P-225: Uninvolved hypogammaglobulinemia was not a predictor of the infection risk in anti-CD38 chemotherapy

Introduction: Infections are major causes of morbidity with myeloma treatment. Anti-CD38 antibodies are important additions to myeloma chemotherapy. Through their effect on normal plasma cells, they can reduce the levels of uninvolved immunoglobulins (Ig). We sought to investigate if there was any relationship between the changes in the uninvolved Ig levels during anti-CD38 therapy and the risk of developing infections. We also compared the levels of Ig across different anti-CD38 therapies at different time points.

Methods: This is a retrospective single-centre analysis of myeloma patients treated with anti-CD38 therapy (single agent or in combination). The levels of the uninvolved Ig and lymphocytes were recorded at baseline and at three-monthly intervals in addition to the number of infections requiring inpatient treatment for up to 18 chemotherapy cycles. Patients who had less than 3 months of treatment and those in the induction phase were excluded.

Results: Between March 2018 and March 2023, 108 patients were treated with Isatuximab-Pomalidomide-Dexamethasone (IPD), Daratumumab-Bortezomib-Dexamethasone (DVd) and Daratumumab monotherapy (n=28, 49 and 32 respectively). The median age at the start of treatment was 73 (range 41 to 90). The median number of treatment cycles was 11 (range 3 to 45). Thirty-two patients (29%) had at least one infective episode requiring hospitalisation during the course of treatment. Of those, 25% had Daratumumab monotherapy and the rest were equally divided between the IPD and the DVd groups (37.5% each). The median uninvolved Ig levels were subnormal at all the time points assessed within the three regimens. There was no significant change in the uninvolved IgG and IgM levels in the infected group at all time points. The non-infected group showed a significant drop in the uninvolved IgM and IgG levels at the 3 and 6-month time points after which the levels were not significantly different from the baseline. There was a significant drop in the uninvolved IgA levels which persisted at all the time points in both the infected and the non-infected groups. The lymphocyte count changes were not significant in both groups at all time points. The analysis of the uninvolved Ig changes subgrouped by the treatment regimen showed a significant and maintained drop in the IgA levels with Daratumumab monotherapy and DVd and a significant drop in the IgM levels at the early time points with DVd. Comparing the levels of the Ig isotypes across regimens revealed no significant changes attributable to the underlying chemotherapy treatment.

Conclusions: Uninvolved hypogammaglobulinemia was universal with various anti-CD38 regimens but did not predict infection risk. Uninvolved Ig reduction was most pronounced with IgA but did not correlate with the development of infections. We conclude that the uninvolved hypogammaglobulinemia is a poor predictor of the infection risk with anti-CD38 therapy.