Long-Term Data Demonstrating Inferior Survival Outcomes in Patients with Myeloid Sarcoma

Background and objectives

Myeloid sarcoma (MS) is a tumour mass consisting of myeloid blasts occurring in any other anatomical site other than the bone marrow. It may precede or coincide with acute myeloid leukemia (AML) or constitute acute blastic transformation of an antecedent myeloid malignancy. Data on the prognostic significance of MS is limited. We aim to study outcomes and characteristics of MS with AML (MS-AML) and without concomitant AML (isolated MS) and compare this to a cohort of AML patients with no extramedullary disease (AML only)

Methods

Patients aged 18 years or older meeting the criteria for MS according to the WHO 2016 classification in Princess Margaret Cancer Centre between January 1, 2000 and December 31, 2019 were studied. Overall survival (OS) and relapse-free survival (RFS) rates were calculated using the Kaplan-Meier method and the impact of covariates were assessed using the Log-rank test. We then compared the outcomes of patients with MS with a group of patients treated at our centre with AML with no extramedullary disease between 2015-2018.

Results

Among 158 patients that were identified with a diagnosis of MS, 115 (72%) had a concomitant diagnosis of AML (MS-AML). M4 and M5 morphology was seen in 48 (42%) of these patients with CD56 positivity in 22 (19%) of patients. Cytogenetics by MRC criteria is shown in table 2. 18% of patients had unavailable cytogenetics and the majority of these patients had isolated MS without AML involvement meaning cytogenetics was not performed. The most common abnormal karyotype was trisomy 8 (9%) followed by t(8;21)(q22;q22) (5%). NPM1 was mutated in 24 (19%) patients and FLT3 ITD mutation was seen in 17 (13%) (Table 1). The most common site of MS was skin (37%), lymphatic organs (23%) and abdominal organs (12%). Of the 41 (26%) who did not have a diagnosis of AML, 21 (13%) had an isolated MS with no other haematological malignancy and 23 (14%) had an associated myeloid malignancy which included MPN (32%), MDS (27%), CMML (18%) and CML (9%).

125 (79%) patients received induction with 74% receiving ‘3 + 7’. CR/CRi was achieved in 75 (60%) patients. Induction mortality was 8% and of these patients average age was 66 and over 50% had multi-organ involvement of MS. Of the 50 who did not achieve CR, 25 (50%) were re-induced (10 FLAG-IDA, 8 NOVE-HIDAC, 7 others) with CR in 12 (48%) (Total CR rate 69%). (Table 1)

OS and RFS of all patients with MS at 2 years was 28% (95% CI 0.21-0.36) and 36% (CI 95% 0.26-0.46) respectively. Stratified by cytogenetic risk, OS was 66% for favourable risk, 34% for intermediate risk and 11% for poor risk (p=<0.001). Secondary disease (Therapy related or secondary disease from antecedent MDS/MPN) had an OS of 11% compared to de novo MS-AML of 35% (p=<0.001). We then compared these outcomes with a group of AML (AML-only) patients (n=377) without extramedullary disease treated from 2014-2018 (Table 2). OS at 2 years was lower when compared with the AML-only group (28% vs 51% p=<0.001) (Figure 1). RFS was also significant when compared to this group (36% vs 19% p=0.001). There was no difference in OS or RFS when comparing isolated myeloid sarcoma with AML-only (OS 44% vs 51% p=0.06, RFS 13% vs 19% p=0.74) suggesting that isolated MS has a similar prognosis to AML-only. There was an increased number of patients with secondary disease in the MS group suggesting a cause of the poorer OS and DFS (38% vs 18% p=<0.001).

Conclusion

This retrospective analysis found MS with concomitant AML is a poor prognostic marker. In the era of new therapy for AML, every attempt must be made to perform genomic profiling to understand the unusual tropism of blasts for extramedullary tissue and understand the role of targeted therapy.