Study of NY-ESO-1 ISCOMATRIX® in Patients With Measurable Stage III or IV Melanoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of NY-ESO-1 ISCOMATRIX® in Patients With Measurable Stage III or IV Melanoma (NCT00518206), a non-randomized phase 2 study. The primary outcome is Number of Subjects With Best Overall Tumor Response.

2. Clinical Question

The study also set out to assess cellular and humoral immunogenicity, delayed-type hypersensitivity reactions, safety, and tolerability. The registry record identifies these objectives and lists several secondary outcomes, but it does not provide numerical results for the primary response endpoint. Thus, the clinical question can be described from the record, but the posted information does not establish how many participants had complete response, partial response, stable disease, or progressive disease.

3. Trial Design

This was a Phase 2, open-label, non-randomized, parallel study sponsored by the Ludwig Institute for Cancer Research. The registry lists enrollment of 46 and a completed study. The recorded start date is 2003-11-28, and primary completion is dated 2010-01-22. The study title identifies people with measurable Stage III or IV melanoma, and the brief description characterizes the disease as advanced malignant melanoma.

Participants were assigned to one of two cohorts. Cohort 1 received NY-ESO-1 ISCOM vaccine: 100 μg of NY-ESO-1 protein formulated with 120 μg of ISCOM adjuvant, administered intramuscularly every 4 weeks for 3 doses in every cycle. Cohort 2 received cyclophosphamide at 300 mg/m^2 intravenously 1 day before each vaccination, followed by the same vaccine formulation and schedule. The registry describes vaccine administration as occurring every 4 weeks and the primary endpoint timeframe as up to 22 months.

Masking was absent, so the study was open-label. Allocation was non-randomized, and the record explicitly cautions that the arms are not a like-for-like control comparison. The record does not report how cohort membership was decided, whether allocation changed over time, or whether baseline characteristics were balanced. Those details are not reported in the registry record.

The primary endpoint was the number of subjects with each best overall tumor response category. The record also lists cellular and humoral immunogenicity, delayed-type hypersensitivity reactions, and treatment-emergent adverse events as secondary outcomes. Although the enrollment count is recorded, cohort-specific enrollment and the number of participants contributing primary endpoint data are not reported in the registry record.

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:

Analysis population for the primary outcome, as defined in the registry: Subjects with data available from at least 1 post-baseline response assessment.

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.

Population: All enrolled subjects

6. Primary Endpoint

The registry lists the primary outcome as Number of Subjects With Best Overall Tumor Response, measured over Up to 22 months.

The primary endpoint counts participants according to their best overall tumor response. Tumors were assessed using computed tomography and categorized under RECIST version 1.0. The registry describes assessments at baseline, at week 11, between weeks 23 and 25, and every 12 weeks thereafter, with an endpoint timeframe of up to 22 months.

RECIST provides standardized categories for changes in target lesions. Complete response means disappearance of all target lesions, with no evidence of disease. Partial response requires at least a 30% decrease in the sum of the longest diameters of target lesions. Progressive disease requires at least a 20% increase in that sum. Stable disease denotes changes that do not meet the stated criteria for partial response or progressive disease. These definitions describe the classification rules; they do not show which categories participants in this study reached.

Best overall response compresses assessments over follow-up into a participant-level outcome. In general, this can make response patterns easier to summarize than a sequence of measurements, while retaining a clinically interpretable category. Its interpretation depends on the assessment schedule, evaluability, and response rules. The registry specifies that the primary population included subjects with at least one post-baseline response assessment, but does not report the number meeting that condition or the category counts. Consequently, response proportions, comparisons between cohorts, and uncertainty around those results cannot be described from the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The available information describes the outcome definitions and study structure, but does not give the statistical methods used to analyze the primary endpoint. The record lists no posted analyses. The explanations below distinguish what the registry documents from general statistical principles.

What does a count endpoint represent?

The primary outcome is a count of participants in each best overall response category. This is a categorical endpoint: each evaluable participant is classified according to the recorded response rules. If category counts and the evaluable population were reported, they could describe the observed response distribution. They are not reported in the registry record, so no response proportion or cohort contrast can be stated.

In general, categorical outcomes may be summarized using counts and proportions. Comparisons may use methods suited to the design and sample size, while estimates of uncertainty help show how much the observed data can support. The choice of method depends on the question, allocation process, and assumptions. The registry record does not identify a statistical test, model, adjustment, or confidence-interval method for tumor response.

How should the RECIST assessments be understood?

The endpoint is based on serial computed tomography assessments, beginning at baseline and continuing at the recorded visits. RECIST version 1.0 defines response categories through changes in the summed longest diameters of target lesions and the disappearance of target lesions for complete response. “Best overall” uses assessments over follow-up to assign a participant-level category. The registry describes the schedule and criteria, but does not explain details such as how missing or unevaluable scans were handled, how confirmation was applied, or whether review was independent. Those details are not reported in the registry record.

