Carboplatin, Bevacizumab and Pemetrexed in the First-Line Treatment of Patients With Malignant Pleural Mesothelioma (MPM): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Carboplatin, Bevacizumab and Pemetrexed in the First-Line Treatment of Patients With Malignant Pleural Mesothelioma (MPM) (NCT00604461), a na phase 1, phase 2 study. The primary outcome is Number of Participants With Partial Response (PR) of Target Lesions.

2. Clinical Question

3. Trial Design

This was a Phase I/II, single-group study of carboplatin, bevacizumab and pemetrexed for first-line treatment of malignant pleural mesothelioma. The registry lists no masking and allocation as NA. Enrollment was 13, and the study status is TERMINATED. These design features are important when interpreting the results: the study was not a randomized comparison, and its outcomes cannot directly estimate a treatment difference against a concurrent control group.

The intervention description identifies tiered dose escalation followed by maintenance therapy. The listed tiers pair carboplatin AUC 4, AUC 5 or AUC 6 with bevacizumab 15 mg/Kg and pemetrexed 500 mg/m^2. The maintenance description is truncated in the registry record, so its full regimen cannot be reconstructed. The study includes both Phase I and Phase II designations, but the registry record does not report the operational rules for dose escalation, dose-limiting toxicity, or selection of a phase II dose.

The sponsor is H. Lee Moffitt Cancer Center and Research Institute. The recorded start was 2007-10 and primary completion was 2011-01. The brief summary describes a planned 24 months from first patient screened to last patient enrolled, and a planned total study length of 36 months, including enrollment, treatment and follow-up of at least 12 months. These are planned durations, not a statement of the actual duration or follow-up achieved. The registry record does not state why the trial was terminated.

The brief summary gives context that cisplatin and pemetrexed were considered standard for this disease and that off-protocol treatment would typically use cisplatin or carboplatin with pemetrexed. That context does not create a comparator within this trial. The evidence recorded for this study remains a single-group account of participants receiving the combination.

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: All participants with baseline and at least one post-baseline target lesion measurement.

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 Number of Participants With Partial Response (PR) of Target Lesions, measured over Up to 12 Months.

The primary endpoint is the number of participants with partial response of target lesions, with a timeframe of up to 12 Months. It is a binary outcome at the participant level: each assessed participant either meets the stated response definition or does not. RECIST defines partial response as at least a 30% decrease in the sum of the diameters of target lesions, using the baseline sum as the reference. The registry record reports that response was assessed in 12 patients with at least one follow-up CT scan.

A target-lesion response endpoint gives a structured way to describe change in measured tumor burden after treatment. A prespecified response definition helps make assessments consistent, rather than relying on an informal impression of improvement. A count of participants meeting the definition can be interpreted alongside the denominator and an uncertainty interval, when those are reported. The registry record does not give the number with partial response, the corresponding proportion, or an uncertainty interval. Thus, the endpoint definition and assessed population are described, but the primary response result itself is not reported in the registry record.

The registry lists number of months of progression-free survival as a secondary outcome. That is a time-to-event outcome and addresses a different feature of disease course than the binary response count. No PFS result or analysis is reported in the registry record. A response outcome and a time-to-event outcome should not be treated as interchangeable: they summarize different aspects of the observed clinical course.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry describes a participant count for a binary primary outcome: whether a participant had a partial response of target lesions. A binary endpoint classifies each assessed participant into a response category according to the stated RECIST criterion. In general, a response count is most informative when reported with its denominator and an uncertainty interval, since the count alone does not convey how precisely the response frequency is estimated. In this trial record, the response assessment population is described as 12 patients with at least one follow-up CT scan, but the count meeting the partial-response criterion is not reported in the registry record.

What does the response definition capture?

The registry uses RECIST and defines partial response as at least a 30% decrease in the sum of the diameters of target lesions, referenced to the baseline sum. This is a measurement-based classification. It is not the same as a time-to-event endpoint such as PFS, and it does not by itself describe duration of response. The timeframe is stated as up to 12 Months; the registry record does not give further detail about the timing of individual assessments or whether responses had to be confirmed at a later assessment.

In general, a confidence interval describes a range of values compatible with the observed data under a statistical model and indicates uncertainty in an estimate. It is not a guarantee that the true value lies in that range. No confidence interval is reported for the primary response outcome in the registry record. No p-value or inferential test for that outcome is reported either. Without the response count and its denominator as an outcome result, neither the size of the observed response frequency nor its statistical uncertainty can be described.

What can a single-group design estimate?

A single-group design can describe outcomes among participants enrolled and treated in that group. It does not provide a contemporaneous treatment contrast. In general, an observed outcome in a single group can be affected by participant selection, disease course, assessment timing, and other care, as well as the intervention. Random allocation to concurrent groups can help support a causal comparison by balancing factors on average, but this study has allocation listed as NA and is identified as a single-group study. The record does not report a controlled comparison.

