This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Capmatinib, Ceritinib, Regorafenib, or Entrectinib in Treating Patients With BRAF/NRAS Wild-Type Stage III-IV Melanoma (NCT02587650), a non-randomized phase 2 study. The primary outcome is Confirmed Overall Response Rate (ORR).
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02587650 |
| Phase | Phase 2 |
| Condition | ALK Fusion Protein Expression, BRAF wt Allele, Invasive Skin Melanoma |
| Allocation | Non-Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 1 |
| Arms in the primary outcome | n/a |
| Lead sponsor | University of California, San Francisco |
| Results posted | Yes |
2. Clinical Question
Population
The trial concerned patients with unresectable stage III/IV melanoma described as BRAF/NRAS wild-type. Treatment groups were defined by the kinase fusion present: MET, ALK, RET or BRAF, and NTRK1, NTRK2, NTRK3 or ROS1. The registry's primary-endpoint description refers to patients with rearrangement-positive tumors.
Intervention
The study evaluated capmatinib, ceritinib, regorafenib and entrectinib. Which drug was used depended on the fusion group: capmatinib for MET, ceritinib for ALK, regorafenib for RET or BRAF, and entrectinib for NTRK1, NTRK2, NTRK3 or ROS1.
Comparator
There was no randomized control group. The registry describes a nonrandomized, parallel-arm design, with treatment assigned according to fusion group. These arms therefore do not constitute a like-for-like comparison of the drugs.
Primary question
The primary endpoint was confirmed overall response rate at 24 weeks. It asked whether an evaluable participant had a complete or partial response under RECIST version 1.1, confirmed by a measurement at least 4 weeks after the response-defining assessment. No participants met the registry's evaluability criteria, so the record contains no evaluable primary-endpoint result.
3. Trial Design
This phase II study, NCT02587650, was sponsored by the University of California, San Francisco. Its official title describes a trial of targeted kinase fusion inhibition in unresectable stage III/IV BRAF/NRAS wild-type melanoma. The registry lists the study as nonrandomized, unmasked and parallel-arm. It records enrollment of 1 and a status of terminated. The recorded start date is 2015-03-26, and the primary completion date is 2018-07-12.
Four treatment groups were specified according to tumor fusion. Participants with MET fusion were assigned to capmatinib, taken by mouth twice daily on days 1-28 of repeating 28-day courses. The ALK fusion group received ceritinib once daily on days 1-28, with courses repeating every 28 days. The RET or BRAF fusion group received regorafenib once daily on days 1-21 of repeating 28-day courses. Participants with NTRK1, NTRK2, NTRK3 or ROS1 fusion received entrectinib once daily on days 1-28, with courses repeating every 28 days. The descriptions specify continuing courses in the absence of disease progression or unacceptable toxicity. Laboratory biomarker analysis was also listed as an intervention.
Because treatment depended on fusion group and assignment was not randomized, the arms were not designed as direct, balanced comparisons of the drugs. Differences between groups could reflect differences in the fusion-defined populations as well as treatment. The registry record does not report a control arm, masking procedures beyond none, or a rationale for termination.
The primary endpoint was confirmed overall response rate at 24 weeks. Secondary outcomes were clinical benefit rate, overall survival, progression-free survival and evaluation of each kinase inhibitor's adverse-effect profile. The registry lists five outcome measures, but no outcome analysis results. Its primary-endpoint description states that no participants met the criteria to be evaluable for that analysis.
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
| Stage | Arm A (Capmatinib) | Arm B (Ceritinib) | Arm C (Regorafenib) | Arm D (Entrectinib) |
|---|---|---|---|---|
| STARTED | 0 | 0 | 1 | 0 |
| COMPLETED | 0 | 0 | 1 | 0 |
| NOT COMPLETED | 0 | 0 | 0 | 0 |
Analysis population for the primary outcome, as defined in the registry: Evaluable participants would be those with baseline staging, treatment for at least 8 weeks, and at least one post-baseline staging scan while on treatment. No participants meet the criteria to be considered evaluable for this analysis.
