A Study of Avastin (Bevacizumab) in Combination With Fotemustine in Patients With Metastatic Melanoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Avastin (Bevacizumab) in Combination With Fotemustine in Patients With Metastatic Melanoma (NCT01069627), a na phase 2 study. The primary outcome is Percentage of Participants With Complete Response (CR) or Partial Response (PR).

2. Clinical Question

3. Trial Design

This was a phase 2, open-label, single-group study of bevacizumab plus fotemustine in metastatic melanoma. The masking field is listed as none, and allocation is listed as NA. No separate control arm is described, so the design does not support a randomized, like-for-like comparison between treatment groups. The registry record gives enrollment as 20 and study status as completed.

The official title describes the combination as first-line therapy in patients with metastatic melanoma. The eligibility description in the brief summary further specifies stage IV melanoma and no previous chemotherapy or immunotherapy for metastatic disease. The registry does not report other eligibility criteria in the registry record for this account.

The treatment description specifies bevacizumab 15mg/kg intravenously on Day 1 of every 3 week cycle. Fotemustine is described at 100mg/m² intravenously on Days 1, 8 and 15, followed by 4 weeks rest, then 100mg/m² every 3 weeks for 4-6 cycles. The anticipated time on treatment was until disease progression. These are descriptions of the regimen, not evidence about the treatment's effect.

The study start is recorded as 2006-12, and primary completion as 2009-07. The registry record does not report a statistical analysis plan, details of how participants entered the study, or reasons for the enrollment size. It also does not give a comparison group against which to interpret response or safety outcomes.

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:

Analysis population for the primary outcome, as defined in the registry: Intent-to-Treat (ITT) Population: all participants who signed the informed consent form and were assigned a study patient number; data for 1 participant were not assessable as the participant was discontinued from the study due to a protocol violation.

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: Safety Population: all participants who signed the informed consent form, were assigned a study patient number, and took at least one dose of each drug of study combination.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With Complete Response (CR) or Partial Response (PR), measured over Baseline, every 9 weeks during study treatment, and every 3 months during follow-up, up to 36 months.

The primary endpoint was the percentage of participants with a complete response or partial response. It describes the proportion meeting either of those objective tumor-response categories under the Response Evaluation Criteria In Solid Tumors (RECIST). A complete response was defined as disappearance of all clinical and radiological evidence of tumor, involving both target and non-target disease. A partial response was defined as at least a 30 percent decrease in the sum of the longest diameters of target lesions, using the baseline sum as the reference.

A response percentage is a compact way to describe how often participants met a prespecified tumor-response definition in the study population. It does not by itself describe how long a response lasted, whether disease later progressed, or how outcomes compare with another treatment. Those questions require other outcomes or a suitable comparison.

The recorded assessment schedule was baseline, every 9 weeks during study treatment, and every 3 months during follow-up, up to 36 months. The endpoint lists a 95% confidence interval as its measure of precision. However, the numerical response estimate and the confidence interval are not reported in the registry record, so the frequency of response and its statistical precision cannot be assessed from the posted results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a single-group design estimate?

In a single-group study, outcomes are described for participants who receive the study regimen. For a response endpoint, the basic target is the proportion of the analyzed group meeting the response definition. This can describe what happened in that group, but it does not directly estimate a treatment contrast against a control group. Differences between outcomes in separate studies or patient groups may reflect differences in eligibility, disease characteristics, assessment, follow-up, or other factors rather than a treatment effect.

This registry identifies the study as single-group and gives allocation as NA. It therefore does not record a randomized allocation process or a control arm. Randomization, when used, helps make treatment groups comparable on average; it is not part of the design. The record also lists no masking. In general, lack of masking can matter when outcomes involve judgment, because knowledge of treatment may influence care or assessment. The registry does not describe how RECIST assessments were conducted or whether assessment was blinded.

What does the response percentage represent?

The primary outcome combines complete response and partial response into one category: an objective response under RECIST. The criteria distinguish complete disappearance of clinical and radiological tumor evidence from a partial reduction in the summed longest diameters of target lesions. Combining categories gives a single summary of response frequency, but it does not retain all detail about the depth of response. It also does not express response duration or time until progression.

To estimate a percentage, an analysis needs a numerator, representing participants who meet the response definition, and a denominator, representing the population included in the calculation. The registry identifies an intent-to-treat population definition and notes that data for 1 participant were not assessable following discontinuation due to a protocol violation. The numerical response count, percentage, and exact denominator used for the primary endpoint are not reported in the registry record. No independent calculation can therefore reproduce the endpoint from the available results.

How should a confidence interval be read?

