Chemotherapy, Irradiation, Cell Infusions, and Interleukin-2 to Treat Metastatic Melanoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Chemotherapy, Irradiation, Cell Infusions, and Interleukin-2 to Treat Metastatic Melanoma (NCT00314106), a non-randomized phase 2 study. The primary outcome is Complete Response.

2. Clinical Question

Statistically, the study is a description of outcomes under a complex treatment strategy, organized by prior IL-2 exposure. It does not establish what would have happened to the same patients under a different treatment. Any difference between the groups could reflect prior therapy or other differences between patients, rather than the regimen itself. The record does not report details of how the response assessment was conducted or how patients with incomplete follow-up were handled.

3. Trial Design

The registry describes a completed, non-randomized, parallel-group Phase II study of a treatment strategy for metastatic melanoma. Enrollment was 26. The two groups were defined by prior IL-2 therapy: one group had received prior IL-2 and the other had not. Both group descriptions include TBI 1200 cGy, TIL and high-dose IL-2. The record gives 18 at risk for adverse-event reporting in the prior-IL-2 group and 8 in the group without prior IL-2.

The treatment plan included a preparative regimen intended to deplete lymphocytes: cyclophosphamide (60 mg/kg/day x 2 days intravenous), fludarabine (25mg/m^2/day IV X 5 days) and 1200 cGy total body irradiation. The record also lists melanoma-reactive TIL, IL-2 and stem-cell collection and reconstitution in its description of the study approach. The brief description states that stem-cell collection followed five days of filgrastim. The posted arm descriptions are truncated, so the complete treatment sequence and any further operational details are not reported in the registry record.

Allocation was non-randomized and masking was none. Thus, group membership was not assigned by random allocation, and the record does not describe blinded assessment. The two groups reflect prior IL-2 exposure, not two randomized alternatives. Prior treatment can be associated with other patient or disease characteristics, and those characteristics may also relate to response or adverse events. The registry record does not report adjustment for such differences.

The primary outcome was complete response, defined as disappearance of all target lesions, with a timeframe of 33 months. The primary outcome description does not give the observed response count. The record lists a secondary outcome of participants with adverse events. Although the study status is completed, the primary result and statistical analyses are not reported in the registry record. The study start was 2006-04 and primary completion was 2009-02.

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

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 Complete Response, measured over 33 months.

The primary endpoint is complete response (CR), recorded as a number of participants. The registry definition is disappearance of all target lesions. This endpoint asks whether a patient achieved a substantial tumor-response state under the study’s treatment approach; it is not a continuous measurement of tumor size or a time-to-event endpoint. The stated timeframe is 33 months.

In general, a binary response endpoint reduces each participant’s outcome to whether the specified response occurred. Reporting the count alongside the number assessed would allow readers to understand how often the endpoint was observed in the study population. A response fraction can be useful in an early-phase study because it describes the observed activity of a treatment approach and can inform questions for later research. It does not, by itself, establish that treatment caused the response, especially when there is no randomized comparator.

The registry record does not report the number of complete responses, the number of participants evaluated for response, or a response estimate with uncertainty. It also does not report whether response was confirmed, how lesions were selected or assessed, or how death, progression, withdrawal, or missing assessments were treated in the endpoint definition. Those details matter because different response-assessment and missing-data rules can change which participants count in the endpoint denominator and whether an apparent response is considered complete.

Accordingly, the endpoint is identifiable from the record, but its observed result and precision cannot be interpreted from the registry information. The endpoint description states the question the investigators intended to address, not the answer.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record identifies a binary primary outcome and lists participant counts for adverse events. It does not report a primary-outcome analysis, an effect estimate, a confidence interval, or a p-value. The methods below explain general statistical ideas relevant to this kind of study; they are not claims about analyses performed in the trial.

What does a binary endpoint represent?

In general, a binary endpoint records whether a prespecified event occurred for each participant. For complete response, the event is defined as disappearance of all target lesions. A count of participants meeting that definition is informative only when interpreted with the number whose endpoint status was assessed. The ratio of those quantities is commonly used to express the observed response fraction, but neither the response count nor the assessment denominator is reported in the registry record. No response fraction can therefore be interpreted from this record.

A binary endpoint also depends on its operational definition. The record supplies the complete-response definition and timeframe, but does not state how target lesions were selected, how assessments were performed, whether confirmation was required, or what rules applied to participants without an assessable result. These details affect how the event is classified and how the analysis population is formed. They are not reported in the registry record.

What would a confidence interval add?

In general, a confidence interval describes uncertainty around an estimate under a statistical model and its assumptions. For a response fraction, it would help distinguish a precise estimate from one compatible with a broad range of underlying response probabilities. It does not mean that the true value is known to lie within one particular interval with a specified probability after the data have been observed. It also does not correct for selection bias, confounding, incomplete follow-up, or errors in endpoint classification. The record does not report a confidence interval for complete response.

With small groups, uncertainty is often important because each outcome contributes meaningfully to the estimate. A point estimate alone can give an overly definite impression. The enrollment and adverse-event group counts in the registry do not substitute for a reported primary-outcome estimate, because the number assessed for complete response is not stated.

