Abnormal Serum Fixation Patterns Are a Common Event Following BCMA CAR-T Therapy in Relapsed Refractory Multiple Myeloma and Are Often Preceded By Immunological Events

Introduction

Abnormal serum immunofixation pattern (ASIP), defined as the appearance of a monoclonal band with a different isotype than that of the original multiple myeloma (MM) isotype, is seen in 30-60% of MM patients who undergo autologous stem cell transplantation (ASCT) and other treatments such as immunomodulators or proteasome inhibitors. Importantly, the presence of ASIP has been associated with improved outcomes following ASCT. We describe the presence of ASIP in MM patients after BCMA CAR-T-cell therapy, which has not yet been characterized, as well as the relationship to immunologic events and hypogammaglobulinemia.

Methods

Patients with relapsed MM who received BCMA CAR-T at two institutions were included. ASIP was defined as a new serum monoclonal band on serum immunofixation different than the underlying MM, including those with free light chains of a new isotype, or a biclonal band on immunofixation with either the same or different isotype as the underlying MM, indicating a new clone. Patients with ASIP after progression of disease were excluded for analysis due to new clones likely being a result of evolving disease and not therapy. Clinical characteristics, serum protein electrophoresis/immunofixation (SPEP/sIFE), quantitative immunoglobulins, presence of immune-related events such as vaccinations, infections, and outcomes at the time of ASIP all after CAR-T therapy were collected.

Results

Fifty-eight patients were included (15 idecabtagene vicleucel and 43 ciltacabtagene autoleucel). ASIP was seen in 12 patients (21%) before the progression of the disease, including 1 patient with 2 and 2 patients with 3 independent ASIP events. There were an additional 7 patients with ASIP after progression of disease. The median time to ASIP was 9 months and lasted a median of 2 consecutive SPEP/sIFE assessments (range 1-12). Seven out of 12 (58%) of ASIP were preceded within 60 days by an immunological event (6 infections, 2 vaccinations). All ASIP cases occurred after achieving the patient’s best response (11 patients CR, 1 pts VGPR). Patients with ASIP had longer PFS when compared to those without pre-progression ASIP (median not reached vs 20.9 months), however this was not statistically significant (p = 0.056).

There were no differences in the proportion of patients who had an immunological event (58% vs 74%, p = 0.3) or the median number of events (1 (1 – 2) vs 2 (1 – 3), p =0.3) in those with ASIP vs no ASIP. Similarly, there were no differences in the proportion or median number of infectious events (any infection: 58% vs 59%, p > 0.9, median 1 (1 – 1) vs 2 (1 – 2) p= 0.074) or vaccination events (any vaccination: 25% vs 31%, p > 0.9, median 1 (1 – 1) vs 2 (1 – 2), p 0.14).

When only including patients with more than 6 months of follow-up at the time of ASIP, 83%, 60%, and 73% of patients had serum IgA, IgG, and IgM below the lower limit of normal (LLN). However, when comparing to patients without ASIP, those with ASIP had a higher median IgG (576 vs 443 mg/dL, p = 0.017) and IgM (27 vs 10 mg/dL, p = 0.005), whereas IgA was higher but not significant (17 vs 5 mg/dL, p = 0.12). Throughout their course, patients with ASIP had lower rates of immunoglobulins below the LLN (IgG: 60% vs 88% , IgA: 83% vs 100%, IgM: 73% vs 100%, percentage represent proportion of immunoglobulin < LLN out of all measurments for each specific immunoglobulin). Additionally, patients with ASIP had more SPEP/sIFE assessments total (7 vs 21 < 0.1).

Discussion

ASIP was a frequent event following BCMA CAR-T, with 19 (33%) ASIP events in the cohort, including 12 (21%) with ASIP pre-progression. ASIP uniformly occurred after achieving a deep response (≥ VGPR) and was often associated with hypogammaglobulinemia at the time of ASIP appearance. However, patients with ASIP had lower rates of hypogammaglobulinemia throughout their course, along with a similar rate of immunological events. This may suggest that ASIP events occur in patients who have a deeper response and a relatively more preserved residual non-neoplastic plasma cell compartment in response to an immunological stimulus. The effects of ASIP on the loss of CAR-T on-target effect or outcomes cannot be assessed due to the sample size, but this represents an important next step given ASIP being associated with better outcomes in other treatment modalities. Association of ASIP with PFS is confounded by the selection bias introduced by ASIP, and landmark analysis should be considered for CAR-T and other settings.