Rollover Study of BMS-354825 in Patients With CML and Ph+ALL: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Rollover Study of BMS-354825 in Patients With CML and Ph+ALL (NCT01030718), a registered phase 1, phase 2 study. The primary outcome is Number of Participants With Adverse Events (AEs), Serious Adverse Events (SAEs), Deaths, and Discontinuation.

2. Clinical Question

The study asked about long-term safety and efficacy of dasatinib in participants with chronic myelogenous leukemia (CML) or Philadelphia chromosome positive acute lymphoblastic leukemia (Ph+ ALL) who were resistant or intolerant to previous treatment and had completed the previous Phase I/II protocol, CA180-031/NCT00337454. The primary outcome focused on safety: counts of participants with adverse events, serious adverse events, deaths, and discontinuation. The record also lists cytogenetic response and time-to-response outcomes as secondary measures.

3. Trial Design

This was a completed Phase I/II rollover study sponsored by Bristol-Myers Squibb. It enrolled 54 participants and studied dasatinib in three disease groups: CML in chronic phase (CML-CP), CML in accelerated or blast phase (CML-AP/BP), and Ph+ ALL. The rollover design continued study drug after participants completed the previous Phase I/II protocol, CA180-031/NCT00337454. It was not a randomized comparison, and the allocation method is not stated. The record describes a parallel model and no masking.

The reported start was 2006-01 and primary completion was 2009-06. The primary outcome population was all treated participants. The safety follow-up schedule included baseline and recurring assessments: every 4 weeks for participants on study < 6 months, including the earlier protocol; every 12 weeks for those on study >=6 months and <=2 years; every 24 weeks for those on study >2 years; and assessment at discontinuation. The registry also states that reported deaths include deaths after the study end.

The secondary outcomes include cytogenetic response percentages for each disease group, time to complete cytogenetic response (CCyR), duration of CCyR, and time to major cytogenetic response for CML-CP. Although the record lists 24 outcome measures, it reports zero analyses. Therefore, the design and outcome list can be described, but results for those secondary measures are 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.

Before assignment: 55 participants were randomized to study CA180-031; 1 participant with chronic myelogenous leukemia-chronic phase (CML-CP) was withdrawn prior to dosing due to thrombocytopenia. 54 subjects were treated; 44 subjects finished the study period of CA180-031 and transferred CA180-036. Analyses were done on the population enrolled into CA180-031 study.

Studies CA180-031 and -036 Combined

Reasons for not completing this period:

Study CA180-036 Only

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All treated participants. Number of deaths represents all reported deaths, including after the study end. For AEs leading to discontinuation: 4 in CML-CP=1 insufficient effect (IE) +3 AEs in Participant Flow (PF); 7 in CML-AP/BP= 1 death + 5 AEs + 1 IE in PF; 4 in Ph+ALL=3 IE + 1AE in PF.

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 (AEs), Serious Adverse Events (SAEs), Deaths, and Discontinuation, measured over baseline; every 4 weeks (if on study < 6 months, including CA180-031(NCT00337454); every 12 weeks (if on study >=6 months and <=2 years); every 24 weeks (if on study >2 years); at discontinuation.

The primary endpoint is a set of counts: participants with adverse events (AEs), serious adverse events (SAEs), deaths, and discontinuation. An AE is defined as a new untoward medical occurrence or worsening of an existing medical condition; the definition does not require a causal relationship with treatment. A related AE is one classified as certain, probable, possible, or missing. An SAE is an event meeting at least one listed criterion, including death, a life-threatening event, inpatient hospitalization or prolonged hospitalization, persistent or significant disability, a congenital anomaly or birth defect, drug dependency or abuse, or another important medical event.

Counts of participants are useful for describing how many people had events during safety monitoring. They do not, by themselves, show the number of separate episodes, how long events lasted, or whether treatment caused them. The registry specifies all treated participants as the primary safety population. It also describes follow-up at baseline, at intervals that depend on time on study, and at discontinuation, with deaths counted even when reported after the study end.

The posted results report 33 of 54 participants with serious adverse events and 54 of 54 with other adverse events. The death count is not reported in the registry record. These counts describe the recorded safety experience in this study population; without a control group or a comparative analysis, they do not estimate a treatment-attributable difference.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation starts with the study question and the way participants entered the study. This was a rollover study in people who had completed an earlier Phase I/II protocol, rather than a randomized comparison. The record describes a parallel model, no masking, and allocation as not stated. A parallel model describes groups followed in the same study framework; it does not itself establish random assignment or make the groups comparable. Since the record does not describe a control arm suitable for a like-for-like comparison, the reported safety counts are descriptive.

What a participant count represents

The primary outcome is expressed as the number of participants with specified events. A participant count describes people who experienced an event, rather than the total number of event episodes. A person may have more than one event, so participant counts and event counts answer different questions. The registry lists 33 participants with serious adverse events among 54 at risk and 54 participants with other adverse events among 54 at risk. These are the posted counts; they do not imply that each participant had only one event.

