Gemcitabine/Taxotere/Xeloda (GTX) With Cisplatin in Subjects With Metastatic Pancreatic Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Gemcitabine/Taxotere/Xeloda (GTX) With Cisplatin in Subjects With Metastatic Pancreatic Cancer (NCT01459614), a na phase 2 study. The primary outcome is Percentage of Participants Without Disease Progression (Progression-Free Survival) at 6 Months.

2. Clinical Question

3. Trial Design

This was a single-arm phase II study of GTX with cisplatin for metastatic pancreatic cancer. The registry lists the phase as PHASE2, the allocation as NA, masking as NONE, and the model as SINGLE GROUP. These features mean that the study was not a blinded randomized comparison. Its central purpose was to estimate outcomes for participants treated on the study regimen rather than to establish a contrast against a concurrent control group.

The registry describes an initial cohort and an expansion cohort. The planned enrollment described in the study summary was approximately 38 evaluable participants, with 28 in the initial cohort and 10 in the expansion cohort. A safety run-in was planned for 6 participants at DL1, with observation for limiting toxicity during the first 2 cycles before further accrual. The record also says that adverse events would be monitored continuously after the run-in. Whether these planned procedures were followed exactly is not reported in the registry record.

The primary cohort used 21-day cycles, consisting of two weeks of treatment followed by one week off. Xeloda was given on days 1-14, while gemcitabine, Taxotere, and cisplatin were given on days 4 and 11. For the expansion cohort, the cycle was 28 days, with two weeks of treatment followed by two weeks off. Since the schedules differed by cohort and there was no random allocation, outcomes across cohorts should not be read as a controlled comparison of schedules.

The registry lists the study as completed and gives 44 as the enrollment. That overall figure is not interchangeable with the planned evaluable population or the number contributing to a particular outcome. The number included in the primary endpoint analysis is not reported in the registry record. The start and primary completion dates are listed as 2011-11 and 2016-11.

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.

Before assignment: 44 patients were consented and screened. Of these, 5 were screen failures and 39 were eligible and assigned to receive treatment.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: One patient was consented and enrolled, but was not considered evaluable per protocol, as he came off study prior to completing a cycle of treatment for reasons other than disease progression or death.

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 Percentage of Participants Without Disease Progression (Progression-Free Survival) at 6 Months, measured over 6 months.

The primary endpoint is titled as the percentage of participants without disease progression at 6 months, with progression-free survival defined in the study summary as being alive and free of progression at that time. Such an endpoint captures whether participants remain alive without documented worsening during a specified follow-up period. It is often used in oncology studies because it focuses on the time before progression or death, rather than only on survival.

The registry’s endpoint description is not fully consistent with that interpretation. It says that PFS is the percentage of patients with disease progression or relapse, or death due to any cause, at 6 months. That wording describes events rather than participants who remain progression-free. The endpoint title and the study summary point toward a progression-free proportion, but the description creates uncertainty about the direction of the reported percentage. The registry record does not clarify this inconsistency.

The description refers to RECIST 1.1 criteria for assessing response and progression, and says that the estimate is based on the Kaplan-Meier curve. Kaplan-Meier estimation is a time-to-event method that can account for participants whose event status is not yet known at the end of their observed follow-up. The registry record lists a 95% confidence interval as the endpoint’s dispersion measure. However, neither the numerical estimate nor its confidence interval is reported in the registry record, so the magnitude and precision of the endpoint cannot be evaluated from this record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The methods described in the registry record include a single-arm design, a time-to-event estimation approach for the primary endpoint, and a stated confidence interval. Important details needed to fully evaluate the analysis are not reported in the registry record. The explanations below describe the statistical ideas generally, while distinguishing them from methods explicitly recorded for this study.

What does a single-arm design estimate?

In general, a single-arm study estimates outcomes among participants receiving the study regimen. Without a concurrent control group, it cannot directly estimate a treatment contrast under the same study conditions. Comparing the result with outcomes from other studies or historical expectations requires assumptions about differences in participants, clinical practice, follow-up, and outcome assessment. The registry record does not report a randomized comparison or a formal external-control analysis.

