Paclitaxel and Carboplatin or Temozolomide in Treating Patients With Stage IV Melanoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Paclitaxel and Carboplatin or Temozolomide in Treating Patients With Stage IV Melanoma (NCT00568451), a non-randomized phase 2 study. The primary outcome is Number of Participants With an Objective Tumor Status of Either a Complete Response(CR) or Partial Response (PR), According to RECIST (Response Evaluation Criteria in Solid Tumors) Criteria.

2. Clinical Question

The trial’s stated purpose was to study side effects and how well paclitaxel with carboplatin or temozolomide worked in people with stage IV melanoma. Its rationale also noted that it was not known whether the combination was more effective than temozolomide alone. That uncertainty describes the question motivating the study; it is not a result.

The four cohorts distinguish prior chemotherapy status as well as treatment. Those distinctions matter when interpreting any observed outcomes: a difference between cohorts could reflect differences in treatment, treatment history, or other differences between participants. Without random allocation, the registry record cannot support treating the cohorts as interchangeable groups differing only in treatment.

In addition to objective response, the listed outcomes include time to disease progression, survival time, duration of response among evaluable participants with an objective response, and changes in several immunologic measures. Numerical results for those outcomes are not reported in the registry record. The primary response count is also not reported, so no response estimate or treatment contrast can be described.

3. Trial Design

This was a phase II, parallel, non-randomized study with no masking. The registry lists enrollment of 12 and status as terminated. The start was recorded as 2006-06, and primary completion as 2009-05. The sponsor was Mayo Clinic.

The treatment groups were organized into four cohorts: PC among previously chemotherapy-treated participants, PC among chemotherapy-naive participants, TMZ among previously treated participants, and TMZ among chemotherapy-naive participants. PC refers to paclitaxel and carboplatin; TMZ refers to temozolomide. The record identifies carboplatin, paclitaxel, and temozolomide as drug interventions.

Parallel means that the trial record describes groups receiving different treatment regimens rather than a sequence in which every participant receives each regimen. Non-randomized means assignment was not described as random. Without random assignment, characteristics that influence treatment assignment may also be related to outcomes. The registry does not report the allocation procedure or enough information to determine how participants came to be in each cohort.

The record gives a response assessment schedule of every other cycle for the first 6 cycles, with a cycle defined as 4 weeks. It defines an objective response as a response noted on 2 consecutive evaluations at least 4 weeks apart. The response population includes enrolled participants who met eligibility criteria, signed consent, and began treatment. The registry does not report how many participants met that definition or how many contributed to each outcome.

The trial status is listed as terminated, but the reason for termination is not reported in the registry record. The protocol’s planned enrollment, sample-size justification, detailed analysis plan, and any interim analyses are not reported in the registry record. These omissions limit assessment of whether the observed enrollment or reporting matches the original study plan.

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: One patient canceled participation in the trial prior to starting Temozolomide therapy. This patient was excluded from all analysis.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All subjects enrolled, met the eligibility criteria who have signed a consent form and have begun their study treatment were evaluable for response.

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 an Objective Tumor Status of Either a Complete Response(CR) or Partial Response (PR), According to RECIST (Response Evaluation Criteria in Solid Tumors) Criteria, measured over Every other cycle of therapy (cycle=4 weeks) for the first 6 cycles of treatment.

The primary endpoint counts participants with an objective tumor response: either a complete response (CR) or a partial response (PR), according to RECIST criteria. It is a count of participants meeting either response definition, not a measure of the size of an individual tumor or the time until an event.

The registry defines CR as disappearance of all target lesions. PR requires at least a 30 percent decrease in the sum of the longest diameters of target lesions, using the baseline sum as the reference. Target lesions are measurable lesions, up to a maximum of 10, chosen to represent involved organs. The response must be noted on 2 consecutive evaluations at least 4 weeks apart.

A binary response endpoint can be useful because it gives a defined way to classify whether each participant met a response criterion during the assessment period. Counts can then describe how many participants in a group met that criterion. In general, the count alone does not show how precisely the underlying response probability is estimated; that also depends on the number of participants evaluated and on uncertainty around the estimate.

