STEVIE: A Study of Vismodegib in Patients With Locally Advanced or Metastatic Basal Cell Carcinoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for STEVIE: A Study of Vismodegib in Patients With Locally Advanced or Metastatic Basal Cell Carcinoma (NCT01367665), a non-randomized phase 2 study. The primary outcome is Percentage of Participants Who Experienced Any Adverse Events (AEs), AEs Grade 3 or 4, AEs Leading to Drug Interruptions or Discontinuations and Any Serious Adverse Events (SAEs).

2. Clinical Question

STEVIE examined safety and efficacy of vismodegib in people with locally advanced or metastatic basal cell carcinoma. The registry describes an open-label, single-group study: it does not report a randomized control group or a comparator treatment. The primary outcome concerned adverse events, while response, disease control, survival, symptoms, and quality of life were listed as secondary outcomes.

The study also listed Best Overall Response Rate, duration and time to response, progression-free survival, overall survival, questionnaire scores, and symptom reductions as secondary outcomes. Results for those outcomes are not reported in the registry record. Thus, the central question that can be read from the posted results is descriptive safety: which adverse-event categories were recorded and how many participants in each disease subgroup were affected. Without a control group, the design cannot by itself separate events associated with treatment from events that might have occurred for other reasons.

3. Trial Design

STEVIE (NCT01367665) is recorded as a completed, multicentre, open-label phase II study sponsored by Hoffmann-La Roche. The registry records a start date of 2011-07-01 and primary completion on 2017-06-14. Enrollment is listed as 1232. The study is classified as non-randomized and single-group, with no masking.

Participants received vismodegib 150 mg orally once daily. The two safety groups are named Vismodegib - Locally Advanced and Vismodegib - Metastatic. They distinguish disease setting, rather than assigning participants to different interventions. Treatment was to continue until disease progression or another stopping circumstance: intolerable toxicity most probably attributable to vismodegib, consent withdrawal, death, study termination by the sponsor, or another reason deemed appropriate by the investigator. The registry does not report how long each participant received treatment.

The primary outcome was framed as percentages of participants experiencing adverse events, including events of grade 3 or 4 and events leading to drug interruption or discontinuation, as well as serious adverse events. The outcome timeframe is baseline to the data cut-off of 14 June 2017, up to 6 years. Events were graded using National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03. The registry lists secondary outcomes spanning tumor response, time-to-event outcomes, symptoms, and questionnaire domains.

This design is suitable for describing outcomes observed among participants treated in the study. Because allocation was non-randomized and no control group is listed, it does not support a direct treatment-versus-control estimate. The record does not report the protocol’s sample-size justification, a prespecified statistical analysis plan, or interim analyses. Those design details should not be inferred from the enrollment or completion information.

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: A total of 1232 participants were enrolled in the study, but only 1215 received at least one dose of any study treatment. Results include only the treated 1215 participants. Not Done category includes participants who did not complete the final end of study completion CRF page, including those who withdrew due to termination of the study.

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 Percentage of Participants Who Experienced Any Adverse Events (AEs), AEs Grade 3 or 4, AEs Leading to Drug Interruptions or Discontinuations and Any Serious Adverse Events (SAEs), measured over Baseline to the data cut-off of 14 June 2017 (up to 6 years).

The primary outcome groups several safety measures: the percentage of participants with any adverse event, an adverse event graded 3 or 4, an adverse event leading to drug interruption or discontinuation, and any serious adverse event. These measures capture different aspects of safety. An event may be recorded regardless of severity; grade 3 or 4 identifies higher severity categories; treatment interruption or discontinuation describes an effect on treatment delivery; and seriousness is a separate classification. Seriousness should not be treated as another name for severity grade.

The registry states that adverse events were graded according to National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03. It defines grade 3 as severe or medically significant, potentially involving hospitalization, disability, or limitations in self-care, and grade 4 as an immediate threat to life requiring urgent medical intervention to maintain survival. The endpoint therefore includes both broad occurrence and clinically consequential categories, rather than relying on a single summary of all events.

