Vorinostat in Patients With Primary Cutaneous T-Cell Lymphoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Vorinostat in Patients With Primary Cutaneous T-Cell Lymphoma (NCT00958074), a non-randomized phase 2 study. The primary outcome is Objective Response.

2. Clinical Question

3. Trial Design

This was a phase II, prospective, non-randomized clinical trial of dose-adjusted vorinostat in primary cutaneous T-cell lymphoma. The registry lists a parallel model and no masking. The University of Washington was the sponsor. Enrollment was 11, and the record lists disease stages I through IV. The study began in 2009-07, had a primary completion date of 2013-02, and is marked terminated. The reason for termination is not reported in the registry record.

Participants were organized by age into two cohorts. Cohort I, for participants aged 65 years or older, began with 200 mg of vorinostat orally once daily on days 1–28. Cohort II, for participants younger than 65 years, began with 400 mg on the same schedule. In either cohort, the dose could be escalated in increments of 100 mg per day, up to 500 mg per day, in the absence of dose-limiting toxicity. The treatment schedule specified repeating each 28-day course for 6 courses. Whether participants actually received all scheduled courses, and how many reached each dose level, are not reported in the registry record.

These age-based groups are not randomized treatment arms and should not be interpreted as a controlled comparison of doses. Participants were not assigned at random, and age as well as starting dose differed between cohorts. The record does not describe a separate control group, a blinding procedure, or a comparator treatment. The primary endpoint focused on objective response in skin assessment; additional listed outcomes addressed extracutaneous manifestations, toxicity-related dose adjustments, Sezary cell count, and erythroderma assessment.

The registry lists 11 enrolled participants overall. Its safety information separates 4 participants in Cohort I and 7 in Cohort II. For the primary endpoint, the record says that 3 of the 4 Cohort I participants who consented were used for analysis; one person completed 3 days of treatment, withdrew consent, and was considered unevaluable because the initial evaluation timeframe was 14 days. The primary endpoint analysis population for Cohort II is 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: In cohort 1, only 3 of the 4 subjects consented were used for analysis. 1 patient in this cohort only completed 3 days of treatment before removing themselves from treatment and withdrawing consent. This patient was deemed inevaluable, as the timeframe for initial evaluation is 14 days.

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 Objective Response, measured over After at least 14 days. With Confirmation after additional 28 days..

The primary endpoint was objective response, expressed as a percentage of the total, with the assessment scheduled after at least 14 days and confirmation after an additional 28 days. The registry defines response as either no evidence of clinical disease or marked improvement, with marked improvement requiring a decrease of at least 50% in the modified Severity-Weighted Assessment Tool (mSWAT) skin assessment score compared with baseline.

This endpoint combines a complete clinical state, as described by no evidence of clinical disease, with a threshold for improvement in a measured skin score. The score-based part gives a stated criterion for judging change relative to each participant’s baseline, rather than relying only on an unstructured impression. The endpoint is intended to capture change in skin disease; it is not, by itself, a measure of all disease manifestations or of long-term outcomes. The record separately lists extracutaneous manifestations, Sezary cell count, dose adjustments related to safety or toxicity, and physicians’ serial assessment of erythroderma as secondary outcomes.

The response definition and assessment timing are reported, but the numerical primary response result is not reported in the registry record. No confidence interval, response count, or analysis estimate is available in the record. The cohort I population note also identifies one participant as unevaluable after withdrawing following 3 days of treatment, before the stated initial evaluation timeframe. The endpoint’s role in the study can therefore be described, but the posted information does not permit interpretation of the magnitude or precision of the observed primary response.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation begins with the endpoint and its analysis population. The primary endpoint is a binary response classification based on the registry’s stated criteria: no evidence of clinical disease or a decrease of at least 50% in mSWAT skin assessment score compared with baseline. In general, a binary endpoint is summarized by counting participants who meet the definition and relating that count to a specified analysis population. The registry record does not report the primary response count or percentage, so neither can be interpreted. It also does not report a confidence interval or a statistical test for the primary endpoint.

What the response definition captures

The endpoint combines a clinical category with a threshold for change in a skin assessment score. The score-based criterion refers to change from baseline, so it depends on both the baseline assessment and the follow-up assessment. In general, a prespecified threshold can make a response definition more reproducible than a purely informal judgment, although the registry record does not provide additional information about score assessment procedures, assessor training, or agreement between assessors. The stated timing is an initial assessment after at least 14 days, with confirmation after an additional 28 days. The record does not clarify the handling of participants who did not reach confirmation or the exact analysis rule for those cases.

Why precision matters

In general, a confidence interval describes the range of values compatible with an estimate under the statistical model and assumptions used. It conveys uncertainty as well as the point estimate. With a limited number of participants, estimates of a response proportion can be imprecise, but no response estimate or confidence interval is reported in the registry record. It would therefore be inappropriate to infer a response rate, uncertainty range, or statistical significance from the enrollment count alone.

