GANNET53: Ganetespib in Metastatic, p53-mutant, Platinum-resistant Ovarian Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for GANNET53: Ganetespib in Metastatic, p53-mutant, Platinum-resistant Ovarian Cancer (NCT02012192), a randomized phase 1, phase 2 study. The primary outcome is Progression Free Survival (PFS).

2. Clinical Question

The trial’s rationale focused on platinum-resistant disease and ganetespib’s proposed targeting of mutant p53 through Hsp90 inhibition. That rationale explains the treatment strategy being investigated; it does not establish what the trial results demonstrate. The posted medians describe the observed time-to-progression outcome in each group, while the confidence intervals indicate uncertainty around each median. Because the registry does not report a comparative estimate or test, the medians alone do not establish whether the groups differed statistically. The record also does not report why the trial was terminated or whether the planned Phase II comparison was completed as intended.

3. Trial Design

GANNET53 is registered as a multicentre, international Phase I and II trial, with randomized allocation, no masking, and a parallel-group model. Its official title identifies the population as women with high-grade serous, high-grade endometrioid, or undifferentiated platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer. The registry lists an enrollment of 133 and identifies the Medical University Innsbruck as sponsor. The study start is recorded as 2014-07-04, primary completion as 2017-11-30, and status as terminated. The reason for termination is not reported in the registry record.

The brief summary describes two parts. Phase I was intended to examine the safety of ganetespib combined with weekly paclitaxel, with the ganetespib dose dependent on Phase I results. The second part was a randomized Phase II comparison of the combination against weekly paclitaxel alone. In both arms, paclitaxel was listed as 80 mg/m2 intravenously once weekly for 3 out of 4 weeks, until progression. The combination arm’s ganetespib schedule was also listed as intravenous treatment once weekly for 3 out of 4 weeks.

The primary outcome is progression-free survival in the ITT population. The posted result groups contain 90 for ganetespib plus paclitaxel and 43 for paclitaxel; the registry does not explain how these outcome groups relate to enrollment across both trial parts. It does not report the random sequence method, concealment, stratification, or the practical procedures used to implement allocation. Masking is recorded as none, but the record does not describe who knew treatment assignments or how outcome assessments were conducted under the open-label design. Those omissions limit assessment of how the design was implemented, without changing the registry entries of the registered comparison.

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

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

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 Progression Free Survival (PFS), measured over Time until progression (median w/o new drug 4 months).

The primary endpoint is progression-free survival (PFS), reported as a median in months. It measures the time until progression, as stated in the registry timeframe. The study description says response or progression was to be evaluated using Response Evaluation Criteria in Solid Tumors Criteria version 1.1, CA-125 according to published GCIG criteria, and investigator assessment based on physical and/or gynaecological examinations. The registry does not provide further detail about how those sources were combined when they differed.

A median is the point in the observed time-to-event distribution at which half of the observations have had the event and half have not yet had it, under the usual interpretation of a median. This is a general statistical explanation. Time-to-event endpoints are useful when the timing of progression matters and participants may have different follow-up durations. They can account for the fact that some participants have not experienced the event by the end of their observed follow-up, if censoring is handled in the analysis. The registry does not report the censoring rules or analysis details used for these results.

The posted median PFS is 3.5 months (95% CI 3.092 to 3.882) for ganetespib plus paclitaxel, and 5.3 months (95% CI 4.006 to 6.652) for paclitaxel. The timeframe also mentions “median w/o new drug 4 months”; the registry does not clarify this phrase as an observed result for the trial’s control group. The posted control result is 5.3 months. A median and its interval describe each group, but do not by themselves provide a direct estimate or test of the difference between groups.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Progression-free survival is a time-to-event endpoint: the outcome concerns how long it takes for a defined event to occur. The registry calls the event progression and reports the endpoint as a median in months. In time-to-event analysis, participants who have not had the event by their last known follow-up may contribute information up to that point; this is generally handled through censoring. This is a general explanation, not a description of the specific analysis in GANNET53. The registry record does not report its censoring rules, follow-up procedures, or the statistical method used to estimate the medians.

What does a median summarize?

A median summarizes the middle of a distribution, rather than its arithmetic average. For a time-to-event outcome, the median is the time at which the estimated event-free proportion reaches the midpoint, using the analysis method applied. It can be easier to interpret than a model coefficient, but it compresses the event-time pattern into one point. It does not describe the full shape of the event curve or indicate whether a treatment effect is constant over time. The registry posts only the medians and their confidence intervals for the primary outcome; it does not post event curves or other distributional summaries in the record details.

How should the confidence intervals be read?

The reported 95% confidence intervals are 3.092 to 3.882 months for the combination group and 4.006 to 6.652 months for the paclitaxel group. In general, a confidence interval conveys the precision of an estimate under the statistical procedure used. A narrower interval indicates less uncertainty on that scale than a wider interval, while neither interval removes uncertainty or describes the range of outcomes for individual participants. The record does not identify the interval-estimation method, so the exact assumptions behind these intervals are not reported in the registry record.

