Gemcitabine and Pemetrexed Disodium in Treating Patients With Advanced Mycosis Fungoides or Sézary Syndrome: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Gemcitabine and Pemetrexed Disodium in Treating Patients With Advanced Mycosis Fungoides or Sézary Syndrome (NCT00369629), a non-randomized phase 1 study. The primary outcome is Maximum Tolerated Dose as Measured by the Number of Dose Limiting Toxicities Seen in Cohort..

2. Clinical Question

The study was planned to examine side effects and determine a dose for the gemcitabine–pemetrexed combination in patients with advanced cutaneous T-cell lymphoma. The primary endpoint focused on dose-limiting toxicity, or DLT, rather than a comparison of clinical benefit between treatments. This focus is consistent with the stated phase I purpose: characterizing toxicity across dose levels to inform a dose selection.

The study record describes a rule under which a cohort with one DLT among its first three patients would enroll three more patients. It defines the maximum tolerated dose as the lowest dose level at which two or more patients experience a DLT. However, the record says the study closed before enough patients were enrolled, and the endpoint result is not reported in the registry record. The central question therefore remains a planned question rather than a reported conclusion. The posted safety figures give some context about reported events, but they do not establish the primary endpoint or answer a comparative treatment question.

3. Trial Design

This was a phase I, non-randomized, single-group trial sponsored by Northwestern University. The registry lists enrollment of 14 and a terminated status. The official title describes a phase I/II trial, but the brief summary states that the phase II portion was never conducted because of a lack of funding. The record therefore describes a phase I study, not a completed phase I/II evaluation.

The intervention combined gemcitabine and pemetrexed. The four planned cohorts varied the dose levels. Cohort 1 paired pemetrexed at 400 mg/m2 with gemcitabine at 800 mg/m2. Cohort 2 paired pemetrexed at 500 mg/m2 with gemcitabine at 800 mg/m2. Cohort 3 paired pemetrexed at 500 mg/m2 with gemcitabine at 1000 mg/m2, and Cohort 4 paired pemetrexed at 500 mg/m2 with gemcitabine at 1200 mg/m2. Each drug was scheduled on day 1 and day 15 of a 28-day cycle, with gemcitabine following pemetrexed. These are planned regimens; the registry record does not establish that every planned cohort or dose was administered as described.

There was no random allocation, masking, or independent comparator. Cohort differences are dose differences within the study, not randomized treatment contrasts. The record gives a cohort enrollment rule tied to early DLT observations, but a complete protocol or further details of dose escalation are not reported in the registry record. Recruitment began on 2006-08-28, and primary completion was recorded as 2012-07-16. The study is listed as terminated. The reason given in the brief summary for not proceeding to phase II is lack of funding; further details about termination are not reported in the registry record.

Because the study stopped before the planned work was complete, its design should be read as an early dose-finding investigation with incomplete reporting of its primary result. It was not designed as a randomized efficacy comparison, and no phase II assessment is available in the posted 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.

Treated 2 Cycles/Reached First Response

Reasons for not completing this period:

Moved on to 4 More Cycles

Reasons for not completing this period:

Completed 4 More Cycles

Reasons for not completing this period:

Follow up for 1 Year

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Patients that experienced a DLT in the first cycle in each cohort. If one DLT is seen in the first 3 patients then 3 more patients are enrolled in that cohort. MTD and is defined as the lowest dose level at which two or more patients experience a DLT. The study closed before enough patients were enrolled to determine the MTD or define DLTs.

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 Maximum Tolerated Dose as Measured by the Number of Dose Limiting Toxicities Seen in Cohort., measured over From the day that the first treatment is given through the first 28 day period for each patient..

The primary endpoint is titled maximum tolerated dose, measured by the number of dose-limiting toxicities seen in each cohort. The registry specifies that DLTs were collected during the first treatment cycle, from the first treatment through the first 28-day period for each patient. The endpoint therefore captures early, protocol-defined toxicity at the dose levels under study. It is not a general count of every adverse event throughout treatment, and it does not directly measure tumor response or survival.

The record gives toxicity criteria based on the National Cancer Institute's Common Toxicity Criteria for Adverse Events version 3.0. Grade 1 is mild, Grade 2 moderate, Grade 3 severe, Grade 4 life-threatening or disabling, and Grade 5 death related to an adverse event. Specified events counted as DLT include Grade 3 and/or 4 non-hematologic toxicity, with stated exceptions; unexpected Grade 3 and/or 4 non-hematologic toxicities; Grade 4 vomiting despite maximal antiemetic support; febrile Grade 4 neutropenia during the first cycle; and specified Grade 4 neutropenia or thrombocytopenia on day 1 of the second cycle despite the stated support conditions.