The posted safety counts are descriptive. They identify how many participants were at risk and how many were recorded with serious or other adverse events in each cohort. Those counts do not provide a randomized comparison, and they do not by themselves establish a causal relationship between treatment and an event. A comparison of groups would also be difficult to interpret because allocation was non-randomized and the cohorts differ by age and starting dose. The record does not report a comparative safety model or adjusted estimates.

How to read the cohort structure

The study had a parallel model, but its cohorts were not randomized arms. Cohort I began at 200 mg per day for participants aged 65 years or older, while Cohort II began at 400 mg per day for participants younger than 65 years. The dose could increase in increments of 100 mg per day, up to 500 mg per day, in the absence of dose-limiting toxicity. These are protocol schedule descriptions. The record does not state how many participants escalated, the dose each participant ultimately received, or whether dose escalation occurred in practice. Consequently, the cohort labels should not be treated as observed exposure groups at every dose level.

In general, randomization helps balance measured and unmeasured characteristics across treatment groups, while non-randomized group comparisons can be confounded by differences between participants or by the reasons treatment was selected. Here, age defines the cohorts and the initial dose also differs. Thus, even if outcome counts by cohort were available, a difference could not be attributed to dose alone from this design. The record does not report an adjustment strategy intended to address such differences.

Planning illustrations and observed data

The statistical planning illustrations are based on the observed effect and use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is described as a stated fraction of the observed effect. These are planning assumptions, not a report of the original trial’s sample-size justification. The original sample-size justification is not reported in the registry record. Such illustrations are not observed power and do not show that the trial had a particular probability of detecting an effect. They are conditional calculations that depend on the effect used for planning and the stated assumptions. The numerical planning results are not reported in the registry record information.

More generally, power is a design-stage probability defined for a specified effect, variability, sample size, and analysis procedure. It is not a measure of whether a result is true, and post hoc calculations based on an observed effect should not be interpreted as a direct measure of evidential strength. In this study, no primary estimate or inferential analysis is reported, so the record does not permit a trial-specific assessment of power or precision.

Multiplicity and analysis choices

The registry lists 5 outcome measures, including the primary endpoint and secondary outcomes concerning extracutaneous manifestations, dose adjustment, Sezary cell count, and erythroderma score. When multiple outcomes or repeated analyses are considered, the chance of at least one apparently notable result can increase; in general, statistical methods may address this through prespecified testing strategies or adjustment. The registry record does not report a multiplicity plan, a hierarchy among the outcomes, or an adjustment procedure. It lists 0 analyses. That means the record does not present an analysis result for these measures, rather than establishing that no analyses were conducted outside the posted record.

Likewise, methods for missing observations, endpoint confirmation, and analysis populations are not fully described. The record identifies one Cohort I participant who withdrew after 3 days and was deemed unevaluable for the primary endpoint, but it does not describe broader missing-data rules or the Cohort II primary analysis population. These omissions constrain interpretation because analysis denominators and handling of unavailable assessments affect the meaning of endpoint estimates.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 5 outcome measures: the primary objective response endpoint and secondary measures covering extracutaneous manifestations, toxicity-related dose adjustment, Sezary cell count, and physicians’ serial assessment of erythroderma. Considering multiple outcomes can create multiple opportunities for a result to appear notable. In general, investigators may address this with a prespecified hierarchy, a defined primary testing strategy, or an adjustment for multiple comparisons. The registry record does not report such a strategy, an adjustment procedure, or a planned hierarchy beyond identifying the primary endpoint.

The record lists 0 analyses. Accordingly, no endpoint-specific statistical analysis results are posted in the registry record. This does not establish that no analyses took place; it means the record does not report them. The numerical primary response result, its uncertainty, and results for the secondary outcomes are not reported in the registry record.

Some analysis-population information is available. For Cohort I, 3 of the 4 participants who consented were used for analysis. One participant completed 3 days of treatment, withdrew consent, and was considered unevaluable because the initial evaluation timeframe was 14 days. The Cohort II primary endpoint analysis population and broader rules for missing or unconfirmed assessments are not reported in the registry record. The record does not describe whether the analysis was intention-to-treat, per-protocol, or based on another defined population. These gaps prevent a full assessment of the endpoint denominator and how incomplete follow-up was handled.

11. 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: pulmonary embolism (14.3%).

The safety information reports counts by cohort. In Cohort I, 0 of 4 participants were recorded as having a serious adverse event; in Cohort II, 1 of 7 was recorded with a serious adverse event. For other adverse events, the record lists 4 of 4 participants in Cohort I and 7 of 7 in Cohort II. These are participant counts, not event rates per unit of time, and the record does not describe the severity distribution or event-level timing for the other adverse events.

Pulmonary embolism is listed as the top serious adverse event term, with 14.3% recorded. The registry record does not provide enough context to interpret this percentage’s denominator or whether it refers to participants or events, so it should be read as a reported registry figure rather than recalculated from the cohort totals. The death count is not reported in the registry record.

These figures are descriptive observations from a non-randomized study without a control group. They do not establish whether an event was caused by treatment, and they do not support a comparative safety conclusion between the age-defined cohorts. The cohorts differ in age and starting dose, and the record does not report an adjusted safety analysis.

