Stem Cell Injection in Cancer Survivors: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Stem Cell Injection in Cancer Survivors (NCT02509156), a randomized phase 1 study. The primary outcome is Proportion of Major Adverse Cardiac Events (MACE).

2. Clinical Question

The study’s stated primary purpose was to examine the safety and feasibility of delivering allo-MSCs by transendocardial injection to cancer survivors with left ventricular dysfunction related to anthracycline-induced cardiomyopathy. Its secondary purpose was to obtain preliminary evidence about therapeutic efficacy. Those purposes are distinct: the primary endpoint concerns adverse cardiac events, while secondary outcomes include changes in cardiac imaging measures.

The posted randomized arm results report 3 events for Allo-MSCs and 6 for Placebo. These are event counts, not reported proportions with confidence intervals. The registry identifies the endpoint as a proportion but does not provide an uncertainty interval or a formal between-arm test for these counts. The record also describes the safety and feasibility population as including the open-label lead-in as well as randomized subjects; the listed arm counts are for the randomized arms.

The trial therefore poses a safety-oriented primary question alongside exploratory questions about cardiac function and strain. The registry record does not report a definitive efficacy question with a prespecified effect size or a detailed hypothesis for each secondary measure.

3. Trial Design

This was a Phase I, first-in-human, multicenter, randomized, double-blinded, placebo-controlled, parallel-group study. It compared transendocardial injection of allo-MSCs with placebo, described as Buminate solution. The registry lists 46 participants as the enrollment and identifies the study status as completed.

The intervention arm’s description specifies a target dose of 100 million allo-MSCs. The primary safety and feasibility population was defined to include open-label lead-in participants, n=6, as well as randomized subjects, n=31. The primary outcome results list 20 participants in the Allo-MSCs arm and 17 in the Placebo arm. The registry record does not explain the relationship between the overall enrollment figure, the safety population, and the arm-specific outcome counts, or report reasons for any differences.

The primary endpoint timeframe was baseline to 12 months. Secondary outcomes included changes from baseline in left ventricular ejection fraction, global and regional strain measured with HARP MRI, and left ventricular end diastolic volume index. Some were also defined as trajectories over time. The registry lists 28 outcome measures and 20 analyses, while 21 secondary outcomes are posted.

Randomization and quadruple masking are reported, but details of the randomization sequence, allocation concealment, and which roles were masked are not reported in the registry record. The record gives a study start of 2016-08 and primary completion of 2019-11. It does not report a detailed protocol or a full statistical analysis plan, so the rationale for specific modeling choices cannot be established from the registry information alone.

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: 46 subjects consented to participate; 37 completed baseline testing and met eligibility criteria. This includes 6 (open label) and 31 (randomized) subjects. Reasons for failed eligibility (n=9) include elevated LVEF, failure to complete baseline testing, MRI contraindications, and investigator discretion.

Open-Label Lead-In Phase

Randomized Phase

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Population for safety and feasibility measures include open label lead-in participants (n=6) as well as randomized subjects (n=31).

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 Proportion of Major Adverse Cardiac Events (MACE), measured over Baseline to 12 months.

The primary endpoint is the proportion of major adverse cardiac events from baseline to 12 months. The registry description includes adjudicated events comprising death, hospitalization for worsening heart failure, and/or another heart failure exacerbation that did not result in hospitalization. These components combine serious clinical events with worsening heart failure managed without hospitalization.

Such a composite endpoint can capture several clinically important forms of worsening in one safety measure. Adjudication is intended to apply defined criteria when deciding whether reported events meet the endpoint definition. However, the registry record does not describe the adjudication process, the criteria used for each event component, or whether components were weighted equally in the analysis.

The endpoint is labeled as a proportion, but its posted arm results are presented as event counts: 3 in Allo-MSCs and 6 in Placebo. No arm-specific proportions, confidence intervals, or formal comparison for this primary result are reported in the registry record. Consequently, the counts convey the observed event totals but do not by themselves show the precision of an estimated difference between arms. They also do not establish how the event totals were distributed across the composite components.

