Phase I Dose Escalation Trial of Efavirenz in Solid Tumours or Non-Hodgkin Lymphoma in Therapeutic Failure.: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Phase I Dose Escalation Trial of Efavirenz in Solid Tumours or Non-Hodgkin Lymphoma in Therapeutic Failure. (NCT01878890), a non-randomized phase 1 study. The primary outcome is Number of Participants Who Experienced Dose-Limiting Toxicities (DLTs).

2. Clinical Question

The trial is NCT01878890, a phase I study sponsored by Institut Bergonié. The registry record describes a single-center study for patients whose disease was in therapeutic failure. Its stated objective was to evaluate efavirenz doses above 600 mg per day and characterize safety. This is a dose-escalation question: the study examines tolerability across successive dose cohorts, rather than estimating the effect of efavirenz against an untreated or active control group.

The brief summary also describes pharmacokinetic assessments at 2, 4, and 12 weeks, objective response at 12 weeks, progression-free survival at 6 months, and biological progression-free survival at 6 months for prostate tumours only. The secondary outcomes listed in the registry include 12-week objective response rate and 12-week non-progression rate. Results for these outcomes are not reported in the registry record.

3. Trial Design

NCT01878890 is listed as a phase I, non-randomized, unmasked, sequential trial. It enrolled 25 participants. The registry record gives a start date of 2011-09-05, a primary completion date of 2014-09-16, and a completed status. The sponsor is Institut Bergonié.

The study used dose escalation, with cohorts receiving efavirenz 600 mg, 1200 mg, 1800 mg, or 2200 mg. The cohort descriptions specify oral administration once a day, continuing until progression or toxicity. The brief summary describes the escalation scheme as the continual reassessment method likelihood approach (CRML). The registry record does not report further details of how the method was implemented, including its decision rules or the criteria for moving between doses.

Because allocation was non-randomized, dose groups are not randomized arms intended for direct causal comparison. Dose assignment is part of the escalation process. Differences in outcomes across cohorts could reflect characteristics of the sequential dose-finding design as well as dose, and should not be interpreted as if the groups had been made comparable by random assignment. The record does not report a control group.

The primary objective concerned safety, particularly the maximum tolerated dose. The primary outcome was the number of participants who experienced dose-limiting toxicities (DLTs), assessed up to 28 days for each dosing cohort. The record also describes secondary assessments of pharmacokinetics, objective response, progression-free survival, and biological progression-free survival for prostate tumours only. Specific analysis methods and results for these outcomes are not reported in the registry record.

The registry record does not report a sample-size justification, detailed stopping rules, or a statistical analysis plan. Its study description names CRML, but that alone does not establish the exact algorithm settings, model assumptions, or operational decisions used in the trial.

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 Who Experienced Dose-Limiting Toxicities (DLTs), measured over Up to 28 days for each dosing cohort.

The primary endpoint is the number of participants who experienced dose-limiting toxicities. It is a count-based safety outcome, assessed up to 28 days for each dosing cohort. The endpoint description defines a DLT as drug-related toxicity with grade >= 3 according to NCI-CTCAE v4.0, except alopecia, nausea, and vomiting regardless of grade. It also includes drug-related toxicity, regardless of grade, that led to a treatment delay > 14 days, and a Hospital Anxiety And Depression Scale (HAD) score >= 19 during treatment.

In general, a DLT endpoint is used in dose-finding research to identify toxicities that limit treatment during a specified observation window. It focuses attention on events considered important for dose escalation, rather than treating all adverse events as equivalent. The listed definition combines toxicity grade, treatment delay, and a scale-score criterion. The registry record does not report whether any participants met these criteria, how the outcome was handled if assessment was incomplete, or what dose was judged to be the maximum tolerated dose.

The endpoint is a participant count, not a measure of the total number of toxicity episodes. The registry record does not provide the event-level results or explain whether participants could meet more than one DLT criterion. Without the primary outcome results, the endpoint definition can be described, but the observed safety findings for this outcome cannot be interpreted.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several statistical ideas help distinguish what a dose-escalation study can address from what a randomized comparison can address. The explanations below are general unless they describe a fact specific to this trial.

