A Study of BBI608 Administered With Paclitaxel in Adult Patients With Advanced Malignancies: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of BBI608 Administered With Paclitaxel in Adult Patients With Advanced Malignancies (NCT01325441), a non-randomized phase 1, phase 2 study. The primary outcome is Number of Participants With Adverse Events and Serious Adverse Events.

2. Clinical Question

3. Trial Design

The registry describes a phase Ib/II, open-label study of BBI608 administered with paclitaxel in adults with advanced malignancies. The allocation is non-randomized and the study model is sequential. Its brief description identifies a phase 1 dose-escalation study followed by a phase 2 study. Enrollment was 565, and the record lists the study as completed.

There is no masking and no randomized comparator. The listed treatment groups correspond to BBI608 doses of 200mg BID, 240mg BID, 400mg BID, 480mg BID, and 500mg BID, each given with paclitaxel. The arm descriptions state that BBI608 was taken orally twice daily and continuously. Paclitaxel was administered by a 1 hour infusion on days 3, 10, and 17 of each 28 day cycle. Cycles were to continue until progression of disease.

The record gives a start date of 2011-04 and a primary completion date of 2021-06-01. It does not report how participants were assigned among the dose groups, whether dose groups correspond exactly to the phase 1 and phase 2 portions, or the rules governing movement between those portions. The protocol’s rationale for the sequence and the dose-selection procedure are not reported in the registry record.

Because allocation was not randomized, differences between dose groups cannot be interpreted as causal treatment effects. The groups may differ in ways that are not captured in the posted results, and the design does not supply a concurrent randomized control arm. The study can describe safety observations and other listed outcomes in its enrolled participants, but it cannot by itself establish how those observations compare with an alternative treatment.

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: Participants who died, withdrew consent to survival follow up or were lost to follow up were considered to have completed the study.

Overall Study

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 Number of Participants With Adverse Events and Serious Adverse Events, measured over The time from the date of first treatment, while the patient is taking napabucasin, and for 30 days after stopping therapy, an average of 4 months..

The primary endpoint counts participants with adverse events and serious adverse events. It is a safety endpoint: it records whether participants experienced events during the specified observation interval. The registry defines that interval as beginning with the first treatment, continuing while the participant is taking napabucasin, and extending for 30 days after therapy stops. The average duration is stated as 4 months.

A participant count is a direct, descriptive way to report how many people in a treatment group had an event. It does not, by itself, describe event severity, duration, recurrence, or the number of events per participant. Those details require separate reporting. Nor does a count establish that treatment caused an event; attribution and causal interpretation require information beyond simple occurrence.

The endpoint is useful in a study involving multiple dose groups because safety observations are important to describe alongside the treatment regimen. However, the registry does not report a randomized comparison, a prespecified statistical contrast, or a formal hypothesis test for this endpoint. The posted counts should therefore be read as safety summaries for the observed groups, not as evidence of comparative risk. The record also does not specify an analysis population or explain how participants with incomplete follow-up were handled.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a participant count describe?

A count reports the number of participants who met an outcome definition. For safety reporting, a participant can be counted as affected if at least one qualifying event occurred during the defined observation period. A count is easy to interpret as a number of people, but it does not give a complete account of event burden. It does not show how many separate events occurred in each person, how long events lasted, or how severe they were unless those features are separately reported.

Counts also depend on how many participants were observed. When group sizes differ, the same count can represent different proportions of those at risk. A proportion can help describe occurrence relative to a denominator, but its precision depends in part on the amount of information available. Very small groups can produce especially uncertain summaries. In this study the groups range from 2 at risk to 317 at risk, so a count in a small group should not be read as equally informative as the same count from a much larger group.

Why does the design affect interpretation?

In general, random assignment helps balance both measured and unmeasured participant characteristics across treatment groups, making a treatment comparison more interpretable. This study is non-randomized and has no randomized control group. The dose groups are therefore not equivalent to randomized arms. Differences in their event counts can reflect group size, participant characteristics, treatment exposure, follow-up, or other factors, not necessarily a dose effect.

In general, masking can reduce the influence of expectations on some assessments and decisions. This study is described as open-label, with no masking. That design feature does not make the event counts unusable, but it is relevant to interpretation, especially for outcomes whose identification or reporting may involve judgment. The registry does not report additional procedures used to reduce such influences.

What would uncertainty measures add?

In general, a confidence interval gives a range of values compatible with the observed data under a specified statistical method. It helps distinguish a precise estimate from one with substantial uncertainty. A p-value, when used for a defined hypothesis test, describes how incompatible the observed data are with a specified null model; it is not the probability that a treatment works or that a result is due to chance. These measures are not reported for the posted primary outcome results.

No statistical outcome analyses are listed in the registry record. It therefore does not report a model, adjusted estimate, confidence interval, or hypothesis test for these safety counts. Without those analyses, the counts support descriptive statements about the participants observed, not claims of statistical significance or comparative effect. The registry record also does not report the planned statistical analysis for the primary endpoint.

