Study of the Efficacy and Safety of Parsaclisib in Participants With Primary Warm Autoimmune Hemolytic Anemia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of the Efficacy and Safety of Parsaclisib in Participants With Primary Warm Autoimmune Hemolytic Anemia (NCT05073458), a randomized phase 3 study. The primary outcome is Percentage of Participants Attaining a Durable Hemoglobin Response.

2. Clinical Question

3. Trial Design

This was a Phase 3, randomized, triple-masked, parallel-group, placebo-controlled study of parsaclisib in participants with primary warm autoimmune hemolytic anemia. The registry lists an enrollment of 13 and a terminated status. The study start date is 2022-03-15, and the primary completion date is 2023-10-17. The reason for termination is not reported in the registry record.

During the double-blind period, one group was assigned to parsaclisib and the other to placebo for 24 weeks. The placebo group was then described as receiving parsaclisib in a 24-week open-label period. Participants in the parsaclisib group could enter the open-label period after completing the double-blind period if, in the investigator’s opinion, they were tolerating treatment. Both descriptions mention possible continued treatment in a long-term extension; the full extension details are not reported in the registry record.

The primary endpoint was the percentage of participants attaining a durable hemoglobin response up to Week 24. The reported analysis population was the Safety Analysis Set, defined in the record as randomized participants who received at least 1 dose of study drug. Group membership for this analysis followed actual Day 1 treatment, regardless of assigned study treatment, and only participants with available data were analyzed. The primary outcome has posted results for 3 participants in the parsaclisib group and 4 in the placebo group. Those outcome denominators differ from the adverse-event groups, so the endpoint and safety figures should be read as separate reported summaries.

The registry record lists 13 outcome measures and indicates 12 secondary outcomes posted. Secondary measures include fatigue scores, walking distance, hemoglobin changes, transfusions, and corticosteroid dose. The record does not report a detailed statistical analysis plan, the reason for stopping the study, or the planned recruitment rationale.

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: This study was designed to evaluate parsaclisib 2.5 mg QD compared with placebo over a 24-week double-blind treatment period followed by a 24-week open-label treatment period with parsaclisib. Participants could then continue to receive parsaclisib in a long-term extension period.

24-week Double-blind Period

Reasons for not completing this period:

24-week Open-label Period

Reasons for not completing this period:

Long-term Extension Period (~2 Years)

Analysis population for the primary outcome, as defined in the registry: Safety Analysis Set: all randomized participants who received at least 1 dose of study drug. Treatment groups were determined according to the actual treatment the participant received on Day 1 regardless of assigned study treatment. Only participants with available data were analyzed.

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 Percentage of Participants Attaining a Durable Hemoglobin Response, measured over up to Week 24.

The primary endpoint is the percentage of participants attaining a durable hemoglobin response up to Week 24. The definition required hemoglobin ≥10 grams per deciliter (g/dL), together with an increase from baseline of ≥2 g/dL that was not attributed to rescue therapy. This response had to be present at ≥3 of the 4 available visits at Week 12 and/or later during the 24-week double-blind treatment period.

This endpoint combines the hemoglobin level reached with the change from each participant’s baseline. Requiring the response at multiple visits makes the definition reflect a sustained pattern over the assessment period rather than a single qualifying measurement. Excluding responses attributed to rescue therapy is part of the registry’s endpoint definition; the record does not provide further details about how rescue therapy was handled in the analysis.

In general, a binary endpoint such as this classifies each analyzed participant as either meeting or not meeting the response definition, then reports the percentage meeting it. That percentage is easy to describe clinically, but its precision depends on the number of analyzed participants and on how many meet the definition. The registry reports 33.3% for parsaclisib among 3 participants and 25.0% for placebo among 4 participants. It does not report confidence intervals for these percentages, and the posted data do not support an estimable two-arm comparison.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: each analyzed participant either met the durable response definition or did not. The registry reports the percentage meeting the definition for each group. A percentage is an intuitive summary, but when its denominator is small, a change in the classification of an individual can have a noticeable influence on the reported value. This is a general statistical feature of proportions, not evidence about any participant’s outcome beyond what the registry reports.

