High-dose Chemotherapy for Poor-Prognosis Relapsed Germ-Cell Tumors: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for High-dose Chemotherapy for Poor-Prognosis Relapsed Germ-Cell Tumors (NCT00936936), a non-randomized phase 2 study. The primary outcome is Number of Participants With 2-year Event-Free Survival (EFS).

2. Clinical Question

3. Trial Design

This is a phase 2 clinical research study, labeled nonrandomized and single-group in the registry. The record lists 64 enrolled participants and a planned enrollment of up to 67 patients in the study summary. It gives a status of completed, a start date of 2009-06-02, and a primary completion date of 2024-01-11. Those record details describe the study’s registration and conduct; they do not supply an efficacy result.

The intervention is described as two cycles of high-dose chemotherapy, each followed by stem-cell infusion. The first cycle is Gemcitabine/Docetaxel/Melphalan/Carboplatin plus PBPC. The second is Ifosfamide/Carboplatin/Etoposide plus PBPC. The intervention list also includes mesna and stem-cell transplant. The brief summary describes the study as investigational and says the safety of the combinations will be studied.

Although the overall design is labeled single-group, the serious adverse-event data are presented under two cohort titles: one for HD GeM-DMC+ ICE + Bevacizumab for Refractory GCT and one for HD GeM-DMC+ ICE for Refractory GCT. Their presence does not establish a randomized comparison. The registry explicitly warns that the arms are not a like-for-like control comparison. The record does not explain how cohort assignment occurred or provide enough information to interpret cohort differences as causal effects.

The primary outcome is a count of participants with 2-year event-free survival. Overall Survival is listed as a secondary outcome. The registry record reports no outcome analyses and does not state the methods used to estimate or compare these outcomes.

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

Reasons for not completing this period:

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 2-year Event-Free Survival (EFS), measured over 2 Years.

The primary endpoint is titled “Number of Participants With 2-year Event-Free Survival (EFS),” with the unit listed as participants and the measure described as a count. Its timeframe is 2 Years. The event-free survival description starts the clock at the first day of High-Dose Course Cycle #1 (Day -6) and defines an event as tumor progression, relapse, or death from any cause.

Event-free survival is used in clinical research to capture the time until a specified unfavorable event, combining several possible ways that a participant may experience disease progression or die. In general, such an endpoint can describe how long participants remain free of the events included in its definition. A count of participants meeting an event-free survival definition at a specified time is a simpler summary than a full time-to-event curve: it reports how many meet that definition at the stated time, but does not by itself show the timing of events or the pattern of follow-up.

There is a distinction within the registry description: the endpoint is recorded as a participant count, while its text says event-free survival is estimated from a starting day until an event. The record does not report the count, an estimate, or an uncertainty interval. It therefore establishes what the endpoint was intended to measure, but does not show the observed 2-year result. Overall Survival is listed as a secondary outcome, but its result is also not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry labels this study phase 2, nonrandomized, and single-group. These labels describe the registered design, but do not tell readers which statistical procedures were used to analyze outcomes. The record lists zero analyses. It does not report an analysis plan, model, adjustment strategy, or method for handling follow-up. The absence of those details matters because an endpoint’s definition alone does not tell readers how the estimate was calculated or how uncertainty was assessed.

What does the primary endpoint count?

The primary outcome is recorded as a count of participants with 2-year event-free survival. Its text defines the starting point as the first day of High-Dose Course Cycle #1 (Day -6), and the events as tumor progression, relapse, or death from any cause. A participant would meet the event-free definition only if none of the specified events occurred during the stated period, under the endpoint’s intended rules. The record does not report the number meeting that definition.

In general, a fixed-time count and a time-to-event analysis answer related but different questions. A count at a specified time classifies participants by whether an event has occurred by that point. A time-to-event approach uses event timing and can account for differing lengths of observation, depending on the chosen method and assumptions. The registry’s endpoint description refers to event-free survival being estimated, but its measure is coded as a count. It does not state which estimation method was used, how participants with incomplete observation were handled, or whether any survival curve was calculated.

What can the design establish?

Random allocation can help balance known and unknown participant characteristics across treatment groups, on average. This study is labeled nonrandomized, and its overall model is single-group. The registry separately lists serious adverse-event data for two cohorts and explicitly cautions that they are not a like-for-like control comparison. General comparisons between nonrandomized groups are vulnerable to confounding: differences in outcomes can arise from differences in participants or their circumstances rather than treatment alone. The record does not describe an adjustment method that would address this concern.

The cohort labels include different treatment descriptions, but the registry does not report how people entered each cohort or whether cohort membership was determined by a prespecified rule. It also does not report a statistical comparison between them. Their counts are therefore descriptive entries, not evidence from a randomized head-to-head test.

How should uncertainty be understood?