The initial and expansion cohorts used different schedules. Because allocation was not randomized and the cohorts were not described as a controlled comparison, a contrast between their outcomes would not isolate the causal effect of schedule. Cohort differences could have other explanations. The registry record does not report a planned statistical comparison between these cohorts.

How does Kaplan-Meier estimation work?

In general, Kaplan-Meier estimation uses the observed times to an event together with information from participants whose event status is known only up to their last observed follow-up. Those participants are censored at that point rather than being treated as though they experienced the event or remained event-free indefinitely. The method estimates the probability of remaining event-free over time, provided that censoring is appropriately handled.

For this study, the registry says the primary estimate is based on a Kaplan-Meier curve and gives a 6-month timeframe. The record does not report the numbers at risk over follow-up, the number of events, the censoring distribution, or the precise calculation used to derive the percentage at 6 months. Those details affect how readers judge the stability of a time-specific estimate.

A confidence interval describes statistical uncertainty around an estimate under the assumptions of the analysis. In general, a narrower interval indicates greater precision, while a wider interval indicates greater uncertainty. The registry lists a 95% confidence interval for the primary endpoint, but neither its limits nor the endpoint estimate are reported in the registry record. Precision therefore cannot be assessed.

Why does the endpoint wording matter?

A time-to-event analysis requires a clear definition of the event. The study summary describes the primary outcome as the proportion alive and free of progression at 6 months. The endpoint title similarly refers to participants without progression. In contrast, the endpoint description defines PFS as the percentage with progression or relapse, or death, at 6 months. This is an internal ambiguity in the registry record. It is not possible to resolve it by statistical convention alone; the actual event coding and reported value are not reported in the registry record.

The description refers to RECIST 1.1 criteria. In general, standardized response criteria aim to make classification of tumor response and progression more consistent. The record describes complete response, partial response, progressive disease, and stable disease, but the endpoint wording does not make clear whether the posted percentage is a Kaplan-Meier estimate of remaining progression-free or a percentage experiencing an event by the specified time.

What does a confidence interval add?

A point estimate alone does not show how much uncertainty is associated with the result. A confidence interval provides a range of values compatible with the observed data and the method’s assumptions. It is not a guarantee that the true value lies inside the interval, and it does not account for every source of uncertainty, such as bias from participant selection or differences in assessment. Since no numerical interval is reported, neither statistical precision nor compatibility with a benchmark can be examined.

How should planning illustrations be read?

General sample-size planning translates assumptions about a target effect, variability, significance level, and desired power into a proposed enrollment. Such calculations are planning illustrations, not observed evidence about whether a study’s result is statistically persuasive. The registry record describes a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It also says that the effect to detect is a stated fraction of the observed effect. These assumptions do not describe the single-arm trial’s own design, and no sample-size justification for the trial is reported in the registry record. Any planning illustration should not be described as observed power.

For time-to-event settings, planning methods may be based on a target number of events, whereas other outcome types may use participants per arm. That distinction is general and does not turn this single-arm study into a two-arm comparison. The registry record does not report a randomized allocation or a control-arm sample for this trial.

What analysis details are absent?

The registry record does not report the statistical analysis plan, the endpoint estimate, the confidence interval limits, the analysis population, or detailed handling of missing and censored observations. It does not report whether any interim efficacy analyses were conducted or whether any multiplicity adjustment was planned. The record lists no posted analyses. In the absence of these details, interpretation should remain limited to the study design and endpoint definitions that are documented.

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("Primary Cohort (21 Day Cycle)", "Expansion Cohort (28 Day Cycle)")
n      <- c(28, 10)
events <- c(22, 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  Primary Cohort (21 Day Cycle)  28  22  78.6  59.0  91.7  60.5  89.8  60.3  89.7
ARM  Expansion Cohort (28 Day Cycle)  10  3  30.0  6.7  65.2  10.8  60.3  10.9  61.0

11. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates several outcomes, comparisons, time points, or analysis populations. In general, testing many hypotheses can increase the chance of at least one apparently notable finding unless the analysis plan accounts for that testing structure. The registry lists a primary endpoint and secondary outcomes that include grade 3 or above treatment-related toxicity, disease control rate, progression-free survival, and overall survival. Whether and how multiplicity was addressed is not reported in the registry record.