Assessments were scheduled every other cycle during the first 6 cycles of treatment, with each cycle lasting 4 weeks. The registry defines the response-evaluable population as all enrolled subjects who met eligibility criteria, signed consent, and began study treatment. Although the endpoint definition and assessment schedule are recorded, the primary response counts are not reported in the registry record. The number evaluated in each cohort and the handling of missing or incomplete assessments are also not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods determine how trial observations are converted into estimates and how uncertainty is represented. For this study, the registry records a binary primary outcome: each evaluable participant either met the objective-response definition or did not. The registry record does not report the statistical analysis plan or the primary response results, so the method used for an efficacy comparison is not reported in the registry record.

What does a binary endpoint represent?

A binary endpoint classifies an observation into one of two categories. Here, the categories are objective response and no objective response according to the RECIST-based definition. A group’s response proportion would ordinarily relate the number of participants meeting the definition to the number evaluated. That proportion is not reported in the registry record, and no proportion can be inferred without the response count and the relevant denominator.

In general, a confidence interval describes a range of values compatible with the data under a specified statistical method. For a binary endpoint, an interval around a response proportion conveys uncertainty in the estimate, with small denominators commonly yielding less precise estimates. No confidence interval for the primary response endpoint is reported in the registry record. A p-value, which under a specified model describes how unusual data would be under a stated null hypothesis, is also not reported for the response endpoint.

Why does allocation matter?

Random allocation is used in trials to make treatment groups comparable in expectation with respect to measured and unmeasured participant characteristics. This study is recorded as non-randomized. Therefore, any difference between cohorts could be associated with treatment assignment, prior chemotherapy, or other group differences. Statistical adjustment can sometimes address measured differences, but it cannot automatically remove bias from unmeasured factors. The registry record does not report an adjustment plan or adjusted treatment estimates.

The parallel structure means the listed cohorts represent different treatment groups, rather than repeated treatment periods in the same participants. The four groups are also separated by prior chemotherapy status. In a non-randomized study, those categories need to be considered when describing results; pooling or comparing them without adequate details could obscure differences in who was treated and how outcomes were assessed.

How should counts and uncertainty be read?

Counts are direct descriptions of participants meeting an outcome definition. They do not, by themselves, describe precision or establish a treatment effect. For interpretation, counts should be read alongside the number of participants assessed, the amount of missing outcome information, the timing of assessment, and an uncertainty measure such as a confidence interval. The registry record reports the planned response assessment timing and evaluability definition but does not report the primary response counts, group-specific response denominators, or confidence intervals.

For adverse events, the record gives numbers at risk and numbers affected for serious and other events in the TMZ groups, along with zero participants at risk in both PC entries. These are descriptive safety counts. In general, a comparison of adverse-event rates needs appropriate denominators and comparable exposure and follow-up. The registry information does not establish that those conditions are met, and its PC entries do not provide participant safety data for comparison.

What is over-dispersion, and is it relevant?

Over-dispersion is a general statistical issue in count data: the observed variability can exceed what a simple model assumes. It can lead to uncertainty estimates that are too narrow if the model does not account for the extra variability. The primary endpoint is recorded as a count of participants with an event, and the registry does not report a model-based analysis of that endpoint. There is therefore no basis in the registry record to determine whether over-dispersion was assessed or whether it affected any result. This general concept should not be read as a claim that over-dispersion occurred in this trial.

What does the sample-size planning information mean?

The registry record includes general assumptions for sample-size planning illustrations: a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect defined as a stated fraction of the observed effect. These are described as planning illustrations computed from the observed effect, not as the trial’s own design. For non-time-to-event outcomes, the stated convention is participants per arm; for time-to-event outcomes, the stated convention is total events using the Schoenfeld approach.