5. Baseline Characteristics
The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.
Population: Only 1 participant was enrolled on this protocol to Arm C (regorafenib). Evaluable patients would be those with baseline staging, treatment for at least 8 weeks, and at least one post-baseline staging scan while on treatment. No participants meet the criteria to be considered evaluable for this protocol.
| Characteristic | Unit | Arm C (Regorafenib) |
|---|---|---|
| Participants analysed | 1 | |
| Age, Customized: 60-65 years | Participants | 1 |
| Sex: Female, Male: Female | Participants | 0 |
| Sex: Female, Male: Male | Participants | 1 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 0 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 1 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 |
| Race (NIH/OMB): Asian | Participants | 0 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 |
| Race (NIH/OMB): Black or African American | Participants | 0 |
| Race (NIH/OMB): White | Participants | 1 |
| Race (NIH/OMB): More than one race | Participants | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 0 |
| Region of Enrollment: United States | participants | 1 |
6. Primary Endpoint
The registry lists the primary outcome as Confirmed Overall Response Rate (ORR), measured over 24 weeks.
| Element | Registry entry |
|---|---|
| Result type | N/A |
| Measure of spread | not applicable |
| Unit | not stated |
| Endpoint classification | not reduced to a single comparison |
Overall response rate measures the proportion of evaluable participants whose tumors meet the criteria for complete or partial response. In this study, a response had to be assessed using Response Evaluation Criteria in Solid Tumors version 1.1 and confirmed by a further measurement at least 4 weeks after the response-defining determination. The endpoint's timeframe was 24 weeks.
Confirmation is intended to reduce the chance that a single assessment, which may be uncertain or transient, is counted as a response. A standardized response framework also gives researchers a consistent way to classify tumor changes across participants. In general, response rate is useful in early-phase studies because it describes tumor response without requiring follow-up until a time-to-event outcome occurs. It does not, by itself, describe how long a response lasts or establish a comparison with another treatment.
Interpretation depends on having participants who can be assessed according to the endpoint definition. The registry specified that evaluable participants needed baseline staging, at least 8 weeks of treatment and at least one post-baseline staging scan while on treatment. It reports that no participants met those criteria. Consequently, the record does not yield an evaluable confirmed response rate, and the endpoint cannot be interpreted as evidence for or against activity in the intended population.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The central statistical fact is that the registry lists enrollment of 1 but reports no participants evaluable for the primary endpoint. It therefore gives no confirmed overall response rate estimate for the study's intended analysis population. There is no response count or rate to interpret, and the registry lists zero analysis results. The absence of an evaluable endpoint result is not the same as an observed rate of zero: it means the stated analysis could not be conducted on an evaluable population according to the registry's criteria.
The design also limits between-treatment interpretation. This was a nonrandomized study with parallel arms, and the drug received depended on the tumor's fusion group. There was no like-for-like randomized control comparison. Even if outcome estimates were available, a direct ranking of the drugs would not follow from contrasts across these groups: molecular eligibility and treatment assignment are linked, and the populations could differ. The record does not report an adjusted comparative analysis.
The endpoint was defined as a confirmed complete or partial response at 24 weeks under RECIST version 1.1. Requiring confirmation at least 4 weeks after the response-defining assessment is part of the endpoint definition, not an observed finding. The registry does not provide a response estimate, confidence interval or test result. Without an estimate and its uncertainty, readers cannot assess the precision of an effect or distinguish a potentially informative result from one too imprecise to guide interpretation.
The record lists enrollment of 1 and status as terminated, but it does not report the reason for termination. It does not establish why the primary analysis population was unevaluable, whether all planned groups began treatment, or what happened to participants beyond the posted information. Those explanations should not be inferred from the design or status alone.
Safety information is also limited. The listed Arm C figures include one participant at risk for serious events, one affected by a serious event, one at risk for other events, one affected by an other event, and one death. Sepsis is listed among the serious-event terms at 100.0%. These are small-count registry descriptions, not comparative safety estimates. The record does not establish that a reported event or death was caused by regorafenib or the study.
