Systemic Chemotherapy and Subtenon Carboplatin, and Local Ophthalmic Therapy in Children With Intraocular Retinoblastoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Systemic Chemotherapy and Subtenon Carboplatin, and Local Ophthalmic Therapy in Children With Intraocular Retinoblastoma (NCT00072384), a na phase 3 study. The primary outcome is Group D Eyes - Treatment Failure Within One Year.

2. Clinical Question

3. Trial Design

This was a completed Phase III, single-group trial sponsored by Children's Oncology Group. The registry record lists enrollment of 30 and describes the study as a trial of systemic and subtenon chemotherapy for Group C and Group D intraocular retinoblastoma. Allocation was not applicable, and masking was listed as none. These features mean the study was not a randomized comparison and did not include a like-for-like control arm.

The treatment approach combined systemic chemotherapy, subtenon carboplatin and ophthalmic therapy. The registry lists liposomal vincristine sulfate, carboplatin, etoposide and filgrastim as drug or biological interventions, and cryosurgery and laser surgery as procedures. The arm description states that patients receive intravenous liposomal vincristine sulfate on day 1, intravenous carboplatin and etoposide on days 1 and 2, and daily subcutaneous filgrastim beginning on day 3 until blood counts recover. The description of subtenon treatment is cut off in the registry record, so further details are not reported there.

The primary endpoint concerns Group D eyes rather than all enrolled participants. Twenty eligible patients had at least one Group D eye, and 25 such eyes contributed to the analysis. That distinction matters: the number of eyes is not the same as the number of independent patients, and the registry description explicitly notes possible dependence between eyes belonging to the same patient.

The listed study dates are 2007-04-16 for start and 2013-02-01 for primary completion; the status is completed. A protocol-level statistical plan, sample-size justification and details of interim analyses are not reported in the registry record. The trial's single-group design should guide interpretation: its outcome can describe results under the study treatment approach, but it cannot by itself establish how those results compare with another treatment.

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: Twenty eligible patients had at least one Group D eye to contribute to the analysis. A total of 25 Group D eyes were available for analysis.

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 Group D Eyes - Treatment Failure Within One Year, measured over One year.

The primary endpoint is the probability of treatment failure within one year for Group D eyes. Each eye was classified as having experienced failure within one year after treatment started or not. This yes-or-no classification gives the endpoint a clinical interpretation tied to an event over a specified follow-up period, rather than a measurement of change on a continuous scale.

A one-year probability can describe how often the endpoint occurred in the analysed group over the defined period. It is important to distinguish that measure from a comparison between treatments: this trial has one treatment group and no concurrent comparator, so the endpoint does not directly estimate a relative treatment effect.

The record says that the point estimate is reported as a mean. Although the analysis metadata classifies the measure as continuous, the endpoint itself is an eye-level binary outcome; a mean of binary indicators can represent the proportion, or estimated probability, of eyes with failure. The record also says that the method of Rosner et al. will be used to model possible dependence between eyes from the same patient. That adjustment is relevant because eyes from one patient may not provide as much independent information as eyes from different patients.

The numerical one-year estimate, its standard deviation and any confidence interval are not reported in the registry record. As a result, the endpoint definition is available, but its observed magnitude and precision cannot be assessed from the posted results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the endpoint estimate?

Each Group D eye was classified according to whether treatment failure occurred within one year after treatment began. For a binary outcome, a mean can be interpreted as an estimated probability when the eye-level classifications are coded as event or no event. The registry record identifies the point estimate as a mean, but does not give its value. It also names standard deviation as the measure of dispersion, without reporting a numerical value for it.

In general, an estimate describes the observed or modeled outcome in the analysis population, while a measure of uncertainty describes how precisely that estimate is determined. A confidence interval is a general statistical tool for expressing uncertainty around an estimate; it gives a range of values compatible with the data and the method under specified assumptions. No confidence interval for this primary endpoint is reported in the registry record, so neither its width nor the precision of the estimate can be assessed.

Why account for multiple eyes from one patient?

The primary analysis includes 25 eyes from twenty patients. Eyes belonging to the same person may have related outcomes because they share patient-level characteristics and care. In general, this is called clustering or dependence: observations are not necessarily independent simply because they are recorded separately. When outcomes are correlated within clusters, methods that treat every observation as independent can misstate uncertainty.

