Ruxolitinib Phosphate, Paclitaxel, and Carboplatin in Treating Patients With Stage III-IV Epithelial Ovarian, Fallopian Tube, or Primary Peritoneal Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Ruxolitinib Phosphate, Paclitaxel, and Carboplatin in Treating Patients With Stage III-IV Epithelial Ovarian, Fallopian Tube, or Primary Peritoneal Cancer (NCT02713386), a randomized phase 1, phase 2 study. The primary outcome is Dose-limiting Toxicities (Phase I).

2. Clinical Question

The registry title and brief description frame the broader question as whether adding ruxolitinib to paclitaxel and carboplatin can be studied in patients with advanced disease. The phase II arms align with that comparison: Arm I received paclitaxel and carboplatin, while Arm II received ruxolitinib, paclitaxel, and carboplatin. Surgery is also listed as an intervention, and several secondary outcomes concern surgery or tumor response.

These questions are related but not interchangeable. Dose-limiting toxicity is assessed in phase I to inform evaluation of the regimen’s tolerability at dose levels. The phase II outcomes address matters such as adverse events, resection, and pathological response. A result from the phase I dose-level cohorts does not itself answer whether the phase II combination performs differently from chemotherapy alone.

3. Trial Design

NCT02713386 is a completed phase I/II study sponsored by NRG Oncology. The registry lists enrollment as 147 and identifies the allocation as randomized, the masking as none, and the model as parallel. The registry does not report the randomization procedure, allocation ratio, or which phase-specific participants were included in any particular analysis.

The two treatment arms in phase II were paclitaxel and carboplatin alone (Arm I), and ruxolitinib with paclitaxel and carboplatin (Arm II). The phase I results instead identify two dose-level cohorts of Arm II: DL1 and DL2. The registry does not state the actual ruxolitinib doses in the posted arm descriptions. It would therefore be inappropriate to infer the doses or treat the phase I cohorts as randomized arms.

The official title describes front-line neoadjuvant and post-surgical therapy. Conventional surgery is listed among the interventions. The total gross resection outcome was assessed at surgery, which the registry says occurred within 6 weeks after completion of cycle 3, as soon as nadir counts permitted and the investigator deemed surgery safe. Complete pathological response was assessed at tumor reductive surgery or biopsy after 3 cycles (i.e. after 63 days).

The primary outcome is phase I dose-limiting toxicity over 42 days (2 cycles). Secondary outcomes span adverse events, surgery, pathological response, progression-free survival, and overall survival. The registry lists the study as completed, but completion status does not establish that every listed outcome has a posted result or that all analyses are available. The statistical plan, sample-size justification, and analysis populations 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.

Phase 1 - Dose Level 1

Phase 1 - Dose Level 2

Phase 2

Analysis population for the primary outcome, as defined in the registry: Eligible

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: Eligible and evaluable

6. Primary Endpoint

The registry lists the primary outcome as Dose-limiting Toxicities (Phase I), measured over 42 days (2 cycles).

The primary endpoint counts participants with dose-limiting toxicities (DLTs) during 42 days (2 cycles). DLTs are used in early-phase studies to identify toxicities that may limit treatment delivery and to guide evaluation of dose levels. The endpoint is therefore focused on tolerability under the study’s stated criteria. It is not a direct measure of tumor response or a randomized comparison of the combination against chemotherapy alone.

The registry describes DLTs as hematologic or non-hematologic toxicity, or treatment delays caused by toxicity, meeting specified criteria. These include a toxicity-related dose delay > 7 days, omission of day 8 or day 15 paclitaxel, or a treatment-related death. Hematologic criteria include study-treatment-related febrile neutropenia, grade 4 neutropenia lasting > 7 days, and study-treatment-related grade 4 thrombocytopenia or bleeding associated with grade 3 thrombocytopenia. Non-hematologic criteria include study-treatment-related grade 3 or grade 4 toxicity.