What does non-randomized allocation imply?

Randomization, in general, aims to make treatment groups comparable on average with respect to measured and unmeasured characteristics. It is not part of this study: allocation is listed as non-randomized. As a result, cohort differences may reflect participant characteristics, selection, or other factors as well as the differing treatment sequences. Statistical adjustment can sometimes address measured differences, but it cannot automatically remove unmeasured confounding. The record does not report an adjusted analysis or the information needed to assess baseline balance.

The study was also unmasked. In general, awareness of treatment assignment can influence participant reporting, clinical decisions, and outcome assessment, although the extent depends on the outcome and procedures. The primary endpoint uses imaging and defined criteria, but the record does not report whether image assessments were blinded or centrally reviewed.

How should planning estimates be read?

The record describes sample-size illustrations as planning exercises based on an observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The effect to detect is described as a stated fraction of the observed effect. These are planning assumptions, not evidence about the study's actual power or its observed results. Sample-size illustrations are hypothetical estimates for a specified design and effect; they are not observed power. The registry record does not report the trial's own sample-size justification, and the primary response counts needed to evaluate the effect underlying such planning are not reported in the registry record.

In general, power is a design property calculated under assumptions about the effect, variability, allocation, and analysis. After a study, an observed power calculation based on the observed effect does not replace estimates and confidence intervals. It is more useful to inspect the effect estimate and its uncertainty, if reported. Those inferential results for the primary endpoint are not reported in the registry record.

Why does over-dispersion matter in general?

Over-dispersion occurs when count data vary more than a statistical model assumes. A model that underestimates variability can produce confidence intervals that are too narrow and tests that are too optimistic. This is a general consideration for count outcomes. The primary endpoint is described as participant counts by response category, and the registry does not report a model that would permit an assessment of over-dispersion. No conclusion about this issue can be drawn for the study analysis.

In short, the statistical methods that can be reconstructed from the record are limited to the response classification and the descriptive design details. The analysis plan, endpoint estimates, inferential procedures, and handling of incomplete assessments are not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists several scientific aims, including tumor response, cellular and humoral immunogenicity, delayed-type hypersensitivity reactions, and treatment-emergent adverse events. When a study evaluates multiple outcomes or makes multiple comparisons, the chance of at least one apparently unusual result can increase. In general, multiplicity procedures may control that risk through a prespecified hierarchy, adjusted significance thresholds, or other methods. Whether such procedures were planned or used is not reported in the registry record.

The primary endpoint population is defined as subjects with data available from at least one post-baseline response assessment. This makes evaluability relevant: enrolled participants without an assessment may not contribute to the endpoint as defined. The registry does not report how many participants were evaluated, how many lacked post-baseline data, why data were missing, or how missing assessments affected best overall response. These details are not reported in the registry record.

In general, missing outcome data can affect both the size and interpretation of an estimate. Excluding participants with missing assessments can introduce bias if missingness relates to prognosis, treatment, or response. Sensitivity analyses can explore how conclusions depend on assumptions about missing outcomes, but the registry record does not report such analyses.

The record lists no posted statistical analyses for the outcomes. It also does not report an analysis population beyond the primary endpoint's stated requirement for at least one post-baseline response assessment. Intention-to-treat, per-protocol, or safety-population definitions are not reported in the registry record. Without these details and the outcome counts, the extent of missingness, multiplicity handling, and analysis population cannot be evaluated.

11. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: metastases to central nervous system (10.5%), chest pain (5.3%), abdominal pain (5.3%), metastatic malignant melanoma (5.3%), neuralgia (5.3%).

The registry reports serious adverse events in two non-randomized cohorts. In Cohort 1, 27 participants were at risk and 5 were affected by a serious adverse event. In Cohort 2, 19 were at risk and 7 were affected. The record lists 0 deaths in each cohort. For other adverse events, 27 of 27 participants in Cohort 1 and 19 of 19 in Cohort 2 were recorded as affected. These are descriptive counts; they do not establish that the cohorts are comparable.

The listed serious-event terms and percentages are metastases to central nervous system, 10.5%; chest pain, 5.3%; abdominal pain, 5.3%; metastatic malignant melanoma, 5.3%; neuralgia, 5.3%; and headache, 5.3%. The registry does not provide term-specific participant counts alongside these percentages, nor does it state whether participants could contribute more than one term. The percentages should therefore be read as reported term summaries, not converted into counts or treated as mutually exclusive categories.

In general, serious adverse event counts describe observed events or affected participants under the record's reporting conventions; they do not by themselves establish causation. Attribution, timing, severity details, and the methods used to collect and code events are not reported in the registry record.