Safety outcomes are important in a study that records ongoing drug exposure and stopping due to toxicity among the possible reasons to end treatment. However, a participant-level percentage does not describe the number of episodes, their duration, or the timing of events. The registry’s listed figures provide counts for serious events and other adverse events in the two disease groups. The separate percentages for every component of the primary endpoint are not reported in the registry record, so the full endpoint cannot be reconstructed from the listed counts alone.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several statistical ideas help distinguish what the registry results can show from what they cannot. The study records binary participant-level safety outcomes: whether a participant was recorded as having an event in a category. A proportion or percentage for such an outcome summarizes the share of participants classified as having that event. The posted safety groups include numbers at risk and numbers affected for serious adverse events and other adverse events. The primary outcome is framed in percentage terms, but the record does not supply every component percentage.

What does a participant-level count mean?

A participant count is not an event count. A person can experience more than one event, and a participant-level summary may count that person once within a category even when multiple episodes occurred. The registry figures therefore describe how many participants were recorded as affected in the specified categories, not the total number of episodes, the burden over time, or the distribution of event frequency among participants. Those details are not reported in the registry record.

The denominator also matters. The safety groups state 1119 at risk for each of the locally advanced group’s listed event categories and 96 at risk for each of the metastatic group’s listed categories. These are the denominators attached to those safety summaries. They should not be replaced with overall enrollment when interpreting the subgroup counts. The registry does not provide further detail on how participants entered or left the analysis population for each outcome.

Why distinguish event severity and seriousness?

Severity grading describes how intense or medically consequential an event is under the stated grading system. Seriousness is a separate regulatory and clinical classification. An event can be serious because of features such as its consequences or required care, while grade records its severity according to the specified criteria. The registry describes grade 3 as severe or medically significant, with possible hospitalization, disability, or limitations in self-care, and grade 4 as an immediate threat to life requiring urgent intervention to maintain survival. The endpoint separately names serious adverse events and grade 3 or 4 adverse events, so the categories should not be collapsed into one.

How should uncertainty be assessed?

In general, a confidence interval gives a range of values reasonably compatible with the observed data under a statistical model. It communicates sampling uncertainty around an estimate; it does not resolve limitations in study design or establish causality. Confidence intervals for the safety summaries are not reported in the registry record. Nor are p-values or adjusted comparisons reported. Without these, readers cannot assess the precision or formal statistical significance of subgroup contrasts from the record.

In general, randomized comparisons help balance known and unknown participant characteristics across treatment groups. This study is non-randomized and has no control group listed, so its subgroup summaries cannot supply that protection or estimate a controlled difference between treatments. Comparing locally advanced and metastatic groups would also be a comparison of disease settings, not a randomized treatment contrast. No statistical adjustment for subgroup differences is reported in the registry record.

What do the planning illustrations represent?

The planning illustrations are described as calculations based on an observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. They are illustrations for planning, not observed power and not evidence about the actual study’s design. The record describes the effect to detect as a stated fraction of the observed effect. For time-to-event outcomes, the planning quantity is a total number of events; for other outcomes, it is participants per arm. These assumptions do not transform this single-group study into a two-arm randomized trial. The registry does not report the study’s own sample-size justification, so the illustrations should not be mistaken for that justification.

General statistical practice also distinguishes a prespecified analysis from an analysis chosen after seeing results. The registry record does not report the statistical plan, the handling of missing data, interim analyses, or an analysis population definition. It is therefore not possible to verify from this record which models, adjustments, or decision rules were planned or used. No recalculated estimate is presented as a substitute for sponsor analysis; the available summaries are described as recorded.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 13 outcome measures, including the primary safety outcome and secondary outcomes concerning response, duration and time to response, progression-free survival, overall survival, Skindex-16 questionnaire domains, and symptom reductions according to the MDASI scale. When a study evaluates many outcomes, the chance of obtaining at least one apparently notable result by chance can increase if each is tested separately. In general, multiplicity procedures can control that risk, but they involve choices about which outcomes form a testing family and how error rates are controlled.

