be compared with a rate calculated under another, because the measurement rules and disease context differ.

How should the groups be compared?

In general, random allocation helps balance measured and unmeasured characteristics between treatment groups, supporting a fairer comparison of treatment strategies. This trial was non-randomized, had no masking and assigned participants to parallel groups that represented different diseases. Group I and Group II are therefore not interchangeable comparison arms. Any difference between their outcomes could reflect differences in the participants, disease, assessment criteria, treatment schedule or other factors, rather than a treatment effect. The record does not report a randomized comparison or a control group.

What do the planning assumptions mean?

General sample-size planning asks how many participants or events would be needed to detect a specified effect with a chosen statistical method and assumptions. The planning illustrations associated with the results use a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation; the effect to detect is expressed as a stated fraction of the observed effect. These are planning assumptions, not the trial’s reported design rationale. The registry record does not report the protocol’s sample-size justification. Any planning illustration is not observed power and should not be read as a measure of the probability that the completed trial found a real effect. The registry entries do not include numerical planning outputs or an observed primary endpoint estimate.

Why can over-dispersion matter?

In general, a simple binomial model assumes that participants’ binary outcomes follow the model’s specified probability structure. If the actual outcomes vary more than the model assumes, over-dispersion can make standard errors too small and confidence intervals too narrow. That can make an estimate appear more precise than it is. Whether over-dispersion was assessed or addressed in this trial is not reported in the registry record. The stated use of a binomial proportion describes the primary endpoint summary, but does not by itself establish what additional modeling or uncertainty procedures were used.

What does a time-to-event analysis add?

Time-to-response, duration of response, progression-free survival and event-free survival are time-to-event outcomes, unlike the binary primary response classification. In general, time-to-event methods account for both when an event occurs and participants whose follow-up ends before the event is observed. Such methods depend on event definitions, follow-up information and assumptions about censoring. Although these outcomes are listed as secondary outcomes, their analysis methods, definitions beyond their names, and numerical results are not reported in the registry record. The record therefore does not support an interpretation of their estimates or uncertainty.

Finally, the adverse-event summaries are counts of participants affected in each group, alongside counts at risk. Counts help describe the reported safety experience, but do not alone establish causality, account for differing exposure time, or provide a comparison with an untreated or alternative-treatment group. Such conclusions require details and comparisons 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("Group I (Pembrolizumab, Interferon Gamma-1b)", "Group II (Pembrolizumab, Interferon Gamma-1b)")
n      <- c(16, 12)
events <- c(6, 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  Group I (Pembrolizumab, Interferon Gamma-1b)  16  6  37.5  15.2  64.6  18.5  61.4  18.4  61.7
ARM  Group II (Pembrolizumab, Interferon Gamma-1b)  12  0  0.0  0.0  26.5  0.0  24.2  0.2  24.7

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 8 outcome measures and 5 analyses, and names secondary outcomes that include adverse events, time to response, duration of response, progression-free survival, event-free survival and response duration beyond 12 months. It does not report in the available record how these outcomes were organized into analyses, which were confirmatory, or whether any adjustment for multiple testing was planned or used. When several outcomes or comparisons are examined, the chance of at least one apparently notable result can increase. In general, multiplicity procedures can address that issue, but the choice depends on the analysis plan. Whether such procedures were used is not reported in the registry record.

The record says that not all participants reached the first disease assessment timepoint specified in the protocol. The complete response-analysis population, the number without an assessment and the method for handling missing response data are not reported in the registry record. In general, excluding participants without an assessment, treating them as nonresponders, or using another approach can lead to different estimates. The analysis approach therefore needs to be known to interpret a response proportion properly.

The record reports that every participant in both groups received at least one dose of pembrolizumab and that 27 participants received at least one dose of interferon gamma. It does not specify the analysis population for every endpoint, whether all enrolled participants were analyzed for each outcome, or the handling of participants who stopped treatment or had incomplete follow-up. Interim analyses and the broader statistical analysis plan are not reported in the registry record. These gaps limit assessment of missing-data assumptions, analysis selection and the extent to which the posted results would reflect the originally planned analyses.

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: pericardial tamponade (8.3%), cardiac arrest (8.3%), abdominal pain (8.3%), vomiting (8.3%), nausea (8.3%).

The registry’s serious adverse-event summary gives 16 participants at risk in Group I, with 4 affected, and 12 at risk in Group II, with 5 affected. For other adverse events, it gives 16 at risk and 15 affected in Group I, and 12 at risk and 12 affected in Group II. These are descriptive counts of participants, not counts of all event episodes.

The record also lists 4 deaths in Group I and 6 in Group II. The registry entries available do not state whether deaths were considered related to treatment, what the causes were, or when they occurred. A death reported in a safety summary should not be assumed to have been caused by the study treatment.

Among the listed serious-event terms, pericardial tamponade, cardiac arrest, abdominal pain, vomiting, nausea and disease progression are each reported as 8.3%. The registry entries do not identify a group for each term or specify the denominators behind those percentages. They therefore should not be used to infer a group-specific risk or to compare the groups.

In general, safety counts need context about exposure, event definitions, timing, severity and attribution. The non-randomized design and distinct populations make direct group comparisons especially difficult. The available adverse-event summary describes reported events but does not establish that the combination caused them or that another treatment would have a different safety profile.