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

The record shows that a Phase 2 study evaluated NY-ESO-1 ISCOM vaccine in people with measurable advanced melanoma, either without the listed cyclophosphamide sequence or after cyclophosphamide before vaccination. It gives the intended response measurement, RECIST version, assessment schedule, and selected safety summaries. It does not report the primary endpoint category counts or statistical analysis results.

For that reason, the posted information cannot quantify tumor response, its precision, or a difference between cohorts. It cannot establish whether one cohort had more favorable outcomes, and it cannot determine the contribution of cyclophosphamide. These limitations arise both from missing outcome results and from the non-randomized design. The open-label status and the lack of reported details about cohort assignment, baseline balance, and outcome review further constrain interpretation.

The safety summaries describe participants affected by serious and other adverse events and list several serious-event terms. They do not alone determine whether an event was caused by study treatment or provide a complete comparison of safety between cohorts. The registry record does not report the detailed information required for those judgments.

In general, a trial endpoint definition tells readers what was measured, while outcome data and an analysis are needed to estimate what happened. Statistical uncertainty is also important: an estimate without an interval or other uncertainty measure can give an incomplete picture of the evidence. For this study, the response estimates and their uncertainty are not reported in the registry record.

The appropriate reading is therefore bounded. The record documents a clinical study design and some descriptive safety results, but it does not provide the numerical primary results needed to assess tumor response. It should not be interpreted as evidence either for or against an effect of the vaccine or the addition of cyclophosphamide.

13. Why This Trial Matters Statistically

This study illustrates why the wording of a trial question and the strength of its comparison are separate statistical issues. The primary question concerns a defined imaging-based response outcome, but the two cohorts were non-randomized. Even if response counts were available, a difference between cohorts would need cautious interpretation because allocation may be associated with participant characteristics or other factors.

It also shows the difference between endpoint specification and outcome reporting. The registry gives RECIST categories, assessment timing, and a population definition, which make the intended measurement more transparent. Yet without the category counts, evaluable denominator, and analysis results, those definitions cannot yield a numerical estimate of the observed response pattern.

The safety summaries offer another statistical lesson: event counts and named terms answer different descriptive questions. Cohort-level counts indicate how many participants were recorded as affected, while the term percentages describe selected serious-event categories. Without term-specific counts and details about overlap or coding, those figures should not be combined or read as a complete account of event burden.

Finally, planning calculations are not results. The record describes assumptions for sample-size illustrations, but such calculations depend on an effect and design assumptions and are not a substitute for the observed endpoint data. They also are not observed power. Separating planning, observed outcomes, and inferential uncertainty helps prevent a design illustration from being mistaken for evidence about the study's findings.

These distinctions are broadly relevant to clinical research: a clear outcome definition, a credible comparison, transparent handling of missing observations, and complete reporting of estimates and uncertainty are all needed for a sound statistical interpretation.

14. 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 primary limitation for interpreting this trial from its registry record is the absence of numerical results for the primary tumor-response endpoint. The record gives the response definitions and assessment schedule, but not the number of evaluable participants or the counts in each response category. It also does not report estimates of uncertainty or an inferential comparison.

The non-randomized allocation means that the cohorts cannot be treated as equivalent groups by design. The record cautions that they are not a like-for-like control comparison, and it does not report the method of assigning participants to cohorts or comparative baseline information. Any cohort contrast would therefore be vulnerable to confounding, even if outcome data were available.

The study was open-label. The record does not report whether imaging was reviewed independently or blinded to cohort. It also does not give details on response confirmation, handling of non-evaluable scans, or how missing assessments were treated. Although the primary population is defined by availability of at least one post-baseline assessment, the number meeting that definition is not reported in the registry record.

For safety interpretation, the record gives cohort-level counts and selected serious-event terms, but not term-specific counts or details of event timing, attribution, coding, or overlap between terms. The protocol's statistical analysis plan, the sample-size justification, interim analyses, and any adjustment for multiple comparisons are not reported in the registry record. These gaps limit the assessment of both the methods and the completeness of the reported evidence.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Study of NY-ESO-1 ISCOMATRIX® in Patients With Measurable Stage III or IV Melanoma — NCT00518206.
• Registry record last updated 2022-10-25; status last verified 2022-10.
• Results first posted 2017-06-23.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

18. Record Summary

The registry describes a completed Phase 2, open-label, non-randomized, parallel study of NY-ESO-1 ISCOM vaccine in people with measurable Stage III or IV melanoma. One cohort received vaccine alone, and the other received cyclophosphamide before vaccination. Enrollment was 46.

The primary endpoint was the number of subjects by best overall tumor response under RECIST version 1.0, assessed over a timeframe of up to 22 months. The record supplies response definitions and an assessment schedule but does not report response category counts or statistical analyses. Selected adverse-event summaries include serious-event counts of 5 among 27 at risk in Cohort 1 and 7 among 19 at risk in Cohort 2, with 0 deaths reported in either cohort. The non-randomized cohorts are not a like-for-like control comparison.