How should planning illustrations be understood?

In general, sample-size planning asks how many participants may be needed to detect a defined effect under specified assumptions. A planning illustration is not a result from the enrolled participants and is not observed power. The planning assumptions described for this record use a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of an observed effect. Those assumptions do not make the actual trial randomized or turn its dose groups into a valid control comparison.

The illustration is described as a calculation based on the observed effect, not as the trial’s own design. Its assumptions should not be mistaken for the protocol’s original sample-size justification. That justification is not reported in the registry record. A planning calculation also depends on the effect selected for detection and the assumptions used; changing those assumptions changes the planning result. The calculation does not show that the trial had a particular probability of finding an effect.

Why is follow-up and population definition important?

For a safety count to be interpreted consistently, readers need to know who was included, how the event window was applied, and what happened when follow-up was incomplete. The record describes the treatment-related observation window, but the analysis population, missing-data handling, and detailed participant disposition for the endpoint are not reported in the registry record. These omissions constrain assessment of whether participants had comparable opportunities for events to be detected and counted.

In general, unequal observation time can affect event counts because participants followed longer have more opportunity to experience an event. Time-to-event analyses can account for timing and censoring when their assumptions and data are appropriate, but no such primary safety analysis is listed. The registry’s average duration of 4 months does not provide individual follow-up times or explain the distribution of observation periods.

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("Napabucasin 200mg BID Plus Paclitaxel", "Napabucasin 240mg BID Plus Paclitaxel", "Napabucasin 400mg BID Plus Paclitaxel", "Napabucasin 480mg BID Plus Paclitaxel", "Napabucasin 500mg BID Plus Paclitaxel")
n      <- c(2, 185, 4, 317, 57)
events <- c(2, 183, 4, 317, 57)
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  Napabucasin 200mg BID Plus Paclitaxel  2  2  100.0  15.8  100.0  34.2  100.0  29.2  99.2
ARM  Napabucasin 240mg BID Plus Paclitaxel  185  183  98.9  96.1  99.9  96.1  99.7  96.2  99.7
ARM  Napabucasin 400mg BID Plus Paclitaxel  4  4  100.0  39.8  100.0  51.0  100.0  47.8  99.5
ARM  Napabucasin 480mg BID Plus Paclitaxel  317  317  100.0  98.8  100.0  98.8  100.0  98.8  100.0
ARM  Napabucasin 500mg BID Plus Paclitaxel  57  57  100.0  93.7  100.0  93.7  100.0  93.8  100.0

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 9 outcome measures, including pharmacokinetic measures, preliminary anti-tumor activity, objective response rate, disease control rate, progression-free survival, overall survival, and pharmacodynamics. It does not report how these outcomes were prioritized beyond identifying the primary safety endpoint, or how the possibility of testing multiple outcomes was handled. A multiplicity procedure, such as an adjustment to preserve a specified overall error rate, is not reported in the registry record.

No statistical outcome analyses are listed. Consequently, there are no posted analysis results from which to assess whether outcomes were tested, whether analyses were prespecified, or whether any results were selected for reporting. The registry record does not report interim analyses, a statistical analysis plan, or a protocol-specific sample-size justification.

The record does not state which analysis population was used for the adverse-event counts. It also does not report how missing observations, early discontinuation, or incomplete follow-up were handled. These details matter because a participant’s opportunity to have an event recorded depends on exposure and observation time. The endpoint window is described, but the handling of participants who do not complete that window is not.

These omissions do not invalidate the counts as registry-posted descriptions, but they limit interpretation of completeness and inferential strength. The results cannot establish that all relevant outcomes were observed or analyzed according to a prespecified plan. They should be read as the posted summaries, with the analysis population and missing-data procedures not reported in the registry record.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: diarrhoea (50.0%), fatigue (50.0%), dehydration (3.5%), vomiting (1.8%), colitis (1.8%).

For each dose group, the serious adverse-event figures give the number at risk and the number of participants affected. The groups list 2 at risk and 1 affected at 200mg BID; 185 and 15 at 240mg BID; 4 and 2 at 400mg BID; 317 and 28 at 480mg BID; and 57 and 7 at 500mg BID. The figures are participant counts, not counts of all separate serious events.

The small groups need particular care: a count based on 2 or 4 participants is less informative than one based on a larger group, and no confidence intervals are reported. Because allocation was not randomized, differences among groups do not provide a controlled estimate of the effect of dose. The record does not report a formal statistical comparison of these counts.

Deaths are also listed as counts for each group, but the registry does not state in these figures whether deaths were treatment-related. Serious adverse-event terms are reported with percentages, but the denominators for those term percentages are not reported in the registry record. These figures describe what the record reports; they do not establish causation or comparative safety.