What does “durable response” mean statistically?

The endpoint requires two hemoglobin conditions and repeated qualifying assessments. Hemoglobin must be ≥10 g/dL and must have increased from baseline by ≥2 g/dL, without attribution to rescue therapy. The response must be present at ≥3 of the 4 available visits at Week 12 and/or later during the 24-week double-blind period. Statistically, this is a rule that converts measurements taken over time into one participant-level response classification. The registry does not report how unavailable visits were treated when determining whether the rule was met, beyond stating that only participants with available data were analyzed.

Why is a two-arm comparison not available?

The posted primary values are percentages for groups with small analyzed denominators. The record indicates that values are missing or not estimable for a two-arm comparison of this binary percentage endpoint. Consequently, there is no posted inferential estimate of the treatment contrast, no confidence interval, and no p-value to interpret. A difference should not be calculated from the displayed percentages as if it were a prespecified or inferential result. The posted record does not establish that the groups differ, nor does it establish that they are equivalent.

In general, a confidence interval gives a range of estimates compatible with the observed data and the chosen statistical method. Its width reflects statistical precision as well as the method’s assumptions. A confidence interval is not reported for the primary percentages. In general, a hypothesis test evaluates compatibility between the data and a stated null hypothesis. The registry record does not post a test result for the primary comparison, so the outcome should be read descriptively rather than through a significance threshold.

What does the analysis population imply?

The record defines the Safety Analysis Set as all randomized participants who received at least 1 dose of study drug. It assigns treatment groups using actual treatment received on Day 1, regardless of assigned treatment, and analyzes only participants with available data. This is the population definition reported for the primary endpoint. It is not an analysis explicitly described as intention-to-treat, and the registry record does not report an intention-to-treat analysis for this result. The primary outcome denominators are 3 for parsaclisib and 4 for placebo; these are the denominators attached to the posted endpoint values.

The group assignment rule can affect the interpretation of a randomized study because assigned treatment and treatment actually received are conceptually different. The registry’s description tells readers which rule was used for this endpoint, but does not report whether any participant received a treatment other than the one assigned. No such occurrence should be inferred from the rule alone.

How should the planning illustrations be read?

In general, sample-size calculations estimate the number of participants needed to detect a specified effect under stated assumptions. The planning illustrations associated with these results are based on the observed effect, a two-sided significance level of 0.05, equal allocation, a normal approximation, and effects defined as stated fractions of the observed effect. They are illustrations, not the trial’s documented design rationale. The registry record does not report the study’s own sample-size justification. Such calculations are not observed power and do not supply evidence that the observed result is reliable or that a future study would produce the same effect.

For binary outcomes, the analysis method also depends on how the response proportions and their uncertainty are modeled. General methods may use approximations or exact procedures, but the registry record does not report a method for an inferential comparison of this endpoint. It would therefore be inappropriate to attribute a particular model, adjustment, or testing procedure to the study.

What can the registry support?

The record supports a description of the endpoint definition, the analysis population, and the posted response percentages. It does not support a precise estimate of the difference between treatment groups. General statistical reasoning can explain why small denominators and unavailable inferential estimates limit precision, but it cannot fill in missing analyses or establish results not reported by the registry.

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("Parsaclisib", "Placebo")
n      <- c(3, 4)
events <- c(1, 1)
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  Parsaclisib  3  1  33.3  0.8  90.6  6.1  79.2  6.8  80.6
ARM  Placebo  4  1  25.0  0.6  80.6  4.6  69.9  5.3  71.6

11. Multiplicity, Missing Data and Analysis Populations

The registry indicates 12 secondary outcomes posted, in addition to the primary endpoint. The listed secondary measures cover fatigue, walking distance, hemoglobin, transfusions, and corticosteroid dose, including changes over visits and percentages of participants meeting specified criteria. Assessing multiple outcomes or repeated time points raises a general multiplicity issue: the more comparisons considered, the greater the possibility of finding an apparently notable result by chance if all are interpreted without an appropriate plan. Whether and how multiplicity was controlled is not reported in the registry record.