The registry description says the method of Rosner et al. will model possible dependence between eyes from the same patient. This establishes that the analysis was designed to recognize the patient-level grouping. The record does not provide further model details or the fitted result, so the strength of the modeled dependence and its effect on the estimate are not reported.

What can a single-group design establish?

A single-group study observes outcomes in participants receiving the study treatment approach. In general, this design can describe outcomes in the analysed group, but without a concurrent control group it does not isolate the effect of treatment from other explanations for the observed outcomes. This trial has no randomized comparator; allocation is listed as not applicable. The primary endpoint is therefore a one-year outcome probability, not a randomized estimate of a difference between treatments.

Masking is listed as none. In general, lack of masking can matter when decisions or assessments may be influenced by knowledge of treatment, although the registry record does not describe whether or how this affected the endpoint assessment. The endpoint is a yes-or-no classification, but the detailed criteria for treatment failure are not reported in the registry record beyond the one-year timing and classification.

How should planning illustrations be read?

General sample-size calculations ask how many participants or events a future study might need under specified assumptions. The planning assumptions described for this analysis use a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. Those assumptions describe hypothetical planning, not the design of this single-group trial. The record does not report the trial's own sample-size justification, so the planning assumptions should not be attributed to its protocol.

Planning illustrations are not observed power. In general, power is a design quantity tied to assumptions about an effect and a study plan; a calculation based on an observed effect is not evidence that the completed study had a particular probability of detecting that effect. Time-to-event illustrations use total event counts under the stated Schoenfeld approach, while other illustrations use participants per arm. Those conventions do not create a control arm or alter the actual single-group design.

The registry record reports zero analyses, and no numerical primary endpoint estimate or uncertainty measure is included in the available results. General statistical methods can explain what such estimates would mean, but cannot substitute for the missing numerical findings.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Treatment (Chemotherapy, Surgery)")
n   <- c(20)
m   <- c(0.52)
sp  <- c(0.102299)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Treatment (Chemotherapy, Surgery)  20  0.520  0.102  0.472  0.568

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists multiple outcomes: the primary one-year treatment-failure endpoint for Group D eyes and secondary outcomes involving event-free survival, chemotherapy-associated toxicity and patterns of treatment failure. It reports six outcome measures but zero analyses. Numerical results for the primary endpoint and the listed secondary outcomes are not reported in the registry record.

Multiplicity arises in general when several outcomes or statistical tests are considered: the chance of at least one apparently unusual result can increase as the number of tests grows. A statistical plan may address that issue through outcome hierarchy or adjustments, but such a plan is not reported in the registry record. No multiplicity adjustment or testing hierarchy should therefore be inferred.

The record identifies twenty eligible patients with at least one Group D eye and 25 Group D eyes for the primary analysis. It does not report a broader account of exclusions, missing observations, follow-up completeness or how missing endpoint classifications were handled. Those details are not reported in the registry record; it is not possible to determine from the posted information whether missing data occurred or how they might affect the estimate.

The analysis population is described specifically for the primary endpoint, but additional analysis-population definitions are not reported in the registry record. The listed enrollment of 30 and the endpoint population counts describe different aspects of the study and should not be treated as though every enrolled participant contributed a Group D eye to this analysis. No conclusions about the effect of missing data or the representativeness of the analysed group can be drawn without further 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 non-serious adverse event terms

The registry record reports serious adverse events for one treatment group. There were zero participants affected among 22 at risk in the serious adverse-event category. In the separate other adverse-event category, 18 participants were affected among 22 at risk. The record does not list any serious adverse-event terms. Deaths are not reported in the registry record.

These are counts within the study group, not a comparison of safety between treatments. The trial has no concurrent control group, so the figures do not estimate how the event counts compare with another treatment or with no treatment. The registry record also does not provide further detail about event definitions, severity breakdowns for the other category or the timing and duration of safety observation.

Zero reported serious adverse events does not establish that serious events were impossible or that the treatment approach is free of risk. It means that none were recorded among the stated number at risk in this registry entry. Likewise, the other adverse-event count is a descriptive report and cannot by itself determine the cause of an event. These safety counts should be read separately from the unreported numerical primary endpoint.

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

The registry record describes a Phase III, single-group study of systemic chemotherapy, subtenon carboplatin and ophthalmic therapy in children with intraocular retinoblastoma. For the primary outcome, it identifies a one-year treatment-failure classification for Group D eyes and an analysis population of twenty patients contributing 25 eyes. It also describes an approach to account for possible dependence between eyes from the same patient.