A multiplicity strategy, endpoint hierarchy, or adjustment method is not reported in the registry record. The record also lists 0 analyses. This means the registry entry does not provide inferential results for these outcomes, such as hypothesis tests or adjusted estimates. It is not possible to determine from the record whether analyses were conducted elsewhere or whether a formal multiplicity procedure was used.

Missing-data methods and analysis populations are also not reported in the registry record. The entry does not explain how unavailable outcome measurements, incomplete follow-up, or participants who stopped treatment were handled in analyses. Nor does it define an intention-to-treat, safety, or other analysis population for the outcome summaries. These details matter because the included participants and the handling of missing observations can affect descriptive percentages and time-to-event estimates.

For that reason, the registry’s reported counts should be read as the recorded summaries, without assuming a missing-data method or a particular analysis population. The treatment stopping criteria do not themselves describe how participants were handled statistically after stopping. Any account of imputation, censoring, or sensitivity analysis would be speculation; those approaches are 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: renal failure (2.1%), pneumonia (1.5%), anaemia (1.0%), myocardial infarction (1.0%), sinus node dysfunction (1.0%).

The safety record separates participants with locally advanced disease from those with metastatic disease. Serious adverse events were recorded for 272 of 1119 at risk in the locally advanced group and 31 of 96 at risk in the metastatic group. Other adverse events were recorded for 1065 of 1119 and 90 of 96, respectively. These are participant counts with the corresponding at-risk denominators, not counts of individual event episodes.

The registry lists deaths separately, as 103 in the locally advanced group and 19 in the metastatic group. It does not state in these figures whether deaths were considered adverse events, whether they were judged related to vismodegib, or when they occurred. They should not be interpreted as treatment-caused deaths. Serious adverse-event terms are also listed with reported percentages: RENAL FAILURE, 2.1%; PNEUMONIA, 1.5%; ANAEMIA, 1.0%; MYOCARDIAL INFARCTION, 1.0%; SINUS NODE DYSFUNCTION, 1.0%; and VOLVULUS, 1.0%. Details of the denominators for these term-specific percentages are not reported in the registry record.

These summaries do not provide event timing, duration, recurrence, or a control-group rate. They describe what was recorded during the stated study timeframe, baseline to the data cut-off of 14 June 2017, up to 6 years. Seriousness, severity grade, and treatment attribution should be interpreted as distinct concepts.

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

The record establishes that STEVIE was a large, completed phase II study with a single-group, open-label design and that it recorded safety outcomes for participants with locally advanced or metastatic basal cell carcinoma who received vismodegib. The listed counts show that adverse events were recorded in both disease subgroups, including serious adverse events. The registry also gives selected serious adverse-event terms and their reported percentages. These observations are useful as descriptions of the safety experience recorded in the study.

They do not, on their own, establish how much of the observed event burden was caused by vismodegib. Without a control group, events cannot be contrasted against an untreated or differently treated population under the same study conditions. The death counts likewise do not establish treatment attribution. Differences between the locally advanced and metastatic groups would not represent a randomized comparison, and the record does not report adjusted subgroup analyses.

The primary outcome includes multiple safety categories, but the available counts do not provide every component percentage or uncertainty interval. Therefore, the precision of the reported subgroup summaries and the full pattern of grade 3 or 4 events or treatment interruptions cannot be assessed from the registry record. The selected term-specific percentages also lack denominator detail in this record.

Secondary outcomes were listed for tumor response, time-to-event outcomes, symptoms, and quality of life. Their inclusion indicates topics the study planned to assess, but it does not supply results for those measures. Such results are not reported in the registry record. The study record therefore supports a bounded conclusion: it documents the design and selected safety counts, while leaving the comparative effect, causal attribution, and full range of efficacy and patient-reported results unresolved.

In general, an observed frequency describes events in a particular study population and follow-up context; it is not automatically a prediction for another population. Differences in setting, participant characteristics, follow-up, outcome definitions, and completeness of recording can all matter. This is a statistical interpretation of the study record, not medical advice or a recommendation about treatment.