General: a proportion is often used to place a participant count in context by relating the count to the number assessed. A percentage alone still does not describe event severity, duration, timing, or attribution. Here the registry also lists percentages for selected serious-event terms. Those entries should be read as the registry’s reported term frequencies, not as comparative treatment effects.

Why event definitions matter

The AE definition includes a new medical occurrence or worsening of a pre-existing condition, without requiring a causal relationship to treatment. The related-AE classification includes certain, probable, possible, or missing relationships. SAE status is defined by specified seriousness criteria, including hospitalization, life-threatening events, and death, as well as other listed outcomes. Consequently, “serious” describes the event classification, while “related” describes a separate judgment about its relationship to treatment. These categories should not be treated as interchangeable.

The record’s endpoint includes all treated participants. It also gives a monitoring schedule that varies with time on study, and notes that the death count includes reported deaths after study end. Follow-up duration and assessment timing matter in safety interpretation because participants observed for different periods may have different opportunities for events to be recorded. The registry record does not provide participant-level observation times or an analysis adjusting for differences in follow-up.

What statistical comparisons would require

General: a treatment comparison ordinarily requires an appropriate comparator and an analysis that accounts for how participants were assigned and followed. Randomization can help balance measured and unmeasured prognostic factors between groups, although it does not guarantee identical groups. Without a like-for-like control comparison, an observed event count cannot isolate the effect of treatment from the underlying disease, previous therapies, or other influences. No comparative estimate or analysis is reported in this record.

General: an uncertainty interval, such as a confidence interval, expresses the range of values compatible with an estimate under the analysis assumptions. It is not a guarantee that the true value lies in that range. The registry record does not report confidence intervals for the primary safety counts. It also does not report a statistical test for the primary endpoint. Thus, the counts can be described, but their statistical precision beyond the reported denominators is not quantified in the record.

Secondary outcomes and time-to-event methods

The secondary outcome list includes cytogenetic response proportions and time to or duration of responses. A response proportion concerns whether participants met a defined response criterion. Time-to-event outcomes additionally depend on the time origin, event definition, censoring rules, and follow-up. General: censoring occurs when an individual’s event time is not observed in full, for example because follow-up ends before the event. Standard time-to-event methods can use partial follow-up, but their interpretation depends on assumptions and a clearly specified analysis plan. Those analysis details and results are not reported in the registry record.

The record lists 24 outcome measures and zero analyses. This means the outcome list is not accompanied by reported analysis results in the registry record. It does not establish which methods were used outside the registry, or whether analyses were conducted elsewhere. No interim analyses, sample-size justification, or detailed statistical analysis plan are reported in the registry record.

Planning illustrations are not trial findings

General: sample-size planning asks how many participants would be needed to detect an assumed effect under specified statistical assumptions. Such calculations depend on the effect size, significance level, allocation, and chosen method. The planning assumptions described for this study use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect defined as a fraction of an observed effect. For time-to-event outcomes, the stated planning approach uses a total event count; other calculations are expressed as participants per arm.

These are planning illustrations, not observed power and not evidence that the trial was designed around those assumptions. The registry does not report that the original trial used this planning procedure. The record does not include numerical planning outputs in the registry record described for interpretation, so no required sample size can be stated. No recalculated treatment-effect estimate is reported.

10. Multiplicity, Missing Data and Analysis Populations

The record lists 24 outcome measures, including several cytogenetic response and time-to-response outcomes, but reports zero analyses. It does not describe a statistical analysis plan that identifies a primary hypothesis-testing strategy, a hierarchy among outcomes, or an adjustment for multiple comparisons. A multiplicity procedure is therefore not reported in the registry record. General: when many outcomes are examined, the chance of at least one apparently unusual result can increase; pre-specification and suitable adjustment can help address that issue. In this record, there are no reported analysis results from which to assess how multiplicity was handled.

Missing-data methods are also not reported in the registry record. The posted description does not state how participants with incomplete assessments, withdrawal, or unavailable outcome measurements were handled for the listed cytogenetic and time-to-event outcomes. For safety, the stated population is all treated participants, but the registry record does not provide further detail on missing safety observations or how follow-up gaps were treated.

The record does not report interim analyses, a sample-size justification, or a detailed statistical analysis plan. It is therefore not possible to establish whether these issues were addressed in a protocol or in analyses outside the registry. These gaps limit assessment of the prespecification and robustness of any inferential claims; the record itself reports no analyses to evaluate.

11. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pneumonia (11.1%), acute lymphocytic leukaemia recurrent (11.1%), platelet count decreased (7.4%), pyrexia (5.6%), cerebral haemorrhage (5.6%).

The safety population is all treated participants, with 54 participants listed at risk. The registry reports 33 participants with serious adverse events and 54 with other adverse events. “Serious” follows the registry’s event criteria, which include outcomes such as death, life-threatening events, hospitalization, persistent or significant disability, and other important medical events. An AE does not necessarily have a causal relationship to treatment.

The serious-event terms listed most prominently include pneumonia and recurrent acute lymphocytic leukaemia at 11.1% each; platelet count decreased at 7.4%; and pyrexia, cerebral haemorrhage, and pleural effusion at 5.6% each. These percentages are the registry’s reported values. The death count is not reported in the registry record, although the primary endpoint includes deaths and the record says reported deaths after study end are included.