The primary outcome matches the stated safety and feasibility purpose. It does not directly measure whether cardiac function improved; that question is addressed through secondary imaging outcomes. The distinction matters because an observed pattern in an imaging measure cannot substitute for the primary safety endpoint, and the event counts alone do not answer the secondary efficacy questions.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Several types of statistical summaries appear in the registry results. They answer different questions, and they should not be treated as interchangeable. General explanation: an arm-specific mean describes the average observed value in that group, while a mean difference compares the two arm averages. A confidence interval expresses uncertainty around an estimate under the assumptions of the analysis; it is not a range containing every individual participant’s response.

What does the adjusted comparison estimate?

For LVEF change and global strain, the registry reports ANCOVA results described as mean differences in final values. In general, analysis of covariance compares groups on an outcome while accounting for one or more covariates. The registry does not identify the covariates used, so the precise adjustment cannot be reconstructed from the posted results. The reported LVEF estimate is 0.80, with a 95% confidence interval from -4.52 to 6.11 and p=0.746. The reported global strain estimate is -0.71, with a 95% confidence interval from -2.68 to 1.26 and p=0.328. Both intervals include zero, which means the reported intervals are compatible with no difference as well as with differences in either direction.

For the LVEF change outcome, a simple unadjusted recalculation from the posted arm summaries gives a mean difference of 0.79 and a 95% confidence interval from -4.249 to 5.829. For global strain, the corresponding recalculation gives -0.71 and a 95% confidence interval from -2.561 to 1.141. These recalculations use posted summary statistics and are not substitutes for the sponsor’s adjusted analyses. The sponsor’s own analysis takes precedence when interpreting the trial’s reported analysis.

How does the trajectory analysis differ?

The LVEF trajectory outcome concerns measurements described at baseline, 6 months, and 12 months. The registry reports a repeated measures linear regression result, with a slope of time of 1.53 and p=0.024. In general, repeated measures models account for observations collected over time from the same participants, rather than treating each assessment as unrelated. A slope summarizes a modeled pattern over time; it is not the same estimand as a single baseline-to-12-month change or an adjusted difference in final values. The confidence interval and further model details are not reported in the registry record. The registry’s brief result does not permit a full assessment of model assumptions or the exact interpretation of the slope parameter.

Why are confidence intervals important?

General explanation: a point estimate gives one central summary of an effect, while its confidence interval indicates how much uncertainty remains under the model. A narrow interval gives greater precision than a wide interval. An interval that includes zero for a difference estimate does not prove that the groups are identical; it means the data and model do not exclude zero at the interval’s stated confidence level. The LVEF and global strain ANCOVA intervals extend across zero and include values in both directions, so the point estimates should not be read as precise evidence of a particular direction or magnitude.

What do the p-values say?

General explanation: a p-value measures how incompatible the observed data are with a specified null model, given the analysis assumptions. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. The reported p-values are specific to their analyses. The registry record does not report adjustment for the number of secondary comparisons, so the nominal p=0.024 for the trajectory result should be considered in the context of multiple outcomes and analyses rather than as a stand-alone confirmatory finding.

How should the planning illustrations be read?

The sample-size planning illustrations are based on the observed effect and state assumptions of a two-sided significance level of 0.05, equal allocation, and a normal approximation. They show planning examples for detecting the stated fractions of the observed effect, with participants per arm reported for 80% and 90% power. General explanation: power is a design-stage probability of detecting a specified effect under stated assumptions. These illustrations are not observed power, do not describe the trial’s original design justification, and cannot show that the study had adequate power for its actual endpoints. The registry record does not report its own sample-size justification.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 28 outcome measures, 20 analyses, and 21 posted secondary outcomes. This is a setting in which multiplicity matters: multiple outcomes, timepoints, and analytic approaches can produce several statistical tests. General explanation: if many tests are considered without accounting for their number or a prespecified testing hierarchy, nominal p-values can overstate the strength of evidence. The registry record does not report a multiplicity adjustment, a hierarchy for secondary outcomes, or which analyses were designated confirmatory.