13. Why This Trial Matters Statistically

STEVIE is statistically notable for the mismatch between the breadth of its clinical questions and the limits of its design. It assessed safety and listed several efficacy, survival, symptom, and quality-of-life outcomes, but it was non-randomized, open-label, and single-group. This makes it an example of why endpoint selection and study design determine what conclusions can be supported. Descriptive safety results can characterize events in a treated cohort, while a comparative treatment effect ordinarily requires an appropriate comparison strategy.

The registry separates participants into locally advanced and metastatic groups. That can help describe outcomes by disease setting, but it does not make the groups equivalent to randomized arms. Subgroup differences may reflect characteristics other than treatment, and the record reports no adjustment analysis to address them. The distinction is especially important when interpreting safety counts, death counts, or any apparent contrast between the groups.

The study also demonstrates why endpoint labels should be read closely. Any adverse event, grade 3 or 4 event, interruption or discontinuation, and serious adverse event are not interchangeable measures. Severity, seriousness, and consequences for treatment delivery provide different information. The record’s summary counts cover only some of the primary endpoint components, so it is not possible to reconstruct the full safety profile from those counts alone.

Finally, a registry entry can identify planned outcomes without reporting their results or the methods used to analyze them. Here, statistical analyses, missing-data handling, analysis populations, and multiplicity procedures are not reported in the registry record. That absence limits reproducibility and assessment of uncertainty. The lesson is methodological: interpret each posted result within its design, endpoint definition, denominator, and reporting detail, rather than treating enrollment size or a completed status as a substitute for those elements.

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.

Interpretation is limited by the single-group, non-randomized design and the absence of a control group. The observed safety counts describe participants receiving vismodegib but cannot isolate treatment-associated risk from background events or other influences. The locally advanced and metastatic groups differ by disease setting and are not randomized treatment comparisons. No adjusted subgroup comparison is reported in the registry record.

The posted safety summaries are incomplete for the primary endpoint as defined. They include participant counts for serious and other adverse events, but the registry does not report separate percentages for all listed components, including grade 3 or 4 events and events leading to drug interruption or discontinuation. The listed serious adverse-event terms have percentages, but their denominator details are not reported in this record. Counts do not convey event frequency per participant, timing, duration, recurrence, or severity distribution.

The record lists deaths but does not supply attribution or timing in these figures. It also does not report confidence intervals, p-values, or other uncertainty measures for the safety summaries. The absence of these statistics prevents assessment of how precisely subgroup frequencies are estimated. Secondary outcomes are named, but their results are not reported in the registry record.

Other important analytic details are not reported in the registry record, including the statistical analysis plan, sample-size justification, interim analyses, missing-data handling, and analysis population definitions. The completed status and enrollment do not fill these gaps. The record’s treatment stopping criteria describe when treatment could end, but do not explain how follow-up observations after discontinuation were handled statistically. These limitations constrain causal interpretation, reproducibility, and assessment of the full outcome set.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: STEVIE: A Study of Vismodegib in Patients With Locally Advanced or Metastatic Basal Cell Carcinoma — NCT01367665.
• Registry record last updated 2019-06-06; status last verified 2019-03.
• Results first posted 2015-08-31.
• 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

STEVIE (NCT01367665) is a completed, multicentre, open-label phase II study of vismodegib in locally advanced or metastatic basal cell carcinoma. The registry records non-randomized, single-group allocation and enrollment of 1232. Participants received vismodegib 150 mg orally once daily, with treatment continuing until progression or a listed stopping circumstance.

The primary outcome concerned percentages of participants with adverse-event categories over baseline to the data cut-off of 14 June 2017, up to 6 years. Safety counts include serious adverse events and other adverse events in the locally advanced and metastatic groups, as well as separate death counts. The registry also lists secondary outcomes for response, survival, symptoms, and questionnaire scores, but their results are not reported in the registry record.

No control group, inferential analyses, confidence intervals, missing-data methods, or analysis population definitions are reported in the registry record. The available results support descriptive reading of recorded safety counts, not a controlled estimate of treatment effect or causal attribution.