Pilot Study of Liposomal Doxorubicin Combined With Bevacizumab Followed by Bevacizumab Monotherapy in Adults With Advanced Kaposi s Sarcoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pilot Study of Liposomal Doxorubicin Combined With Bevacizumab Followed by Bevacizumab Monotherapy in Adults With Advanced Kaposi s Sarcoma (NCT00923936), a non-randomized phase 2 study. The primary outcome is Overall Response Rate (ORR) of Six Cycles of Liposomal Doxorubicin Combined With Bevacizumab in Patients With Advanced KS..

2. Clinical Question

The study is identified as NCT00923936 and is listed as completed. Its stated primary objective was to estimate the overall response rate after six cycles, described as an average of 18 weeks. Other objectives included evaluating safety, estimating complete response after the six cycles, describing the median number of cycles needed to obtain a partial response, and assessing 12-month progression-free survival.

This is a question about the response observed in the enrolled study population, not a randomized estimate of the difference between treatments. Because allocation was non-randomized, the two groups should not be treated as interchangeable treatment arms or as a direct control comparison. The registry record does not report the primary response count, percentage, or confidence interval, so it does not establish what response was observed.

3. Trial Design

The registry identifies this as a phase 2, non-randomized, parallel study with no masking. Enrollment is listed as 16. The trial began on 2009-04-23, reached primary completion on 2015-05-13, and is marked completed. The sponsor is the National Cancer Institute (NCI).

The study concerns adults with advanced KS. Its two registry-defined groups distinguish KS that is classic or HIV-positive but not improved on antivirals from other advanced HIV-associated KS. The first group description includes people who are HIV-negative, people with stable disease despite antiretroviral therapy, and people with progressive disease despite antiretroviral therapy. The other group is described as all other patients with advanced AIDS-associated KS. These are clinical categories, not randomized treatment assignments.

The intervention names listed are liposomal doxorubicin and bevacizumab. The title and study objectives describe a combination regimen, with bevacizumab monotherapy following the combination. The primary endpoint concerns response after six cycles of the combination, with the timeframe characterized as an average of 18 weeks.

Non-random allocation means that group membership was not assigned by chance, while no masking means treatment and assessment were not described as blinded. The registry record does not report further details of allocation procedures, masking of any outcome assessment, or the rationale for the group structure. It also does not report a randomized comparator. Consequently, comparisons between the groups, if considered, cannot by themselves separate treatment effects from differences in the clinical characteristics used to define the groups.

Study status and design describe how the investigation was organized; they do not establish the size or direction of any treatment effect. The primary response result is not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Overall Response Rate (ORR) of Six Cycles of Liposomal Doxorubicin Combined With Bevacizumab in Patients With Advanced KS., measured over 6 cycles, an average of 18 weeks.

The primary endpoint is the overall response rate after six cycles of liposomal doxorubicin combined with bevacizumab. The registry describes this timeframe as an average of 18 weeks and states that response was assessed using a modification of the Acquired Immune Deficiency Syndrome Clinical Trial Group Oncology Committee criteria.

Overall response combines complete response, clinical complete response, and partial response. In the registry definitions, complete response and clinical complete response require the absence of detectable residual disease, including tumor-associated edema, persisting for at least 4 weeks. Partial response is defined with reference to the absence of progressive disease. Thus the endpoint summarizes the proportion of participants classified as having one of the specified response categories at the assessment after the planned treatment cycles.

As a general statistical concept, a response-rate endpoint is useful when the clinical question concerns whether participants meet a defined response criterion during a specified assessment period. It condenses several response categories into a single proportion, making the endpoint interpretable as a frequency in the study population. It does not describe how long a response lasts, and it is not equivalent to a comparison with an untreated or randomized control group.

The registry lists the endpoint as a percentage of participants and specifies an 80% confidence interval as its dispersion measure. However, the observed response estimate and its interval are not reported in the registry record. The record also does not report additional operational details that would clarify how borderline or unevaluable responses were handled.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does an overall response rate represent?

The primary endpoint is a proportion: participants classified as having complete response, clinical complete response, or partial response after six cycles of the combination are counted as having an overall response. The denominator would ordinarily be the participants included in the endpoint analysis, according to the analysis definition. That denominator and the response count are not reported in the registry record, so the rate cannot be reconstructed from the enrollment figure. Enrollment is not automatically the same as the number with an evaluable endpoint.

As a general statistical concept, proportions are intuitive because they describe the frequency of a defined event in a group. Their precision depends on both the estimated frequency and the amount of information available. A response percentage without its denominator can be difficult to assess, and even a percentage with a denominator can be uncertain when based on a small group. The registry specifies an 80% confidence interval for the primary endpoint, but the interval itself is not reported in the registry record.

How should the confidence interval be read?

In general, a confidence interval gives a range of values compatible with the observed data under a specified statistical procedure. It is not a guarantee that the true value falls within the particular interval, and it does not express the probability that a treatment works. The interval is most useful alongside the estimate and its analysis population. Because neither the endpoint estimate nor its interval is reported, the registry record does not allow readers to judge the primary result's precision.