The primary endpoint is described at 6 months, but the record does not provide a full statistical analysis plan or explain whether other time points were examined. It does not report interim efficacy analyses. The registry record lists no analyses, and no numerical efficacy results are included in the record available for this assessment. A correction for multiple testing should not be assumed either to have been used or omitted in the actual analysis without documentation.

Missing data and censoring require separate consideration. In general, a Kaplan-Meier analysis can accommodate right-censoring when participants’ follow-up ends before the event is observed, under assumptions about the censoring process. It does not remove the need to describe why follow-up ended or to assess whether loss to follow-up could be related to outcome. The registry record does not report the number censored, reasons for censoring, missing endpoint data, or sensitivity analyses.

The study summary notes that one enrolled participant was not considered evaluable per protocol after leaving before completing a cycle for reasons other than progression or death. The record does not say how this participant was treated in each analysis, whether additional participants were excluded, or how the primary analysis population was defined. Enrollment, evaluability, and inclusion in a time-to-event analysis are distinct concepts. The analysis population and missing-data procedures 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: febrile neutropenia (10.3%), intracranial hemorrhage (10.0%), mucositis (3.4%), nausea (3.4%), neutrophil count decreased (3.4%).

The safety listing separates the primary cohort, which used a 21-day cycle, from the expansion cohort, which used a 28-day cycle. The registry lists 29 participants at risk for serious adverse events in the primary cohort and 5 affected, and 10 at risk in the expansion cohort and 1 affected. These are counts recorded for the two cohorts; they do not establish that the schedules differ in safety, because the cohorts were not randomized comparisons and their size and composition may differ.

The record also lists other adverse events for 29 of 29 participants in the primary cohort and 10 of 10 participants in the expansion cohort. Deaths are listed as 24 in the primary cohort and 7 in the expansion cohort. The registry record does not clarify the relationship between these death counts, the adverse-event listings, and the outcome follow-up. They should not be interpreted as treatment-related deaths without a stated attribution.

Listed serious adverse-event terms include febrile neutropenia at 10.3%, intracranial hemorrhage at 10.0%, and mucositis, nausea, neutrophil count decreased, and vomiting at 3.4% each. The record does not provide the denominators or detailed event-counting rules for each term in the information. Percentages for individual terms may not correspond to the proportion of participants with any serious event, since a participant can experience more than one event. The registry’s toxicity listing is descriptive and does not by itself establish causation.

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

The study was designed to estimate activity and characterize toxicities for GTX with cisplatin in metastatic pancreatic cancer. The registry identifies 6-month progression-free survival as the primary endpoint and says it was estimated using a Kaplan-Meier curve. However, the numerical endpoint estimate and its 95% confidence interval are not reported in the registry record. This prevents assessment of the observed outcome’s size or precision.

The record’s endpoint wording also needs cautious interpretation. The endpoint title and study summary refer to participants without progression, while the description refers to participants with progression or death. Those are different event interpretations. Without a clarified definition and a reported result, the record does not support a conclusion about the direction or magnitude of the primary endpoint.

Even if an estimate were available, the single-arm design would describe outcomes in the treated study population rather than establish a comparative treatment effect. The study has no concurrent control group, and the two cohorts used different schedules without randomized allocation. Comparisons with external results or between cohorts would need assumptions that are not documented in the registry record.

The study summary states an activity threshold of 50% or less as indicating insufficient activity for further study, but the remainder of the decision criterion is incomplete. Because the primary estimate is absent and the full threshold is not recorded, it is not possible to assess whether the study met its planned activity criterion. A planning illustration based on an observed effect, if considered, is not a substitute for an observed analysis and is not observed power.