These are intervals for each group’s median. They do not directly estimate the contrast between groups. A formal comparison might use a time-to-event model or a test based on event-time distributions; which method is appropriate depends on design and assumptions. For example, a hazard ratio is a relative event-rate measure from a model, and its interpretation can depend on whether the relative hazard is reasonably stable over time. The registry does not post a sponsor-reported hazard ratio, its interval, or a primary-outcome p-value. Therefore, none should be inferred from the group medians or their separate intervals.

Why can over-dispersion matter?

Over-dispersion is a general issue for count outcomes: it occurs when observed variability exceeds what a simple model assumes. It can make standard errors too small if the model does not account for the extra variability. It is not a direct feature of a median time-to-event endpoint, and the posted primary result is not a count outcome. There is no basis in the registry record to say that an over-dispersion adjustment was used or needed for the PFS analysis.

What do recalculations and planning illustrations mean?

The registry analysis notes that a two-arm recalculation is not available for this median time endpoint. No recalculated comparative estimate should therefore be read into the results. The record’s planning assumptions describe illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation; the effect to detect is defined as a stated fraction of the observed effect. Those descriptions are planning assumptions, not evidence that the trial had a particular power or that an effect was detected. The registry record does not post the resulting planning values in the registry record for this analysis, and they should not be treated as observed results or observed power.

For time-to-event planning, the record says event totals are the relevant quantity in its planning framework. In general, this reflects that information for time-to-event comparisons depends substantially on the number of events, rather than enrollment alone. The registry does not provide the event count for the primary PFS analysis or a sample-size justification. It also does not report whether interim analyses were planned or conducted, what design assumptions informed the trial, or what statistical analysis plan governed the comparison.

The listed endpoint is median PFS in the ITT population, and the posted results report group medians with 95% confidence intervals. The distinction between a descriptive group estimate and a comparative model result is important: one cannot recreate the latter from the two medians alone. No recalculated result is presented as an alternative to the sponsor’s analysis. The registry’s record does not include a sponsor comparative estimate to take precedence, and it does not report the analysis details needed to independently assess one.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure, PFS, and no posted secondary outcomes. It also records zero analyses. These registry entries mean the record contains no reported set of multiple outcome tests from which to assess how multiplicity was handled. They do not establish that no other analyses were conducted outside the posted record. Any prespecified hierarchy, multiplicity adjustment, or strategy for controlling false-positive findings is not reported in the registry record.

The primary outcome is identified as being assessed in the ITT population. In general, an intention-to-treat analysis aims to retain participants in the groups to which they were assigned, regardless of treatment received, preserving the randomized comparison. The registry does not define the ITT population, provide participant flow, or state how deviations, treatment discontinuations, or missing outcome information were handled. The number of records in the posted outcome groups is reported as 90 for ganetespib plus paclitaxel and 43 for paclitaxel; the record does not explain how those analysis groups relate to enrollment across the Phase I and Phase II parts.

Missing-data procedures are not reported in the registry record. For a time-to-event endpoint, follow-up that ends before documented progression can affect how much information is available and how censoring is applied. That is a general statistical concern; no claim can be made about the amount or handling of missing follow-up in this trial from the registry entries. The record also does not report sensitivity analyses addressing missingness, outcome ascertainment, or departures from the stated analysis population. These gaps limit evaluation of robustness, but do not justify assuming that data were missing or mishandled.

11. 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: small intestinal obstruction (8.9%), erysipelas (7.0%), sepsis (6.7%), urinary tract infection (5.6%), vomiting (4.7%).

The registry reports serious adverse events as affecting 34 of 90 in the ganetespib-plus-paclitaxel group and 10 of 43 in the paclitaxel group. It reports other adverse events as affecting 88 of 90 and 41 of 43, respectively. The record also lists deaths as 66 and 29 in the corresponding groups. These are reported counts, not newly calculated rates. The registry does not state in the registry record whether the deaths were attributed to treatment, disease progression, or another cause, so they should not be interpreted as treatment-related deaths.

The serious adverse-event terms listed with percentages are Small intestinal obstruction, 8.9%; Erysipelas, 7.0%; Sepsis, 6.7%; Urinary tract infection, 5.6%; Vomiting, 4.7%; and Urinary tract obstruction, 4.7%. The record does not specify the denominator or arm attribution for these listed terms in the registry record, so those details should not be inferred. A serious adverse event is a regulatory category, not simply a synonym for a severe symptom; the registry’s posted figures should be read according to its event classification.

Adverse-event counts do not by themselves establish causality, comparative risk, or a treatment effect. Interpretation also depends on definitions, collection period, exposure, and event-level details, which are not reported in the registry record. The counts describe what the registry recorded for the groups, with these limitations.

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

The posted results establish that the registry reports median PFS of 3.5 months (95% CI 3.092 to 3.882) for ganetespib plus paclitaxel and 5.3 months (95% CI 4.006 to 6.652) for paclitaxel alone. They also establish that the posted outcome group sizes are 90 and 43, respectively. These are the trial’s recorded group-level outcome summaries, rather than a direct estimate of the comparative treatment effect.