In general, a dose-finding endpoint based on early DLTs helps distinguish tolerability across planned dose levels. The registry defines the maximum tolerated dose as the lowest dose level at which two or more patients experience a DLT. It also describes enrolling additional patients in a cohort if one DLT occurs among the first three. The completed endpoint count and resulting dose determination are not reported in the registry record, so the stated definition should not be mistaken for a reported finding.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

This discussion explains the statistical ideas relevant to the design and the posted results. The explanations of statistical concepts are general; the registry record does not report a complete statistical analysis plan.

What does a single-group dose-finding design show?

In general, an early dose-finding study examines toxicity at planned dose levels to support selection of a dose for later study. A single-group design does not include a concurrent control group, and non-random allocation means participants are not randomly assigned between dose levels. Consequently, observed differences across cohorts are descriptive. They do not have the protection against selection differences that randomization can provide, and they are not automatically causal estimates of dose effects.

Here, the cohorts differed in their planned gemcitabine and pemetrexed doses. The registry describes a rule for adding patients to a cohort after an early DLT observation and defines the maximum tolerated dose using a cohort-level DLT threshold. This explains the intended role of the cohort counts: they were meant to inform a dose decision, not to estimate a comparative treatment effect. The actual result of that decision is not reported in the registry record.

Why define a dose-limiting toxicity?

A DLT is a protocol-defined event, not simply any adverse event. In general, defining which events qualify and the period in which they count makes the endpoint more consistent across participants. This record specifies a first-cycle period and lists toxicity grades and qualifying events. The definition includes exceptions and conditions, so the event classification depends on the full protocol rules and clinical context. A serious adverse event and a DLT are not interchangeable categories: seriousness and protocol-defined dose limitation address different questions. The registry does not provide a crosswalk showing which displayed serious events also met the DLT definition.

The endpoint is a count of DLTs by cohort, with the maximum tolerated dose derived according to the stated threshold. A count is useful for applying that rule, but it is less informative when the number of observed participants is limited or when cohort observation is incomplete. In general, a count without a denominator, time-at-risk detail, and uncertainty estimate gives only a partial picture of how frequently events occurred. The posted record does not include a completed primary endpoint analysis or an interval estimate.

What does the absence of analyses mean?

The registry lists one outcome measure and zero analyses. This means the posted record does not include a reported inferential analysis of the primary endpoint. In general, an inferential analysis may summarize uncertainty with a confidence interval or assess compatibility with a statistical hypothesis using a test. A confidence interval describes a range of values compatible with the data under the analysis assumptions; it is not a guarantee that the true value lies in that range. A p-value, when used, is a measure tied to a specified statistical model and null hypothesis, not a measure of clinical importance. No such estimates are posted here, so none can be interpreted.

The absence of an analysis does not establish that no analysis was ever performed outside the registry. It means that these results are not reported in the registry record. The record also does not report a statistical plan, a model for the cohort counts, or a rationale for the planned enrollment. Those details should not be inferred from the study title, endpoint definition, or the cohort rule.

Why are cohort comparisons limited?

In general, comparing raw counts across groups can be misleading when group sizes or follow-up differ. The registry's adverse-event display includes different at-risk counts across cohorts, including zero for Cohort 4. That pattern underscores why raw numbers alone do not represent a controlled comparison. A formal comparison would require a prespecified estimand, a clear definition of the analysis population and observation period, and methods appropriate to the data. Those elements are not reported as a completed analysis.

The trial's non-randomized design and dose-finding purpose are important context. If participants entered cohorts at different stages or under a rule responding to earlier toxicity, the groups would not necessarily be exchangeable. The record does not provide enough information to assess all allocation details or adjust for differences. It is therefore safer to describe each cohort's posted counts without ranking dose levels or treating a numeric contrast as an effect estimate.

How should readers think about the early stopping?

The brief summary says the phase II portion was not conducted because of a lack of funding, and the record says the study closed before enough patients were enrolled for the stated endpoint purpose. In general, an incomplete study can leave estimates less precise and may prevent the planned decision rule from being applied. It can also leave planned outcomes unavailable. This does not justify calculating a replacement result from the safety display: the DLT endpoint has specific criteria and timing, and the posted adverse-event figures do not supply the required classification and analysis.

