A Study of Vemurafenib in Participants With BRAF V600 Mutation-Positive Cancers: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Vemurafenib in Participants With BRAF V600 Mutation-Positive Cancers (NCT01524978), a non-randomized phase 2 study. The primary outcome is Confirmed Best Overall Response Rate (BORR).

2. Clinical Question

This question is descriptive of activity across a group of selected, molecularly defined cancers, not a randomized comparison of treatments. Because the cohorts covered different malignancies and included both monotherapy and combination therapy, a pooled response measure would need careful interpretation. The registry record describes separate cohorts, but does not report enough outcome detail to establish the response in each cancer type or regimen. It also does not report an estimate that would quantify how outcomes compare with an alternative treatment.

3. Trial Design

This was an open-label, phase II, multi-center study with non-randomized allocation and a parallel-group model. The registry lists an enrollment of 208 and a completed status. Its start date is 2012-04-12, and its primary completion date is 2016-10-28. These record-level details describe the study, but do not establish how many participants contributed to each outcome analysis.

The cohorts covered non-small cell lung cancer, ovarian cancer, colorectal cancer, cholangiocarcinoma, multiple myeloma, and other solid tumors. Vemurafenib monotherapy was used in most cohorts. One colorectal cancer cohort received vemurafenib with cetuximab, while another received vemurafenib alone. The registry describes separate analysis of some subcohorts if they had 7 or more participants, but does not provide a complete account of how every cohort contributed to each outcome.

Open-label means that treatment assignment was not masked. Non-randomized means that participants were not assigned to groups through random allocation. Therefore, treatment arms should not be interpreted as balanced comparison groups. The registry warns that the arms are not a like-for-like control comparison. Differences between cohorts could reflect cancer type, participant selection, or treatment regimen; the record does not report an adjustment strategy that would separate these influences.

The study was designed to assess efficacy and safety. The listed secondary outcomes include clinical benefit, overall response, duration of response, time to response, time to tumor progression, progression-free survival, overall survival, and maximum tolerated dose for the vemurafenib-cetuximab combination. The registry record does not report analysis results for these outcomes in the registry record for this account.

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.

Before assignment: One participant with breast cancer was screened shortly after Cohort 5 (Breast Cancer) had been closed. This participant was allowed to enter the study in Cohort 7: Other BRAF V600-positive tumors. For analysis purposes Cohort 7 was split into sub-cohorts for indications with sufficient participants.

Only the first 6 of 11 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: ITT population included all participants enrolled in the study irrespective of whether they had received study drug or not.

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: Intent-to-treat (ITT) population included all participants enrolled in the study irrespective of whether they had received study medication or not.

Only the first 6 of 12 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Confirmed Best Overall Response Rate (BORR), measured over Up to approximately 3 years.

The primary endpoint was confirmed best overall response rate, expressed as the percentage of participants with an objective response. For solid tumors, the investigator used RECIST v1.1; for multiple myeloma, the investigator used International Myeloma Working Group criteria. The response categories included complete response and partial response, with stringent complete response and very good partial response also included for myeloma. A qualifying response had to be confirmed on two occasions at least 4 weeks apart. The timeframe was up to approximately 3 years.

These criteria provide structured rules for deciding whether disease measurements meet definitions of response. As a general statistical point, a response rate summarizes the proportion of an analysis population that meets a prespecified outcome definition; it does not describe how long responses last or compare treatments by itself. Requiring confirmation helps distinguish a response that persists on a subsequent assessment from an isolated observation. The registry record does not report the primary response percentage or its 95% confidence interval, so the magnitude and precision of the endpoint cannot be assessed from the reported details.

The endpoint combines response categories and applies different disease-specific criteria. That makes the result clinically interpretable within each assessment framework, but cautions against assuming that the same measurement process applies identically across all cancer types. The registry record does not report cohort-level response results that would permit that comparison.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a response percentage represent?

In general, a response percentage is a proportion: the count meeting the response definition divided by the relevant analysis population, then expressed as a percentage. It depends on both the endpoint rules and the population used as the denominator. Here, the registry defines the primary analysis population as all enrolled participants, whether or not they received study drug. However, the outcome numerator and denominator are not reported in the registry record, so the response percentage cannot be reconstructed or checked.