Why does non-random allocation matter?

In general, random allocation is used to make treatment groups comparable, on average, with respect to both measured and unmeasured characteristics. Here, the record describes groups according to prior IL-2 exposure and states that allocation was non-randomized. Patients in these groups may differ in ways that relate to response or safety. A comparison between the groups could therefore mix treatment history, patient selection and other factors. The record does not report an adjustment method or a causal analysis that would account for these potential differences.

Because both groups are described as receiving the treatment strategy, their comparison is not a randomized test of the regimen against a control treatment. Observed outcomes can describe what occurred among participants in each group, but cannot isolate the effect of the treatment combination from other influences. No comparative primary-outcome estimate is reported in any case.

What can be learned from the safety counts?

In general, event counts are descriptive. The registry lists the number at risk and the number affected by serious adverse events and other events in each group. A count of affected participants differs from a count of event episodes: one participant may experience more than one event, and the record does not specify whether its counts refer to unique participants or episodes beyond the labels shown. The event-collection window and rules for attributing events to treatment are also not reported.

Over-dispersion is a general concern when an analysis models counts of repeated events. It occurs when the observed variability exceeds what a simple count model assumes, often because some participants experience more events than others. A model that ignores over-dispersion can understate uncertainty. The registry record does not provide repeated-event data or describe a count model, so this issue cannot be evaluated for this study.

What do planning illustrations mean?

In general, a sample-size calculation describes how many participants might be needed under stated assumptions to detect a specified effect with a chosen error rate and power. The planning approach described for the posted calculations uses a two-sided significance level of 0.05, equal allocation, a normal approximation and effects expressed as fractions of the observed effect. These are planning illustrations based on the observed effect, not the trial’s own design justification and not observed power. The registry record does not report a sample-size justification for the study, and no planning values are stated in the record. Such calculations cannot recover an unreported endpoint result or establish that the study was adequately powered.

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("TBI 1200 cGy + TIL +HD IL-2, Prior IL-2", "TBI 1200 cGy + TIL +HD IL-2, no Prior IL-2")
n      <- c(18, 8)
events <- c(7, 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  TBI 1200 cGy + TIL +HD IL-2, Prior IL-2  18  7  38.9  17.3  64.3  20.3  61.4  20.3  61.6
ARM  TBI 1200 cGy + TIL +HD IL-2, no Prior IL-2  8  3  37.5  8.5  75.5  13.7  69.4  13.7  70.1

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists a primary complete-response outcome and a secondary outcome of participants with adverse events. It does not report a statistical analysis plan, the number of planned comparisons, a multiplicity adjustment, or an interim-analysis plan. The record lists no analyses. It is therefore not possible to determine whether other outcomes or comparisons were prespecified, or how any multiple testing was handled.

In general, evaluating several outcomes or comparisons can increase the chance of a false-positive finding if each is tested without accounting for the full set of tests. A prespecified primary endpoint and analysis help clarify which result is confirmatory and how other findings should be interpreted. Since the registry record does not provide the analysis plan or a primary endpoint result, no claim can be made about control of multiplicity or the statistical status of any finding.

The registry record also does not describe missing-data rules or analysis populations for complete response. It does not state how participants who lacked an assessment, left treatment, died, or had an otherwise unknown response status were handled. The record gives enrollment, but it does not report the primary-outcome denominator or the number with an assessable endpoint. The meaning of any response percentage would depend on whether its denominator included all enrolled participants or only those assessed, and that information is not reported.

In general, missing outcomes can bias an estimate when the chance of being missing relates to a participant’s outcome or health status. Different assumptions and analysis methods can produce different results. A complete-case analysis, for example, describes participants with observed results but may not represent everyone enrolled. No missing-data method or analysis population is reported in the registry record, so the effect of missing information cannot be evaluated.

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: thrombotic microangiopathy (25.0%), somnolence/depressed level of consciousness (25.0%), anorexia (16.7%), hemoglobin decreased (11.1%), weight loss (11.1%).

The registry reports serious adverse-event counts separately for the groups defined by prior IL-2 exposure. In the group with prior IL-2, 7 of 18 participants at risk were affected; in the group without prior IL-2, 4 of 8 were affected. For other adverse events, the record lists 18 of 18 and 8 of 8 affected, respectively. These are descriptive counts, not evidence that one group had a higher or lower treatment-related risk.

The registry also lists serious-event terms with reported percentages: thrombotic microangiopathy, 25.0%; somnolence/depressed level of consciousness, 25.0%; anorexia, 16.7%; hemoglobin decreased, 11.1%; weight loss, 11.1%; and peripheral nerve infection, 11.1%. The denominator and whether these percentages refer to participants or event episodes are not reported in the registry record. They should be read as registry-reported figures, not recalculated rates.

The record does not report deaths in the adverse-event data. The event-collection period, severity criteria, attribution to treatment and handling of repeated events are not reported in the registry record. Serious-event counts therefore describe what the registry lists, but do not establish cause, comparative risk or the full safety profile.

