A National Platform and Scoring System Allocates CAR-T Treatment Slots for Multiple Myeloma to Patients with High Likelihood of Complete Remission

Introduction

Therapy of relapsed and refractory multiple myeloma (MM) with Ide-Cel is superior to alternative treatments, but availability is lower than the number of patients in need. Treatment slots are distributed by the manufacturer per country. Their allocation to individual patients is a medical, logistical and ethical challenge. Ideally, such allocation should be inclusive, fair, transparent, effective, and base on prospectively confirmed objective outcome criteria.

Methods

Swiss Blood Stem Cell Transplantation and Cellular Therapy (SBST) has established the national SBST-CAR-T MM board for allocation of Ide-Cel treatment slots. Allocation process and triage criteria were developed by clinical experts in exchange with the ethical board of the Swiss academy for medical sciences and health insurers, approved by all JACIE-accredited national centers for cellular therapies, and agreed that Ide-Cel therapy in Switzerland is only available through this allocation system. Patients are registered in parallel with request for insurance coverage in one key document. Patients had to meet qualifying approval criteria for Ide-Cel in Switzerland (relapsed and refractory MM, >2 prior therapy lines, triple-class-exposed).

Reaching CR after Ide-Cel is associated with superior PFS. We therefore developed an algorithm to select patients with high likelihood for CR. Patient factors negatively associated with CR derived from the KarMMa trial were complemented by factors prospectively associated with inferior overall survival after CAR-T in malignant lymphoma. High tumor burden (bone marrow infiltration > 50% or high tumor load by PET-CT), inflammatory activity (increasede ferritin or D-dimers), ECOG patient status and comorbidities were used as negative predictive factors and combined to a sum score of 0-4. All registered patients are presented at the monthly virtual meeting and ranked according to this score. Ranking is refined based on availability of alternative treatment options, resulting in a final national ranking agreed by all board members. Because no patients with severe comorbidities were registered, this dimension was later excluded.

Results

Between May 2022 and April 2023 all 48 Ide-Cel regular treatment slots in Switzerland were allocated in 12 virtual meetings lasting median 55 minutes (25-70) with the participation of median 7 centers (6-8), 8 physicians (6-15), discussing in total 72 patients (median 5.5 per meeting, range 4-10) to distribute 4 slots monthly. All therapy slots were successfully allocated, including short-notice changes in availability of slots or patients requiring immediate allocation of substituting patients. 74.2% of all patients registered with the SBST CART MM board received Ide-Cel. Patient, treatment and outcome parameters were centrally collected to SBST database from 37 treated patients per April 2023. (1 patient each had cell collection failure, Ide-Cel production failure, missing outcome data). Patients were 65% triple refractory, 65 % penta refractory, 14% received prior BCMA-targeted therapy, 8% prior bispecifc antibodies and 67% had high risk cytogenetics. Best response achieved so far was CR or better in 57% of patients, PR-VGPR 16%, PD 13%, not stated/too early 13%. A low SBST-MM CAR-T Score was prospectively associated with CR (table 1, p=0.048). Each scoring dimension (ECOG, tumor load, inflammation) significantly correlated inversely with CR, while conventional characteristics for poor MM outcomes (HR CG, penta refractory disease) did not correlate. PFS and OS data is pending. Cross cohort comparison with the Real-World Experience from the US Myeloma CAR-T Consortium (Hansen et al., JCO 2023) revealed a strong trend towards higher CR rates (57% vs. 23%, p=0.10).

Conclusion

This algorithm prospectively selects MM patients for BCMA-CAR-T treatment to maximize CR rate. Also, this national real world cohort demonstrates the feasibility of triaging the allocation of

CAR-T therapy slots based on a, fair, objective, transparent, consensual algorithm using predefined criteria. The SBST scoring system prospectively reflects differential likelyhood of CR (100>64>53>33%), and the scoring dimensions (ECOG, tumor load, inflammation) prospectively correlate with a CR rate of 57 %, supporting the selection process. The SBST CAR-T scoring system may be used to allocate CAR-T treatment slots to maximize benefit from this treatment.