Steroid-toxicity in newly diagnosed patients with multiple myeloma treated with a limited dexamethasone regimen; Results from the nmsg rest study

Abstract Introduction: For decades corticosteroids (steroids) have been essential in treating patients with multiple myeloma (MM). However, the role of steroids in the era of novel therapies is uncertain. Considering the unfavourable toxicity profile of steroids, the phase 2 REST study investigated the efficacy of a quadruplet regimen with limited steroids (dexamethasone) in patients with newly diagnosed MM who were ineligible for autologous stem cell transplantation. Patients were treated with isatuximab-bortezomib-lenalidomide-dexamethasone; dexamethasone was omitted after two cycles, bortezomib was omitted after eight cycles and isatuximab was omitted after 18 cycles. Efficacy was comparable with full-dose steroid regimens. This study aimed to investigate the impact of omitting dexamethasone on patient-reported steroid toxicity. Methods: Patients completed two patient-reported outcomes questionnaires: the Steroid-Symptom Questionnaire for patients with MM (SSQ-MM) and the European Organization for Research and Treatment of Cancer (EORTC) Quality of life – Core 30 (QLQ-C30) at day 1 and 22 of cycles 1 and 2 (with dexamethasone) and cycles 4 and 5 (without dexamethasone). Frailty was assessed for each patient using the patient-reported frailty phenotype based on responses to five of the EORTC QLQ-C30 items. The SSQ-MM Total score calculated by the patients´ responses to 19 of the 20 items of SSQ-MM assessed steroid toxicity. Mean change from baseline and mean score differences between time points with dexamethasone compared to without dexamethasone were estimated using linear mixed model of repeated measures (LMRM). Statistically significant estimates (p-value <0.05) were interpreted using Benjamini-Hoechberg procedure to avoid type I errors and minimal important difference to ensure clinical relevance. The proportion of patients developing steroid toxicity during cycles with and without dexamethasone were compared using mixed effect logistic regression, and statistical significance was assessed with a p-value of <0.05. The impact of steroid toxicity on other health-related quality of life (HRQL) domains was investigated with LMRM including a Gaussian random intercept for each patient with adjustments using Benjamini-Hoechberg procedure. Results: 51 patients were included with a median age of 77 years (range 70-88). 16 patients (31%) were 80 years or older. Eight patients (16%) were ECOG 2 or more and 21 patients (41%) were frail according to the patient-reported frailty phenotype. Questionnaire completion rate was 88%. Assessed by SSQ-MM Total, we found an overall impact of dexamethasone on steroid toxicity as the overall p-value of <0.001 was significant after adjustment using Benjamini-Hoechberg. Mean score differences in steroid toxicity between time points with dexamethasone compared to time points without dexamethasone were not statistically significant (Cycle 1 day 22 (C1D22) compared to C4D1; p-value 0.07, C2D1 compared to C5D1; p-value 0.25 and C2D22 compared to C5D22; p-value 0.21). At C1D22 and C2D22, patients reported clinically meaningful greater steroid toxicity compared to baseline (p-values <0.01), and sensitivity analysis revealed significantly more patients reported steroid toxicity at C1D22 (16 out of 38) compared to C4D22 (2 out of 29, p-value 0.004). Steroid toxicity was associated with impaired physical, emotional and social functioning and greater fatigue, pain, appetite loss and insomnia. Conclusions: Steroids are included in most myeloma-targeted therapies. However, growing evidence shows similar disease responses with steroid-limited treatments. Overall, we found a statistical significant difference in steroid toxicity during cycles with dexamethasone compared to cycles without dexamethasone. A clinically meaningful increase in steroid toxicity from C1D1 (baseline) to C1D22 (also confirmed with sensitivity analysis) and C2D22 was observed. Steroid toxicity also negatively impacted several HRQL domains.