BCMA-Directed mRNA CAR T Cell Therapy for Myasthenia Gravis Trial Results

Okay, here’s a breakdown of the statistical analysis methods used in this study, extracted from the provided text. I’ve organized it for clarity:

1. Sample Size & Power calculation:

* Target: 15 participants per treatment group.
* Power: 80% power to detect a difference of 47% in responders.
* Expected Response Rates: 87% in the Descartes-08 group, 40% in the placebo group.
* Test: Two-sample proportion test for self-reliant samples.
* Significance Level: 0.05 (non-directional).
* Justification: Based on Phase 1b/2a Descartes-08 data and ancient placebo controls. Verified with Monte Carlo simulations (100,000 trials) to assess expected proportions and standard deviations in the placebo group across diffrent cutoff values for MG-ADL, MGC, QMG, and MG-QoL-15r scores.
* Enrollment: Up to 50 participants to account for potential dropouts.

2. Populations for Analysis:

* Primary Analysis Population: Modified Intention-to-Treat (mITT) – all participants enrolled at an academic medical center with at least one post-baseline follow-up.
* All Enrolled Participants: Used for baseline demographics and safety endpoint analysis.

3. Descriptive Statistics:

* Categorical Variables: Percentages.
* Continuous Variables: Means and standard deviations (SD) or medians and ranges (used for skewed distributions – mean-to-SD ratio < 2).

4. Primary Efficacy Analysis:

* Endpoint: Response status at month 3 (day 85) – defined as a ≥5 point reduction in MGC total score from baseline.
* Test: Two-independent-sample proportion test (two-sided, 5% significance level).
* Test Statistic: Wald chi-squared test to assess association between treatment group and response status.
* Missing Data Handling:

* Participants requiring rescue medication before day 85: MGC scores set to missing and imputed using Baseline Observation Carried Forward (BOCF), and classified as non-responders.
* Other missing data: Multiple imputation (assuming data are missing at random). (Note: the text states this wasn’t required for the mITT population, as data were complete).
* Reporting: Percentage of responders, difference in proportions, and Wald 95% Confidence interval (CI).

5. Multiple imputation Details:

* Method: Fully conditional specification.
* Variables Included in Imputation Model: sex,age,race,treatment group,and baseline scores.
* Number of Datasets: 50 complete datasets generated.
* Software: PROC MI and PROC MIANALYZE (SAS).

6. Subgroup Analyses:

* Subgroups Investigated:

* Anti-AChR antibody-positive vs.negative.
* Early-onset vs. late-onset disease.
* Prior exposure to complement and FcRn inhibitors vs. no prior exposure.
* Development of fever vs. no fever.
* Analysis within each subgroup: Differences in proportions of responders were determined within each treatment group.
* Test: Wald chi-squared test for association between response status and subgroup (performed separately within each treatment group).

7. Software Used:

* SAS (version 9.2 or higher)
* Mathematica (version 11.0 or higher)

Let me know if you’d like me to elaborate on any specific aspect of these methods!