A Study Comparing the Blood Levels of Both Pegaspargase (S95014) Formulations (Liquid vs Lyophilized) in the Treatment of Paediatric Patients With Acute Lymphoblastic Leukemia (ALL): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study Comparing the Blood Levels of Both Pegaspargase (S95014) Formulations (Liquid vs Lyophilized) in the Treatment of Paediatric Patients With Acute Lymphoblastic Leukemia (ALL) (NCT04954326), a randomized phase 2 study. The primary outcome is Pharmacokinetics Measurement.

2. Clinical Question

The clinical question was whether the liquid and lyophilized formulations of S95014 produced comparable pharmacokinetic exposure during induction in paediatric patients with newly diagnosed acute lymphoblastic leukemia (ALL). The study compared the area under the plasma asparaginase activity–time curve from time zero to infinity after a single intravenous dose. This is a comparison of drug exposure, not a direct comparison of leukemia response or other clinical outcomes.

The registry reports a geometric mean ratio of 102.8%, with a 90% confidence interval from 91.4% to 115.6%. These values summarize the estimated relative exposure and its precision. Whether that interval establishes formal equivalence depends on prespecified equivalence limits, which are not reported in the registry record.

3. Trial Design

This was a multicentre, phase II, randomized, open-label study with two parallel groups. It compared the pharmacokinetics of the liquid and lyophilized formulations of S95014 in paediatric patients with newly diagnosed ALL. The registry lists the study as completed and records enrollment of 89. The start date was 2021-05-07 and the primary completion date was 2022-05-20.

Participants were assigned to receive one of the two formulations during the induction phase. The brief description specifies a single intravenous dose. The lyophilized product was reconstituted before use; both formulation descriptions identify a 750 U/mL solution for single use. The registry does not describe masking, and records the study as unmasked. In an open-label study, treatment assignment is known rather than concealed from participants and study personnel; that design feature does not, by itself, determine whether a laboratory measurement is biased.

The primary pharmacokinetic analysis set contained 41 participants in the lyophilizate group and 39 in the liquid group. The registry defines this set as patients who received at least one dose and were evaluable for pharmacokinetic analysis. Evaluability required enough samples for interpretable results and no deviations that might affect pharmacokinetic interpretation. Enrollment and analysis-set counts refer to different trial populations, and the registry record does not explain their relationship in detail.

The registry also lists activity and immunogenicity measurements among the secondary outcome categories. Detailed results for those outcomes are not included in the registry record reported for this analysis. The study design is therefore best understood as a randomized formulation comparison centered on measured drug exposure, with additional outcome categories listed but not fully.

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: There was a sceening period (from Day-14 to Day-1) to check the eligibility of the patient in the study

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Pharmacokinetic (PK) Analysis Set included the patients who have received at least one dose of IMP and are evaluable for PK analysis : the patients who had enough samples collected to provide interpretable PK results with no deviations that might have affected the PK interpretation (e.g. infusion interrupted for any reason, deviation in the theoretical administered dose \> 10%, at least one missing PK sample during the 48 first hours, ≥ 2 missing PK samples after the 48-hour time point).

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.

Population: One patient in the S95014 liquid (LIQ) group was prematurely withdrawn from the study on the day of randomization due to a major deviation/protocol violation (patient with B-cell Burkitt Acute Lymphoblastic Leukemia) and did not receive the study treatment. Another patient who was randomized in the lyophilizate (LYO) group but received the LIQ formulation in error, was analyzed according to the treatment actually received, LIQ, and counted here under LIQ.

6. Primary Endpoint

The registry lists the primary outcome as Pharmacokinetics Measurement, measured over Predose up to 600 hours.

The primary endpoint was the area under the plasma asparaginase activity–time curve from time zero to infinity (AUCinf), measured from predose up to 600 hours. Its unit was mU*day/mL. The endpoint summarizes activity over time rather than relying on a single blood measurement. In general, an area-under-the-curve measure is used to represent cumulative exposure across the measurement period and the modeled time beyond it.

The registry reports the endpoint as a geometric mean and gives geometric coefficients of variation for each group. Geometric summaries are commonly used for pharmacokinetic measurements because such measurements can be positively skewed: a small number of relatively high values may make an arithmetic mean less representative. A geometric mean summarizes values on a multiplicative scale, while a geometric coefficient of variation describes relative variability on that scale. These are general statistical interpretations; the registry record does not give further details of the transformation or model used to produce the group summaries.