Such calculations are hypothetical planning exercises. They do not describe the number of participants actually enrolled, and they are not observed power. A planning calculation based on an observed effect also depends on that effect estimate and its uncertainty. Numerical planning results are not included in the registry record used for this explanation. The registry’s protocol-specific sample-size justification and design assumptions are not reported in the registry record.

The record lists 7 outcome measures and 1 analysis. Those counts do not disclose the analysis procedures, hypotheses, model choices, or multiplicity strategy. The statistical analysis plan, including whether estimates were adjusted, how missing outcomes were handled, and whether any interim analysis was conducted, is not reported in the registry record. The results cannot be reconstructed from the outcome definitions alone.

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("TMZ (Previously Treated)", "TMZ (Chemo Naive)")
n      <- c(2, 9)
events <- c(0, 2)
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  TMZ (Previously Treated)  2  0  0.0  0.0  84.2  0.0  65.8  0.8  70.8
ARM  TMZ (Chemo Naive)  9  2  22.2  2.8  60.0  6.3  54.7  6.7  55.6

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 7 outcome measures, including the primary objective-response count and secondary measures of disease progression, survival, duration of response, and immunologic changes. Considering multiple outcomes can raise multiplicity concerns: when many hypotheses are tested, the chance of at least one apparently notable result can increase. In general, a prespecified testing hierarchy or adjustment method can clarify how multiple tests are handled. The multiplicity plan and any adjustment method are not reported in the registry record.

The record reports 1 analysis, but that number does not specify which outcome or comparison was analyzed, what model was used, or whether multiple cohorts or outcomes were considered together. The analysis plan is not reported in the registry record. No reported p-values or confidence intervals are available to assess the inferential results.

The response population is defined as enrolled subjects who met eligibility criteria, signed consent, and began study treatment. The record does not state the number who met that definition, the number evaluated for response, or how participants with missing assessments were treated in analysis. The missing-data method and analysis populations for the secondary outcomes are not reported in the registry record.

Missing outcomes can matter because the participants with observed data may differ from those without observed data. In general, the consequences depend on why data are missing and on the method used to handle them. No assumptions or sensitivity analyses for missing data are reported in the registry record. The reason the trial was terminated is also not reported, so the available information does not establish whether termination affected follow-up or outcome reporting.

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: anemia (11.1%).

The registry separates serious adverse events from other adverse events and reports numbers at risk and numbers affected for the TMZ cohorts. Among previously treated TMZ participants, 0 of 2 had a serious event recorded, and 1 of 2 had another event recorded. Among chemotherapy-naive TMZ participants, 1 of 9 had a serious event recorded, and 8 of 9 had another event recorded.

The PC cohorts each have 0 participants at risk in the listed safety entries. Those entries therefore do not describe adverse-event experience among participants receiving PC and cannot be used as a comparison with TMZ. The registry lists anemia at 11.1% among the top serious terms; the listed information does not attribute that term to a particular cohort. Death counts are not reported in the registry record.

These figures are descriptive counts, not evidence that one regimen is safer than another. The denominators are small, and the record does not provide the event-level details or exposure and follow-up information needed for a fuller comparison. It also does not report how the adverse events were collected or whether every participant had comparable observation time. The counts should be interpreted within those limits and not as a complete account of safety.

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

The registry record establishes that this phase II study examined paclitaxel with carboplatin and temozolomide in cohorts of people with stage IV melanoma, separated by prior chemotherapy status. It also provides a defined primary response outcome, a response assessment schedule, and some safety counts for the TMZ cohorts. These registry entries describe the study and its recorded outcomes; they do not establish the treatments’ effects.

The primary response results are not reported in the registry record. It is therefore not possible to state how many participants had a complete or partial response, estimate response proportions, or compare the regimens on the primary endpoint. The listed secondary outcomes also lack numerical results in the registry record. The record does not support a conclusion that either regimen improved an outcome or failed to do so.

Even if cohort results were available, the non-randomized allocation would limit causal interpretation. Differences between groups could reflect factors other than treatment, including the prior-chemotherapy distinction used to define the cohorts. The record does not describe an adjustment strategy that would resolve those concerns. The listed safety counts are also insufficient for a comparative safety conclusion, particularly because the PC safety entries have no participants at risk.