In general, an estimate describes the observed data, while a confidence interval communicates the statistical uncertainty around that estimate under a specified method. A narrower interval generally indicates greater precision, whereas a wider interval indicates greater uncertainty. Neither an estimate nor its interval can be interpreted without knowing what population and endpoint they refer to and how the data were analyzed. The registry record does not report an efficacy estimate or confidence interval for the primary outcome.

A p-value is a measure tied to a statistical test and its assumptions; it is not a measure of effect size or clinical importance. No efficacy p-value is reported in the registry record. Since no analysis is listed, readers should not infer that a test was conducted or infer a result from the endpoint’s presence in the registration.

What do the planning assumptions mean?

The planning assumptions describe illustrations based on a stated fraction of an observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. For time-to-event tables, the planning quantity is total events; for other outcomes, it is participants per arm. These are planning illustrations, not observed power and not the study’s reported enrollment rationale. The registry record does not report a sample-size justification or the statistical plan used to design the trial. It also does not report an efficacy estimate from which the stated observed effect could be assessed.

In general, a sample-size calculation depends on assumptions about the effect to detect, variability or event frequency, the analysis method, and the chosen error rates. Changing those assumptions changes the planning requirement. An illustration based on an observed effect is especially not a substitute for a prospective design rationale: it describes a hypothetical planning exercise conditional on that effect, rather than proving that the completed study had a particular probability of detecting an effect. The planning figures should not be described as observed power.

Why does over-dispersion matter in general?

For count outcomes, a simple statistical model may assume that the variability is tied closely to the average count. If the actual counts vary more than that model expects, the data are over-dispersed. Ignoring over-dispersion can make standard errors too small and uncertainty intervals too narrow. This is a general consideration for count-data methods, not a reported feature of this study. The registry identifies the primary result as a participant count but does not report a count model or evidence about over-dispersion.

Overall, the record gives endpoint definitions and design labels but not the analysis methods or efficacy results needed for a full statistical assessment. Methodological explanations can clarify what the endpoint and planning assumptions mean, but they cannot recover unreported estimates or establish which analysis was actually performed.

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("Cohort 1: HD GeM-DMC+ ICE + Bevacizumab for Refrac", "Cohort 2: HD GeM-DMC+ ICE for Refractory GCT")
n      <- c(41, 23)
events <- c(28, 11)
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  Cohort 1: HD GeM-DMC+ ICE + Bevacizumab for Refrac  41  28  68.3  51.9  81.9  53.0  80.4  52.9  80.4
ARM  Cohort 2: HD GeM-DMC+ ICE for Refractory GCT  23  11  47.8  26.8  69.4  29.2  67.0  29.1  67.2

11. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures, with Number of Participants With 2-year Event-Free Survival as the primary outcome and Overall Survival as a secondary outcome. It reports zero analyses. No analysis plan is reported in the registry record, including whether multiple outcomes or multiple analyses were considered together. The record does not describe a multiplicity adjustment or explain how the primary and secondary outcomes were prioritized in statistical testing.

In general, testing several outcomes can increase the chance of obtaining an apparently notable result by chance if each is assessed separately. A prespecified strategy can clarify which outcome is primary and whether adjustments or other decision rules are intended. That general issue cannot be resolved for this trial from the endpoint list alone. No outcome estimates, p-values, or analysis details are reported for assessing how multiplicity was handled.

The registry record does not report missing-data procedures, censoring rules, or how incomplete follow-up was handled for event-free survival or Overall Survival. It does not describe the analysis population, and the population field for the primary endpoint is blank. The record therefore does not specify whether analyses were intended to include all enrolled participants, only participants meeting particular criteria, or another defined group.

These omissions limit interpretation. In general, missing observations can affect an estimate when the chance of missingness is related to a participant’s outcome or circumstances. The direction and size of any such effect depend on the data and assumptions; they cannot be inferred from this record. The listed enrollment and safety counts do not resolve who contributed to the efficacy endpoint or how incomplete outcome data were treated.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: bacterial (17.1%), fungal (14.6%), bleeding (no gi no pul) (9.8%), creatinine increased (8.7%), gastrointestinal bleeding (4.9%).

The registry lists serious adverse-event data for two cohorts. For Cohort 1: HD GeM-DMC+ ICE + Bevacizumab for Refractory GCT, it reports 41 at risk, 7 affected, and 22 deaths. For Cohort 2: HD GeM-DMC+ ICE for Refractory GCT, it reports 23 at risk, 2 affected, and 12 deaths. These are cohort-specific entries and should be read as descriptive safety information, not as a randomized treatment comparison. The record warns that the arms are not a like-for-like control comparison.

The registry also reports selected serious adverse-event terms with percentages: Bacterial, 17.1%; Fungal, 14.6%; Bleeding (no GI no PUL), 9.8%; Creatinine increased, 8.7%; Gastrointestinal bleeding, 4.9%; and Oral Mucositis, 4.9%. The record does not provide enough detail to determine how these terms overlap, what time period they cover, or whether the percentages refer to a particular cohort or combined population. They should not be added or treated as mutually exclusive categories.