The endpoint lists a 95% confidence interval. In general, a confidence interval communicates uncertainty around an estimate, under the assumptions and calculation method used. A narrower interval indicates greater precision than a wider interval for the same type of estimate. It does not mean that the true value has a 95% probability of lying in one particular calculated interval under the usual frequentist interpretation. The actual interval and the method used to calculate it are not reported in the registry record. Its width, limits, and compatibility with different response rates therefore cannot be discussed.

For a percentage based on a finite group, uncertainty is influenced by the number of observations and the observed frequency. General methods for proportions can use different approximations or exact procedures, especially when counts are small or the estimated percentage is near an endpoint of its possible range. Which method, if any, was used for this study is not reported. This matters because alternative methods can yield different interval limits.

Why does missing outcome information matter?

The registry says that 1 participant's data were not assessable because the participant was discontinued following a protocol violation. This identifies an issue for endpoint interpretation, but does not explain how that participant was treated in the response analysis. The intent-to-treat population description includes participants who signed consent and were assigned a study patient number; it does not, by itself, specify the handling of unassessable response data.

In general, excluding participants with unavailable outcome data can change an estimated response percentage if the chance of missing data is related to their likely outcome. Counting unassessable participants as nonresponders is another possible approach, but the registry does not state that this was done. Imputation or sensitivity analyses are also not described. The actual missing-data strategy is not reported in the registry record, so the direction or size of any influence cannot be determined.

What can planning calculations say?

Any sample-size illustrations associated with this record are planning exercises, not observed evidence and not observed power. The stated assumptions describe two-sided significance level 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. The record says time-to-event illustrations give total events, while other illustrations give participants per arm. These assumptions describe hypothetical planning, not the design actually used in this single-group study.

A planning illustration based on an observed effect depends on that effect estimate and on the specified assumptions. It does not establish that the original study was designed around those assumptions, and it does not replace an analysis of the observed data. The registry does not report the study's own sample-size justification, and no numerical planning estimates are stated in the registry record details. The enrollment of 20 should therefore be understood as the recorded enrollment, not as evidence of a particular power or a design rationale.

How does multiplicity affect interpretation?

The registry lists 18 outcome measures, including the primary response endpoint and secondary outcomes involving time to progression, response duration, and stable disease. Examining multiple outcomes can be useful for describing different aspects of disease course, but it can also increase the chance of finding an apparently notable result by chance. Statistical methods may address this through prespecification, adjustment, or cautious interpretation, but the registry does not report a multiplicity strategy or the analysis plan. Because it lists no analysis results, there are no posted outcome estimates to evaluate for consistency or selectivity.

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("Bevacizumab + Fotemustine")
n      <- c(20)
events <- c(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  Bevacizumab + Fotemustine  20  3  15.0  3.2  37.9  5.2  36.0  5.4  36.3

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 18 outcome measures. The primary measure is the percentage of participants with complete or partial response; the listed secondary measures include time to progression and duration measures for response and stable disease. Considering multiple outcomes can provide a broader description of the study, but it also raises the general issue of multiplicity: the more statistical comparisons or outcome tests examined, the greater the opportunity for chance findings. The registry record does not report a multiplicity adjustment or analysis hierarchy, and it lists no analysis results.

The registry defines the intent-to-treat population as participants who signed informed consent and were assigned a study patient number. It notes that data for 1 participant were not assessable because the participant was discontinued due to a protocol violation. The record does not state how this participant's response outcome was handled, whether any further participants had missing endpoint data, or whether sensitivity analyses were performed. The numerical primary endpoint result and denominator are also not reported.

The statistical analysis plan, including the handling of missing data and multiplicity, is not reported in the registry record. As a result, readers cannot determine whether the reported population definition was applied directly to the endpoint analysis or whether additional analysis rules were used. These gaps limit assessment of the robustness and reproducibility of any endpoint result.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: thrombocytopenia (5.0%), metrorrhagia (5.0%), dyspnoea (5.0%), pleural effusion (5.0%), pulmonary embolism (5.0%).

For the bevacizumab plus fotemustine group, the registry reports 4 participants affected by serious adverse events among 20 at risk, and 18 affected by other adverse events among 20 at risk. These are participant counts in separate reported categories, not a measure of event severity for every event and not evidence that the drugs caused the events. The counts should not be added: participants may be represented in more than one category.

The serious adverse-event terms listed most prominently are thrombocytopenia, metrorrhagia, dyspnoea, pleural effusion, and pulmonary embolism, each recorded as 5.0%. The registry does not provide further detail in the registry entries about event timing, duration, resolution, or attribution. The death count is not reported in the registry record.

These safety figures describe what was recorded in this study group. With no comparator group, they cannot show whether event frequency differed from another treatment or from an untreated comparison. The record also does not provide enough information to infer the likelihood of an event for other patients.