The Phase I/II designation and the dose-escalation description indicate a study that includes dose evaluation as well as a phase II component. The listed carboplatin tiers are AUC 4, AUC 5 and AUC 6; bevacizumab is listed at 15 mg/Kg and pemetrexed at 500 mg/m^2. General dose-escalation methods use observed tolerability to guide movement among dose levels, but the registry record does not specify the rules used, the toxicity window, the decision procedure, or a selected dose. Those methods and decisions should not be inferred from the tier list alone.

How should the planning assumptions be read?

The planning illustration is described as using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It describes an effect to detect as a stated fraction of the observed effect. For time-to-event planning, the stated quantity is a total number of events using the Schoenfeld approach; other planning quantities are participants per arm. These are planning assumptions, not observed trial outcomes and not observed power. They also do not describe the actual design, which is single-group. The registry record does not report the sponsor's sample-size justification or an original two-arm design. No planning estimate should be interpreted as the trial's enrolled sample size or as evidence of the probability that the study detected an effect.

For a binary outcome, a normal approximation is a general statistical method that can be used in some settings to approximate the sampling distribution of an estimate. Its adequacy depends on the amount and distribution of data. The registry record does not identify a method used to calculate a confidence interval or test for the primary endpoint, and it reports no such analysis. Over-dispersion is a concern in models where variability exceeds that expected under the assumed distribution; it is not a substitute for the missing response count, nor does the record identify an over-dispersion model for this binary endpoint.

PFS is listed as a secondary time-to-event outcome. In general, time-to-event methods account for the time until an event and can incorporate participants whose event status is not known through the end of follow-up by censoring their observed time. A PFS estimate requires a definition of the event, follow-up information, and analysis choices. The registry record does not report the PFS definition in detail, the event data, censoring information, estimate, or analysis method. Consequently, no time-to-event result can be interpreted from the registry entry.

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("Dose Escalation Followed by Maintenance Therapy")
n      <- c(12)
events <- c(4)
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  Dose Escalation Followed by Maintenance Therapy  12  4  33.3  9.9  65.1  13.8  60.9  13.9  61.4

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 2 outcome measures, including the primary partial-response count and the secondary outcome of number of months of PFS. It records 0 analyses. No statistical analysis plan, multiplicity adjustment, hierarchy among secondary outcomes, or interim-analysis plan is reported in the registry record. It is therefore not possible to determine whether multiple outcomes were formally tested, how any tests were ordered, or whether a reported significance level was adjusted for multiple comparisons. No inferential results are reported that would permit that assessment.

The response population is described as participants with baseline and at least one post-baseline target-lesion measurement. The registry says response was assessed in 12 patients with at least one follow-up CT scan, while enrollment was 13. It does not explain why the response assessment population is smaller than enrollment, identify the outcome status of participants without an assessment, or describe any method for handling missing measurements. No assumption about missing outcomes, imputation procedure, or sensitivity analysis is reported in the registry record.

The record also does not describe whether all enrolled participants were included in a particular analysis population, how treatment discontinuation was handled, or what follow-up was available for PFS. These details affect how outcomes should be interpreted, especially when the group assessed for an endpoint may differ from the full enrolled group. The registry record does not report an intention-to-treat analysis or another named analysis population. The response result and PFS analysis therefore cannot be reconstructed from the recorded information.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: fatigue (15.4%), anemia (7.7%), death (7.7%), infection (7.7%), leukopenia (7.7%).

The safety summary lists serious adverse events for a group with 13 participants at risk: 4 participants were affected. Other adverse events are listed for 13 participants at risk, with 13 affected. These are participant counts in the trial group, not comparisons between treatments. The record also lists serious-event terms and their percentages: Fatigue 15.4%, Anemia 7.7%, Death 7.7%, Infection 7.7%, Leukopenia 7.7% and Thromboembolism 7.7%.

Such entries should be read as descriptions of reported events in the enrolled group. A serious adverse event is not necessarily an event caused by the study treatment; the registry record does not establish causality for each listed term. The term-level percentages do not provide timing, duration, severity beyond the serious classification, or details about whether participants had multiple events. Although Death is listed among serious terms, the deaths field is null, so a separate death count is not reported in the registry record.

The group size and lack of a control group limit what these safety counts can establish. They describe the events recorded among participants in this study but do not provide a comparative risk estimate or identify excess risk relative to another regimen. The registry record also does not provide a complete narrative of adverse-event ascertainment or follow-up.