Missing-data methods are also not reported in the registry record. The arm sizes differ across posted outcomes; for example, the LVEF change analysis lists 12 Allo-MSCs and 13 Placebo participants, while the global strain analysis lists 11 and 13. These outcome-specific denominators describe the posted analyses, but the record does not explain why they differ from the primary arm counts or whether missing assessments were handled through imputation, complete-case analysis, or another method. It also does not provide a detailed account of participant flow or reasons for unavailable measurements.

The analysis populations are not fully specified for every outcome. The primary endpoint description states that safety and feasibility measures include open-label lead-in participants and randomized subjects. The arm-specific results, by contrast, list randomized treatment groups. The record does not explain whether all other endpoints use an intention-to-treat population, a per-protocol population, or another defined set of participants. This limits assessment of how exclusions and missing observations may affect the estimates.

The protocol-level statistical plan, interim analysis procedures, and sample-size justification are not reported in the registry record. The planning illustrations based on observed effects are not a substitute for these design details. Without these details, it is not possible to determine whether the posted analyses follow a prespecified plan or how many analyses were considered during the study.

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: cardiac failure acute (11.8%), cardiac failure (11.8%), atrial fibrillation (5.9%), ventricular tachycardia (5.9%), arrhythmia (5.9%).

The posted serious adverse event figures report 5 participants affected among 20 at risk in the Allo-MSCs group and 11 among 17 at risk in the Placebo group. The registry also reports 1 death in Allo-MSCs and 0 in Placebo. These are descriptive safety figures; the record does not provide confidence intervals or a formal comparison for them. They should not be interpreted as proof that the groups differ, or that they are equivalent.

The registry lists serious event terms including acute cardiac failure, cardiac failure, atrial fibrillation, ventricular tachycardia, arrhythmia, and atrial flutter, with the posted percentages 11.8%, 11.8%, 5.9%, 5.9%, 5.9%, and 5.9%, respectively. The record does not provide enough detail in the listed terms to determine overlap between participants or to infer that each term represents a distinct person. Event terms may also describe different categories of reported events; the registry does not explain their coding or attribution in this summary.

General explanation: adverse event counts are descriptive and can be affected by the population included, follow-up, event definitions, and the way events are counted. The registry does not report whether serious events were considered related to the intervention. The open-label lead-in is included in the stated safety population, while the listed arm-level serious event figures correspond to the randomized arms. These differences in population description should be kept in mind when reading the numbers.

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

The results describe a small Phase I study designed primarily to examine safety and feasibility, with preliminary efficacy evidence as a secondary purpose. The registry reports 3 primary endpoint events in the Allo-MSCs arm and 6 in the Placebo arm. Because no primary endpoint confidence interval or formal comparison is reported, those counts alone do not establish the size or precision of a between-arm difference.

The secondary cardiac measurements likewise do not provide a definitive answer. The posted adjusted LVEF and global strain comparisons have confidence intervals spanning zero, while one trajectory analysis has a reported p-value but no confidence interval. These findings are compatible with uncertainty and should not be compressed into a simple claim that the intervention did or did not have an effect. The available registry information does not support a definitive efficacy conclusion.

General explanation: a statistically non-significant result is not evidence that two treatments are identical. It may reflect a small or uncertain effect, limited precision, or other features of the design and data. Likewise, a nominally small p-value in one of multiple secondary analyses does not by itself establish a reproducible effect. The registry record does not report enough detail about multiplicity, missing data, or the full analysis plan to resolve these issues.

These results also do not answer questions beyond the study population and outcome definitions. The enrolled group was cancer survivors with left ventricular dysfunction secondary to anthracycline-induced cardiomyopathy, and the primary timeframe was baseline to 12 months. The posted findings should not be generalized automatically to other populations, other follow-up periods, or other interventions. They are best understood as registry-reported results from an early-phase study, with substantial uncertainty around comparative interpretation.