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

The registry record describes a Phase II, non-randomized study of a complex treatment approach for metastatic melanoma and defines complete response as disappearance of all target lesions. It also provides adverse-event counts by prior IL-2 group. These details make the study design and the intended primary outcome understandable, but the record does not report the observed complete-response result or its statistical uncertainty.

As a result, the record cannot establish how many participants achieved complete response, how frequent that outcome was, or how precisely it was estimated. It cannot show whether the response fraction differed between groups, because no response counts or comparative analysis are reported. Completion status does not resolve these gaps. The primary result is not reported in the registry record.

Even if group-specific response results were available, the design would limit causal interpretation. Participants were not randomized, and group membership reflected prior IL-2 exposure. Differences between groups could be related to their prior treatment or other characteristics. The record does not report methods to adjust for such differences. The comparison would not be equivalent to a randomized comparison of this treatment strategy with an alternative.

The adverse-event figures offer descriptive safety information, but they do not establish that the treatment caused an event or quantify comparative risk. The registry does not define the event-collection window or the denominators for the listed serious-event terms. Nor does it report death counts in the adverse-event data.

In general, a study result is most useful when the endpoint definition, observed count, denominator, analysis method and uncertainty measure can be considered together. Here, the endpoint definition and timeframe are stated, but the result and analysis are not. The record therefore supports an account of what the study set out to assess and what safety counts it registered; it does not support a conclusion about the magnitude, certainty or cause of complete response.

14. Why This Trial Matters Statistically

This study raises several statistical issues common to early-phase research on complex treatment strategies. The primary endpoint is a clinically interpretable binary outcome, but interpretation depends on knowing both the number of complete responses and the number of participants assessed. The registry record supplies the endpoint definition and timeframe, yet does not report those outcome data. This illustrates why a clearly stated endpoint and a posted result are both necessary for readers to evaluate evidence.

The non-randomized design also shapes the questions the data can answer. The groups were defined by prior IL-2 exposure, and both were assigned the treatment combination described in the registry. This allows the study to organize observations by treatment history, but does not create a randomized control comparison. Statistical adjustment, if used, would not automatically remove all confounding, and the record does not report an adjustment analysis.

Safety reporting provides another useful distinction: participant counts affected by events are not necessarily counts of event episodes, and percentages need a clear denominator to be interpreted. The record lists serious events and other events separately, but does not specify all of the definitions needed to assess their rates or compare them across groups.

Finally, study completion, enrollment and outcome definitions do not substitute for an analysis plan or results. The record does not report a sample-size justification, interim analyses, missing-data rules or the primary endpoint analysis. This makes it impossible to assess the precision of the intended efficacy outcome or the role of any planned statistical decisions. The case underscores the importance of transparent endpoint denominators, analysis methods and uncertainty estimates, particularly when allocation is not randomized.

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.

Several study-specific limits constrain interpretation. First, the registry record does not report the primary complete-response count, the response denominator or a statistical analysis. The endpoint is defined, but its result and precision cannot be assessed. No response estimate, confidence interval or p-value is available in the record.

Second, the study was non-randomized. Its parallel groups differ by prior IL-2 exposure rather than by random assignment to the treatment combination or a control. This leaves open the possibility that group differences reflect patient selection, prior treatment or other characteristics. The registry record does not report adjusted comparisons or methods intended to address confounding.

Third, the treatment is a combination of components, including chemotherapy, total body irradiation, TIL, stem-cell reconstitution and high-dose IL-2. The design described in the record does not isolate the contribution of an individual component. The posted arm descriptions are truncated, and the complete treatment sequence is not reported in the registry record.

Fourth, outcome-assessment procedures and missing-data rules are not reported. The record does not explain how response was assessed, whether responses required confirmation, or how participants without an assessable result were handled. These details are needed to judge the endpoint’s reproducibility and the applicability of any result.

Finally, the adverse-event record does not report the collection period, attribution rules or denominator definition for the serious-event terms. The record does not report deaths in the adverse-event data. It also does not report a statistical plan, sample-size justification, interim-analysis plan or multiplicity approach. These gaps limit assessment of both the safety summaries and the intended efficacy analysis.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Chemotherapy, Irradiation, Cell Infusions, and Interleukin-2 to Treat Metastatic Melanoma — NCT00314106.
• Registry record last updated 2012-10-18; status last verified 2012-09.
• Results first posted 2012-10-18.
• 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, non-randomized Phase II study of a treatment strategy for metastatic melanoma, with enrollment of 26. The parallel groups were defined by prior IL-2 exposure. Both group descriptions include TBI 1200 cGy, TIL and high-dose IL-2, along with a preparative regimen of cyclophosphamide and fludarabine.

The primary endpoint was complete response, defined as disappearance of all target lesions, with a timeframe of 33 months. The registry record does not report the complete-response count, denominator or analysis. It lists adverse-event counts by group and serious-event terms with percentages, but does not report the event-period or the denominator definition for those terms. The record lists no analyses and does not describe the statistical plan, missing-data methods or sample-size justification. Thus, the design and endpoint are described, while the primary result and its uncertainty are not reported in the registry record.