The available primary outcome description says that only participants with available data were analyzed. It does not explain why data might be unavailable, how missing visits affected the durable response classification, whether missing outcomes were imputed, or whether sensitivity analyses were conducted. Those missing-data procedures are not reported in the registry record. The endpoint definition refers to 4 available visits, but that wording alone does not specify how incomplete participant records were handled in the analysis.

The primary endpoint population is described as the Safety Analysis Set: randomized participants who received at least 1 dose of study drug. Treatment groups were based on actual treatment received on Day 1, regardless of assigned treatment. The posted endpoint denominators are 3 for parsaclisib and 4 for placebo. The record does not describe a separate intention-to-treat primary analysis or explain how participants who did not receive treatment were handled for efficacy outcomes.

The registry record lists no analysis records. A prespecified testing hierarchy, multiplicity adjustment, missing-data model, sensitivity analysis, and detailed statistical analysis plan are not reported. These omissions limit assessment of how outcome selection, repeated measurements, and incomplete data were addressed.

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: cholecystitis (9.1%), colitis (9.1%), diarrhoea (9.1%), haemoglobin decreased (9.1%), immune-mediated enterocolitis (9.1%).

The safety summaries report serious adverse events among 6 participants at risk in the parsaclisib group, with 6 affected, and among 6 participants at risk in the placebo group, with 0 affected. The record reports other adverse events among 11 participants at risk in the parsaclisib group, with 10 affected, and among 6 participants at risk in the placebo group, with 4 affected. Deaths are reported as 0 in each group. These are descriptive counts; the registry does not provide a comparative statistical analysis of them.

The serious-event terms listed are cholecystitis, colitis, diarrhoea, haemoglobin decreased, immune-mediated enterocolitis, and pyrexia, each reported as 9.1%. The available record summary does not attach those term percentages to a treatment group, so they should not be interpreted as arm-specific rates. The term list also does not establish whether an individual participant experienced more than one event.

In general, an adverse-event count describes events or affected participants according to the registry’s reporting convention; it does not by itself establish that treatment caused an event. The group sizes and exposure periods are not interchangeable with the primary endpoint denominators. The registry does not report further detail in these summary figures about event timing, duration, severity grading, or causal attribution.

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

The posted result describes the percentage meeting a specified durable hemoglobin response definition among participants included in the primary outcome analysis. That percentage is 33.3% for parsaclisib among 3 participants and 25.0% for placebo among 4 participants. The definition requires the specified hemoglobin level and increase at repeated assessments, with responses attributed to rescue therapy excluded.

These percentages do not establish that parsaclisib is effective or ineffective. The registry record does not provide a confidence interval, p-value, or estimable two-arm comparison for the primary outcome. Without those inferential results, the comparative precision is not reported. The descriptive percentages alone cannot determine whether any apparent difference is compatible with chance variation or reflects a treatment effect.

The record also does not establish equivalence between the groups. Failure to report an estimable comparison is not evidence that outcomes are the same. Similarly, the safety summaries are descriptive and do not establish that treatment caused or did not cause the listed events. The registry lists the trial as terminated but does not report why it ended, so termination should not be interpreted as evidence about efficacy or safety.

In general, a clinical trial’s result must be interpreted in light of its endpoint definition, analysis population, missing data, precision, and analysis plan. For this study, the record specifies the endpoint and a Safety Analysis Set using treatment received on Day 1, but does not report the detailed handling of missing data or the analysis procedures for an inferential comparison. The small endpoint denominators and absence of interval estimates mean that the posted result is best understood as a limited descriptive summary rather than a definitive comparative finding.

The secondary outcomes and safety summaries provide additional types of measurements in the registry, but their presence does not replace an inferential analysis of the primary endpoint. The registry record does not report enough detail to reconstruct an analysis plan or draw conclusions beyond the posted descriptive results.