13. Why This Trial Matters Statistically

This trial is statistically informative because it illustrates the difference between a safety-oriented primary endpoint and preliminary secondary efficacy measures. The primary endpoint combines several types of cardiac events, but the registry posts counts without a formal comparative estimate or uncertainty interval. That makes clear why a point count alone is not enough to characterize the precision of an estimated treatment contrast.

It also illustrates how different analyses answer different questions. A baseline-to-12-month change, an ANCOVA comparison of final values, and a repeated-measures trajectory are not interchangeable summaries. Even when all concern LVEF, their estimands and assumptions differ. Reading their results together requires attention to the outcome definition and model, rather than treating all estimates as measures of one identical effect.

Small outcome-specific analysis groups and a large collection of registered measures make uncertainty and multiplicity particularly relevant. The reported trajectory p-value is one finding among multiple secondary analyses, and the registry does not report whether multiplicity was addressed. Confidence intervals for other analyses show why effect estimates need a precision assessment alongside p-values. Planning illustrations based on observed effects can show how assumed effects relate to hypothetical sample sizes, but they are not evidence of the study’s original power or a substitute for a prospective design rationale.

Finally, the record’s limited reporting of missing-data handling, analysis populations, and model details shows how registry results can constrain statistical interpretation even when numerical estimates are available. Transparent reporting of those features is essential for readers to assess how robust an analysis may be.

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.

Several limitations specific to this trial restrict interpretation of the posted results. The primary endpoint is described as a proportion, but the arm results are counts, and no primary comparative estimate, confidence interval, or p-value is reported. The event counts also combine multiple components, while component-specific primary endpoint results are not reported in the registry record. This prevents assessment of how much each component contributed to the composite.

The primary safety population includes open-label lead-in participants as well as randomized subjects, but the posted arm results are organized by randomized group. The record does not fully explain the analysis populations for all outcomes or the relationship between the various participant counts. Outcome-specific analysis groups are smaller than the listed primary arm groups for some secondary measures, but reasons and missing-data methods are not reported in the registry record.

The secondary outcomes are numerous, and the registry does not report whether multiplicity was controlled or which comparisons were prespecified as confirmatory. Some analyses have confidence intervals and others do not. The LVEF trajectory result reports a p-value without a confidence interval or detailed model specification. The exact covariates in the ANCOVA analyses and further assumptions of the repeated measures model are not reported in the registry record.

The record also does not report detailed eligibility criteria, randomization procedures, allocation concealment, the roles included in quadruple masking, interim analyses, or the trial’s own sample-size justification. These are not reasons to presume a flaw, but their absence limits appraisal of design and analysis choices. The planning illustrations use the observed effect and should not be mistaken for a prospective justification or observed power.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Stem Cell Injection in Cancer Survivors — NCT02509156.
• Registry record last updated 2020-11-05; status last verified 2020-10.
• Results first posted 2020-11-05.
• 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

NCT02509156 is a completed, randomized, double-blinded, placebo-controlled Phase I study of allo-MSCs in cancer survivors with anthracycline-induced cardiomyopathy. The intervention was transendocardial injection of a target dose of 100 million allo-MSCs; the comparator was Buminate solution. The primary outcome concerned adjudicated major adverse cardiac events from baseline to 12 months.

The primary results list 3 events in Allo-MSCs and 6 in Placebo, without a reported comparative interval or test. Secondary findings include ANCOVA comparisons for LVEF and global strain and a repeated measures trajectory analysis for LVEF. Their interpretation is limited by uncertainty, multiple analyses, and incomplete reporting of missing-data handling and analysis populations.

The registry describes safety and feasibility as the primary purpose and efficacy as preliminary. Its posted results are useful for describing observed outcomes, but the available information does not support a definitive comparative conclusion.