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

The registry results describe safety observations among participants treated with BBI608 and paclitaxel across five listed BBI608 dose groups. They report how many participants were at risk, how many were affected by serious and other adverse events, and how many deaths were recorded in each group. The primary endpoint is a participant count over a treatment-related observation window, not a randomized estimate of comparative safety.

The study’s non-randomized, open-label design is central to what can be concluded. There is no randomized control group against which to compare the combination. The dose groups cannot be assumed to be alike apart from dose, and their counts do not isolate a causal dose effect. The posted summaries also do not include confidence intervals, adjusted estimates, or statistical tests.

The event counts do not establish why an event occurred. In particular, the death counts alone do not identify cause or treatment relationship. The registry’s lists of serious adverse-event terms add descriptive detail, but the denominators and detailed counting rules for the reported percentages are not stated.

These results therefore support a bounded interpretation: they describe the adverse-event information posted for participants in this study. They do not establish that the combination is safer or less safe than another treatment, nor do they answer comparative efficacy questions. Other outcomes are listed in the registry, but outcome analyses are not listed. The clinical meaning of those outcomes cannot be inferred from their inclusion in the study record alone.

In general, conclusions from a clinical study should reflect its design, the precision of its estimates, and how completely outcomes were recorded. In this case, the absence of randomized comparison and inferential analyses means that descriptive counts should not be converted into claims of treatment effect. The registry record also does not report missing-data procedures or the analysis population, which further limits evaluation of completeness.

14. Why This Trial Matters Statistically

This study illustrates why design and analysis details matter when reading safety data. The registry reports five dose groups, but the allocation is non-randomized. A reader cannot treat these groups as interchangeable or interpret differences in their observed counts as the effect of changing dose. Randomization, when used, supports a more credible comparison by reducing systematic differences between groups; it is not present in this study.

Group size is another important feature. The number at risk ranges from 2 to 317, so counts from the smaller groups carry limited information and are sensitive to individual participants’ outcomes. The largest counts are not directly comparable with smaller-group counts without considering denominators and follow-up. Even then, a descriptive comparison would not remove the design limitations.

The registry identifies a safety endpoint and a treatment-related observation window, which give readers a framework for understanding what was counted. However, the analysis population, missing-data approach, and statistical analysis plan are not reported. No outcome analyses are listed. That makes the distinction between reporting an outcome and statistically evaluating it especially important.

The record also lists several secondary outcomes. Their presence does not mean that all were analyzed, or that any observed effect was established. Multiplicity handling and interim analyses are not reported. The study is therefore a useful example of how clinical trial registry data can provide important descriptive information while leaving unanswered questions about inference, precision, and the strength of comparisons.

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 principal limitation for interpretation is the non-randomized design. There is no randomized control group, so dose groups cannot serve as like-for-like comparators. Differences in participant characteristics, treatment exposure, or follow-up could contribute to differences in event counts. The record does not report an adjustment strategy to address such differences.

The open-label design is another consideration. No masking is reported, and the registry does not describe additional procedures to limit the influence of awareness of treatment assignment on event assessment or reporting. The reported participant counts remain descriptive, but their interpretation depends on how events were identified and classified.

Group sizes vary substantially, including groups with 2 and 4 participants at risk. Those small groups offer very limited information about the range or frequency of uncommon events. The registry does not report confidence intervals or other uncertainty measures for the safety results. It also does not report the denominators and event-counting rules for the percentages attached to serious-event terms.

The analysis population, missing-data procedures, participant disposition for the endpoint, and individual follow-up durations are not reported in the registry record. The primary observation period is described as an average of 4 months, but an average does not explain variation in follow-up. The registry lists 9 outcome measures and 0 outcome analyses, leaving no posted inferential results for those outcomes. Protocol details about interim analyses, multiplicity, and the original sample-size justification are also not reported.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of BBI608 Administered With Paclitaxel in Adult Patients With Advanced Malignancies — NCT01325441.
• Registry record last updated 2023-11-15; status last verified 2023-11.
• Results first posted 2023-02-03.
• 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

NCT01325441 is a completed, open-label, non-randomized phase Ib/II study of BBI608 with paclitaxel in adults with advanced malignancies. The registry lists 565 participants and five dose groups: BBI608 at 200mg BID, 240mg BID, 400mg BID, 480mg BID, or 500mg BID, with paclitaxel. Treatment was described in 28 day cycles, with paclitaxel infusions on days 3, 10, and 17.

The primary endpoint counts participants with adverse events and serious adverse events during treatment and for 30 days after stopping therapy. The posted group counts include participants at risk, those affected, and deaths. The record lists 9 outcome measures and 0 outcome analyses. It does not report a randomized comparison, confidence intervals, statistical tests, the analysis population, or missing-data procedures. The safety results are descriptive and do not establish comparative treatment effects or event causation.