The safety listings describe reported events in the two cohorts, but they do not prove that the regimen caused each event or establish comparative risks. The event denominators, attribution details, and fuller safety analysis are not reported in the registry record. Overall, the available record defines the study question and design, but does not provide enough numerical outcome information to judge the primary efficacy estimate or its statistical precision.

14. Why This Trial Matters Statistically

This study is statistically notable for the way it combines a single-arm phase II design with a time-specific survival endpoint. A single-arm design can provide an estimate of outcomes among treated participants, but it cannot directly separate the effect of treatment from the underlying course of disease, participant selection, or other influences. That distinction is central when interpreting any apparent activity in an uncontrolled study.

The endpoint is also a useful example of why outcome definitions must be internally consistent. The registry title and study summary describe the percentage alive and free of progression at 6 months, while the endpoint description refers to the percentage with progression or death. A statistical estimate cannot be interpreted reliably without knowing which event or status the percentage represents.

The study includes an initial cohort and an expansion cohort with different cycle schedules. Since the cohorts were not randomized against each other, the design does not support a clean statistical comparison of the schedules. The record’s description of a safety run-in further illustrates that early safety monitoring and later outcome estimation can serve different purposes; the details of the analysis plan are not reported in the registry record.

Finally, the registry lists a primary endpoint and multiple secondary outcomes but does not report a numerical primary result, a confidence interval, or a statistical analysis plan. This limits conclusions about precision, missing data, and multiplicity. For clinical researchers, the distinction between a study’s intended question and the information available to answer it is important: design and endpoint definitions can be described, but an efficacy estimate cannot be judged without the outcome data and analysis details.

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 of interpreting this study from the registry record is the absence of the numerical primary endpoint result and its 95% confidence interval. Without them, the estimated 6-month progression-free outcome and its precision cannot be assessed. The endpoint description is also internally ambiguous: the title and summary refer to being without progression, while the description refers to having progression or death.

The study was single arm, with no concurrent control group. It therefore cannot provide a direct comparison with another treatment or isolate a treatment effect from other explanations. The initial and expansion cohorts followed different cycle schedules and were not randomized comparisons. Cohort-to-cohort outcome differences would not establish a schedule effect.

The registry lists enrollment of 44 and describes approximately 38 evaluable participants as planned. It identifies one participant who was not considered evaluable per protocol, but does not state the analysis population used for the primary endpoint or explain how incomplete observations were handled. The numbers at risk, event counts, censoring details, and follow-up completeness for the efficacy endpoint are not reported in the registry record.

Other limitations include the lack of a reported statistical analysis plan, the absence of detail on multiplicity and interim efficacy analyses, and the incomplete description of the activity threshold. The safety listing does not include enough detail to determine how event percentages were calculated or to establish attribution. These gaps limit interpretation of both efficacy and safety beyond the descriptive information recorded.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Gemcitabine/Taxotere/Xeloda (GTX) With Cisplatin in Subjects With Metastatic Pancreatic Cancer — NCT01459614.
• Registry record last updated 2021-05-19; status last verified 2021-04.
• Results first posted 2019-01-23.
• 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 completed, single-arm phase II study of gemcitabine, Taxotere, and Xeloda with cisplatin in metastatic pancreatic cancer. Enrollment is listed as 44. The primary endpoint was progression-free survival at 6 months, estimated using a Kaplan-Meier curve, with a 95% confidence interval specified as the dispersion measure.

The study summary describes an initial cohort and an expansion cohort using different cycle schedules, as well as a safety run-in. Secondary outcomes included serious treatment-related toxicity, disease control rate, progression-free survival, and overall survival. The registry lists adverse-event counts and selected serious adverse-event terms.

The numerical primary endpoint and confidence interval are not reported in the registry record. The event definition is not fully consistent across the endpoint title, summary, and description. The analysis population, detailed missing-data handling, and statistical analysis plan are also not reported. The design and outcome definitions can be described, but the primary result’s magnitude and precision cannot be evaluated from this record.