14. Why This Trial Matters Statistically

This study illustrates the distinction between a trial’s design label and the strength of the result that can be interpreted from its posted outcome. It is a randomized, triple-masked, placebo-controlled Phase 3 study, yet the primary response percentages are reported for 3 participants in one group and 4 in the other. The study’s phase and design describe its intended framework; they do not by themselves guarantee precise estimates or an interpretable comparison.

The primary endpoint also shows how a clinical concept becomes a statistical outcome. A durable response is defined through a threshold hemoglobin level, a threshold increase from baseline, exclusion of responses attributed to rescue therapy, and repeated assessments during the double-blind period. Readers need that operational definition to understand what the percentages count. The response classification does not summarize every aspect of disease status or participant experience.

Another statistical issue is the analysis population. The registry describes a safety set that includes randomized participants who received at least 1 dose and classifies them by actual Day 1 treatment, with available data analyzed. This should be distinguished from a result based strictly on assigned group. Clear population definitions are essential because denominators and group labels determine what a percentage represents.

Finally, the record demonstrates why uncertainty reporting matters. Without a confidence interval or an estimable two-arm comparison, the percentages cannot support a precise comparative conclusion. A sample-size planning illustration based on an observed effect is not a substitute for a trial’s documented design rationale, nor is it observed power. The posted information supports careful description, while leaving important questions about precision and analysis unresolved.

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 primary outcome is available for 3 participants in the parsaclisib group and 4 in the placebo group, and the registry does not provide confidence intervals or an estimable two-arm comparison. The resulting percentages therefore have no reported measure of precision. The small endpoint denominators also mean that the displayed values should not be treated as stable estimates of outcomes in a broader population.

The registry lists enrollment of 13 and a terminated status, but does not report the reason for termination. It is not possible to infer from those details whether termination was related to recruitment, operations, safety, efficacy, or another consideration. The record also does not report the trial’s sample-size justification, a detailed statistical analysis plan, or the planned approach to multiplicity.

The outcome analysis used the Safety Analysis Set and assigned participants by actual treatment received on Day 1, regardless of randomized assignment. Only participants with available data were analyzed. The registry record does not describe the extent or reasons for missing data, the handling of missing visits in the repeated-visit response definition, or any imputation or sensitivity analyses. It does not report an intention-to-treat analysis for the primary result.

The primary endpoint is a specific composite rule based on hemoglobin measurements and rescue-therapy attribution. The record does not provide enough detail to evaluate the measurement schedule beyond the stated visit criteria or the operational assessment of rescue-therapy attribution. Secondary outcomes are listed, but multiplicity procedures and detailed inferential results are not reported. Safety summaries are descriptive, and the listed serious-event term percentages are not assigned to treatment groups in the available summary. These limitations constrain interpretation to the definitions and results explicitly posted in the registry.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Study of the Efficacy and Safety of Parsaclisib in Participants With Primary Warm Autoimmune Hemolytic Anemia — NCT05073458.
• Registry record last updated 2025-11-03; status last verified 2025-10.
• Results first posted 2025-01-15.
• 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

The registry describes a Phase 3, randomized, triple-masked, parallel-group, placebo-controlled study of parsaclisib in primary warm autoimmune hemolytic anemia. It lists enrollment of 13, a start date of 2022-03-15, a primary completion date of 2023-10-17, and a terminated status. The reason for termination is not reported in the registry record.

The primary endpoint was the percentage attaining a durable hemoglobin response up to Week 24. The posted percentages are 33.3% among 3 participants receiving parsaclisib and 25.0% among 4 receiving placebo. The registry does not provide a confidence interval, p-value, or estimable two-arm comparison. The primary analysis population was the Safety Analysis Set, grouped by actual Day 1 treatment, with only available data analyzed.

The record also includes secondary outcome and safety summaries. Details of the statistical plan, multiplicity control, missing-data methods, and termination rationale are not reported in the registry record.