Acute myeloid leukemia (AML) after myeloproliferative neoplasm (MPN): Real-world outcomes in the new AML treatment era Free

MPNs have a variable risk of progression to AML. With a median overall survival (OS) of 3-6 months (ms), patients (pts) with post-MPN AML are often excluded from clinical trials and have limited treatment options. Moreover, real-world data is scarce due to the rarity of the disease. We conducted a retrospective analysis to assess the impact of novel AML therapies, including venetoclax (VEN)-based regimens, on treatment patterns and outcomes in post-MPN AML.

We used the Flatiron Health Research Database, an electronic health record-derived deidentified database from ~280 oncology practices in the US, to select post-MPN AML pts diagnosed in 2014-2024, who received treatment within 6 ms of diagnosis. Treatments were classified as intensive chemotherapy (IC; cytarabine + anthracycline-based) or lower-intensity therapy (LIT; hypomethylating agents [HMA] as monotherapy [mono] or in combinations including VEN, ruxolitinib, and targeted therapies such as IDH1/2 and FLT3 inhibitors). Pts were grouped into pre-VEN (2014-2018) and post-VEN (2019-2024) treatment eras. We performed logistic and Cox regression analyses to identify variables associated with choice of first line of treatment (LOT1) and OS. Subgroup analyses were conducted for pts undergoing allogeneic hematopoietic cell transplant (allo-HCT).

The study included 392 pts with post-MPN AML (143 pre-VEN and 249 post-VEN) with a median age at diagnosis of 72 (interquartile range: 66-78; 21.2% <65) years. Most pts were male (61.5%), White (73.0%) and had Medicare (47.7%) or commercial insurance (38.8%). No significant differences in patient characteristics were observed between the two eras.

Most pts (n=264, 67.3%) received LIT as LOT1, increasing from 52.4% (n=75) pre-VEN to 75.9% (n=189) post-VEN. However, HMA-mono usage decreased from 39.9% (n=57) pre-VEN to 18.9% (n=47) post-VEN. A total of 73 pts (18.6%) received IC, decreasing from 30.1% (n=43) pre-VEN to 12.0% (n=30) post-VEN. The multivariable model indicated that pts diagnosed in the post-VEN era (odds ratio [OR]=2.74, 95% confidence interval [CI]: 1.65-4.54, p<.01) and older age groups (65-74 years OR=3.54, 95% CI 1.82-6.89, p<.01; 75-85 years OR=7.97, 95% CI 3.81-16.66, p<.01; vs 40-64 years) were more likely to receive LIT (vs all other treatments).

In the post-VEN era, HMA/VEN was the most used LOT1 regimen (n=110; 44.2%). These pts (vs all other) were more likely to be older (65-74 vs 40-64: OR=3.63, 95% CI 1.51-8.71; p<.01), fitter (ECOG 2-4 vs 0-1: OR=0.32, 95% CI 0.12-0.84; p=.02) and male (OR=2.00, 95% CI 1.13-3.55; p=.02), but less likely to have myelofibrosis as the antecedent MPN (OR=0.49, 95% CI 0.25-0.98; p=.04). Only 24 (9.6%) pts received targeted agents alone or in combination during follow-up.

The median OS for our cohort was 7.24 ms (95% CI 6.22-9.21) and did not differ by era (pre-VEN 7.1 ms, 95% CI 5.7-9.5; post-VEN 7.6 ms, 95% CI 5.8-10.1; p=.39). Favorable OS was associated with male sex (hazard ratio [HR]= 0.70, 95% CI 0.55-0.89; p<.01) and higher socioeconomic status (5^th vs 1^st quintile HR=0.61, 95% CI 0.40-0.93; p=.02), while worse baseline performance status was linked with poorer OS (ECOG 2-4 vs 0-1 HR=2.17,95% CI 1.49-3.17; p<.01).

Pts treated with IC had longer median OS (9.2 ms) compared to pts treated with HMA/VEN (8.9 ms; p=.03) or HMA-mono (6.0 ms; p<.01). Multivariable analysis confirmed that pts treated with HMA/VEN had worse OS than those with IC (HR=1.95, 95% CI 1.12-3.41; p=.02).

Among 59 pts (15% of cohort) who underwent allo-HCT, 50.9% (n=30) received it after LOT1 and 39.0% (n=23) after LOT2 with a median OS of 22.1 (95% CI 13.4–59.0) and 16.2 (95% CI 10.2-41.5) ms, respectively. For bridging treatment, 13 (43.3%) and 9 (30.0%) of 30 pre-VEN pts received IC and LIT, yet of 29 post-VEN pts, 11 (37.3%) received IC and 11 (37.3%; 8 after HMA/VEN) received LIT. Pts who received IC and LIT as bridging treatment had comparable outcomes (median OS: IC 22.1 [95% CI 10.2-98.1) ms, LIT 16.9 [95% CI 12.5–36.2] ms; p=.37).

This large, real-world analysis shows that outcomes among pts with post-MPN AML remain poor and without improvement of OS in the new AML treatment era. IC was associated with better OS than LIT but the pts receiving IC and LIT as a bridge to allo-HCT had comparable OS. Our study highlights allo-HCT as the only potentially curative option, currently accessible to a minority of patients, and underscores the urgent need for novel, more effective therapies for post-MPN AML.