AUCinf is relevant to a formulation comparison because it captures measured asparaginase activity across time, allowing the exposure associated with each formulation to be compared. It is still a pharmacokinetic endpoint. It does not directly establish a difference in clinical benefit, disease control, or harms. Those questions require outcomes and analyses designed to assess them.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary comparison uses geometric summaries of AUCinf and an analysis identified as ANOVA. The registry reports a geometric mean ratio, a confidence interval, and a p-value, and labels the analysis as an equivalence comparison. These elements answer related but different statistical questions: the ratio gives the estimated relative exposure, the interval describes uncertainty around that estimate, and the test evaluates a specified statistical hypothesis. The registry record does not report enough detail to reconstruct every step of the model.

What does the geometric mean summarize?

A geometric mean is calculated on a multiplicative scale. In general, it is useful when measurements are positive and their variation is more naturally expressed as relative changes than as fixed additive differences. Pharmacokinetic measurements often have this pattern. The geometric mean is not the same quantity as an arithmetic average, and it should be interpreted together with the measure of relative spread. The registry gives geometric mean AUCinf values of 362000.0 mU*day/mL and 352200.0 mU*day/mL, with geometric coefficients of variation of 33.55 and 31.08.

A geometric coefficient of variation describes dispersion relative to the geometric mean rather than giving an absolute spread in the original measurement units. In general, relative dispersion can be important in pharmacokinetic data because variability may increase along with the size of the measurement. The record names this measure but does not provide further computational details.

How should the ratio be read?

A geometric mean ratio expresses one group’s geometric mean relative to the other group on a percentage scale. A value at the no-difference ratio corresponds to equal geometric means; values on opposite sides indicate different directions of the estimated relative exposure. The registry reports a ratio of 102.8% and a 90% confidence interval from 91.4% to 115.6%. The stated groups are the lyophilizate and liquid formulations. The exact model parameterization and any covariates are not described in the registry record, so the ratio should be interpreted as the reported comparison rather than as an independently reconstructed model result.

What does the confidence interval convey?

In general, a confidence interval gives a range of estimates compatible with the observed data and the assumptions of the analysis, at the stated confidence level. It conveys precision: a narrower interval indicates less uncertainty in the estimate than a wider interval, all else equal. The reported 90% interval includes relative exposures both below and above the no-difference ratio. It therefore does not establish that one formulation has greater AUCinf. Nor does an interval alone establish equivalence: a formal equivalence analysis needs prespecified acceptable limits, and the registry record does not report those limits.

Why does equivalence require a margin?

In a general equivalence design, the purpose is to assess whether the uncertainty around a treatment contrast lies within boundaries considered acceptably small. Those boundaries must be selected and specified before interpreting the results. A failure to find a conventional difference is not, by itself, proof of equivalence, because a wide interval may also be compatible with meaningful differences in either direction. The study is labelled as an equivalence analysis and reports p = 0.05, but the margin and the precise hypothesis tested are not reported. Consequently, the registry values permit description of the estimate and its uncertainty, but not independent verification that a formal equivalence rule was met.

What does ANOVA contribute?

ANOVA is a general family of methods for comparing group means. For pharmacokinetic outcomes, analyses may use transformed measurements and then express the comparison as a ratio after returning to the original scale. The registry identifies ANOVA and reports a geometric mean ratio, but does not give the transformation, model terms, assumptions checks, or handling of any covariates. Those details are not reported in the registry record. It is therefore not possible to reproduce the sponsor’s model from the listed summary statistics alone.

What is the role of the analysis population?

The primary pharmacokinetic set includes participants who received at least one dose and had enough samples for interpretable pharmacokinetic results, without deviations that might affect interpretation. This is an endpoint-specific analysis population. Such rules can be appropriate when the outcome depends on timed samples, since inadequate or affected measurements may not support a valid exposure estimate. However, restricting analysis to evaluable participants can make the results differ from an analysis of all randomized participants. The registry record does not report a separate analysis of all randomized participants or a sensitivity analysis addressing excluded or incomplete data.

Planning illustrations, if considered alongside these results, should be understood as hypothetical sample-size calculations rather than observed power. In general, such illustrations depend on assumptions about the effect to detect, allocation, significance level, and statistical approximation. They do not change the uncertainty in the completed study’s observed estimate, and they are not evidence that the trial had a particular probability of detecting an effect. The registry record does not report the trial’s own sample-size justification or detailed statistical plan.

10. Analyses Reported by the Sponsor

The registry reports the comparison as an ANOVA-based geometric mean ratio of 102.8%, with a 90% confidence interval from 91.4% to 115.6% and p = 0.05. It labels the analysis as an equivalence comparison. These are the reported analysis results and should be the reference point for describing the study’s statistical finding.

