Outcomes of patients with mayo stage IIIb systemic AL amyloidosis treated with daratumumab-based frontline therapy

Abstract Introduction Patients with Mayo stage IIIb AL amyloidosis (defined as NT-proBNP8500 pg/mL and high-sensitivity troponin-T hs-TnT≥50 ng/L) face dismal prognosis and are commonly excluded from clinical trials. The prospective phase 2 EMN study in patients with stage IIIb demonstrated a heme-CR rate of 27.5% with single-agent daratumumab, with an estimated overall survival (OS) at 1 year being 45%. While the pivotal ANDROMEDA trial excluded these patients, prior real-world data suggest that daratumumab plus-VCd (Dara-VCd) is feasible and safe in these patients and induces deep hematologic responses. We report hematologic and cardiac responses at long-term follow-up and overall survival (OS) in 36 consecutive patients with Mayo stage IIIb, treated with frontline Dara-VCd/Dara-Vd. Methods Clinical and treatment data were extracted retrospectively. Hematologic and organ responses were assessed by consensus guidelines and reported on an intention-to-treat basis. OS was analyzed using Kaplan-Meier method, and predictors of OS were evaluated using univariate Cox regression.Results 36 patients (median age 63.4 years) with Mayo 2004 Stage IIIb were included in this study. Median dFLC (difference between involved and uninvolved free light chain) was 65.3 mg/dL (range, 9.6-906.9), median NT-proBNP was 13,044 pg/mL (range, 8565-70,000), and median hs-TnT was 119 ng/L (range, 41-3119). Among 35 patients with available data, 25 (72%) had NYHA class III/IV at presentation. The median interventricular septal thickness was 1.5 cm, and 21 patients (58.3%) had left ventricular ejection fraction less than 50%. Kidney involvement was present in 22 patients (61.1%), with a median 24-hour urine protein of 2.7 g (range, 0.2-28.1), median eGFR of 42 mL/min/1.73 m2; 13.6% required renal replacement therapy at diagnosis. First-line therapy was Dara-VCd in 33 patients and Dara-Vd in 3 patients. No patient received frontline auto-transplant. Over a median follow-up of 38.6 months (95% CI, 26.6-58.4), the best hematologic responses were complete response (heme-CR) in 17 (47.2%), very good partial response (VGPR) in 5 (13.9%), PR in 9 (25%), and no response (NR) in 5 (13.9%). VGPR or better was achieved in 35.5%, 54.5%, and 52.9% at 1, 3 and 6 months respectively. One patient received a heart transplant within 3 months. Of 35 evaluable patients, cardiac response at 3, 6, and 12 months was 38.7%, 38.7%, and 42.4% respectively. At best cardiac response: 11.4 % achieved cardiac CR, 28.6% VGPR, 2.8% PR, and cardiac NR/progression was seen in 57.1%. The median time to best cardiac response was 26.3 months (range, 3.8-63.1). Over a median follow up of 38.6 months, 23/36 patients died. The median OS was 16.3 months, with estimated OS at 1, 2, and 3 years being 55.6%, 46.7%, and 35.2% respectively. NYHA class at diagnosis was the only factor significantly associated with worse OS HR for NYHA III vs II: 5.4, p=0.028; HR for NYHA IV vs II: 12.0; p=0.002; HR for NYHA IV vs III: 2.2; p=0.075. The estimated 2-year OS in patients with NYHA II was 90.0% vs 53.3% for NYHA class III and 27.3% for NYHA class IV (p=0.0007). Notably, concurrent kidney involvement was significantly associated with a superior OS HR: 0.04; p=0.036, indicating potential overestimation of NT-proBNP due to reduced eGFR in this subgroup of patients. Second line clone-directed therapy was received by 41.7% of patients due to suboptimal response to Dara-VCd/Dara-Vd (n=9) or hematologic progression/relapse (n=6).Conclusions The heme-CR rate with Dara-VCd/Dara-Vd in patients with stage IIIb compare favorably approximately two-fold increase to single-agent daratumumab. However, 3-year survival remains limited. Importantly, NYHA class I/II subgroup have favorable OS despite high NT-proBNP - supporting their inclusion in clinical trials. These findings underscore the need for prospective trials of daratumumab-based combination regimens and anti-amyloid strategies targeting this high-risk population.