The record says toxicity was assessed according to CTEP CTCAE version 4.03, with CTCAE version 5.0 used starting 04/01/2018. Its DLT definition also refers to CTEP CTCAE Version 4.0. The posted phase I counts are 2 participants in DL1 and 3 participants in DL2. The registry does not provide a comparative estimate or inferential analysis for these phase I counts, and the cohort counts alone do not establish which dose is preferable.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Most posted outcomes are counts of participants with an event. For these outcomes, a proportion describes the observed count relative to the participants represented in that group. The phase I primary endpoint is reported as participant counts, and the secondary adverse-event, resection, and pathological-response outcomes are also binary in form: a participant either meets the outcome definition or does not. Counts are useful descriptions, but they do not by themselves quantify uncertainty or establish a causal treatment effect.

What does a risk difference describe?

A risk difference compares the observed event proportions between two groups by subtracting one group’s proportion from the other’s. A positive value means the event proportion is higher in the group designated as the intervention in that comparison; a negative value means it is lower. Its interpretation depends on which groups are being compared and on whether the comparison is suitable for causal inference. For phase II Arm I versus Arm II, random allocation is listed for the study, but the registry does not describe the analysis population, missing-data handling, or whether the posted comparison follows the randomized assignment. For comparisons that combine phase I and phase II cohorts, the groups are not a straightforward randomized contrast.

How should a confidence interval be read?

A confidence interval expresses statistical uncertainty around an estimate under the method’s assumptions. A wider interval indicates less precision, while a narrower interval indicates more precision. For a risk difference, an interval that includes zero is compatible with no difference under that interval procedure; an interval that excludes zero indicates the estimate is less compatible with zero under the same assumptions. Neither pattern answers whether the difference is clinically important. The phase II resection and pathological-response intervals include zero; the phase II grade 3 or higher adverse-event interval excludes zero. These intervals are simple recalculations from posted counts, not sponsor-reported estimates.

What does an unadjusted p-value contribute?

An unadjusted p-value evaluates how unusual data at least as inconsistent with a null hypothesis would be if that null model and the calculation assumptions held. The posted recalculations report unadjusted p-values of 0.012 and 0.005 for the adverse-event comparisons, 0.512 for total gross resection, and 0.498 for complete pathological response. These values do not describe the probability that the null hypothesis is true. They do not account for multiple outcome testing, confounding, or analysis choices unless those features are explicitly included in the method. The registry does not report a multiplicity strategy or a sponsor analysis for these comparisons.

Why do cohort and outcome definitions matter?

Phase I DL1 and DL2 are separate dose-level groups, and the primary DLT outcome is defined for a short assessment window. The phase II outcomes have their own populations and assessment times. For example, resection is assessed at surgery, while complete pathological response is assessed at surgery or biopsy after 3 cycles (i.e. after 63 days). The adverse-event outcome’s timeframe is baseline through 30 days after last treatment, with an average of 10 months. These are not interchangeable observation windows. Differences in group composition, follow-up, and outcome ascertainment can affect comparisons.

The registry’s reported counts allow simple unadjusted re-calculations for some secondary comparisons. Such calculations describe contrasts implied by the posted counts and are not a substitute for a prespecified analysis. The registry lists no reported analyses. If a sponsor analysis were available, it would be the primary source for the trial’s inferential results; the record does not report one for the outcomes.

What are the planning illustrations?

The posted planning illustrations use the observed effect as a starting point and show participant requirements per arm for a two-sided significance level of 0.05, equal allocation, and a normal approximation. They examine planning for 100%, 75%, and 50% of the observed effect. For the phase II grade 3 or higher adverse-event contrast, the illustrations give 83 participants per arm for 80% power and 110 for 90% power when planning for 100% of the observed effect; for 75%, they give 162 and 216; for 50%, 397 and 530. These are planning scenarios, not observed power and not the trial’s sample-size justification.

For total gross resection, the corresponding illustrations are 890 participants per arm for 80% power and 1,191 for 90% power at 100% of the observed effect; 1,593 and 2,132 at 75%; and 3,608 and 4,829 at 50%. These values illustrate how planning requirements can increase when a smaller effect is chosen for planning. They do not establish that the observed effect is stable or that a future study would reproduce it. The registry does not report that these illustrations informed the design.

