Imaging, QoL and Patient-Reported Outcome and Supportive Care
Hira Mian, MD (she/her/hers)
Associate Professor
McMaster University, Hamilton, ON, Canada
Hamilton, Ontario, Canada
MAIA and CEPHEUS showed daratumumab (Dara) plus standard of care improved treatment (tx) outcomes, including PFS, in nontransplanted patients (pts) with NDMM, regardless of baseline (BL) frailty. Recent studies suggest that frailty is a dynamic state and that the degree of frailty over time appears to influence PFS more than age or Charlson Comorbidity Index score. Further understanding of changes in pts’ frailty over their tx course may have important clinical considerations in tx delivery based on each pt’s changing fitness status. For the first time, we assess outcomes based on dynamic frailty (post hoc) in the phase 3 MAIA and CEPHEUS trials (TIE pts).
Methods: Pts were randomized 1:1 to receive DRd/Rd (MAIA) and DVRd/VRd (CEPHEUS). Frailty (IFM simplified frailty score; nonfrail, score 0/1; frail, score ≥2; ultrafrail, score ≥3) was assessed in MAIA and CEPHEUS TIE pts at BL and year (y) 1, 2, 3, and 4. PFS and overall MRD negativity (neg) (MRD-neg [10-5] with ≥CR) rate were assessed.
Results:
Median follow-up was 64.5 months (mo) in MAIA (median age 73 [range 45–90] y; 122 [17%] pts ECOG PS ≥2) and 58.7 mo in CEPHEUS (median age 70 [31–80] y; 37 [9%] pts ECOG PS 2). At BL, 172/368 (47%) pts assigned to DRd and 169/369 (46%) to Rd in MAIA and 48/144 (33%) TIE pts assigned to DVRd and 35/145 (24%) to VRd in CEPHEUS were frail; ultrafrail pts accounted for 42% and 43% of frail pts in the DRd and Rd arms of MAIA, respectively.
In pts with evaluable post-BL scores, frail and nonfrail status changed at y1 compared with BL in 136/658 (21%) pts in MAIA (67 [10%] improved; 69 [10%] deteriorated) and 50/261 (19%) in CEPHEUS (19 [7%] improved; 31 [12%] deteriorated). Frailty status changed at y4 in 89/354 (25%) pts in MAIA (18 [5%] improved; 71 [20%] deteriorated) and 63/188 (34%) in CEPHEUS (10 [5%] improved; 53 [28%] deteriorated).
In nonfrail pts at landmark times (1, 2, 3, and 4 y post BL), MRD-neg rates were improved with DRd (41–55%) vs Rd (16–28%) in MAIA and with DVRd (68–80%) vs VRd (47–59%) in CEPHEUS. MRD-neg rates in frail pts were also higher with DRd (26–43%) vs Rd (9–15%) and with DVRd (54–60%) vs VRd (33–44%) and in ultrafrail pts in MAIA with DRd (28–45%) vs Rd (8–20%).
Dara consistently improved PFS in both trials across frailty groups at landmark times. In nonfrail pts, DRd vs Rd led to a 34–50% decrease in the risk of disease progression/death and DVRd vs VRd to a 54–69% decrease. In frail pts, DRd reduced risk by 46–58% vs Rd, and DVRd by 51–65% vs VRd. In ultrafrail pts in MAIA, DRd reduced risk by 56–72%.
Future analyses will assess safety across dynamic frailty subgroups.
Conclusions: Our analyses demonstrate that frailty status is dynamic. Moreover, the changing degree of frailty over time influenced PFS, and D(V)Rd vs (V)Rd improved both MRD-neg rate and PFS in frail and nonfrail pts at landmark times. Furthermore, our MAIA analyses include the first phase 3 dynamic frailty data in ultrafrail pts, in whom DRd vs Rd provided clinical benefit.