Survival Outcomes of Patients Diagnosed with Myeloid Sarcoma Versus Acute Myeloid Leukemia: A Retrospective Cohort Study

Background:

Myeloid sarcoma (MS) is a rare subtype of acute myeloid leukemia (AML) and presents as an extramedullary tumor of immature granulocyte cells. It is most often found either concurrently or following a previously recognized AML. Less often, it can occur de novo. Given the rarity of the disorder, most of the literature comprises of case reports or retrospective studies with small sample sizes. Herein, we aim to compare the overall survival (OS) and event-free survival (EFS) of MS versus AML.

Methods:

We conducted a single-center retrospective cohort study to compare OS, EFS, and the clinical response rate for MS versus AML in adult patients (≥18 years) at Cleveland Clinic from 5/2017-9/2023. Comparison groups included MS alone or in combination with bone marrow (BM) disease (MS-AML) and AML based on BM disease but without extramedullary disease (BM-AML). CNS involvement was not considered part of the MS definition. We collected baseline characteristics such as age, gender, ethnicity, comorbidities, AML subtypes, cytogenetics, FLT3 status, mutational profiles (via next-generation sequencing), and ECOG performance status. Responses to first-line treatment were categorized as composite complete response (CCR) (complete response [CR] or complete response with incomplete count recovery [CRi]) and minimal residual disease (MRD) assessed by flow cytometry or real-time PCR. For extramedullary disease, FDG-PET/CT scans were used to evaluate response. OS was defined as the time from the date of diagnosis to the date of death or last follow-up, whichever came first. EFS was defined from the date of diagnosis to the date of first refractory disease or relapse or death from any cause. Survival probabilities were estimated using the Kaplan-Meier method, with differences assessed by the log-rank test. Multivariable Cox regression and propensity score (PS) analysis were used to adjust for confounders, including comorbidities.

Results:

Among 749 patients with AML, there were 24 (3.2%) MS-AML and 725 (97%) BM-AML cases. The median ages were 64 years (IQR: 44-68) for the MS-AML group and 66 years (IQR: 57-75) for the BM-AML group (P=0.06). High-risk cytogenetics were seen in 8.3% of the MS-AML group and 29% of the BM-AML group (P=0.01). The three most common mutations in MS-AML were NPM1 (38%), TET2 (31%), and RUNX1 (15%) compared to DNMT3A (29%), NPM1 (26%), and TET2 (26%) in BM-AML (all P>0.05).

656 patients contributed to the analysis of the AML response, which was significantly associated with the AML subtype. The CCR rate was 35% in MS-AML and 53% in BM-AML. On multivariable logistic regression, the MS-AML group had lower odds of achieving CCR than BM-AML (odds ratio: 0.25, 95% CI: 0.08-0.69, P=0.007). Only 15 patients contributed to the analysis of the MRD response, which showed no significant difference between the two groups (95% CI: 0.18-1.20, P=0.12; reference: BM-AML). Similarly, there was no significant difference in cytogenetic response between the two groups (95% CI: 0.18-1.81, P=0.4; reference: BM-AML). The median follow-up periods for OS and EFS were 34 months (range: 0.49 – 193.31) and 34 months (range: 0.36 – 90.18) with 546 deaths and 581 cases of progression or death, respectively. On Cox proportional hazards model (CPH), there was no significant difference in OS (mortality hazard ratio (HR) = 0.78, 95% CI: 0.44-1.38, P=0.4) in the MS-AML group compared to the BM-AML group. Similarly, there was no significant difference in EFS (HR = 0.95, 95% CI: 0.55-1.65, P=0.9) in the MS-AML group compared to the BM-AML group.

Conclusion:

Patients who were diagnosed with MS-AML had a lower response rate to first-line treatment compared to those diagnosed with BM-AML. However, this was not translated into a significant difference in OS or EFS for patients in the MS group compared to those in the AML group.