Treatment Intensity in Acute Myeloid Leukemia with Extramedullary Disease

Background: Extramedullary disease (EMD) occurs in 5-10% of patients (pts) with acute myeloid leukemia (AML). The extent to which treatment intensity influences outcomes in EMD is unclear. We sought to determine the cytomolecular characteristics and outcomes of AML with EMD.

Methods: This was a retrospective study of all pts with newly diagnosed AML (excluding APL) treated at our institution from 2012 to 2023. Pts with core binding factor AML were excluded. The presence of EMD at diagnosis, as well as whether it was isolated, was recorded. Baseline risk was assigned per the ELN2022 recommendations, with routine cytogenetics and next-generation sequencing performed in all pts. Treatment intensity was classified into intensive chemotherapy (IC) and non-intensive (NI) approaches.

Overall survival (OS) was measured from treatment start to death from any cause and event-free survival (EFS) from treatment start to death, relapse or non-response. The Kaplan-Meier method was used for time-to-event analyses.

Results: 2764 pts were included. Median age was 68 (range, 17-95). 110 (4%) had known EMD at diagnosis, of which 21 (19%; 1% of total AML) had isolated EMD and 89 (21%, 3% of total AML) had synchronous EMD. The most frequently involved sites were skin (51%), lymph nodes (17%) and central nervous system (17%). Pts with EMD were younger, with a median age of 59 vs 68 (p <0.01). EMD was associated with more favorable (16% vs 8%) and less adverse (45% vs 63%) ELN risk disease at diagnosis compared with no EMD (p <0.01), which was predominantly driven by a higher proportion of NPM1^mut (24% vs 15%, p = 0.01) and a lower proportion of TP53^mut (16% vs 25%, p = 0.05). IDH1^mut was less frequent in pts with EMD (3% vs 8%, p = 0.05). Frequencies of FLT3^ITD (14% vs 15%), IDH2^mut (9% vs 13%) and KMT2A rearrangements (6% vs 3%), and RAS pathway mutations (38% vs 36%) were similar between pts with and without EMD.

Treatment intensity differed between the two groups (p <0.01). 59% of pts with EMD received IC. For pts without EMD, 31% received IC regimens. The proportion of pts receiving venetoclax-based treatments was similar (44% vs 38%, p = 0.32), as was the proportion of pts undergoing stem cell transplant (SCT) in each group (27% EMD, 20% non-EMD, p = 0.08).

The overall CR/CRi rate was similar between pts with and without EMD (62% vs 55%, p = 0.15). At a median follow-up of 51 mo (95% CI, 48-55), median OS (mOS) was 8 mo (95% CI, 6-13) for pts with EMD and 10 mo (95% CI, 9-11) for those without (p = 0.35). The median EFS (mEFS) was 9 (95% CI, 5-NE) and 11 (95% CI, 10-13) mo, respectively (p = 0.17). In a landmark analysis of the 29 pts with EMD who underwent SCT, mOS was not reached, 3-year OS 65% (95% CI, 50 – 85%).

To account for differences in baseline characteristics between pts with and without EMD, multivariate analysis (MVA) including EMD status, age, ELN2022 risk, WBC at diagnosis, consolidative SCT, treatment intensity, and use of venetoclax were performed. In MVA, the presence of EMD was associated with increased hazard of death (HR 1.3 [95% CI, 1.0-1.6], p=0.03) and hazard of death/relapse/non-response (HR 1.2 [95% CI, 1.0-1.5], p=0.07).

When sub-divided by treatment intensity, there was no difference between pts with extramedullary disease treated intensively (mOS 12 mo [95% CI, 7-36] for EMD vs 18 mo [95% CI, 16-23] for no EMD, p = 0.26) or non-intensively (mOS 7 mo [95% CI, 3-12] for EMD vs 8 mo [95% CI, 7-9], p = 0.78). A similar trend was seen for EFS. Specifically for pts receiving treatment with HMA + VEN regimens, mOS was 9 mo (95% CI, 5-23) and EFS was 8 mo (95% CI, 4-13) in pts with EMD, compared to a mOS of 8 mo (95% CI, 8-9) and EFS of 4 (95% CI, 4-5) in pts without EMD. Pts with EMD treated with IC + venetoclax (n=12) had a mOS of 36 mo (95% CI, 8-NR) compared to a mOS of 8 mo (95% CI, 5-35) with IC alone (n=41) (p=0.11).

In a MVA of pts with EMD only, treatment intensity did not significantly impact OS (p = 0.9) or EFS (p = 0.5). Factors influencing OS in pts with EMD were: ELN2022 adverse risk (HR 3.8 [95% CI, 1.7-8.3], p <0.01), consolidative SCT (HR 0.4 [95% CI, 0.2-0.7], p < 0.01) and use of venetoclax (HR 0.5, [95% CI, 0.3-0.8], p < 0.01).

Conclusions: EMD occurs rarely in AML. Treatment intensity does not appear to impact OS and EFS in pts with EMD in multivariate models, now that low-intensity venetoclax-containing regimens are standard. In MVA, the use of venetoclax and consolidative SCT improved outcomes in pts with EMD. Larger studies are needed to confirm these findings.