The response definition is based on investigator assessment under RECIST v1.1 or International Myeloma Working Group criteria. For RECIST, complete response requires disappearance of target lesions, and partial response requires a reduction of at least 30% in the sum of target-lesion diameters. For myeloma, the registry defines complete response using immunofixation, soft tissue plasmacytomas, and bone marrow findings. Partial response requires at least a 50% reduction in serum M-protein and a reduction in 24-hour urinary M-protein by at least 90% or to less than 200 mg per 24 hours. Very good partial response and stringent complete response have additional criteria. Requiring a response on two occasions at least 4 weeks apart adds a confirmation rule. These criteria are part of the endpoint definition, not observed results.

Why does the confidence interval matter?

In general, a confidence interval conveys uncertainty around an estimated response proportion. A narrow interval indicates greater statistical precision than a wide interval, given the assumptions used to calculate it. The registry lists 95% confidence interval as the measure of dispersion for the primary endpoint, but no interval or point estimate is reported in the registry record for this account. Therefore, precision cannot be evaluated.

For a percentage outcome, the amount of information depends on the number of participants who contribute to the calculation and on how many meet the endpoint definition. The overall enrollment alone does not give that information. Cohort size, the analysis denominator, and missing outcome assessments can all matter. None of the primary endpoint counts by cohort is reported in the registry record.

What does non-randomized allocation mean for comparisons?

As a general statistical principle, random allocation can help make groups comparable at the start of a study, on average, for measured and unmeasured characteristics. This study was non-randomized. Its cohorts also represented different malignancies and included distinct regimens, so an observed difference between groups could reflect factors other than the treatment. The registry explicitly says the arms are not a like-for-like control comparison. A comparison across these arms should therefore not be interpreted as a randomized estimate of relative treatment effect.

Open-label design means treatment was not masked. In general, knowledge of treatment can influence subjective aspects of care or assessment. The primary endpoint used named response criteria, which structure assessment, but the registry record does not describe independent review or blinded assessment. It states that the investigator assessed response.

How should the safety counts be read?

The serious adverse event summaries give participants at risk and participants affected for two groups. In the colorectal combination cohort, 27 were at risk and 11 were affected; in the pooled vemurafenib arm, 181 were at risk and 91 were affected. These are descriptive counts, not randomized comparisons. They do not indicate the timing, duration, severity distribution, or causal relationship of each event. The registry also lists serious adverse event terms with percentages, including squamous cell carcinoma of skin at 13.8% and keratoacanthoma at 9.9%. These terms and percentages should be understood as registry summaries; the details needed to compare risks across cohorts are not reported.

As a general concept, adverse event counts can differ from counts of events because one participant may experience more than one event. The registry summaries identify participants affected, but the registry entries available do not specify event-level counts or provide the complete safety analysis rules. The death field in the listed group summaries is null, so no death count is reported in those summaries.

What does a planning illustration tell a reader?

In general, a sample-size planning calculation asks how many participants would be needed under stated assumptions to detect an effect with a chosen error rate and power. The registry’s planning assumptions specify a two-sided significance level of 0.05, equal allocation, a normal approximation, and an effect expressed as a fraction of the observed effect. Such calculations are illustrations based on assumptions, not observed trial outcomes and not observed power. They do not replace the trial’s own design justification. The registry record does not report a sample-size justification or give the resulting planning figures in the registry record for this account.

For time-to-event outcomes, general planning methods often use the number of events as the key information quantity; for other outcomes, a planning figure may refer to participants per arm. These distinctions are relevant only when interpreting a stated planning calculation. No trial-specific planning result or rationale is reported in the registry record.

