Dexamethasone dose intensity does not improve 6-month responses in newly diagnosed multiple myeloma and may contribute to hyperglycemia Free

Introduction The role of dexamethasone in newly diagnosed multiple myeloma (NDMM) is increasingly being challenged given that higher dose intensities during induction do not improve PFS or OS (Banerjee et al, Blood 2025) and appear unnecessary to achieve deep responses (Askeland et al, Lancet Haem 2025). The impact of weekly dexamethasone (dex) on cycle-to-cycle responses during induction, which may be clinically meaningful to patients, is not understood. Similarly, the impact of dex on glycemic control has not been analyzed in NDMM.

Methods We conducted a single-center retrospective analysis of patients with NDMM who underwent induction between April 2021 and April 2024. Patients were grouped by their exposure to dex 40 mg weekly or more: those who never received this dose (DEX₀), those who started dex 40 mg weekly but dose-reduced or stopped during Cycles 1-2 (DEX_1-2), those who dose-reduced or stopped during Cycles 3-4 (DEX_1-4), and those who remained on dex 40mg weekly for 6 cycles (DEX_1-6). Responses (≥VGPR) at 2, 4, and 6 months were compared using Chi-squared tests. Sensitivity analyses including types of induction regimens (doublet, triplet, quadruplet) were performed. Separately, we analyzed mean blood glucoses (MBG in mg/dL, not necessarily fasting) from any metabolic panels at pre-treatment (up to 3 months prior), 0-2 months, 2-4 months, and 4-6 months after treatment initiation. When available, pre- and post-treatment hemoglobin A1c were compared.

Results We analyzed 264 patients (median age 65), of whom 37% (n=98) received Dara-VRd, 21% (n=57) VRd, and 9% (n=26) Dara-Rd. This population included 40% DEX₀ (n=105), 15% DEX_1-2 (n=40), 16% DEX_1-4 (n=42), and 29% DEX_1-6 (n=77). Of note, DEX_1-6 comprised 1% of Dara-Rd recipients, 36% of VRd recipients, and 42% of Dara-VRd recipients. At month 2, DEX_1-2 / DEX_1-4 / DEX_1-6 were pooled as they all had similar dex dosing at this time point. Rates of ≥VGPR were higher for this pooled group than for DEX₀ (47% vs 34%, p=0.034); however, there were no other significant intergroup differences at later timepoints. The 6-month ≥VGPR rate was 82% in the DEX_1-6 group versus 75% in the other groups (p=0.2). As expected, there was a strong association between induction regimen and response: 6-month ≥VGPR rates of 83% for quadruplets versus 73% for triplets and 29% for doublets (p=0.003). However, induction type did not alter the lack of significant association between dex dose intensity and response. Multivariate analyses for 6-month ≥VGPR comparing DEX_1-6 versus DEX₀ showed an odds ratio (OR) of 1.8 (95% CI 0.87-3.91, p=0.12). When DEX_1-2 and DEX_1-4 were pooled, the corresponding OR versus DEX₀ was 1.58 (95% CI 0.79-3.26, p=0.2).

MBG in the 0-2 month, 2-4 month, 4-6 month periods (114, 112, and 108, respectively) were all significantly elevated compared to the pre-treatment phase (100; p < 0.01). Analysis of pre- and post-treatment A1c (available for 22% of patients, n=58) showed a significant increase in post-treatment values compared to pre-treatment (6.15% vs 5.8%, p=0.03). Interestingly, MBG from the DEX_1-6 group was not statistically significant compared to MBG of other groups at all time points.

ConclusionIn our real-world analysis of over 250 patients, fewer than half maintained dex 40 mg weekly for all of induction, even among those receiving quadruplet therapy. While the 2-month ≥VGPR rate was higher in patients who initially started dex 40 mg weekly versus those who did not, dex dose reductions did not negatively affect 6-month ≥VGPR rates. Treatment initiation coincided with increases in MBG and A1c values, which may represent dex toxicity or physiologic stress. Contrary to our hypothesis, we did not see a dose-dependent increase in MBG based on dex doses. This may be due to the limited physiologic effects of corticosteroids here or, alternatively, increased glycemic control efforts in the DEX_1-6 group. Limitations of our study include induction regimen heterogeneity and missing A1c values; however, MBG was routinely available for all patients and likely would have similar fasting versus non-fasting distributions regardless of dex dose. Overall, our findings reinforce prior research demonstrating that dex dose reductions do not worsen outcomes in MM. While we could not demonstrate a clear dose-dependent relationship with hyperglycemia, dex likely partially explains the clear rise in MBG following NDMM treatment.