Overall, the registry documents a planned molecularly targeted study and its endpoint definitions, but its posted results do not support an efficacy estimate or comparative treatment conclusion. The most defensible reading is that the primary endpoint was not evaluable in the registry record and that the available safety figures are descriptive for a very limited recorded population.
9. Statistical Methods Explained
What does an overall response rate represent?
Overall response rate is an endpoint based on response categories. For this study, the qualifying categories were complete response and partial response under RECIST version 1.1. In general, complete response means that the assessment meets the framework's criteria for disappearance of target disease, while partial response means that it meets criteria for a sufficient reduction in measured disease. The registry does not give additional operational details for this study beyond naming RECIST version 1.1.
A response rate is a proportion among participants who can be assessed. Its interpretation depends on defining the analysis population consistently and determining how participants without follow-up assessments are handled. The registry's stated evaluability rule required baseline staging, treatment for at least 8 weeks and at least one post-baseline staging scan while on treatment. It reports that no participants qualified. Thus there is no response proportion from which to draw an estimate or calculate precision.
Why require confirmation?
The endpoint required a confirmatory measurement at least 4 weeks after the determination that first defined a response. In general, confirmation can help guard against classifying a response on the basis of a single assessment that may not reflect a sustained or reproducible change. Confirmation is part of how the outcome is defined; it does not itself tell readers whether any participant responded. In this record, no participant was evaluable for the primary analysis.
What would statistical uncertainty ordinarily describe?
In general, a confidence interval gives a range of values compatible with the observed data under a specified statistical model. It communicates precision: a wider interval indicates less precision, while a narrower interval indicates more. A point estimate without an interval can conceal substantial uncertainty, particularly when an analysis includes few participants. Here, the registry provides no primary-endpoint estimate or interval, so neither the magnitude nor the precision of a response rate can be assessed. No p-value for the primary endpoint is reported.
For a proportion, the amount of uncertainty depends in part on the number of evaluable participants and the number meeting the response definition. Since the registry reports no evaluable participants, the primary response proportion is unavailable rather than merely imprecise. Statistical methods cannot recover an endpoint result when the requisite observations are absent from the analysis population.
How does the design affect comparisons?
Random assignment is a general method for making treatment groups comparable, on average, with respect to both measured and unmeasured characteristics. This study was nonrandomized, and treatments were linked to molecular fusion groups. The arms therefore represent different eligibility-defined populations, not randomized alternatives for the same population. A difference between arms, if one were reported, could reflect population composition, treatment, or other factors. The record does not report a comparative statistical model or adjusted treatment effect.
Masking was listed as none. In general, lack of masking can affect how subjective outcomes are assessed or reported, although the registry does not describe whether or how this influenced response assessment in this study. The response framework provides standardized criteria, but it does not replace information about assessment procedures or eliminate all potential sources of bias.
What do the sample-size illustrations mean?
In general, a sample-size calculation asks how many observations might be needed for a planned analysis under chosen assumptions about the effect, significance level, allocation and statistical model. The registry record does not report the trial's own sample-size justification or design calculations. Any planning illustrations based on an observed effect are not evidence that the study achieved a particular power, and they are not the number of participants actually analyzed. They should not be read as observed power or as a substitute for the trial's design rationale.
For this study, the registry reports enrollment of 1 and no evaluable participants for the primary analysis. This distinction matters: planned sample sizes are prospective illustrations, enrollment is a recorded study count, and the evaluable population is defined by endpoint-specific criteria. These quantities answer different questions and should not be conflated.
What can be concluded from the posted analyses?
The registry lists no analysis results. It therefore supplies no estimate, standard error, confidence interval or hypothesis test for the primary endpoint. The interpretation must remain descriptive: the endpoint was defined, the evaluability criteria were stated, and the record says no participants met them. The registry does not report an interim-analysis plan, multiplicity procedure, or detailed statistical analysis plan. General statistical concepts can clarify what those procedures would address, but cannot establish which methods the investigators used.