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("Cohort 1: Non-Small Cell Lung Cancer (NSCLC) - Vem", "Cohort 2: Ovarian Cancer - Vemurafenib", "Cohort 3a: Colorectal Cancer - Vemurafenib", "Cohort 3b: Colorectal Cancer - Vemurafenib + Cetux", "Cohort 4: Cholangiocarcinoma - Vemurafenib", "Cohort 6: Multiple Myeloma - Vemurafenib")
n      <- c(62, 4, 10, 27, 9, 9)
events <- c(23, 2, 0, 2, 2, 2)
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  Cohort 1: Non-Small Cell Lung Cancer (NSCLC) - Vem  62  23  37.1  25.2  50.3  26.2  49.5  26.1  49.6
ARM  Cohort 2: Ovarian Cancer - Vemurafenib  4  2  50.0  6.8  93.2  15.0  85.0  14.7  85.3
ARM  Cohort 3a: Colorectal Cancer - Vemurafenib  10  0  0.0  0.0  30.8  0.0  27.8  0.2  28.5
ARM  Cohort 3b: Colorectal Cancer - Vemurafenib + Cetux  27  2  7.4  0.9  24.3  2.1  23.4  2.3  23.5
ARM  Cohort 4: Cholangiocarcinoma - Vemurafenib  9  2  22.2  2.8  60.0  6.3  54.7  6.7  55.6
ARM  Cohort 6: Multiple Myeloma - Vemurafenib  9  2  22.2  2.8  60.0  6.3  54.7  6.7  55.6

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary response endpoint and multiple secondary outcomes: clinical benefit, overall response, duration of response, time to response, time to tumor progression, progression-free survival, overall survival, and maximum tolerated dose for vemurafenib with cetuximab. When many outcomes are examined, the chance of at least one apparently unusual result can increase. As a general statistical principle, multiplicity procedures can address this issue by adjusting how evidence is interpreted across tests. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting the secondary outcomes.

The registry entries list zero analyses. They do not provide outcome-specific estimates, confidence intervals, hypothesis tests, or analysis populations for the listed secondary endpoints. It is therefore not possible to determine which outcomes were analyzed, whether any analyses were prespecified, or how multiple outcomes were handled. Interim analyses and stopping rules are not reported in the registry record.

The primary endpoint population is described as all enrolled participants, regardless of whether they received study drug. That definition is useful, but it does not disclose how missing response assessments, withdrawals, or participants without an evaluable assessment were handled in the calculation. The registry record does not report a missing-data method, imputation strategy, sensitivity analysis, or the number of participants with missing primary endpoint information. A specified analysis population does not, by itself, explain how unavailable measurements enter an endpoint estimate.

The listed study enrollment should not be assumed to equal the denominator for every outcome. The record does not report the number analyzed by cohort or endpoint. It also does not report a protocol-level sample-size justification. These gaps limit assessment of the analysis population, precision, and robustness of any estimated effect.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: squamous cell carcinoma of skin (13.8%), keratoacanthoma (9.9%), basal cell carcinoma (3.9%), squamous cell carcinoma (3.7%), upper gastrointestinal haemorrhage (3.7%).

The registry reports serious adverse events affecting 11 of 27 participants at risk in the colorectal cancer cohort receiving vemurafenib with cetuximab, and 91 of 181 participants at risk in the pooled vemurafenib arm. These are summaries of participants affected, not a like-for-like comparison between randomized groups. The cohorts differ in regimen and may differ in cancer type and other characteristics, so the figures should be read descriptively rather than as evidence of a relative safety effect.

The registry lists squamous cell carcinoma of skin at 13.8%, keratoacanthoma at 9.9%, basal cell carcinoma at 3.9%, squamous cell carcinoma at 3.7%, upper gastrointestinal haemorrhage at 3.7%, and subileus at 3.7% among the top serious terms. These percentages are registry summaries; the record details do not specify the corresponding term-specific denominators or event counts. They should not be added together, since the record does not establish that the categories are mutually exclusive.

The group summaries do not report deaths. The adverse event information does not establish causation, timing, or comparative risk. It is a selected registry summary rather than a complete account of each participant’s experience.

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

The study addresses a response question in participants with BRAF V600 mutation-positive cancers treated in several disease-specific cohorts. Its primary endpoint uses defined response criteria and a confirmation requirement, which gives the intended outcome a clear operational meaning. The registry also records selected serious adverse event summaries and lists a range of secondary outcomes.

However, the primary response estimate and its confidence interval are not reported in the registry record. The registry entries therefore do not show how many participants had a confirmed response, how the response percentage varied by cancer type, or how precise any estimate was. A completed study status and an enrollment of 208 do not answer those endpoint questions.