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

The registry record establishes that a completed phase 2, open-label, single-group study enrolled 20 participants to receive bevacizumab with fotemustine for metastatic melanoma. It defines an objective-response endpoint using RECIST and describes how complete and partial response were categorized. Those details clarify what the study set out to measure.

The record does not report the numerical primary response result or its 95% confidence interval. It is therefore not possible to determine from the posted results how many participants met the response definition, what percentage they represented, or how precisely that percentage was estimated. The endpoint definition alone is not evidence that a response occurred, nor does it show the size or duration of any response.

The design also determines what comparisons can be made. A single-group study has no randomized control group, so it cannot estimate a randomized difference between the combination and another treatment. Its findings, even if fully reported, would describe outcomes in the enrolled group rather than establish what would have happened to comparable participants under a different treatment. Because both drugs were given together, the design would not separate their individual contributions.

The safety counts are descriptive, and the registry does not report a death count in the registry entries. Adverse-event counts alone cannot establish causation or comparative risk. The missing numerical efficacy results, unreported statistical analysis details, and the single-group design all limit how much can be inferred from this record.

Accordingly, the statistical meaning is bounded: the record identifies a response question, a population, a regimen, and adverse-event counts, but does not provide the primary response estimate needed to quantify the observed endpoint. It also does not provide a controlled comparison capable of supporting a treatment-effect estimate.

14. Why This Trial Matters Statistically

This study is statistically informative as an example of the distinction between defining an endpoint and reporting its result. The primary outcome is specified in operational terms through RECIST categories, an assessment schedule, and a confidence-interval measure. Yet the numerical response estimate and interval are not reported in the registry record. Without those results, readers cannot assess the observed frequency or its precision.

The single-group structure is central to interpretation. It can describe response and adverse events in participants receiving a combination, but it does not provide a randomized estimate of how outcomes compare with a control. This is an important design distinction for clinical readers: an outcome observed after treatment is not, by itself, the same as an effect attributable to treatment.

The record also highlights why analysis populations and missing outcome data matter. It defines an intent-to-treat population and identifies 1 participant whose data were not assessable following a protocol violation, but does not report how that case was handled in the endpoint analysis. That leaves an important part of the path from participant records to an estimated response percentage unspecified.

Finally, the registry lists 18 outcome measures but no analysis results and does not report the statistical analysis plan or multiplicity approach. These omissions make it difficult to evaluate how outcomes were prioritized or how uncertainty and missingness were addressed. The study remains a useful illustration of how design, endpoint definition, analysis transparency, and reported results work together to determine what can be concluded statistically.

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 most consequential limitation for interpreting efficacy is that the numerical primary endpoint result is not reported in the registry record. The record also does not report the confidence interval limits, response count, or the denominator used in the analysis. As a result, the magnitude and precision of the primary endpoint cannot be assessed.

The study is described as single-group, open-label, and without a separate comparator. This limits interpretation to outcomes in the treated group and prevents a randomized comparison. The combination design also does not distinguish the contribution of bevacizumab from that of fotemustine. The record does not describe whether response assessments were blinded or independently reviewed.

The registry identifies 1 participant as not assessable due to discontinuation following a protocol violation, but does not report the endpoint handling of that participant or any missing-data analysis. The statistical analysis plan and multiplicity approach are not reported, despite 18 listed outcome measures. The registry also does not report an interim-analysis plan or the study's sample-size justification.

Safety interpretation is similarly constrained. Counts are available for participants affected by serious and other adverse events, but the record does not report a death count in the registry entries, event attribution, or a comparative group. These figures cannot establish comparative risk or causation. The enrollment of 20 is recorded, but should not be treated as evidence about the precision of an unreported endpoint.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of Avastin (Bevacizumab) in Combination With Fotemustine in Patients With Metastatic Melanoma — NCT01069627.
• Registry record last updated 2018-12-06; status last verified 2018-11.
• Results first posted 2014-06-25.
• 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

NCT01069627 is a completed phase 2, open-label, single-group study sponsored by Hoffmann-La Roche. It evaluated bevacizumab with fotemustine as first-line therapy in stage IV melanoma among people described as not previously treated with chemotherapy or immunotherapy for metastatic disease. Enrollment was 20; the study start is recorded as 2006-12 and primary completion as 2009-07.

The primary endpoint was the percentage of participants with complete or partial response under RECIST, with a 95% confidence interval listed as the precision measure. The numerical primary result and interval are not reported in the registry record. The record lists 18 outcome measures and no analysis results. It identifies 1 participant whose data were not assessable due to a protocol violation.

Safety information reports 4 participants with serious adverse events and 18 with other adverse events among 20 at risk in the treatment group. A death count is not reported in the registry record. The single-group design provides no randomized control comparison.