Sunitinib in Recurrent and Refractory Ovarian, Fallopian Tube and Peritoneal Carcinoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Sunitinib in Recurrent and Refractory Ovarian, Fallopian Tube and Peritoneal Carcinoma (NCT00768144), a na phase 2 study. The primary outcome is Overall Response Rate.

2. Clinical Question

3. Trial Design

This was a phase II, single-group study sponsored by Dana-Farber Cancer Institute. The trial title specifies recurrent and refractory ovarian, fallopian tube and peritoneal carcinoma. The registry lists the study as completed, with enrollment of 36 participants. The recorded start was 2008-09 and primary completion was 2012-05.

All participants received the same intervention: oral sunitinib at a daily dose of 37.5 mg, given continuously over a 28-day treatment cycle. Treatment was intended to continue until clinical or radiological evidence of progressive disease or excessive toxicity. Allocation is recorded as not applicable, masking as none, and the model as single group. There is no concurrent control arm in the registry record.

The primary endpoint was overall response rate, defined using best response on treatment under RECIST 1.0. Assessments were scheduled weekly for the first 4 weeks and every 2 weeks in subsequent cycles. Radiological disease evaluations were planned at baseline, before each odd cycle and at the end of treatment. The record describes response confirmation and requirements for non-target lesions and new lesions.

Secondary outcomes are listed as 16-Week Progression-Free Survival and Progression-Free Survival. The registry record does not report results for these outcomes or the overall response rate. It also does not report an analysis plan, the rationale for the enrollment target, or details of any interim analyses. As a single-group study, the design can describe outcomes among treated participants, but it cannot separate treatment effects from other explanations for observed changes or outcomes.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The analysis dataset is comprised of all treated patients.

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: The analysis dataset is comprised of all enrolled patients.

6. Primary Endpoint

The registry lists the primary outcome as Overall Response Rate, measured over Clinical assessments were performed weekly for first 4 weeks and every 2 wks in subsequent cycles. Disease was evaluated radiologically at baseline, before each odd cycle and at end of trt..

Overall response rate is the proportion of participants whose best response on treatment is a complete response or partial response. The registry defines these categories with RECIST 1.0. For target lesions, complete response requires disappearance of all target lesions; partial response requires at least a 30% decrease in the sum of their longest diameters relative to baseline. A response must be confirmed by repeat assessment no fewer than 4 weeks after the criteria are first met. The definition also requires at least incomplete response or stable disease for non-target lesions and no new lesions.

Response rate is a binary endpoint at the participant level: each treated participant either meets the defined response criteria or does not. As a general statistical concept, this kind of endpoint provides a direct description of how often a specified response was observed in the analysis population. It does not, by itself, describe how long a response lasts, whether disease remains controlled, or how outcomes compare with a different treatment.

The registry specifies a 90% confidence interval for the primary endpoint and states that the analysis dataset comprises all treated patients. However, the response count, estimated proportion and confidence interval are not reported in the registry record. The endpoint definition is therefore available, but the observed response rate and its precision cannot be assessed from the reported results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary participant-level measure. Each treated participant is classified according to whether the best response on treatment met the criteria for complete or partial response. As a general statistical concept, a proportion estimates the frequency of an event in the group being analysed. For this endpoint, the event is a confirmed RECIST 1.0 complete or partial response, subject to the stated requirements for target lesions, non-target lesions and absence of new lesions.

What does the confidence interval express?

The registry specifies a 90% confidence interval for the response-rate endpoint, but does not report the interval or the point estimate. In general, a confidence interval expresses uncertainty around an estimate under the method used to calculate it. A narrower interval indicates greater statistical precision under that method; a wider interval indicates less precision. The interval is not a guarantee that the true response rate lies within its particular bounds. Because neither the observed proportion nor its interval is reported in the registry record, the degree of precision cannot be judged.

The record identifies all treated patients as the analysis dataset for the primary endpoint. This matters because a response proportion depends on who is included in its denominator. The recorded enrollment is 36, but enrollment and the endpoint analysis population are not necessarily interchangeable concepts. The primary-endpoint description says all treated patients are included; the registry does not report a participant-level accounting of treatment initiation, missing assessments or endpoint evaluability.

How should the design affect interpretation?

The study is single-group, with no masking and allocation listed as not applicable. In general, a single-group estimate describes outcomes in the participants who received the intervention. It does not estimate a treatment contrast against a concurrent control group. Without that contrast, the observed outcome cannot isolate the intervention's effect from the natural course of disease, participant selection, other care or measurement-related influences. This is a design limitation, not a flaw in calculating a proportion.