In general, sample-size calculations are planning tools that depend on assumptions about the endpoint, effect or event rate, allocation, and analysis. They are not observed results, and they cannot retrospectively establish what the trial showed. No sample-size justification or complete statistical plan is reported in the registry record. Any planning illustration appearing alongside the posted material should therefore be read as an illustration under stated assumptions, not as an analysis of this trial's observed primary outcome or as observed power.

Overall, the appropriate statistical reading is descriptive and cautious. The record gives the intended dose-finding framework and some event counts, but not the primary endpoint result, uncertainty estimates, or a completed analysis. General statistical concepts help explain why those omissions matter; they do not fill in the missing results.

10. Multiplicity, Missing Data and Analysis Populations

The registry record does not report a multiplicity strategy. It describes one primary outcome measure: the number of DLTs by cohort, used to identify a maximum tolerated dose. Although several cohorts and adverse-event categories appear in the record, it does not report a family of statistical tests, a hierarchy of endpoints, or an adjustment for multiple comparisons. With zero analyses listed, there are no posted p-values or confidence intervals whose multiplicity handling can be assessed. The absence of a reported strategy should not be interpreted as evidence that a particular adjustment was or was not used outside the registry.

Missing-data methods and analysis populations are not reported in the registry record. The primary endpoint is defined for the first treatment period, but the record does not explain how patients who stopped treatment early, lacked complete toxicity assessment, or had other incomplete observations would be handled. It does not specify an intention-to-treat, safety, evaluable, or other analysis population for the primary endpoint. The cohort description refers to patients experiencing DLTs in the first cycle, but that does not provide a complete population definition for analysis.

The study closed before enough patients were enrolled for the stated endpoint purpose, according to the primary outcome description. The phase II portion was not conducted because of a lack of funding, according to the brief summary. These registry entries limit completeness, but the record does not report how incomplete enrollment affected any analysis or whether a final dose decision was made. The posted cohort safety counts should not be assumed to include identical follow-up or complete endpoint ascertainment across participants. Any handling of missing observations, censoring, protocol deviations, or incomplete safety follow-up is not reported in the registry record.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: perforated sigmoid diverticulitis (20.0%), bacteremia cmv (20.0%), febrile neutropenia (20.0%), pneumonia (20.0%), line sepsis and neutropenia (20.0%).

The adverse-event display separates serious adverse events from other adverse events and presents counts by cohort. For serious events, the registry lists 5 at risk and 2 affected in Cohort 1, 3 at risk and 0 affected in Cohort 2, 5 at risk and 1 affected in Cohort 3, and 0 at risk and 0 affected in Cohort 4. The display also lists other adverse events and deaths by cohort. These entries should be read as descriptive registry counts, not as results of a randomized comparison.

The serious-event terms shown include Perforated sigmoid diverticulitis, Bacteremia CMV, Febrile neutropenia, Pneumonia, and Line sepsis and neutropenia, each listed as 20.0%. The record does not clarify the denominator or provide enough context to interpret these term percentages as a comparative risk estimate. Nor does it establish from these entries whether a particular event was treatment-related or met the primary DLT criteria.

In general, adverse-event displays depend on definitions, observation periods, and the number of participants observed. The registry record does not provide the detail needed to determine whether all cohorts had equivalent observation or whether each displayed event occurred in the first treatment cycle. These figures therefore offer a limited description of reported safety information, not a complete safety profile or an estimate of comparative risk.

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

The posted record shows a planned phase I investigation of gemcitabine with pemetrexed, using cohort-specific doses and an early toxicity endpoint. It records enrollment of 14, a terminated status, one outcome measure, and zero analyses. It also reports cohort-level adverse-event counts. These details describe what the study set out to do and what information appears in the registry, but they do not establish the primary endpoint result.

The record defines the maximum tolerated dose as the lowest dose level at which two or more patients experience a DLT. It also specifies that the endpoint was to cover the first 28-day treatment period. However, the outcome description states that the study closed before enough patients were enrolled, and the registry does not report a completed cohort DLT count or a selected dose. The phase II portion was never conducted, according to the brief summary. Thus, the intended dose-finding decision and planned later-stage assessment cannot be treated as completed findings.

The posted event counts do not show a causal effect of dose. The study was non-randomized, had no control group, and allocated patients across dose cohorts rather than creating a randomized comparison. Differences in descriptive counts cannot establish that one dose caused more or fewer events, especially without reported analysis, complete observation details, or adjustment for cohort differences. Serious adverse events also do not automatically equal DLTs, because the primary endpoint used its own toxicity definitions and time window.