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

The registry record establishes that a phase II study of pembrolizumab and interferon gamma-1b was conducted in two separate treatment groups and that overall response rate was its primary endpoint. It also documents that the groups had different disease populations, response criteria and interferon gamma-1b schedules. Those details define what the trial set out to measure and the context needed to interpret any response results.

The available record does not report the response counts or overall response rate. It consequently does not show how many participants had a complete or partial response, how precise a response estimate was, or whether response differed from a comparator. A response rate, even when available, would describe the proportion meeting response criteria; it would not by itself establish how long responses lasted, how participants felt, or what would have happened under another treatment.

The study was non-randomized and had no control group identified in the registry record. Its two groups are not like-for-like comparators: one concerned previously treated mycosis fungoides or Sézary syndrome, and the other concerned advanced synovial sarcoma. The trial therefore cannot be interpreted as a randomized test of the combination against a control, nor can a difference between the groups be attributed to the treatment.

Safety counts offer a descriptive account of participants affected by adverse events, but do not establish cause. The registry’s death counts likewise do not, on their own, identify treatment-related deaths. Additional information about event attribution, exposure and timing is not reported in the registry record.

Overall, the registry entries support a description of the study’s design, endpoint definitions, exposure and reported safety counts. They do not support a conclusion about the combination’s efficacy or comparative safety. Missing response estimates, incomplete assessment participation and unreported analysis details limit how much can be inferred from the record.

14. Why This Trial Matters Statistically

This trial is statistically notable for asking a response question in two different disease settings within one study record. Although both groups received pembrolizumab and interferon gamma-1b, the populations, response criteria and interferon schedules differed. That design makes it important to interpret each group in its own clinical and measurement context rather than treating the groups as a direct comparison.

The endpoint illustrates why a clear operational definition matters. Overall response rate combines complete and partial response, but those categories are determined through disease-specific criteria. The Global Response Score and RECIST 1.1 do not represent a single identical measurement procedure. Statistical summaries are only interpretable when readers know what was classified as a response and which participants were assessed.

The record also highlights the distinction between trial completion and a complete account of results. It reports completed status and names the primary endpoint, yet the response counts and rate are not reported in the available record. Knowing the design and planned outcome does not substitute for the estimate, its precision and the analysis population needed to evaluate the primary result.

Finally, the study illustrates the limits of safety and efficacy interpretation without a randomized control. Counts can describe what occurred among treated participants, but they cannot alone establish treatment causality or comparative benefit and risk. Recognizing these limits is central to reading phase II evidence carefully and keeping descriptive findings separate from causal conclusions.

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 primary limitation for interpreting efficacy is that the registry record does not report the overall response rate or its component response counts. It also does not report a confidence interval or other measure of precision for that endpoint. The primary outcome therefore cannot be quantitatively evaluated from the registry entries.

The trial was non-randomized, unmasked and lacked a control group in the record. Its groups represented different diseases and used distinct response criteria and treatment schedules. Consequently, the groups do not provide a like-for-like comparison, and the design does not isolate the effect of the combination relative to another strategy.

The registry states that not all participants reached the first protocol-specified assessment. It does not report the full analysis population, the number who missed assessment, or how missing outcomes were handled. Those details are necessary to judge whether the response estimate could be affected by incomplete assessment or selection of participants with available data.

The Group I interferon gamma-1b schedule description ends after the 3 weeks on and 3 weeks off wording; further schedule details are not reported in the registry record. The relationship between the listed liposarcoma conditions and the treatment groups is also not reported in the registry entries. The record names multiple secondary outcomes but does not give their numerical results or detailed analysis methods. Interim analyses, multiplicity procedures and the protocol’s sample-size justification are not reported in the registry record.

Safety interpretation is limited by the absence of event attribution, timing and exposure details. The listed serious-event percentages have no group assignment or denominator in the registry entries, and the death counts do not identify cause or treatment relationship. These limitations preclude strong comparative or causal conclusions.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Testing the Combination of Two Experimental Drugs MK-3475 (Pembrolizumab) and Interferon-gamma for the Treatment of Mycosis Fungoides and Sézary Syndrome and Advanced Synovial Sarcoma — NCT03063632.
• Registry record last updated 2023-10-25; status last verified 2023-10.
• Results first posted 2022-08-30.
• 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

NCT03063632 was a phase II, non-randomized, unmasked, parallel-group trial sponsored by the National Cancer Institute. It enrolled 28 participants and is recorded as completed. The study began on 2017-12-14, with primary completion on 2021-04-08.

The trial evaluated pembrolizumab with interferon gamma-1b in separate groups for previously treated mycosis fungoides or Sézary syndrome and advanced synovial sarcoma. Its primary outcome was overall response rate, using disease-specific response criteria and a timeframe of up to 2 years. The record states that all participants received at least one dose of pembrolizumab and 27 received at least one dose of interferon gamma. Not all participants reached the first specified disease assessment.

Serious adverse events affected 4 of 16 participants at risk in Group I and 5 of 12 in Group II. The primary response results, analysis details and further information about the safety events are not reported in the registry record.