The RECIST criteria define the endpoint in a structured way. Complete response requires disappearance of all target lesions, while partial response requires at least a 30% decrease in the sum of longest diameters from baseline. The record also requires confirmation by a repeat assessment no fewer than 4 weeks after response criteria are first met. These rules aim to distinguish a qualifying response from a change that is not confirmed. The record does not report how many participants had assessments at each scheduled time or how missing or unevaluable assessments were handled.

Why can over-dispersion matter?

As a general concept, over-dispersion arises when observed variability exceeds what a statistical model assumes. It is particularly relevant in models for counts or clustered observations, where unaccounted variability can make standard errors too small and confidence intervals too narrow. The primary endpoint is described as a proportion of participants, and the registry record does not identify a model or report an analysis method beyond specifying a 90% confidence interval. It is not possible to assess whether over-dispersion was considered or relevant to the sponsor's analysis.

The registry record lists no analyses. It does not report the interval calculation method, any adjusted analysis, handling of missing endpoint data, or checks of statistical assumptions. General descriptions of confidence intervals and binary endpoints explain how such quantities are commonly interpreted, but they do not establish which exact procedures were used for this trial.

What can planning calculations tell readers?

Planning illustrations based on an observed effect are not the same as the trial's original sample-size justification, and they do not represent observed power. In general, such illustrations estimate the sample size that might be needed to detect an assumed effect under stated design and testing assumptions. Their results depend on those assumptions; changing the assumed effect or allocation changes the planning question. The registry record does not report the trial's own sample-size justification. No observed response estimate is available in the registry record to support interpretation of any planning calculation as a description of the trial's actual result.

For time-to-event outcomes, a general planning approach may focus on the total number of events, whereas a binary-outcome planning approach may express participants per arm. These planning conventions do not turn this single-group trial into a controlled comparison. The record's assumptions describe equal allocation between two arms for planning purposes, while the actual study design is single-group. That distinction is essential: hypothetical planning conditions should not be mistaken for the trial's observed allocation or 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("Sunitinib")
n      <- c(36)
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  Sunitinib  36  0  0.0  0.0  9.7  -0.0  9.6  0.1  9.5

11. Multiplicity, Missing Data and Analysis Populations

The registry record names overall response rate as the primary outcome and lists 16-Week Progression-Free Survival and Progression-Free Survival as secondary outcomes. It does not report results for these outcomes, nor does it describe a statistical analysis plan or a strategy for handling multiple outcome tests. The record also reports no analyses.

As a general statistical concept, multiplicity arises when several outcomes, time points, subgroups or statistical tests are examined. Repeated testing can increase the chance that at least one result appears noteworthy by chance. A prespecified approach may define which outcome is primary and how any additional analyses will be interpreted or adjusted. The registry identifies a primary outcome, but does not report whether other tests were planned, whether multiplicity adjustments were considered, or how any secondary results were intended to support conclusions.

The primary-endpoint population is described as all treated patients. The record does not explain how participants without sufficient follow-up, missing radiological assessments or unconfirmed responses were handled in the analysis. It does not provide a participant flow or an accounting of endpoint data availability. Although the response definition requires confirmation, the registry does not state what classification was assigned when confirmation was unavailable.

These gaps limit assessment of the analysis population and the possible influence of missing data. As a general matter, missing outcomes can affect an estimated response proportion if participants with unavailable assessments differ from those with observed assessments. The direction and size of any such effect cannot be determined without information about missingness and its handling. No interim-analysis plan, multiplicity procedure, or additional analysis-population definition is reported in the registry record.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

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: small bowel obstruction (5.6%), hyperkalemia (2.8%), cerebrovascular accident (2.8%), allergic reaction (2.8%), abdominal pain (2.8%).

The registry lists 36 participants at risk for both serious and other adverse events. It reports 7 participants with a serious adverse event and 36 with other adverse events. These figures describe the reported safety counts in the sunitinib group; they should not be read as a comparison with a control group, because the study had no concurrent comparator.

Serious adverse-event terms listed in the record include small bowel obstruction (5.6%), hyperkalemia (2.8%), cerebrovascular accident (2.8%), allergic reaction (2.8%), abdominal pain (2.8%) and death due to progressive disease (2.8%). The record does not provide further detail about event timing, duration, recurrence, severity grading or the investigators' assessment of relationship to treatment. A listed event is not, on its own, evidence that the intervention caused it.

As a general statistical concept, safety counts and percentages describe observed events in the recorded group, while comparative risk requires an appropriate comparison group and a clearly defined analysis. The registry record does not report deaths as a separate numerical result. The serious-event terms and adverse-event counts therefore offer a limited description of reported safety experience, not a complete account of risk or causality.