Those details establish what the primary endpoint was intended to measure and who contributed to its analysis. They do not establish the observed probability of treatment failure: that numerical result is not reported in the registry record. The standard deviation and any confidence interval are also not reported, limiting assessment of the estimate's precision. Listed secondary outcomes likewise lack numerical findings in the available record.

Even if a numerical outcome estimate were available, the single-group design would describe outcomes in the treated group rather than provide a randomized comparison. Without a concurrent comparator, the registry record cannot show how outcomes would have differed under another treatment approach. The lack of masking is also recorded, but further information about endpoint assessment is not reported.

The safety entry reports zero participants with serious adverse events among 22 at risk, and 18 affected in the other adverse-event category among 22 at risk. These counts are descriptive and do not constitute a comparative safety analysis. They do not replace the missing primary outcome result, and they should not be interpreted as proving the absence of risk.

In short, the record supports a description of the study design, endpoint definition, analysis population and reported safety counts. It does not provide enough numerical outcome information to assess the magnitude or precision of the primary result, nor does its single-group design establish a comparative treatment effect.

14. Why This Trial Matters Statistically

This trial illustrates a key statistical distinction between the number of people and the number of measured units. The primary endpoint is reported for eyes, while eyes are nested within patients: twenty patients contributed 25 Group D eyes. An analysis that recognizes the possibility of within-patient dependence is therefore relevant to how uncertainty should be estimated. The registry record explicitly describes a method intended to model that dependence.

The study also shows why design determines the scope of a conclusion. A single-group study can describe outcomes in its analysed population, but does not provide the concurrent comparison needed for a randomized between-group estimate. The trial's Phase III label does not change its listed single-group model or create a comparator that the record does not contain.

Another statistical lesson is the distinction between an endpoint definition and a result. The registry describes a clear time horizon and yes-or-no treatment-failure classification, yet does not report the numerical estimate or its uncertainty. Without those values, readers cannot judge the observed magnitude or precision. The listed secondary outcomes and safety counts provide additional context, but they do not supply the missing primary endpoint result.

Finally, planning calculations and trial design are not interchangeable. Hypothetical sample-size illustrations based on two-arm assumptions do not describe this single-group study's allocation or supply its missing design justification. Keeping the analysis population, outcome unit, comparison structure and reported results distinct is essential to interpreting the evidence accurately.

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 most consequential limitation in the registry record is that the numerical primary endpoint result is not reported. Although the endpoint definition, follow-up period and analysis population are described, the probability estimate and its uncertainty are unavailable. This prevents assessment of the observed treatment-failure magnitude and its precision.

The design is single-group, with no randomized comparator and no masking. The trial can describe outcomes in the treated group, but the record does not support a direct comparison with another treatment. The extent to which lack of masking could have affected outcome assessment is not reported in the registry record.

The endpoint is eye-based, with 25 Group D eyes from twenty patients. The record states that the method would model possible dependence between eyes from the same patient, but does not provide model details or the fitted estimate. It does not report further criteria for classifying treatment failure beyond the yes-or-no classification within one year.

Details about missing observations, follow-up completeness and handling of missing endpoint data are not reported in the registry record. The treatment description is also incomplete at the point where it begins to describe subtenon treatment, limiting the available detail about the intervention. Although six outcome measures are listed, the record reports zero analyses, and no numerical findings for the listed secondary outcomes are included. The trial's own sample-size justification, interim analyses and full statistical plan are not reported in the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Systemic Chemotherapy and Subtenon Carboplatin, and Local Ophthalmic Therapy in Children With Intraocular Retinoblastoma — NCT00072384.
• Registry record last updated 2021-07-30; status last verified 2021-07.
• Results first posted 2018-09-19.
• 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 identifies NCT00072384 as a completed Phase III, single-group trial sponsored by Children's Oncology Group, with enrollment of 30. It concerns systemic chemotherapy, subtenon carboplatin and ophthalmic therapy for children with intraocular retinoblastoma. Allocation is listed as not applicable, and masking as none.

The primary outcome is the probability of treatment failure within one year for Group D eyes. Twenty eligible patients contributed 25 Group D eyes, and the analysis description says possible dependence between eyes from the same patient would be modeled. The numerical endpoint estimate and its uncertainty are not reported in the registry record.

The safety results list zero participants with serious adverse events among 22 at risk and 18 affected in the other adverse-event category among 22 at risk. Deaths are not reported. Six outcome measures and zero analyses are recorded.