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("Phase 1 - DL1 - Arm II (Ruxolitinib, Paclitaxel, a", "Phase 1 - DL2 - Arm II (Ruxolitinib, Paclitaxel, a")
n      <- c(7, 7)
events <- c(2, 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  Phase 1 - DL1 - Arm II (Ruxolitinib, Paclitaxel, a  7  2  28.6  3.7  71.0  8.2  64.1  8.5  65.1
ARM  Phase 1 - DL2 - Arm II (Ruxolitinib, Paclitaxel, a  7  3  42.9  9.9  81.6  15.8  75.0  15.7  75.5

11. Multiplicity, Missing Data and Analysis Populations

The registry record does not report a multiplicity plan for the multiple secondary outcomes and comparisons. Without a stated hierarchy or adjustment, unadjusted p-values should be read cautiously: testing multiple outcomes can increase the chance that at least one result appears statistically notable by chance. This concern is separate from the size or direction of any individual estimate.

The record does not report how missing outcome data were handled, whether participants were analyzed according to randomized assignment, or which analysis populations were used. It also does not describe exclusions from the posted outcome denominators, reasons for missing assessments, or whether any sensitivity analyses were conducted. These details matter because an outcome comparison may depend on who was assessed and how unavailable results were treated.

The registry lists randomized allocation but does not report the randomization procedure or allocation ratio. It does not report an interim-analysis plan, stopping rules, a sample-size justification, or a full statistical analysis plan. Those items are not reported in the registry record; their absence from the record should not be interpreted as evidence that they were or were not used.

Reported analyses are not listed for the secondary outcomes discussed. The recalculated comparisons are simple, unadjusted calculations from posted counts, and do not account for multiplicity or other possible design and analysis features. No sponsor-reported analysis is available in the registry record for comparison. The sample-size illustrations are planning scenarios based on observed effects, not evidence of the study’s planned power or a retrospective measure of power.

There are also different denominators across outcomes and phases. The registry reports the counts for each outcome, but does not explain every difference in the analyzed populations. It is therefore important not to infer that the overall enrollment applies to every analysis or that the listed denominators represent the same participants across endpoints. The absence of a reported missing-data and population strategy limits how fully the comparisons can be interpreted.

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: anemia (17.0%), white blood cell decreased (14.3%), dizziness (14.3%), hypotension (14.3%), neutrophil count decreased (11.4%).

The serious adverse-event counts are reported separately by phase and group. In phase II, the registry lists 9 participants affected among 42 at risk in Arm I and 42 among 88 at risk in Arm II. In phase I, it lists 1 among 7 in DL1 and 3 among 10 in DL2. These are descriptive counts; they do not establish that treatment caused an event or provide an adjusted comparison of safety.

The adverse-event outcome for grade 3 or higher events is not the same measure as serious adverse events. Seriousness and severity are distinct concepts, and an event can be classified using criteria that differ from a toxicity grade. The registry also lists the serious-adverse-event collection window as baseline through 30 days after last treatment, with an average of 10 months, for the secondary adverse-event outcome. The specific serious-event counts should be read in light of that stated collection period.

Group sizes and exposure can differ, and the registry does not provide a detailed exposure-adjusted safety analysis or a complete account of event-level timing and attribution. The counts are thus useful as a record of reported affected participants, but not as a standalone estimate of comparative risk. They should be considered alongside the endpoint definitions and the limitations of the posted analyses.

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

The primary result describes the number of phase I participants meeting the registry’s DLT criteria during the defined 42-day period: 2 in DL1 and 3 in DL2, with 7 participants reported in each cohort. It provides a descriptive account of DLTs at the two studied dose levels. It does not, by itself, identify a best dose, establish a comparative safety effect, or answer the phase II treatment question. The registry does not post a formal analysis or dose-selection conclusion for these counts.

The phase II secondary comparisons describe posted differences in counts between the paclitaxel-and-carboplatin arm and the combination arm for selected outcomes. The simple unadjusted adverse-event comparison has an interval excluding zero, while the resection and pathological-response intervals include zero. These findings apply to the posted data and the recalculation assumptions. They are not equivalent to a complete analysis of the randomized trial, because analysis populations, missing-data handling, adjustment, and multiplicity procedures are not reported in the registry record.

Neither a statistically notable unadjusted result nor a comparison whose interval includes zero establishes clinical importance, benefit, harm, equivalence, or absence of an effect. Precision and design matter: small event counts and wide uncertainty can leave substantial ambiguity. Likewise, the planning illustrations describe hypothetical future study sizes under stated assumptions; they are not observed power and do not validate the underlying effects.

