GRAPEVINE, TEXAS—Results of a large, retrospective study suggest that allogeneic stem cell transplant (allo-SCT) can overcome the poor prognosis associated with central nervous system (CNS) involvement in acute myeloid leukemia (AML).
By analyzing transplant outcomes in more than 5000 patients, researchers found that subjects with CNS AML had rates of relapse and survival that were similar to those of patients without CNS involvement.
The team also identified factors that can predict for survival in CNS AML, including cytogenetic risk group, the presence of chronic GVHD, and whether a patient was in complete response at transplant.
Jun Aoki, MD, of Tokyo Metropolitan Komagome Hospital in Japan, presented these findings at the 2014 BMT Tandem Meetings as abstract 68.
Dr Aoki pointed out that CNS involvement is rare in adult AML, occurring in about 5% of patients. However, these patients generally have poor prognosis. And although allo-SCT is one of the options used to treat CNS AML, exactly how CNS involvement impacts transplant outcomes remains unclear.
So Dr Aoki and his colleagues conducted a nationwide, retrospective study to gain some insight. They collected data from the registry database of the Japan Society for Hematopoietic Cell Transplantation.
Patients had to be older than 15 years of age, have their first allo-SCT between 2006 and 2011, and not have acute promyelocytic leukemia.
The researchers identified 5068 patients who met these criteria, and 157 of them had CNS AML. CNS involvement was defined as infiltration of leukemia cells into CNS or myeloid sarcoma in CNS that were identified at any time from diagnosis to transplant.
No difference in relapse, survival
There were no significant differences between CNS patients and controls with regard to the estimated overall survival (OS), leukemia-free survival, cumulative incidence of relapse, or non-relapse mortality at 5 years.
OS was 39.9% among controls and 38.5% among CNS patients (P=0.847). Leukemia-free survival was 41.2% and 41.5%, respectively (P=0.82).
The cumulative incidence of relapse was 29.8% among controls and 31.8% among CNS patients (P=0.418). And non-relapse mortality was 22.5% and 26.5%, respectively (P=0.142).
Factors predicting OS
To determine the impact of patient and treatment characteristics on OS, the researchers conducted a multivariate analysis. This confirmed that CNS involvement was not a risk factor for OS.
But it revealed a number of other factors that adversely affect OS, including age of 50 or older (P<0.001), lack of a complete response at allo-SCT (P<0.001), a donor source of unrelated cord blood (P=0.005), having a prognostic score of 2-4 (P<0.001), unfavorable cytogenetics (P<0.001), and the absence of acute or chronic GVHD (P<0.001 for both).
When the researchers analyzed only CNS patients, they discovered that not all of these factors retained significance. Only the absence of chronic GVHD (P=0.002), lack of complete response at transplant (P<0.001), and having either intermediate (P=0.003) or unfavorable cytogenetics (P=0.011) were adversely associated with OS in these patients.