What does sequential dose escalation mean?

In general, a sequential dose-escalation design uses information from participants treated at earlier stages to guide which dose is considered next. This differs from random assignment, where treatment assignment is randomized across groups according to a prespecified allocation process. The registry record identifies this trial as sequential and non-randomized, and names the continual reassessment method likelihood approach (CRML). The record does not report the model specification, dose-toxicity assumptions, or decision rules, so the name alone is not enough to reconstruct the procedure.

Because allocation is not randomized, observed differences between dose cohorts do not have the same interpretation as differences between randomized arms. Cohorts may differ in when participants entered the study or in other characteristics; whether and how that occurred is not reported in the registry record. The cohorts are best understood as dose levels within a dose-finding process, not as a controlled effectiveness comparison.

What does a DLT count measure?

In general, a count of participants with an event records how many people met a defined outcome criterion. It is not the same as the total number of event episodes, because a participant may have more than one event. The primary endpoint in this trial counts participants who experienced a DLT during the specified assessment period. Its definition covers toxicity grade, treatment delay, and a HAD score threshold. Applying this endpoint requires consistent event assessment and attribution to the drug; the registry record does not report operational details beyond the written definition.

A count can be read alongside the number of participants assessed in a cohort to understand how common the outcome was in that cohort. The registry record does not give the primary DLT counts, so a cohort-specific DLT rate cannot be interpreted in the registry record reported. No rate should be inferred from the separate adverse-event summaries: those concern serious and other adverse events, not necessarily the primary DLT definition.

Why do confidence intervals matter?

In general, a confidence interval communicates the uncertainty around an estimate under the analysis assumptions. A narrow interval indicates greater precision than a wide interval, while neither width nor location alone resolves issues such as selection bias, measurement error, or an unsuitable design. The registry record does not report confidence intervals for the primary endpoint or the adverse-event summaries. Without an estimate and its uncertainty, the precision of the observed DLT experience cannot be assessed.

What can the named escalation method tell the reader?

The continual reassessment method is a model-based approach to dose finding in general. It uses a relationship between dose and toxicity to inform dose selection during escalation. The registry record’s naming of the CRML approach indicates the stated method family, but the record does not report enough detail to determine its exact implementation or evaluate whether its assumptions were met. It does not report the target toxicity level, any prior assumptions, the dose-selection calculations, or the final dose recommendation.

In general, model-based dose-finding methods depend on how toxicity is defined and observed, and on how the model links dose to the probability of toxicity. An observation window is important because participants need sufficient follow-up to determine whether an event qualifies. This trial defines DLT collection during the first 28 days (+ / - 7 days) after the first dose in the brief summary, while the primary outcome timeframe is described as up to 28 days for each dosing cohort. The registry record does not explain how those descriptions were reconciled in analysis.

How should sample-size planning be understood?

In general, a sample-size calculation describes a hypothetical design under stated assumptions; it is not a result from the completed trial. Planning illustrations for participants per arm for 80% power or 90% power depend on the effect size and analysis assumptions. They should not be called observed power. The planning assumptions described for this record use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect stated as a fraction of the observed effect. That framework is not the same as the sequential, non-randomized dose-escalation design described for the trial.

For count outcomes, a planning calculation may also depend on the assumed event frequency and the way events are modeled. In general, over-dispersion occurs when variation in counts exceeds what a simpler model assumes; accounting for it can affect uncertainty and planning. The registry record does not report a count model or an over-dispersion assumption for the primary DLT endpoint. The sample-size assumptions do not provide a trial-specific justification in the information recorded, and the record does not report a sample-size calculation used to set enrollment.

These distinctions matter because planning illustrations cannot fill gaps in outcome reporting. They do not supply the missing DLT counts, establish the precision of this study’s safety estimates, or provide evidence of comparative treatment effects.