No separate unadjusted re-calculated estimate is reported in the registry record. The group-level geometric means and coefficients of variation are available, but they do not by themselves reproduce every aspect of the reported ANOVA, including any transformation or model specification. A simple re-calculation from summary values would not replace the analysis used for the trial.

The registry does not state the equivalence margins or explain the exact test associated with the p-value. Therefore, the reported estimate and interval can be quoted and interpreted as measures of relative exposure and precision, while a formal equivalence decision cannot be independently checked against a margin from the record. The reported sponsor analysis is not contradicted by a separate numerical re-calculation in the information available.

11. Multiplicity, Missing Data and Analysis Populations

The registry lists pharmacokinetics, activity, and immunogenicity among the secondary outcome categories. It also records five outcome measures and three analyses. The available information does not specify a multiplicity adjustment, a hierarchy of endpoints, or which secondary analyses were intended to support confirmatory conclusions. If several outcomes or comparisons are tested, the chance of at least one apparently notable result can increase; in general, statistical plans may address this through prespecified testing procedures. Whether and how that was done is not reported in the registry record.

The primary analysis population is defined by receipt of at least one dose and evaluable pharmacokinetic measurements: enough samples for interpretable results and no deviations that might affect interpretation. This definition makes the analysis relevant to participants with usable pharmacokinetic data, but it is not equivalent to analyzing every randomized participant regardless of evaluability. The record lists an enrollment of 89 and primary analysis counts of 41 in the lyophilizate group and 39 in the liquid group. It does not provide a full accounting of screening, randomization, treatment receipt, exclusions, or missing samples, so those counts cannot be reconciled further from the registry information.

Missing-data methods, reasons for missing or unusable measurements, and sensitivity analyses are not reported in the registry record. This matters because pharmacokinetic exposure depends on sample collection over time. If missing samples or deviations relate to participants’ exposure, the analyzed values could differ systematically from values that would have been obtained under complete measurement. The record’s evaluability rule identifies a quality condition for the primary analysis, but does not establish that missingness had no effect.

The registry record also does not report the full statistical analysis plan, the equivalence margin, or whether additional analyses were prespecified. The listed endpoint and reported model provide a basis for interpreting the posted estimate, while these omissions limit assessment of multiplicity control, robustness to missing data, and alignment with the original plan.

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: febrile neutropenia (11.1%), leukopenia (9.3%), white blood cell count decreased (9.3%), lymphocyte count decreased (8.9%), neutropenic colitis (8.9%).

The registry records serious adverse events among participants at risk in each formulation group. For the lyophilizate group, 19 of 43 participants were affected by a serious adverse event; for the liquid group, 18 of 45 were affected. The record also lists 1 death in the lyophilizate group and 2 in the liquid group. These are descriptive counts, not evidence that the formulations caused or prevented the events, and the registry information does not establish a causal attribution.

The serious adverse event terms listed most prominently include febrile neutropenia at 11.1%, leukopenia at 9.3%, white blood cell count decreased at 9.3%, lymphocyte count decreased at 8.9%, neutropenic colitis at 8.9%, and neutropenia at 4.7%. The record does not identify an arm for each listed percentage in the registry record, so they should not be assigned to one formulation. These terms may overlap within participants; the figures should not be added to infer a total number affected.

Safety counts need context such as event definitions, observation time, seriousness criteria, relatedness assessments, and complete event reporting. Those details are not reported in the registry record. The available figures support a descriptive account of recorded serious events, not a comparative safety conclusion.

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

The principal statistical result concerns pharmacokinetic exposure after a single intravenous dose during induction. The registry reports a geometric mean ratio of 102.8%, with a 90% confidence interval from 91.4% to 115.6%. The interval includes relative exposure values on both sides of the no-difference ratio. This describes the estimate’s uncertainty; it does not show that either formulation has definitively higher exposure.

The study is identified as an equivalence analysis, but a formal equivalence conclusion requires specified limits for acceptable differences. Those limits are not reported in the registry record. The p-value of 0.05 also cannot be interpreted in isolation because the exact hypothesis and testing procedure are not stated. The estimate and confidence interval are therefore more informative for describing the size and precision of the reported comparison, while the formal equivalence criterion remains unassessable from the available record.

The endpoint is a pharmacokinetic measure of plasma asparaginase activity over time. It does not directly measure clinical response, remission, survival, quality of life, or other patient-centered outcomes. The fact that a study compares exposure should not be taken as evidence that the formulations have the same clinical effects or safety profile. The activity and immunogenicity categories listed as secondary outcomes do not, by themselves, answer those questions, and detailed results for them are not reported in the registry record.