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

The record establishes that a small, single-group Phase I/II study of carboplatin, bevacizumab and pemetrexed was registered for first-line malignant pleural mesothelioma, enrolled 13 participants and was terminated. It specifies a primary RECIST partial-response endpoint and reports that response was assessed in 12 patients with at least one follow-up CT scan. It does not report how many participants had a partial response. Therefore, the primary endpoint's observed result and precision cannot be stated from the registry record.

The registry also lists PFS as a secondary outcome but does not report a PFS result. No analysis is listed. This means the available entry cannot establish the frequency or timing of progression-free survival events, nor can it support an inferential comparison. The absence of a reported result is not evidence that the outcome was favorable or unfavorable; it is a limit on what can be concluded from the recorded information.

The safety information describes serious adverse events in 4 of 13 participants and other adverse events in 13 of 13, along with several serious-event terms and percentages. These are observations from the trial group. They do not establish treatment causation, and without a concurrent comparator they cannot quantify a difference in safety relative to another treatment. The listed death term should not be interpreted as a definitive death count because the deaths field is null.

More broadly, a single-group trial can characterize what was recorded among its participants, but it cannot by itself show how outcomes would have differed under another treatment. The registry record does not report the protocol's statistical plan, handling of missing outcomes, or reason for termination. Those omissions further constrain interpretation. The posted information supports a description of the trial design, outcome definitions and safety counts, but not a treatment-effect estimate or a conclusion about comparative benefit or harm.

14. Why This Trial Matters Statistically

This study illustrates why outcome definitions and outcome results must be distinguished. The registry defines partial response using a specific RECIST criterion and identifies the timeframe and assessment population, but it does not report the response count. A clearly defined endpoint is necessary for consistent measurement, yet a definition alone cannot show how participants fared or how precise an estimate would be.

The design also matters. The trial is registered as a single group, with allocation listed as NA and no masking. This design can generate descriptive information about treated participants and can be used in a phase I/II setting that includes dose evaluation. It does not provide a randomized contrast that separates the intervention's outcomes from other influences. Readers should not treat the study's outcome descriptions as an estimate of comparative effect.

Safety summaries are also easier to misread without their denominators and context. The serious-event count is recorded among 13 at risk, while the registry separately lists serious-event terms and percentages. Those entries convey observed events, not causal attribution or a comparison with another regimen. A null deaths field and absence of detail about event timing are reminders that a registry summary may not answer every safety question.

Finally, the record shows how missing analysis information limits statistical interpretation. It does not report the analysis plan, missing-data methods, multiplicity approach, or PFS result. The study's terminated status is recorded, but the reason is not reported. Distinguishing recorded details from unreported details is essential when evaluating a clinical trial, especially when the primary outcome result itself is absent.

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 the primary endpoint is that the registry record gives the response definition and assessment population but not the number of participants with partial response. Without that result, neither a response frequency nor its uncertainty can be assessed. The record does not report a confidence interval, p-value, or other primary-outcome analysis.

The trial enrolled 13 participants and had a single-group design. The absence of a concurrent comparator means that the record cannot support a controlled treatment comparison. The registry lists the study as terminated but does not report why it ended, whether termination affected planned enrollment or follow-up, or whether outcome reporting was complete.

Response was assessed in 12 patients with at least one follow-up CT scan, but the registry does not explain the assessment status of all enrolled participants or describe missing-data handling. The time window is given as up to 12 Months, but the assessment schedule and any response-confirmation procedure are not stated. PFS is listed as a secondary outcome without a reported definition in detail, estimate, event data, or analysis.

The intervention description includes dose tiers and refers to maintenance therapy, but the maintenance text is incomplete. The record does not explain the dose-escalation rules, dose-selection results, or analysis plan. It also does not report multiplicity control or interim-analysis procedures. These gaps prevent reconstruction of the trial's statistical methods and restrict interpretation to the information explicitly recorded.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Carboplatin, Bevacizumab and Pemetrexed in the First-Line Treatment of Patients With Malignant Pleural Mesothelioma (MPM) — NCT00604461.
• Registry record last updated 2017-03-23; status last verified 2011-11.
• Results first posted 2011-12-26.
• 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

NCT00604461 is a Phase I/II, single-group trial of carboplatin, bevacizumab and pemetrexed for first-line treatment of malignant pleural mesothelioma. The registry lists 13 enrolled participants, no masking, allocation as NA, and status TERMINATED. The sponsor is H. Lee Moffitt Cancer Center and Research Institute.

The primary endpoint is the number of participants with partial response of target lesions, defined using RECIST and assessed up to 12 Months. Response was assessed in 12 patients with at least one follow-up CT scan, but the response count is not reported in the registry record. PFS is listed as a secondary outcome without a result. Serious adverse events affected 4 of 13 participants, and other adverse events affected 13 of 13. No statistical analyses are listed, and the reason for termination is not reported in the registry record.