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("Efavirenz: 600 mg", "Efavirenz: 1200 mg", "Efavirenz: 1800 mg", "Efavirenz: 2200 mg")
n      <- c(5, 10, 4, 5)
events <- c(0, 3, 1, 3)
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  Efavirenz: 600 mg  5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9
ARM  Efavirenz: 1200 mg  10  3  30.0  6.7  65.2  10.8  60.3  10.9  61.0
ARM  Efavirenz: 1800 mg  4  1  25.0  0.6  80.6  4.6  69.9  5.3  71.6
ARM  Efavirenz: 2200 mg  5  3  60.0  14.7  94.7  23.1  88.2  22.3  88.2

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists 4 outcome measures and 0 analysis entries. The primary outcome is the participant count for DLTs. The listed secondary outcomes are 12-week objective response rate and 12-week non-progression rate. The brief summary also describes pharmacokinetic assessments at 2, 4, and 12 weeks, progression-free survival at 6 months, and biological progression-free survival at 6 months for prostate tumours only. Results for these assessments are not reported in the registry record.

In general, evaluating multiple outcomes can raise multiplicity concerns: when many hypotheses are tested, the chance of at least one apparently unusual result can increase. Researchers may address this through a prespecified hierarchy, adjustment, or careful emphasis on estimation rather than isolated significance tests. The registry record does not report a multiplicity strategy, a hierarchy among secondary outcomes, or inferential analyses for these outcomes. The absence of analysis entries does not establish what analyses may have been performed outside the registry record.

The registry record does not report how missing outcome data were handled, whether any participants were excluded from particular analyses, or how analysis populations were defined. It also does not report the completeness of follow-up for each outcome. These gaps matter because incomplete observation can affect both event counts and estimates, especially when outcomes are assessed at specified time points. The record does not describe imputation, censoring rules, or sensitivity analyses, so none should be assumed.

The stated planning illustrations use equal allocation between two arms, while the trial is described as sequential and non-randomized across dose cohorts. This difference in design means the illustrations should not be taken as an analysis plan for the trial or as observed power.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: other (100.0%), other (20.0%).

The registry safety summary presents, by dose cohort, the number at risk and the number affected for serious adverse events and other adverse events, along with a deaths count. For serious adverse events, the entries are: 600 mg, 5 at risk and 3 affected; 1200 mg, 11 at risk and 4 affected; 1800 mg, 4 at risk and 4 affected; and 2200 mg, 5 at risk and 3 affected. For other adverse events, the entries are: 600 mg, 5 at risk and 5 affected; 1200 mg, 11 at risk and 11 affected; 1800 mg, 4 at risk and 4 affected; and 2200 mg, 5 at risk and 5 affected. The deaths counts listed for these cohorts are 5, 11, 4, and 5, respectively.

These are descriptive counts, not a randomized comparison. The record does not report whether the deaths were related to treatment, the timing or causes of death, or how serious events overlapped with other event categories. The serious-event terms include “Other” at 100.0% and “other” at 20.0%; the registry record does not clarify the differing capitalization or the denominator for these entries. Do not interpret those percentages as cohort-level risks without that context. The safety summaries are also not interchangeable with the primary DLT endpoint, which has its own definition and assessment window.

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

This trial addresses a dose-finding and safety question in patients with solid tumours or non-Hodgkin lymphoma in therapeutic failure. Its sequential, non-randomized design is consistent with examining multiple dose cohorts during escalation. The registry record specifies four efavirenz doses and describes a primary endpoint based on DLTs. It also identifies response and non-progression outcomes, among other planned assessments.

What the registry record does not establish is equally important. The primary DLT results are not reported, so the record does not show the observed DLT count by cohort or identify a maximum tolerated dose. It does not report the results for the listed response outcomes or the other assessments described in the brief summary. The available safety counts cannot substitute for those missing endpoint results, because serious or other adverse-event categories are not identical to the DLT definition.

In general, the design can provide descriptive information about tolerability at the doses studied, but a non-randomized dose-escalation study is not designed as a like-for-like comparison of treatment efficacy against a control. Differences between dose cohorts should not be presented as definitive causal effects of dose. No treatment-effect conclusion can be drawn from the registry record’s design description alone.