This record does not report a treatment effect on disease outcomes, a confidence interval, a p-value, or a formal between-cohort analysis. It therefore cannot support conclusions about efficacy or comparative benefit. In general, statistical interpretation requires both an estimate and information about its uncertainty, alongside the design and endpoint definitions. Here, the central endpoint result and its uncertainty are unavailable in the registry record. The sound conclusion is limited: this was an early dose-finding study that terminated before the registry reported enough information to establish its primary result or proceed to the planned phase II portion.

13. Why This Trial Matters Statistically

This trial illustrates why the distinction between a planned endpoint and a reported result matters. The endpoint was designed around protocol-defined DLTs in the first treatment period, with the aim of identifying a maximum tolerated dose. The registry describes the decision rule, but it does not report the completed cohort-level DLT findings or the resulting dose. A clearly stated endpoint definition is valuable, yet it cannot stand in for observed endpoint data.

It also illustrates the limits of reading dose cohorts as if they were treatment arms in a comparative trial. The study was non-randomized and single-group, with dose combinations varying across cohorts. Those design features are appropriate context for an early dose-finding purpose, but they do not support causal comparisons from raw event counts. Reporting the design alongside the results is essential to prevent readers from interpreting descriptive differences as evidence of dose effects.

The record further shows how trial status and missing analyses shape what can be learned. The phase II portion was not conducted, and the registry lists no analyses. This leaves the intended later-stage evaluation unavailable and the primary dose-finding result unresolved in the posted record. The displayed safety counts remain useful as registry-reported information, but they cannot replace the prespecified endpoint analysis or establish a complete safety comparison.

For researchers and clinicians reading trial registries, the broader statistical lesson is to separate protocol intent, observed data, and analysis. An endpoint definition explains what was meant to be measured. A result reports what was observed. An analysis describes how observations were summarized and what uncertainty remains. In this record, those elements are not all present. Recognizing that distinction prevents overstatement and makes clear which questions remain unanswered.

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 most important limitation is that the primary DLT result is not reported in the registry record. The record states that the study closed before enough patients were enrolled for the endpoint purpose, but it does not provide the completed DLT counts or a maximum tolerated dose determination. The registry lists 14 enrolled participants, but the posted information does not show how many were evaluable for the primary endpoint in each cohort.

The study was terminated, non-randomized, and single-group. There was no control group, and the cohorts were planned dose levels rather than randomized comparators. The phase II portion was not conducted, so the record cannot provide the planned later-stage assessment. The brief summary attributes this to a lack of funding; further details about the termination are not reported in the registry record.

The posted safety display does not resolve these limitations. It gives counts by cohort, but the registry does not report enough detail about timing, completeness of observation, event attribution, or the relationship between serious adverse events and DLT classification. It also does not report a complete statistical plan, missing-data approach, analysis population, or inferential estimates. The listed serious-event terms and percentages should not be treated as a comprehensive or comparative safety analysis.

Finally, the enrollment rationale and statistical assumptions are not reported in the registry record. The posted record does not establish whether planned cohort enrollment or follow-up was completed as intended, nor whether a final analysis was conducted elsewhere. These gaps restrict interpretation to the study's stated design and the descriptive information visible in the registry.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Gemcitabine and Pemetrexed Disodium in Treating Patients With Advanced Mycosis Fungoides or Sézary Syndrome — NCT00369629.
• Registry record last updated 2019-08-28; status last verified 2019-08.
• Results first posted 2018-10-09.
• 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

NCT00369629 is a terminated, non-randomized, single-group phase I trial sponsored by Northwestern University. It studied planned combinations of gemcitabine and pemetrexed for patients with advanced mycosis fungoides or Sézary syndrome, described in the official title as advanced cutaneous T-cell lymphoma. The registry lists enrollment of 14 and four planned dose cohorts.

The primary outcome was the number of dose-limiting toxicities in each cohort during the first 28-day treatment period, used to determine a maximum tolerated dose. The record defines the dose threshold but does not report the completed primary result. It lists one outcome measure and zero analyses. The phase II portion was never conducted because of a lack of funding, according to the brief summary.

The registry also displays serious adverse events, other adverse events, and deaths by cohort. These counts are descriptive and do not constitute a randomized comparison or the missing primary endpoint analysis. The posted record supports interpretation of the trial's intent and reporting limits, but not a completed dose-selection or efficacy conclusion.