Counts of deaths do not establish that deaths were caused by treatment. The record does not report causal attribution for the listed deaths or adverse events. Differences between cohort entries also cannot establish comparative safety because allocation was nonrandomized and the registry cautions against a like-for-like comparison.

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

The registry record documents a study of two cycles of high-dose chemotherapy and identifies 2-year event-free survival as the primary endpoint. It also provides selected serious adverse-event summaries. These details clarify what the study set out to examine and what safety entries were registered, but the record does not report the primary endpoint result or an Overall Survival result.

That absence sets a clear boundary on interpretation. There is no posted efficacy estimate, uncertainty interval, or analysis result in the registry record to describe how event-free survival turned out. Completed status means the record marks the study as completed; it does not establish a particular outcome. Likewise, the fact that an endpoint is named does not show whether it was met or how precisely it was estimated.

The adverse-event entries show that the registry recorded affected counts and deaths by cohort, as well as percentages for selected serious events. They do not establish that treatment caused those events, nor do they provide a valid randomized comparison of the cohort regimens. Nonrandomized cohort data may reflect differences other than treatment. The registry itself cautions that these arms are not a like-for-like control comparison.

The study should therefore be understood as a registered phase 2 investigation with a specified event-free survival endpoint and descriptive safety information, rather than as a source of a reported efficacy estimate in this record. The missing analyses also prevent readers from evaluating precision, statistical significance, or the handling of incomplete outcome data. No conclusion about treatment benefit or lack of benefit follows from the information recorded.

In general, a clinical result is most informative when the outcome definition, analysis population, estimate, uncertainty, and analysis method can be considered together. The registry record provides the endpoint definition and selected design details, but not the efficacy estimate or methods needed for that fuller assessment.

14. Why This Trial Matters Statistically

This trial illustrates why a registered endpoint and a reported result are not the same thing. The record clearly names 2-year event-free survival as the primary outcome and defines the events, but it does not report the outcome count or an analysis. Readers can understand the question the study was designed to address without being able to judge the observed result or its precision.

It also shows why design labels matter when interpreting group data. The overall study is labeled nonrandomized and single-group, while the serious adverse-event record includes two cohort entries. The registry’s warning that the cohorts are not like-for-like controls is important: separate group summaries do not automatically create a valid comparative estimate. Statistical interpretation depends not only on what outcomes are counted, but also on how participants entered groups and how comparisons were planned and analyzed.

Finally, the record distinguishes descriptive safety reporting from efficacy analysis. Serious adverse-event counts and percentages may help characterize what was recorded, but they cannot replace an unreported event-free survival result. Nor do the safety entries establish that a treatment caused an event or that one cohort is safer than another.

For clinical researchers and biostatisticians, the central lesson is to separate design, endpoint definition, observed results, and statistical inference. Each addresses a different question. In this case, the registry gives useful design and endpoint context, while leaving the primary outcome analysis and important methodological details unreported.

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 this trial from the registry record is that the primary endpoint result is not reported. The record gives the endpoint title, timeframe, and event definition, but no count, estimate, or confidence interval. Overall Survival is listed as a secondary outcome, without a result. The record also reports zero analyses, so there is no efficacy analysis to assess.

The statistical methods are not described. The registry does not report an analysis plan, method of estimating event-free survival, approach to incomplete follow-up, analysis population, or missing-data handling. It does not provide a sample-size justification. The population field for the primary endpoint is blank, and the record does not state how many participants contributed to the efficacy outcome.

The nonrandomized, single-group design limits causal interpretation. Serious adverse-event data appear under two cohort titles, but the registry cautions that the arms are not a like-for-like control comparison. The record does not explain cohort assignment or report an adjusted comparison. Differences between cohorts therefore cannot be attributed to treatment from these entries alone.

Safety interpretation is also constrained by the available detail. The record lists at-risk and affected counts, deaths, and percentages for selected serious adverse-event terms, but does not establish causal attribution. It does not clarify in this record how the listed terms relate to one another or provide enough detail to interpret the percentages as comparative risks. These limitations mean the registry provides selected descriptive information, not a complete statistical account of efficacy or comparative safety.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: High-dose Chemotherapy for Poor-Prognosis Relapsed Germ-Cell Tumors — NCT00936936.
• Registry record last updated 2024-08-26; status last verified 2024-08.
• Results first posted 2024-08-26.
• 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

The registry record describes a completed, phase 2, nonrandomized, single-group study of high-dose chemotherapy for poor-prognosis relapsed germ-cell tumors. It lists 64 enrolled participants and describes two chemotherapy cycles followed by stem-cell infusion. The primary outcome is the number of participants with 2-year event-free survival, defined with progression, relapse, or death from any cause as events. Overall Survival is listed as a secondary outcome.

The record reports zero analyses and does not report either outcome result. It lists serious adverse-event counts and selected terms by cohort, while cautioning that the cohorts are not like-for-like controls. The efficacy result, statistical analysis methods, missing-data procedures, and analysis population are not reported in the registry record.