The result does not establish a statistically significant difference between the groups. The registry does not post a hazard ratio, comparative confidence interval, or p-value for the primary endpoint. Separate confidence intervals around medians do not replace a direct comparison. No new comparative calculation is available for this median time endpoint in the registry’s analysis information.

The result also does not establish why the trial was terminated, whether the planned randomized Phase II analysis was completed as intended, or how the posted outcome groups relate to enrollment across the two trial parts. The record does not report a statistical analysis plan, sample-size justification, interim analysis details, censoring rules, or missing-data methods. These omissions constrain what can be concluded about the design and robustness of the posted estimate.

The numbers are not a forecast for an individual participant. A group median does not say how long any one person will remain progression-free, and the confidence intervals describe uncertainty in estimated group medians rather than a range of individual outcomes. The safety counts likewise do not establish that a particular event was caused by treatment. No medical advice follows from these registry summaries.

The appropriate statistical reading is therefore descriptive and bounded: two group medians and their respective intervals are posted for the primary time-to-event outcome, alongside recorded adverse-event counts. The registry record does not provide the comparative analysis needed to quantify or test a treatment effect, and it does not supply enough detail to resolve the design and analysis questions noted above.

13. Why This Trial Matters Statistically

This trial is statistically notable because its design spans a Phase I safety component and a randomized Phase II comparison, while its posted primary result is a time-to-event endpoint. Interpreting the comparison requires distinguishing the intention of the registered design from the analyses actually reported. The record describes a randomized Phase II comparison, but the posted result consists of one median and confidence interval for each group, without a comparative hazard ratio or test.

That distinction is central in clinical research. Group-specific summaries can describe the observed outcomes, but a comparative estimate answers a different question: how the outcome differs between groups, with uncertainty quantified for that contrast. Separate confidence intervals are not a substitute for such an estimate. In this record, the missing comparative result means that the primary endpoint cannot be assessed through a posted formal test or effect estimate.

Time-to-event analyses also rely on details about follow-up and censoring, because participants may be observed for different lengths of time. The registry gives medians but not the method for estimating them, the censoring rules, or the analysis plan. These omissions matter when judging precision and reproducibility. The record’s status is terminated, but does not explain why; that status alone should not be used to infer how complete the analysis was or why the study ended.

The study also illustrates why safety counts need context. Counts of serious and other adverse events, and deaths, are informative descriptions, but are not by themselves comparative risk estimates or evidence of causation. Their interpretation depends on definitions, exposure, attribution, and the time period over which events were collected. The available the registry entries do not supply all of those details.

For biostatisticians and clinical researchers, the key lesson is to keep the estimand, reported statistic, and available evidence distinct. The registered question concerns a comparison of PFS, but the posted values summarize each group separately. Recognizing that gap prevents descriptive medians from being overstated as a formal treatment-effect conclusion.

14. 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 reports the primary PFS medians and their 95% confidence intervals, but does not report the statistical method used to estimate them. It does not give a comparative hazard ratio, comparative confidence interval, or p-value. As a result, the posted summaries do not quantify the between-group effect or its uncertainty directly.

Details needed to assess the time-to-event analysis are also not reported: these include the censoring rules, follow-up procedures, event counts for PFS, and handling of missing outcome information. The endpoint description names criteria and investigator assessments for response or progression, but does not explain how potentially differing assessments were resolved. The record identifies the ITT population but does not define it or provide participant flow.

The registry describes a two-part Phase I and II study and identifies the Phase II comparison as randomized. It does not explain how the posted outcome-group sizes relate to enrollment across both parts or document the randomization procedure. The trial is marked terminated, with no reason recorded. The registry also does not report a sample-size justification, interim analyses, or a statistical analysis plan. These missing details prevent a full assessment of whether the posted analyses followed the intended design.

For safety, event counts and selected serious-event terms are posted, but the registry entries do not specify term denominators or group attribution. Death counts are not accompanied by cause or treatment attribution. These limits restrict interpretation of safety patterns and do not support causal conclusions. The record’s timeframe phrase referring to a median without the new drug is not explained; it should not be substituted for the posted control-group median.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: GANNET53: Ganetespib in Metastatic, p53-mutant, Platinum-resistant Ovarian Cancer — NCT02012192.
• Registry record last updated 2019-08-13; status last verified 2019-06.
• Results first posted 2019-08-13.
• 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.

18. Record Summary

GANNET53 is registered as a multicentre, international Phase I and II trial of ganetespib with weekly paclitaxel in platinum-resistant epithelial ovarian, fallopian tube, or primary peritoneal cancer. The registry describes a randomized Phase II comparison with weekly paclitaxel alone. It lists an enrollment of 133, identifies the Medical University Innsbruck as sponsor, and records the study as terminated. The reason for termination is not reported in the registry record.

The primary outcome is median PFS in the ITT population. The posted median is 3.5 months (95% CI 3.092 to 3.882) for the combination group and 5.3 months (95% CI 4.006 to 6.652) for paclitaxel. No comparative hazard ratio, p-value, or confidence interval for a between-group effect is posted. The record also reports serious and other adverse-event counts and deaths, without enough detail to establish causality or interpret all listed serious-event percentages by group.