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

The record establishes that a phase II single-group trial evaluated sunitinib for recurrent or refractory ovarian, fallopian tube and peritoneal carcinoma, with overall response rate as the primary endpoint. It also documents the response criteria, treatment schedule and planned assessment timing. These details allow readers to understand what response was intended to mean and how it was to be assessed.

The record does not include the primary response count, response proportion or its 90% confidence interval. It also does not report results for the listed progression-free survival outcomes. Consequently, the posted results do not provide the numerical evidence needed to describe the observed efficacy endpoints or their precision. The absence of those results in the registry record should not be interpreted as a result of zero response or as evidence of a particular outcome.

Even if an outcome estimate were available, a single-group design would describe observations among treated participants, not a randomized difference from an alternative. As a general statistical principle, causal interpretation is stronger when a study design supports a credible comparison and limits systematic differences between groups. This registry record contains no concurrent control group, and it does not describe a prespecified historical comparator.

The safety entries provide counts and selected serious-event terms, but do not establish that sunitinib caused those events. They also do not provide a comparative risk estimate. The registry does not report details needed to evaluate event timing, attribution or missing safety information.

Thus, the available record is informative about the trial's question and methods, but limited for judging its findings. It cannot establish the size or precision of the response rate, provide an efficacy comparison, or support conclusions about the causal effect of treatment. Those limits arise from the results and design information reported in the registry record.

14. Why This Trial Matters Statistically

This trial illustrates why an endpoint definition and an endpoint result are different kinds of information. The registry gives a detailed definition of overall response rate, including RECIST 1.0 criteria, confirmation requirements and conditions for non-target lesions. Such detail clarifies what counts as an event. Yet, without the response count, proportion and confidence interval, readers cannot evaluate how often the event occurred or how precisely it was estimated.

The design also matters to the statistical question. A single-group study can describe outcomes among treated participants, but it does not contain a concurrent comparison for estimating a treatment contrast. This distinction remains important regardless of the sophistication of outcome measurement. A carefully defined endpoint does not remove the limitations of an uncontrolled design.

The record further distinguishes enrollment from the stated analysis population: the primary endpoint is defined for all treated patients. Understanding the denominator is fundamental to interpreting a proportion. The record does not provide participant-level accounting that would show how the analysis population relates to enrollment or how unavailable assessments were handled.

Finally, the listed secondary outcomes and safety entries show why registry completeness is important. Outcome names do not substitute for outcome estimates, and event counts without detailed context cannot establish causality. This study is therefore a useful statistical example of how design, analysis population, endpoint definition and reporting completeness jointly shape what can be concluded from a clinical trial record.

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 limitation for interpreting efficacy is that the registry record does not report the primary endpoint estimate or its confidence interval. It also does not report results for either listed secondary outcome. The trial's endpoint definition is detailed, but the absence of outcome values prevents evaluation of the observed response rate and its precision.

The single-group design has no concurrent control group. As a result, the study record cannot supply a randomized comparison or distinguish outcomes associated with treatment from other influences. The lack of masking is also recorded, while the record does not describe procedures that might reduce assessment bias.

The primary analysis population is described as all treated patients, but a participant flow and details of treatment initiation are not reported. The registry does not specify how missing or unevaluable radiological assessments were handled, including situations in which a response required confirmation but a confirmatory assessment was unavailable. It also does not report the number of participants with missing endpoint data.

Additional details that would support appraisal are not reported in the registry record: the sample-size justification, a statistical analysis plan, interim-analysis procedures, multiplicity handling and specific analysis methods. The safety information is limited to event counts and selected terms; attribution, timing and other contextual details are not reported. These omissions do not establish what occurred, but they constrain the conclusions that can be drawn from the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Sunitinib in Recurrent and Refractory Ovarian, Fallopian Tube and Peritoneal Carcinoma — NCT00768144.
• Registry record last updated 2018-08-24; status last verified 2018-07.
• Results first posted 2014-12-02.
• 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

NCT00768144 is a completed phase II, single-group trial sponsored by Dana-Farber Cancer Institute. It enrolled 36 participants with recurrent or refractory ovarian, fallopian tube or peritoneal carcinoma. Participants received oral sunitinib at 37.5 mg daily continuously over a 28-day treatment cycle, with treatment continuing until progression or excessive toxicity.

The primary endpoint was overall response rate under RECIST 1.0, with a 90% confidence interval specified. The primary analysis population is described as all treated patients. Secondary outcomes were 16-Week Progression-Free Survival and Progression-Free Survival. The registry record does not report efficacy results or analyses.

The safety record reports 7 of 36 participants with serious adverse events and 36 of 36 with other adverse events. The single-group design does not provide a concurrent treatment comparison, and the registry record does not report the sample-size justification, analysis plan or handling of missing endpoint data.