10. Multiplicity, Missing Data and Analysis Populations
The study named several treatment groups, multiple fusion-defined populations and several outcomes. In general, testing multiple hypotheses can increase the chance of at least one apparently positive result unless the analysis plan addresses multiplicity. The registry record does not report a multiplicity adjustment, a testing hierarchy, or another procedure for handling multiple comparisons. It also does not report an interim-analysis plan. With no posted outcome analyses, there are no reported tests to assess for multiplicity.
Missing data are important because the primary endpoint required baseline staging, treatment for at least 8 weeks and a post-baseline staging scan while on treatment. The registry states that no participants met these evaluability criteria. It does not report why participants were unevaluable, which specific criterion was unmet, or how missing assessments were handled in any analysis. No imputation method or sensitivity analysis is reported.
The registry describes an evaluable population for the primary endpoint, but does not report results for an intention-to-treat population or another analysis population. Such populations are not interchangeable: one is defined by meeting endpoint-specific assessment criteria, while others can include participants with incomplete follow-up. Without a reported analysis population and endpoint data, the response rate cannot be reconstructed or compared across analysis sets.
Detailed rules for analysis of the secondary outcomes are not reported in the registry record. The posted record therefore does not establish how missing outcomes, multiplicity or analysis-set membership were addressed beyond the primary endpoint's stated evaluability criteria.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Arm C (Regorafenib) | 1 | 1 | 100.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Arm C (Regorafenib) | 1 of 1 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Arm C (Regorafenib) | 1 of 1 |
Most frequent serious adverse event terms
| Event term | Organ system | Arm C (Regorafenib) |
|---|---|---|
| Sepsis | Infections and infestations | 1 of 1 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Arm C (Regorafenib) |
|---|---|---|
| Diarrhea | Gastrointestinal disorders | 1 of 1 |
| Nausea | Gastrointestinal disorders | 1 of 1 |
| Chills | General disorders | 1 of 1 |
| Skin infection | Infections and infestations | 1 of 1 |
| Anorexia | Metabolism and nutrition disorders | 1 of 1 |
| Arthralgia | Musculoskeletal and connective tissue disorders | 1 of 1 |
| Dyspnea | Respiratory, thoracic and mediastinal disorders | 1 of 1 |
| Hot flashes | Vascular disorders | 1 of 1 |
| Hypertension | Vascular disorders | 1 of 1 |
Serious events with the highest arm-level rate: sepsis (100.0%).
The safety record lists Arm C (Regorafenib) with one participant at risk for serious adverse events and one affected. It also lists one participant at risk for other adverse events and one affected, and records one death. Sepsis is identified as a serious adverse-event term at 100.0%. These figures describe the registry's recorded safety group; they are not a comparison with another arm or a population-level estimate of risk.
With one participant at risk, event counts provide little basis for estimating how often an event would occur in a broader population. The percentage attached to sepsis should be read as the registry's reported proportion for that listed term, not as a precise estimate of incidence. The record does not provide enough detail to establish whether the death was related to the serious event, to treatment, or to another cause. It does not report a causal assessment, event timing or additional clinical context. The presence of a serious event or death in a registry entry does not by itself establish causation.
12. What the Result Does, and Does Not, Mean
The record establishes that this was a phase II, nonrandomized study of four kinase inhibitors in fusion-defined groups of patients with unresectable stage III/IV BRAF/NRAS wild-type melanoma. It documents the intended primary endpoint, its response criteria and its evaluability requirements. It also reports enrollment of 1, termination status and no evaluable participants for the primary-endpoint analysis.
Those details do not establish whether any of the drugs produced a response. No evaluable response estimate is reported, so the absence of an estimate must not be interpreted as an observed response rate of zero. Nor does the record establish that treatment was ineffective or effective. It supplies no primary-endpoint confidence interval, hypothesis test or comparative effect estimate.