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

The registry record shows that a phase II, non-randomized study of dose-adjusted vorinostat enrolled 11 participants with primary cutaneous T-cell lymphoma. It describes two age-defined cohorts, their starting doses and dose-escalation schedule, the primary response definition, and participant counts for serious and other adverse events. These details characterize the design and the information posted for the trial.

The record does not report a numerical result for the primary objective response endpoint. It therefore does not show how many participants met the response criteria, what response percentage was observed, or how precisely any response estimate was measured. The endpoint definition alone is not evidence that participants achieved a response. Nor does the record provide the numerical results needed to assess the secondary outcomes.

The non-randomized cohorts are not a control comparison. They differ by age category and initial dose, and there is no randomized comparator group described. Consequently, the information does not isolate a treatment effect or permit a causal comparison between the starting doses. The safety counts are likewise descriptive and do not establish treatment causation.

The posted information also does not explain why the study was terminated, provide a sample-size justification, or report the statistical analysis plan. One Cohort I participant withdrew after 3 days and was excluded from the primary endpoint analysis as unevaluable; the Cohort II analysis population is not reported. These limitations affect the ability to interpret the endpoint denominator and the completeness of follow-up.

In general, a trial’s design, endpoint, analysis population, estimate, and uncertainty all contribute to interpretation. Here, the registry provides some design and safety details but no numerical primary efficacy result or inferential analysis. The appropriate statistical reading is therefore limited to what is documented: the planned response criterion, the reported cohort structure, and the posted safety counts. It does not support a conclusion about the magnitude or certainty of an objective response effect.

13. Why This Trial Matters Statistically

This trial illustrates why a clearly defined endpoint is only one part of an interpretable clinical study. The registry specifies objective response through clinical disease status or a threshold decrease in mSWAT skin assessment score, with an initial evaluation after at least 14 days and confirmation after an additional 28 days. That definition gives readers a framework for understanding what response was intended to mean. However, without the numerical result, analysis denominator, and uncertainty estimate, the endpoint cannot establish the observed response frequency or its precision.

The cohort structure also demonstrates an important design distinction. A parallel study is not necessarily a randomized comparison. Here, age defines the cohorts and starting dose differs between them. The groups therefore cannot be treated as equivalent groups differing only in dose. Recognizing this prevents over-interpretation of any apparent between-cohort differences, even if such outcome figures were available.

The record further shows why enrollment and analysis populations should be distinguished. Although total enrollment was 11, the primary endpoint note says that 3 of 4 Cohort I participants who consented were analyzed, with one withdrawal deemed unevaluable. The Cohort II primary analysis population is not reported in the registry record. Without clear denominators and missing-data rules, a reader cannot determine precisely which participants contributed to a primary outcome estimate.

Finally, registry completeness is part of statistical interpretation. The study is marked terminated, but its reason for termination is not reported. The record lists outcomes and safety counts, but no analyses. Thus, the trial is informative as an example of how design, endpoint definition, analysis population, and result reporting jointly shape what can be learned. It also illustrates the difference between a recorded safety count and a comparative or causal safety analysis.

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 study enrolled 11 participants and had a non-randomized design. Its age-defined cohorts began at different doses, so they do not provide a like-for-like comparison of dose schedules. The record does not describe a control treatment, random allocation, or masking. These design features limit comparative interpretation.

The numerical primary response result is not reported in the registry record. Nor are primary endpoint confidence intervals, test results, or numerical secondary outcome results reported. Although the response criteria and timing are specified, the record does not fully describe assessment procedures, confirmation rules, or how participants without a confirmatory assessment were handled.

The primary endpoint population note identifies one Cohort I participant who withdrew after 3 days and was considered unevaluable, and says 3 of 4 consenting Cohort I participants were analyzed. The Cohort II primary analysis population and general missing-data rules are not reported in the registry record. The reason for termination, original sample-size justification, and planned statistical analysis strategy are also not reported.

The safety information is limited to participant counts and a listed serious adverse event term. The record does not provide a death count or enough context to clarify the denominator for the pulmonary embolism percentage. It also does not report event timing, duration, or comparative safety estimates. Finally, the scheduled dose escalation is described, but the doses actually received and the number of participants reaching each dose are not reported in the registry record.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Vorinostat in Patients With Primary Cutaneous T-Cell Lymphoma — NCT00958074.
• Registry record last updated 2018-10-18; status last verified 2018-09.
• Results first posted 2014-12-15.
• 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

NCT00958074 is a phase II, prospective, non-randomized trial of dose-adjusted vorinostat in primary cutaneous T-cell lymphoma. The record lists 11 participants, parallel age-defined cohorts, no masking, and the University of Washington as sponsor. The study began in 2009-07, reached primary completion in 2013-02, and is marked terminated; the reason for termination is not reported in the registry record.

The primary endpoint was objective response, defined by clinical disease status or a decrease of at least 50% in mSWAT skin assessment score, with confirmation after an additional 28 days. The numerical primary response result and statistical analyses are not reported. Safety counts list serious adverse events in 0 of 4 participants in Cohort I and 1 of 7 in Cohort II, and other adverse events in 4 of 4 and 7 of 7, respectively. The record does not report a death count.