In general, a trial result is most informative when an estimate is accompanied by its denominator, uncertainty, analysis population, and a clear account of the comparison. Those elements are not all reported for this study’s primary endpoint. A missing registry result is not equivalent to a zero effect or to evidence of no response; it is an absence of a reported estimate in the registry record. Interpretation should remain limited to the design, endpoint definitions, and descriptive safety information that are recorded.

14. Why This Trial Matters Statistically

This trial illustrates why design and reporting details are essential to statistical interpretation. Its question compares treatment approaches, but allocation is recorded as non-randomized and the cohorts are additionally divided by prior chemotherapy status. These features mean that a simple difference between groups, if reported, would not automatically represent a treatment effect. The structure of a comparison affects what conclusions can reasonably be drawn from its estimates.

The primary endpoint is clearly defined in terms of RECIST response, including lesion criteria and confirmation on consecutive evaluations. A detailed endpoint definition supports consistency in measurement. However, a definition cannot substitute for reporting the observed counts, denominators, or uncertainty. Because the response results are not reported in the registry record, the central efficacy question remains unresolved by the posted information.

The record also illustrates the distinction between describing safety events and comparing safety. Counts are available for the TMZ groups, while the PC safety entries list no participants at risk. Without comparable observations, the safety entries do not support a direct treatment comparison. Event counts need context, including the number assessed and the observation period.

Finally, the listed outcomes cover tumor response, time-to-event outcomes, and immunologic measures. Multiple outcomes can provide different kinds of evidence, but interpretation depends on a prespecified analysis plan, handling of multiplicity, and transparent reporting of estimates. Those analysis details are not reported in the registry record. The study is therefore useful as an example of how a registry can identify the intended question and endpoint definitions while leaving important result and analysis information unavailable.

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 outcomes is that the registry record does not report the primary response count or numerical results for the listed secondary outcomes. It also does not report response denominators by cohort, confidence intervals, or p-values. Thus, neither the magnitude nor the precision of an efficacy estimate can be assessed.

The non-randomized allocation limits direct comparison between cohorts. The study separates groups by prior chemotherapy status, and the registry record does not report how participants were assigned to treatment or whether analyses adjusted for differences between groups. No causal interpretation can be made from the design description alone.

Enrollment was 12, but the record does not give the number evaluated for each outcome or the flow of participants through treatment and assessment. Although the response-evaluable population is defined, the number meeting that definition and the handling of missing or incomplete assessments are not reported in the registry record. The termination reason is also not reported, leaving its implications for follow-up and outcome availability unclear.

Safety reporting is limited: the PC groups have zero participants at risk in the listed entries, death counts are not reported, and the record does not provide detailed event descriptions or comparable exposure information. Anemia is listed at 11.1% among the top serious terms without cohort attribution in the listed information. Finally, the protocol-specific statistical plan, sample-size justification, multiplicity approach, and any interim analyses are not reported in the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Paclitaxel and Carboplatin or Temozolomide in Treating Patients With Stage IV Melanoma — NCT00568451.
• Registry record last updated 2015-12-30; status last verified 2014-03.
• Results first posted 2012-03-30.
• 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

NCT00568451 is a phase II, parallel, non-randomized study sponsored by Mayo Clinic, with enrollment of 12 and status listed as terminated. The record lists treatment with paclitaxel and carboplatin or temozolomide for people with stage IV melanoma, with cohorts distinguished by prior chemotherapy status. The start was recorded as 2006-06 and primary completion as 2009-05.

The primary endpoint counts participants with complete or partial response according to RECIST criteria. The response schedule and evaluability definition are recorded, but primary response results are not reported in the registry record. Secondary outcomes include progression, survival, response duration, and immunologic changes; their numerical results are also not reported in the registry record. Safety counts are listed for the TMZ cohorts, while the PC entries have zero participants at risk. The record does not support a comparative efficacy or safety conclusion.