The registry's stated 80% confidence level is less stringent than some commonly used confidence levels, but that general observation does not identify why this level was chosen. The rationale for the interval level and the method used to calculate it are not reported in the registry record. It would be inappropriate to infer a protocol rationale or substitute an unreported method.

What does non-random allocation mean for interpretation?

In general, random allocation helps balance measured and unmeasured participant characteristics across treatment groups on average. Non-random allocation does not have that protection. Here, the registry describes groups according to KS and HIV-related characteristics. Those definitions can be associated with differences in participant status or disease features, so a comparison between groups could reflect those differences as well as treatment-related factors. The record does not describe a randomized control comparison, and its own design should not be read as one.

Parallel describes groups followed within the same study period rather than a crossover structure. No masking is listed. In general, masking can reduce the influence of knowledge of treatment assignment on assessment or reporting, depending on who is masked. The registry record does not report additional assessor-masking procedures. This matters particularly when endpoint classification involves clinical judgment, although the record does not provide information to quantify any resulting influence.

Why can over-dispersion matter?

As a general statistical concept, over-dispersion occurs when the observed variability in data is greater than a model assumes. For count outcomes, it can make standard errors too small and confidence intervals too narrow if the model does not account for the extra variability. This study's primary endpoint is a participant-level response proportion, and the registry record does not describe a count model or report evidence of over-dispersion. The concept is relevant to statistical methods generally, but no claim about over-dispersion in this trial is supported by the registry record.

How should planning illustrations be distinguished from results?

General sample-size calculations ask how many participants might be needed under stated assumptions to detect a specified effect with a chosen statistical procedure. The registry's planning assumptions describe a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The illustrated effect is expressed as a stated fraction of the observed effect. These assumptions describe planning illustrations, not the trial's own design or observed power.

Any sample-size illustration based on an observed effect is conditional on that effect and the assumptions used for the calculation. It does not establish that the trial had a particular probability of detecting an effect, and it does not replace an analysis of the observed endpoint. The registry record does not report its own sample-size justification, the effect used in any particular illustration, or an inferential result for the primary endpoint. No planning value should be interpreted as an observed trial outcome.

Finally, the primary endpoint's response categories are based on a modified clinical assessment framework. In general, a composite response endpoint can increase the number of participants counted as meeting the endpoint by including more than one response category, but the combined estimate does not show the separate frequency or durability of each category. The registry lists complete response and clinical complete response separately, and the primary definition combines both with partial response. Their individual primary-endpoint counts are not 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("KS; Classic or HIV+ Not Improved on Antivirals")
n      <- c(10)
events <- c(5)
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  KS; Classic or HIV+ Not Improved on Antivirals  10  5  50.0  18.7  81.3  23.7  76.3  23.4  76.6

11. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple objectives beyond the primary overall response rate. Secondary outcomes include complete response after six cycles, serious and non-serious adverse events, the median number of cycles needed to obtain a partial response, and 12-month progression-free survival. When a study examines several outcomes, there is a general statistical issue of multiplicity: repeated testing can increase the chance of at least one apparently notable result if each test is treated independently. Whether and how multiplicity was addressed depends on a prespecified analysis plan.

The statistical plan for multiplicity, including any hierarchy among outcomes or adjustment for multiple testing, is not reported in the registry record. The record also does not report whether interim analyses were planned or conducted. Neither should be inferred from the outcome list or from the completed status.

Missing data can affect estimates when some participants do not have an assessable outcome. General approaches depend on why data are missing, how much is missing, and the endpoint definition. The registry record does not report the number of participants with missing or unevaluable primary response assessments, the approach to missing endpoint data, or whether participants who did not complete six cycles were included in a particular analysis population.

The record does not report an intention-to-treat population, a per-protocol population, or another defined analysis set for the primary outcome. It also does not provide a participant flow accounting for enrollment, treatment completion, assessment, and analysis. Without those details, the appropriate denominator for the primary response rate cannot be verified. The stated enrollment of 16 should not be assumed to equal the primary endpoint denominator.

These omissions limit assessment of how robust the endpoint result would be to missing assessments or alternative analysis choices. The primary response result itself is not reported in the registry record, so sensitivity analyses or checks against the main analysis cannot be reviewed.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: hemorrhage, gi::rectum (16.7%).

The registry reports serious and non-serious adverse-event counts by the two KS/HIV-related groups. In the group titled “KS; Classic or HIV+ Not Improved on Antivirals,” zero of the 10 participants at risk were recorded with serious adverse events, while 10 of the 10 at risk were recorded with non-serious adverse events. In the group titled “All Other Advanced HIV-associated Kaposi's Sarcoma (KS),” 1 of the 6 participants at risk was recorded with serious adverse events, and 6 of the 6 at risk were recorded with non-serious adverse events. Deaths are listed as zero for each group.

These counts describe whether participants were recorded as affected under the registry's event categories. They do not state how many separate events occurred per participant, how long each event lasted, or whether an event was judged related to treatment. Those details are not reported in the registry record.

