Page 227 - Haematologica Vol. 109 - July 2024

ARTICLE - ICARIA-MM final overall survival analysis
P.G. Richardson et al.
risk reduction in HR versus control, corresponding to an improvement in the true median PFS time from 4.0 to 6.67 months; and a log-rank test at a 1-sided 2.5% signif- icance level, a total of 162 PFS events were calculated to be needed to achieve 90% power for the study. Based on these assumptions, 220 deaths were needed to achieve 80% power for the study.
For the final analysis, a data cutoff date of March 14, 2022, was selected, and data were included up to this date (last patient last visit). Per protocol, the cutoff date for the final OS analysis was to occur when 220 OS events had been observed. The 220th OS event occurred on January 27, 2022, and this date was considered to be the OS cutoff date for the primary OS analysis.
Efficacy analyses were conducted using the inten- tion-to-treat (ITT) population, defined as patients who provided signed informed consent and were allocated a randomization number by the IRT. Safety was assessed in all patients from the ITT population who received at least a partial dose of study treatment. Patients were considered lost to follow-up if the last contact was ≥8 weeks prior to the data cutoff.
OS was analyzed using the Kaplan-Meier method and corresponding 95% CI, which was calculated with log-log transformation of survival function and the method devised by Brookmeyer and Crowley. HR and corresponding 95% CI are calculated from a Cox proportional hazards model, stratified by age and number of previous lines of therapy. Prespecified analyses were completed for updated PFS by investigator assessment (sensitivity analysis), PFS on subsequent therapy or death, and PFS on the first line of subsequent therapy (separate summaries for subsequent therapy with/without daratumumab-based therapy), and TTNT using OS analysis methods. P values for exploratory endpoints are provided for descriptive purposes only. Addi- tional statistical methods have been described previously.14,16 All statistical analyses were performed with SAS (version 9.4; SAS Institute, Cary, NC) or R (version 3.4.3; R Foundation, Vienna, Austria). Since the updated analysis, no additional amendments to the protocol have been made. This study is registered with clinicaltrials gov. Identifier: NCT02990338.
Results
Between January 10, 2017, and February 1, 2018, 387 pa- tients were screened and 307 were randomized (Isa-Pd, 154; control, 153) at 102 sites in 24 countries. Cutoff for the first OS interim analysis, at the time of the primary PFS analysis, was October 11, 2018 (reported previously).13 Cutoff for the preplanned second interim analysis was October 1, 2020 (reported previously).14 Cutoff for the final OS analysis was January 27, 2022, per protocol (Figure 1). Data cutoff for last patient last visit was March 14, 2022.
Baseline characteristics were similar in both arms, as pre-
viously published and briefly shown in Table 1. All patients previously received PI and IMiD agents.
Median treatment duration was 11.0 (range, 2.6-12.4) months among patients receiving Isa-Pd and 5.5 (range, 4.4-21.8) months among patients receiving Pd. At data cutoff (March 14, 2022), 16 (10.4%) patients receiving Isa-Pd and three (2.0%) patients receiving Pd remained on treatment. The most frequent reason for definitive discontinuation was progressive disease (isatuximab, 65.6%; control, 76.5%). After a median follow-up of 52.4 months, median PFS per investigator assessment (ignoring symptomatic deteriora- tion) showed consistent improvement with longer follow-up (Isa-Pd: 11.1 months, 95% CI: 7.8-13.8; Pd: 5.9 months, 95% CI: 4.5-7.9; HR=0.57, 95% CI: 0.44-0.73; 1-sided P<0.0001; Online Supplementary Figure S1).
The prespecified required number of 220 OS events oc- curred on January 27, 2022, which was the analysis cut- off date. Of the 220 deaths, 106 (68.8%) occurred in the Isa-Pd arm and 114 (74.5%) in the Pd arm. Median OS was 24.6 months (95% CI: 20.3-31.3) in the Isa-Pd group and 17.7 months (95% CI: 14.4-26.2) in the Pd group (HR=0.78, 95% CI: 0.59-1.02; 1-sided P=0.0319; Figure 2). Early separation was observed in the OS curves between arms; however, because the 1-sided P value for statistical significance was set to 0.02 based on the previous interim analysis, the current analysis did not cross the level of significance. Censored patients (Isa-Pd: 48/154 [31%]; Pd: 39/153 [26%]) remained alive at data cutoff (Isa-Pd: 39/48 [81%]; Pd: 31/39 [80%]), were alive at the last contact before the cutoff date (1/48 [2%]; 0/39), or were lost to follow-up (8/48 [17%]; 8/39 [21%]). Subgroup analyses of OS are shown in Online Supplementary Figure S2.
Among patients receiving subsequent anti-myeloma ther- apy, daratumumab was given to 23 (23%) of 102 patients in the Isa-Pd group and 71 (60%) of 119 patients in the Pd group (Online Supplementary Table S3). In order to estimate the treatment effect in the absence of a switch to subsequent anti-cancer therapy with daratumumab, sensitivity analy- ses using the RPSFT model were performed, with overall similar results (HR=0.706, 95% CI: 0.538-0.926) to the ITT estimate, in favor of the Isa-Pd arm (HR=0.776, 95% CI: 0.594-1.015; Online Supplementary Appendix, page 17). An OS sensitivity analysis was conducted to assess the impact of death due to coronavirus disease 2019 infection, with similar results to the ITT estimate (Online Supplementary Appendix, page 18).
Overall, 102 (66%) of 154 patients in the Isa-Pd group and 119 (78%) of 153 patients in the Pd group received sub- sequent anti-myeloma therapy. Median TTNT was longer with isatuximab (15.5 months, 95% CI: 12.1-19.8) versus Pd (8.9 months, 95% CI: 6.3-11.5; 1-sided P<0.0001; Figure 3). Median PFS on subsequent therapy or death (PFS2) was longer in the Isa-Pd group (17.5 months, 95% CI: 14.9-19.2) versus Pd (12.9 months, 95% CI: 10.1-16.6; 1-sided P=0.0091; Figure 2).
Haematologica | 109 July 2024
2241