Counts and percentages describe recorded events in the study population; they are not a comparison with another treatment or an estimate of how many events were caused by dasatinib. The record also gives discontinuation tallies by disease subgroup, with reasons including AEs, death, and insufficient effect. These are recorded reasons for discontinuation, not evidence of a comparative treatment effect.

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

The posted results document a safety experience in 54 treated participants enrolled in a rollover study of dasatinib for CML or Ph+ ALL. The registry reports serious adverse events in 33 participants and other adverse events in 54 participants. It also lists selected serious-event terms and percentages, and discontinuation tallies by disease subgroup. These registry entries describe what the registry record reports for this study population.

They do not establish that dasatinib caused every reported AE. The AE definition does not require a causal relationship, and the record does not provide a controlled comparison that would separate treatment effects from disease, prior treatment, or other factors. The study is not described as randomized, its allocation method is not stated, and the registry identifies the arms as not a like-for-like control comparison.

The primary endpoint includes deaths, but a death count is not reported in the registry record. Although the record notes that the death measure includes reported deaths after study end, that note does not supply the missing count. Likewise, the secondary outcome list includes cytogenetic response and time-to-response measures, but zero analyses are reported. The record therefore does not establish response rates, response timing, or comparative efficacy.

General: a descriptive event count can identify the burden observed during study follow-up, while a comparative estimate addresses how outcomes differ between treatment strategies. The latter requires an appropriate comparison and an analysis that accounts for study design and follow-up. No such estimate is reported. Interpretation should remain within the limits of the posted counts, definitions, and study design; the record does not support a causal or comparative conclusion.

13. Why This Trial Matters Statistically

This study illustrates why a rollover safety record must be distinguished from a controlled treatment comparison. Participants continued dasatinib after completing a previous Phase I/II protocol, and the registry lists three disease groups. The central question is therefore descriptive: what adverse events, serious adverse events, deaths, and discontinuations were recorded among treated participants? The record does not describe a like-for-like control comparison.

The endpoint also demonstrates that clinical safety summaries depend on operational definitions. An AE can be recorded without a causal relationship to treatment, while seriousness is determined by specified consequences or clinical importance. Separating those concepts is essential when reading event counts. Similarly, participant counts are not the same as episode counts, and a set of event-term percentages does not itself show comparative risk.

Statistically, the record underscores the difference between listing outcomes and reporting their analyses. It identifies cytogenetic and time-to-response measures, but reports zero analyses. It also does not report a detailed statistical analysis plan, missing-data approach, multiplicity strategy, or interim analyses. These omissions constrain what can be inferred from the registry record, especially for outcomes that would require defined timing, censoring, or comparison methods.

The safety counts remain informative as a description of the recorded experience, but they cannot substitute for a controlled estimate of treatment effect. This distinction is important when clinical researchers interpret registry results: the study’s design, outcome definitions, and reporting completeness determine which questions its results can answer.

14. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The key limitation is the absence of a like-for-like control comparison. The registry states that the study was not randomized and that allocation was not stated. The parallel model and lack of masking do not establish comparable groups. The study is a rollover after an earlier Phase I/II protocol, so the enrolled population had already completed that protocol; the registry record does not describe how that selection affects generalizability.

The record reports 54 enrolled participants and gives the all-treated safety group as 54, but does not provide disease-subgroup denominators in the registry record. The serious-event count is reported as 33 participants, while the death count is not reported. The endpoint includes discontinuation, and the record gives subgroup tallies and reasons, but does not report a comparative discontinuation analysis.

Outcome reporting is incomplete for statistical interpretation. Although 24 outcome measures are listed, zero analyses are reported. The registry record does not report results for the secondary cytogenetic outcomes, confidence intervals, a statistical analysis plan, sample-size justification, interim analyses, multiplicity handling, or missing-data methods. It also does not provide participant-level follow-up information. These limitations prevent assessment of the precision of unreported effects and restrict interpretation to the descriptive safety information that is posted.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Rollover Study of BMS-354825 in Patients With CML and Ph+ALL — NCT01030718.
• Registry record last updated 2010-12-14; status last verified 2010-11.
• Results first posted 2010-12-14.
• 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

The registry record describes a completed Phase I/II rollover study of dasatinib in participants with CML or Ph+ ALL who were resistant or intolerant to previous treatment and had completed an earlier protocol. Bristol-Myers Squibb sponsored the study. Enrollment was 54; the reported start was 2006-01 and primary completion was 2009-06. The study had no masking, was not randomized, and is not described as a like-for-like control comparison.

The primary outcome counted participants with AEs, SAEs, deaths, and discontinuation among all treated participants. The record reports 33 participants with serious adverse events and 54 with other adverse events. It lists selected serious-event terms and percentages, but does not report the death count. Secondary outcomes include cytogenetic response and time-to-response measures; no analyses are reported. The registry record therefore supports a descriptive account of the posted safety counts, but not a comparative estimate of safety or efficacy.