The reported adverse-event counts also do not establish causation. The record does not provide the detail needed to connect deaths or other events to treatment, assess their duration, or account for follow-up time. Likewise, without reported estimates and uncertainty, the precision of any cohort-level safety findings cannot be judged.

In short, the record documents how the trial was structured and what it intended to assess. It does not report enough outcome information to determine the trial’s primary safety finding or draw conclusions about response. The absence of these results in the registry record should not be mistaken for evidence that an outcome did or did not occur.

14. Why This Trial Matters Statistically

This trial illustrates why design and purpose need to be considered before comparing numbers across groups. In a randomized trial, random assignment is intended to support a balanced comparison between treatment groups. Here, allocation was non-randomized and sequential, with cohorts receiving different efavirenz doses. The cohort counts are therefore descriptive dose-finding information, not a controlled estimate of the effect of one dose versus another.

The primary endpoint also shows how an outcome definition shapes interpretation. DLTs are not simply any adverse event: the record specifies toxicity grade, exceptions for named symptoms, treatment delay, and a HAD score criterion. A carefully specified outcome can make dose-escalation decisions more consistent, but readers still need the observed endpoint results and details about assessment to understand what occurred. Those results are not reported in the registry record.

Another statistical distinction is between outcome information and planning information. Sample-size illustrations based on an observed effect rely on assumptions and do not describe the trial’s observed performance. They are not observed power, and their stated equal-allocation, two-arm framework differs from this study’s sequential dose cohorts. This distinction helps prevent planning calculations from being mistaken for trial findings.

Finally, registry summaries may contain safety counts without providing the context needed for causal interpretation or precision assessment. Counts can support a descriptive account, but they do not by themselves show why an event occurred, how long participants were followed, or how uncertain an estimate is. This record is a useful example of why endpoint definitions, allocation methods, missing results, and analysis details all matter when reading clinical trial evidence.

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 registry record does not report the primary DLT outcome results, despite describing the endpoint and its assessment period. It therefore does not provide the principal information needed to understand the safety findings across the dose cohorts or identify the maximum tolerated dose.

The record also does not report the statistical analysis plan, detailed CRML implementation, dose-escalation decision rules, or sample-size justification. The method name alone does not reveal the model assumptions or the criteria used to select a dose. Analysis populations, missing-data handling, and outcome-specific follow-up completeness are not reported in the registry record.

The non-randomized, sequential design limits direct comparisons between dose cohorts. The safety counts are descriptive and do not establish treatment causation. The record does not provide event timing, causes of death, treatment exposure duration, or uncertainty estimates for these summaries. The serious-event term labels “Other” and “other” appear with different percentages, but the registry record does not clarify their denominators or relationship.

Further limitations include the lack of reported results for response and other secondary assessments, and the absence of reported multiplicity procedures. The brief summary describes assessments at specified time points, but the record does not state how incomplete follow-up, censoring, or missing assessments were handled. These omissions prevent a full evaluation of the planned endpoints and the precision or robustness of any findings.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Phase I Dose Escalation Trial of Efavirenz in Solid Tumours or Non-Hodgkin Lymphoma in Therapeutic Failure. — NCT01878890.
• Registry record last updated 2021-01-27; status last verified 2021-01.
• Results first posted 2021-01-27.
• 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

NCT01878890 is a completed phase I, non-randomized, unmasked, sequential dose-escalation trial sponsored by Institut Bergonié. The registry record reports enrollment of 25 and describes oral efavirenz cohorts at 600 mg, 1200 mg, 1800 mg, and 2200 mg. The study population was patients with solid tumours or non-Hodgkin lymphoma in therapeutic failure; pancreatic cancer was excluded in the brief summary.

The primary outcome was the number of participants experiencing DLTs, assessed up to 28 days for each dosing cohort. The record also describes response, non-progression, pharmacokinetic, and survival assessments. Descriptive safety counts are reported by cohort, but primary DLT results and results for the other listed outcomes are not reported in the registry record. The record does not report a maximum tolerated dose, detailed analysis plan, or missing-data procedures.