The registry lists other outcomes, including progression-free survival and overall survival, but their results are not reported in the registry record. Therefore, the posted findings do not provide a complete account of the trial’s longer-term outcomes. No medical recommendation follows from these registry results. Their appropriate meaning is bounded by the endpoints, groups, timeframes, and analyses that are actually reported.

14. Why This Trial Matters Statistically

This study illustrates why phase I and phase II results need to be separated. Phase I asks about dose-limiting toxicity at dose levels, while phase II compares a combination regimen with paclitaxel and carboplatin alone. The phase I counts are important for understanding the early tolerability question, but they cannot be treated as a randomized estimate of the phase II treatment contrast.

It also shows how the same study can contain outcomes with very different statistical properties. A participant count for DLT, a binary adverse-event endpoint, surgical resection, and pathological response each have distinct definitions and assessment windows. Interpreting them together requires attention to denominators, follow-up, and which phase and group contributed the data. A single p-value cannot summarize all these dimensions.

The secondary comparisons demonstrate why effect estimates and uncertainty should be read together. The adverse-event estimates are more distinguishable from zero under the simple unadjusted calculations, whereas the resection and pathological-response intervals include zero. Neither pattern alone determines clinical importance. Multiple secondary outcomes, absent multiplicity details, and no listed sponsor analyses further limit confirmatory interpretation.

Finally, the planning illustrations make visible the dependence of sample-size requirements on the effect chosen for planning. They are hypothetical illustrations based on observed results, not evidence about the original design or proof that a study had a particular probability of detecting an effect. Keeping those distinctions explicit is central to responsible interpretation of registry 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 registry record gives limited detail about the methods needed to assess the posted comparisons. It does not report the randomization procedure, allocation ratio, analysis populations, missing-data handling, or a statistical analysis plan. It also does not report whether the phase II comparisons were analyzed according to randomized assignment or whether any covariate adjustment was used. These omissions constrain causal interpretation of the secondary contrasts.

The primary DLT counts come from small phase I dose-level cohorts, and the record does not provide confidence intervals or a formal comparison between DL1 and DL2. The phase I endpoint also has a specified short assessment window, so it does not represent all possible later toxicities. The registry’s toxicity description refers to CTCAE version 4.03 for assessment, CTCAE version 5.0 beginning 04/01/2018, and CTCAE Version 4.0 in the DLT definition. That wording makes the exact version reference in the definition less uniform.

For phase II, the available secondary results do not cover every listed outcome. Progression-free survival and overall survival results are not reported in the registry record. The available outcome denominators differ, and the record does not explain all reasons for those differences. The adverse-event comparisons also include a phase I cohort comparison against phase II Arm I, which is not a like-for-like randomized comparison.

The secondary calculations are unadjusted and multiple outcomes are listed without a reported multiplicity strategy. The sample-size illustrations are based on observed effects and stated assumptions, not on the study’s reported design justification. Together, these limitations mean that the posted results support descriptive reading and cautious exploratory comparisons, rather than a complete, confirmatory account of treatment effects.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Ruxolitinib Phosphate, Paclitaxel, and Carboplatin in Treating Patients With Stage III-IV Epithelial Ovarian, Fallopian Tube, or Primary Peritoneal Cancer — NCT02713386.
• Registry record last updated 2025-05-25; status last verified 2025-05.
• Results first posted 2024-10-24.
• 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

NCT02713386 is a completed randomized, open-label, parallel phase I/II study sponsored by NRG Oncology. The registry lists enrollment as 147. It concerns ruxolitinib phosphate with paclitaxel and carboplatin, compared in phase II with paclitaxel and carboplatin alone, in advanced epithelial ovarian, fallopian tube, or primary peritoneal cancer.

The primary outcome is phase I dose-limiting toxicity during 42 days (2 cycles), with 2 participants reported in DL1 and 3 in DL2. Selected phase II secondary results include grade 3 or higher adverse events, total gross resection, and complete pathological response. Other listed outcomes include progression-free survival and overall survival, but their results are not reported in the registry record.

The registry lists no reported analyses for the secondary outcomes discussed. The comparisons described are simple unadjusted recalculations from posted counts. The record does not report a statistical analysis plan, missing-data strategy, or multiplicity approach, so the secondary findings require cautious interpretation.