The listed serious term “Hemorrhage, GI::Rectum” is accompanied by 16.7%. The summary does not identify the group associated with this term or provide further details about its timing, severity, or attribution. In general, a safety event observed during a study is not necessarily caused by the study treatment. The counts also do not support a safety comparison with an untreated or randomized control group.

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

This record describes a phase 2, non-randomized study of a liposomal doxorubicin and bevacizumab regimen in adults with advanced KS. It establishes the study's stated primary endpoint: overall response after six cycles, combining complete response, clinical complete response, and partial response. It also reports serious and non-serious adverse-event counts for two clinically defined groups.

The primary response outcome is not reported in the registry record. Therefore, the record does not establish the observed response rate, how precise that estimate was, or how the individual response categories contributed to it. It also does not provide a randomized comparison that could estimate the difference between the regimen and a control treatment. The group categories do not serve as interchangeable treatment arms.

The safety figures show that serious and non-serious adverse events were recorded in the groups as described, with zero deaths listed in each group. These figures do not show that treatment caused or prevented events. The adverse-event summary does not report important context such as event timing, duration, treatment attribution, or a randomized comparator.

As a general statistical principle, a study result is most interpretable when the estimate, uncertainty interval, analysis population, and handling of missing data are reported together. The record gives the planned primary endpoint and its confidence-interval level, but not the estimate, interval, or analysis population. It also does not report the statistical analysis plan or how incomplete endpoint assessments were handled.

The record can therefore support a description of the research question, design, and available safety counts, but not a quantitative conclusion about the primary response outcome or a causal treatment effect. A planning illustration, if considered, is not an observed power result and cannot fill the gap left by the unreported primary endpoint.

14. Why This Trial Matters Statistically

This trial illustrates several statistical distinctions that are important when reading clinical research. First, an endpoint definition is not an endpoint result. The registry specifies that overall response combines complete response, clinical complete response, and partial response after six cycles, but does not report the response estimate or its 80% confidence interval. Knowing what was measured does not reveal what was observed.

Second, design determines the type of inference available. This study is non-randomized, and its two groups are defined using KS and HIV-related characteristics. They are not randomized treatment and control groups. A difference between groups, if reported, could not be attributed to treatment alone without accounting for the clinical differences that define the groups.

Third, the reported safety counts need a clear denominator and context. The record gives the number at risk and the number affected for serious and non-serious adverse events, as well as deaths. These figures are useful descriptions of recorded events, but they do not establish causation or offer a controlled comparative estimate.

Finally, planning calculations and observed evidence answer different questions. A sample-size illustration uses assumptions to describe a hypothetical design requirement; it is not observed power and does not supply missing outcome data. The record's planning assumptions should not be mistaken for the trial's own design justification or for a primary efficacy finding.

These distinctions matter because readers may otherwise mistake an endpoint description for evidence of benefit, a safety count for a causal risk estimate, or a non-randomized group structure for a controlled comparison. Here, the design and safety information can be described, while the primary response result remains unavailable in the registry record.

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.

Several limitations specific to this record restrict statistical interpretation. Most importantly, the primary overall response rate and its 80% confidence interval are not reported in the registry record. The response count, endpoint denominator, and distribution across complete response, clinical complete response, and partial response are also not reported.

The non-randomized design does not provide a randomized control comparison. The groups are defined by KS and HIV-related characteristics, so they may differ in ways relevant to outcomes. The record does not report adjustment methods that might address measured differences, and adjustment could not remove all concerns about unmeasured differences.

The registry lists enrollment as 16 but does not report participant flow through treatment, assessment, and analysis. It is not possible to determine from the record how many participants completed six cycles, had an evaluable response, or were included in each endpoint analysis. The handling of missing or unevaluable outcomes is not reported.

The registry's response criteria are described as a modification of an assessment framework, but further details on assessor procedures and consistency are not reported. No masking is listed. The statistical analysis plan, sample-size justification, interim-analysis procedures, and multiplicity strategy are not reported in the registry record.

The safety summary gives affected-participant counts but limited detail on event timing, duration, attribution, and group assignment for the listed serious-event term. These gaps constrain interpretation of both efficacy and safety. They do not imply that an unreported result was favorable or unfavorable; they mean the registry record does not provide the detail needed to evaluate it.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Pilot Study of Liposomal Doxorubicin Combined With Bevacizumab Followed by Bevacizumab Monotherapy in Adults With Advanced Kaposi s Sarcoma — NCT00923936.
• Registry record last updated 2018-07-24; status last verified 2018-06.
• Results first posted 2017-06-16.
• 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

NCT00923936 is a completed phase 2, non-randomized, parallel study with no masking, sponsored by the National Cancer Institute (NCI). Enrollment is listed as 16. The study examined liposomal doxorubicin combined with bevacizumab in adults with advanced KS, with the primary endpoint defined as overall response after six cycles, an average of 18 weeks.

The primary response estimate and its 80% confidence interval are not reported in the registry record. Safety counts are reported for two KS/HIV-related groups, with zero deaths listed in each. The record does not describe a randomized control comparison, the analysis population, or the handling of missing primary endpoint data.