The design also does not support direct ranking of the four drugs. Treatment was tied to tumor fusion status, and participants were not randomized between drugs. Even with outcome estimates, comparisons across these groups would not automatically isolate treatment effects because the groups were selected by different molecular characteristics.
The safety entry is descriptive and limited to the listed Arm C record. It documents event counts and a death, but does not establish treatment causality or comparative safety. Finally, termination status alone does not explain why the study stopped; the reason is not reported in the registry record.
The statistically appropriate conclusion is narrow: the registry describes the planned study and its endpoints, but its posted results do not provide an evaluable primary outcome or support efficacy conclusions. The available safety counts should be interpreted as limited descriptive information rather than as generalizable rates.
13. Why This Trial Matters Statistically
The study is statistically notable for the gap between a clearly specified primary endpoint and the absence of an evaluable analysis population. The response definition and assessment requirements are stated, but the registry reports no participant meeting those requirements. This illustrates why endpoint definitions alone do not produce an estimate: outcome assessment must also be available for participants who qualify for analysis.
It also illustrates the role of design in determining what comparisons can mean. The four therapies were allocated according to fusion group in a nonrandomized design. Molecularly distinct groups are not equivalent to randomized arms, so an apparent difference between them would not necessarily represent a treatment effect. This is particularly important when a trial explores several targeted therapies in separate biomarker-defined populations.
Registry transparency is consequential for interpretation. The record identifies enrollment, status and endpoint criteria, while leaving the reason for termination, the causes of unevaluability and detailed analysis procedures unreported. Those absences prevent readers from distinguishing among possible explanations. They should remain unknown rather than be filled in through speculation.
Finally, the safety entry demonstrates why event counts and percentages must be read with their denominators and context. A serious event term and a death are recorded, but the small listed population and lack of causal detail limit what can be inferred. The trial record is a useful example of how design, analysis population and completeness of reporting jointly determine the strength of statistical conclusions.
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 principal limitation is that no participants met the stated criteria for the primary endpoint's evaluable population. The registry therefore supplies no confirmed overall response rate, and it does not report why the criteria were unmet. Enrollment of 1 does not, by itself, explain the absence of an evaluable endpoint result.
The nonrandomized design assigned treatment according to fusion group, so the arms were not like-for-like control comparisons. The registry record does not report adjusted comparisons or a common control group. The study is listed as terminated, but the reason for termination is not reported in the registry record. Further details about recruitment, treatment exposure, follow-up and reasons for unevaluability are also not reported in the registry record.
The registry lists five outcome measures but no analysis results. It does not report confidence intervals, hypothesis tests, an interim-analysis plan, a multiplicity procedure, a sample-size justification or a detailed statistical analysis plan. Handling of missing assessments and analysis of secondary outcomes are not reported in the registry record.
Safety interpretation is constrained by the limited Arm C figures and lack of reported causal or clinical context. The listed event counts do not support comparison among treatment groups, and the record does not establish the cause of the death. Together, these limitations leave the record informative about the study's intended design and reported status, but insufficient for efficacy conclusions or broad safety estimates.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Capmatinib, Ceritinib, Regorafenib, or Entrectinib in Treating Patients With BRAF/NRAS Wild-Type Stage III-IV Melanoma — NCT02587650.
- Registry record last updated 2020-01-27; status last verified 2020-01.
- Results first posted 2020-01-27.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT02587650 was a phase II, nonrandomized, unmasked, parallel-arm study sponsored by the University of California, San Francisco. It evaluated capmatinib, ceritinib, regorafenib and entrectinib in fusion-defined groups of patients with unresectable stage III/IV BRAF/NRAS wild-type melanoma. The registry lists enrollment of 1, status as terminated, a start date of 2015-03-26 and a primary completion date of 2018-07-12.
The primary endpoint was confirmed overall response rate at 24 weeks. The registry states that no participants met the endpoint's evaluability criteria and lists no analysis results. It also records limited Arm C safety figures, including one death and sepsis as a serious-event term. The reason for termination and the cause of the reported death are not reported in the registry record.