Marker Chromosomes Are a New Cytogenetic Adverse Risk Factor in AML after Allo-HCT

[Background]

Chromosome analysis is indispensable for the diagnosis and risk classification of acute myeloid leukemia (AML). A marker chromosome (MC) is a fragmented chromosome that cannot be identified as originating from an existing autosomal or sex chromosome. Although often observed, the significance of MC in hematological malignancies is unknown. Recently, MC was found to be the result of chromothripsis (CT). CT is a catastrophic phenomenon by which chromosomes are shattered into tens to hundreds of fragments. Half of all CT cases are related to TP53 mutation. In addition to MC, complex karyotype (CK), monosomal karyotype (MK) and existing sub-clone (SC) have also been noted as the result of CT. Previous studies reported that MC was a poor prognostic factor in AML. These studies included AML patients who underwent allogeneic hematopoietic cell transplantation (allo-HCT). However, the influence of MC on AML after allo-HCT was not clarified.

[Purpose]

In this study, we evaluated the impact of MC on the outcome of AML patients after allo-HCT.

[Method]

This retrospective analysis included 166 AML patients who received allo-HCT at Niigata University Hospital (n=129) or Nagaoka Red Cross Hospital (n=37) between 1990 and 2017. The median age of patients at allo-HCT was 38 years old (range 14-67 y). The median follow-up period was 2.0 y (range 0.0-22.2 y). According to the revised Medical Research Council (rMRC) criteria for cytogenetic risk categories, 26 (15.7%), 104 (62.7%) and 36 (21.7%) patients were categorized as favorable, intermediate and adverse-risk, respectively. Myeloablative conditioning was used for 128 (77.1%) and reduced-intensity was used for 38 (22.9%) patients. Donors were related for 83 (59.3%) and unrelated for 57 (40.7%) patients. The Kaplan-Meier method (log-rank test) and Gray's test were used to estimate the probabilities of overall survival (OS) and cumulated incidence of relapse (CIR). The impact of several variables was assessed by multivariate analysis using a Cox regression model and Fine-Gray test with backward stepwise selection based on the p-value. For the comparison of clinical phenotypes, category variables were evaluated by Fisher's exact test.

[Results]

MC was detected in 14 (8.4%) of 166 patients. Eleven (78.6%) were grouped as adverse-risk by rMRC criteria. CK, MK and SC were observed in 26 (16.3%), 20 (12.0%) and 23 (13.9%) patients, respectively. The median age of AML/MC+ (n=14) and AML/MC-(n=152) patients were similar (38.5 y, range 19-64 y vs 38 y, range 14-64 y, P=0.812). Twelve AML/MC+ patients (85.7%) received allo-HCT at the advanced stage (³3 CR or non-remission) and 10 (71.4%) had primary induction failure (PIF). Compared with AML/MC- patients, those with AML/MC+ had a higher incidence of MK (78.6% vs. 5.9%, p<0.001), sub-clone (64.3% vs. 9.2%, p<0.001) and CK (85.7% vs. 9.9%, p<0.001). On univariate analysis, the 2-y OS and median survival period of AML/MC+ patients were shorter than those of AML/MC- patients (26.8% vs. 61.3% and 0.78 y vs. 4.9 y, p=0.0126). The 2-y CIR of AML/MC+ patients was higher than that of AML/MC- patients (80.4% vs. 37.2%, p<0.01). To further investigate the impact of MC on the outcome, we compared AML patients with/without MC who had the cytogenetic adverse-risk karyotype. The 2-y OS of adverse-risk AML/MC+ patients (n=11) was shorter than that of adverse-risk AML/MC- patients (n=25) (9.1% vs 58.3%, p=0.003). The median survival periods were 0.58 y and 4.0 y, and the 2-y CIR was 89.6% and 44.7% (p=0.002), respectively. The 2-y OS based on the other risk factors in adverse-risk AML patients were as follows: CK+ 33.2%, MK+ 26.9% and SC+ 34.7%. As these results suggested MC to be a poorer risk factor, we performed multivariate analysis for confirmation. In the multivariate analysis adjusted for CK, MK, SC, donor, conditioning and disease stage before allo-HCT, only MC was an independent poor risk factor for OS (HR 3.547, 95%CI: 1.46-8.63, p=0.005) and CIR (HR 3.90, 95%CI: 1.47-10.33, p=0.006) among adverse-risk AML patients using the rMRC criteria.

[Conclusion]

It is very difficult for AML/MC+ patients to achieve complete remission, leading to a markedly poor prognosis after allo-HCT. The outcome for AML/MC+ was inferior to that for AML with CK, MK, SC or the current adverse-risk karyotype by rMRC. This analysis revealed MC as a new independent factor that further indicates a poor prognosis after allo-HCT, especially in high-risk AML patients.