The results apply to the primary pharmacokinetic analysis set, which required sufficient interpretable samples and no deviations considered capable of affecting interpretation. The registry does not report enough information to judge how exclusions or missing measurements may have influenced the results or how well the analyzed participants represent everyone randomized.

In short, the posted data quantify a formulation comparison for AUCinf in the analyzed participants. They support a careful description of the estimated ratio and its uncertainty, but not a broader conclusion about clinical equivalence, treatment benefit, or comparative safety.

14. Why This Trial Matters Statistically

This trial illustrates why a pharmacokinetic comparison should be interpreted on the scale and for the purpose of the endpoint actually measured. AUCinf summarizes plasma asparaginase activity over time, and the geometric mean ratio expresses the relative exposure between formulations. These methods are suited to comparing exposure, but they do not turn a laboratory pharmacokinetic endpoint into a direct measure of clinical benefit.

It also demonstrates the distinction between estimating a contrast and establishing equivalence. The ratio and confidence interval communicate the magnitude and uncertainty of the observed comparison. A formal equivalence interpretation additionally depends on acceptable limits specified in advance. Without those limits in the registry record, readers can understand the reported estimate but cannot verify the decision rule that would define equivalence.

The analysis population is another important statistical feature. Timed pharmacokinetic sampling can make evaluability criteria necessary, yet conclusions based on participants with interpretable measurements may not automatically extend to every randomized participant. Reporting the number analyzed in each group is useful, but a fuller account of missing samples, exclusions, and sensitivity analyses is needed to assess potential selection effects.

Finally, open-label assignment and randomized allocation describe different design features: randomization addresses how groups are assigned, while masking concerns whether assignment is concealed after allocation. Neither feature alone settles questions about measurement quality or missing data. The registry’s account supports a focused reading of the exposure comparison while identifying details needed for a fuller assessment of the statistical evidence.

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 registry record does not report the equivalence margins. This is a central limitation for interpreting an analysis labelled as equivalence, because the acceptable range must be known to assess whether the confidence interval meets the intended criterion. The record also does not specify the exact hypothesis associated with p = 0.05.

Details of the ANOVA model are limited. The transformation, model terms, assumptions checks, covariate handling, and any sensitivity analyses are not reported in the registry record. The reported group summaries and comparison can be described, but the model cannot be fully reconstructed from those summaries alone.

The primary pharmacokinetic analysis set excludes patients without adequate interpretable samples or with deviations considered capable of affecting interpretation. The registry does not provide a complete participant flow or reasons for non-evaluability, nor does it report missing-data methods. As a result, the potential influence of incomplete sampling and exclusions cannot be assessed fully.

The record lists secondary outcome categories but does not give a multiplicity strategy, detailed results for those outcomes, or a hierarchy of statistical claims. The full statistical analysis plan and sample-size justification are also not reported. The enrollment and analysis-set counts do not provide enough information to determine how participants moved through each stage of the study.

Finally, the primary endpoint is pharmacokinetic rather than clinical. The registry information described does not establish how differences in AUCinf relate to clinical outcomes, and the serious adverse event counts are descriptive without detailed attribution or follow-up context. These limitations narrow the conclusion to the reported exposure comparison.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study Comparing the Blood Levels of Both Pegaspargase (S95014) Formulations (Liquid vs Lyophilized) in the Treatment of Paediatric Patients With Acute Lymphoblastic Leukemia (ALL) — NCT04954326.
• Registry record last updated 2023-07-12; status last verified 2023-07.
• Results first posted 2023-07-05.
• 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.

19. Record Summary

NCT04954326 was a completed, multicentre, phase II randomized, open-label, parallel-group study comparing liquid and lyophilized S95014 in paediatric patients with newly diagnosed ALL. The registry lists enrollment of 89 and a single intravenous dose during induction.

The primary endpoint was AUCinf for plasma asparaginase activity, measured from predose up to 600 hours. The pharmacokinetic analysis set included 41 participants receiving lyophilizate and 39 receiving liquid. The reported geometric mean ratio was 102.8%, with a 90% confidence interval from 91.4% to 115.6%; the analysis was labelled as equivalence.

The registry record does not report the equivalence margin, complete model details, missing-data methods, or multiplicity strategy. The findings describe pharmacokinetic exposure in the analyzed population and do not, on their own, establish clinical equivalence or comparative safety.