The non-randomized design also limits causal interpretation. Participants were not assigned randomly to a common treatment comparison, and the cohorts included different cancers and regimens. The registry cautions that the arms are not like-for-like controls. Even if numerical outcomes were available for each arm, differences would need to be interpreted in light of these design features rather than treated as a clean comparison of treatment effects.

The serious adverse event summaries describe affected participants and selected event terms, but do not establish that a treatment caused an event. They also do not report deaths in the listed summaries. These safety data should not be combined with the unreported response outcome to make a broad claim about benefit or harm.

In general, a trial result is most informative when an estimate is accompanied by its analysis population, uncertainty interval, and clear account of the comparison being made. Those elements are not available for the primary endpoint in the registry entries. Accordingly, the record supports describing the study design, outcome definition, and selected safety summaries, but not a conclusion about the size or precision of the primary response result.

14. Why This Trial Matters Statistically

Statistically, this study illustrates the distinction between a clearly defined endpoint and a reported endpoint result. The registry describes confirmed best overall response rate, specifies disease-specific response criteria, identifies the investigator as assessor, and requires confirmation on two occasions at least 4 weeks apart. Yet without the primary estimate, denominator, and confidence interval, the response outcome cannot be quantified or assessed for precision.

The design also makes the comparison question important. The study was non-randomized and included disease-specific cohorts, with both monotherapy and a vemurafenib-cetuximab regimen. Those arms are not like-for-like controls. This illustrates why separate cohorts should not automatically be read as comparative groups: their outcomes may reflect differences in the diseases represented and in the treatment regimens, as well as other factors.

The record further distinguishes the planned scope of measurement from reported analysis. It lists a broad set of secondary outcomes, but the registry entries available do not include their results or analysis details. As a general statistical lesson, outcome lists alone do not establish the number of analyses completed, the handling of multiple tests, or how missing observations were treated.

Finally, the safety summaries show why denominators and context matter. They provide numbers at risk and affected for two groups and list selected serious event terms, but the non-randomized structure and incomplete detail prevent a straightforward comparative reading. The study is therefore useful as an example of how registry records can specify populations and endpoints in detail while leaving important information about estimates, uncertainty, and analysis methods unreported.

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 primary response estimate and its 95% confidence interval are not reported in the registry record. There is no outcome numerator or endpoint denominator by cohort, so neither the magnitude nor the precision of the primary result can be evaluated. The study enrollment of 208 should not be substituted for an outcome-specific analysis count.

The population and regimens were heterogeneous: the cohorts covered different cancers, and most received vemurafenib alone while a colorectal cohort received vemurafenib with cetuximab. Allocation was non-randomized, and the registry warns that arms are not like-for-like control groups. The record does not provide an adjusted analysis that would support a causal comparison between them.

The registry identifies the investigator as the response assessor but does not report whether an independent review was conducted. It does not report how missing or unevaluable response assessments were handled, or provide sensitivity analyses. It also does not report multiplicity procedures for the several secondary outcomes, interim analyses, or the study’s sample-size justification.

Safety reporting is limited in the registry entries to affected-participant summaries for two groups, selected serious adverse event terms and percentages, and no death counts in the listed summaries. The term-specific denominators and full event-level detail are not reported. These gaps constrain interpretation of both the frequency and comparative meaning of the safety summaries.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of Vemurafenib in Participants With BRAF V600 Mutation-Positive Cancers — NCT01524978.
• Registry record last updated 2017-11-20; status last verified 2017-10.
• Results first posted 2017-11-20.
• 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

This completed, open-label, phase II, non-randomized, multi-center study enrolled participants with BRAF V600 mutation-positive cancers across several disease-specific cohorts. Most cohorts received vemurafenib monotherapy; one colorectal cancer cohort received vemurafenib with cetuximab. The primary endpoint was confirmed best overall response rate, assessed by the investigator under RECIST v1.1 or International Myeloma Working Group criteria, with confirmation at least 4 weeks apart.

The registry record does not report the primary response estimate, its confidence interval, or outcome-specific analysis counts. It lists several secondary outcomes and provides selected serious adverse event summaries, but does not report a multiplicity method or missing-data approach in the registry record. Because allocation was non-randomized and the cohorts were not like-for-like controls, the arm summaries are descriptive rather than randomized treatment comparisons.