© ASCO/Rodney White
CHICAGO—Researchers say they have identified 3 leukemia stem cell (LSC) phenotypes that are correlated with cytogenetic/molecular abnormalities and prognosis in acute myeloid leukemia (AML).
The investigators believe this knowledge could aid risk stratification of AML patients, particularly those without identifiable cytogenetic or molecular risk factors.
The findings may also pave the way for scientists to identify novel therapeutic targets on LSCs and monitor LSCs in response to therapy.
Jonathan Michael Gerber, MD, of Levine Cancer Institute in Charlotte, North Carolina, presented the findings at the 2015 ASCO Annual Meeting (abstract 7000*).
Via previous research, Dr Gerber and his colleagues identified 3 different LSC phenotypes in AML:
- LSCs that are CD34-negative
- LSCs that are CD34-positive, CD38-negative, and have intermediate levels of aldehyde dehydrogenase (ALDHint)
- LSCs that are CD34-positive, CD38-negative, and have high levels of ALDH (ALDHhigh).
With the current study, the researchers wanted to determine if these phenotypes correlate with cytogenetic/molecular features and treatment outcomes.
So they analyzed diagnostic samples from 98 patients with newly diagnosed AML who had normal or unfavorable cytogenetics. The patients were enrolled on a phase 2 trial comparing FLAM and 7+3 (Zeidner et al, haematologica 2015).
Dr Gerber and his colleagues identified 22 patients with CD34- LSCs, 43 with ALDHint LSCs, and 33 with ALDHhigh LSCs.
Risk factors
“We found that leukemia stem cell phenotype indeed correlated quite strongly with cytogenetic and molecular risk factors,” Dr Gerber said.
NPM1 mutations were more common in patients with CD34- LSCs (64%) than those with ALDHint LSCs (14%) or ALDHhigh LSCs (6%, P<0.001). NPM1 mutations were the sole abnormality in 50% of patients with CD34- LSCs.
Poor-risk cytogenetics and/or FLT3-ITD mutations were more common in patients with ALDHhigh LSCs (85%) than those with ALDHint LCSs (35%) or CD34- LSCs (18%, P<0.001).
Nine percent of patients in the CD34- LSC group fell into the European Leukemia Network poor-risk category, compared to 73% of patients in the ALDHhigh LSC group.
Only 2 patients had 11q23, and both had CD34- LSCs. Fifty-five percent of patients with ALDHhigh LSCs had prior myelodysplastic syndromes or myeloproliferative neoplasms.
Prognosis
“We found that leukemia stem cell phenotype correlated strongly with outcomes as well,” Dr Gerber said. “It turned out that CD34- patients fared more favorably overall.”
Patients with CD34- LSCs had the highest complete response rate (86%), followed by those with ALDHint LSCs (67%) and ALDHhigh LSCs (45%, P<0.01).
Patients in the CD34- group also had a higher rate of event-free survival at 2 years (46%) than patients in the ALDHint group (26%) or the ALDHhigh group (0%). The median event-free survival was 13 months, 11.3 months, and 2.2 months, respectively (P<0.01).
The rate of overall survival at 2 years was best for the CD34- group (76%), followed by the ALDHint group (38%) and the ALDHhigh group (34%). The median overall survival was not reached, 18.7 months, and 9.4 months, respectively (P=0.02).
Dr Gerber also noted that ALDHhigh patients fared much better if they underwent hematopoietic stem cell transplant.
“There is 0% leukemia-free survival at the 2-year mark for the ALDHhigh patients who were not transplanted,” he said. “Those that were transplanted fared about the same as everyone else in the series. So it was very striking that there were no chemotherapy survivors in that group.”
In closing, Dr Gerber said this research suggests the 3 LSC phenotypes are mutually exclusive and correlate with cytogenetic and molecular risk factors as well as outcomes in patients with AML.
“This [discovery] may allow for rapid risk stratification in this explosive disease, facilitate enrollment onto induction protocols . . . , and allow us to divert those ALDHhigh, very high-risk patients earlier to novel therapies and/or transplant, given that they’re not really helped much by conventional chemotherapy.”
*Information in the abstract differs from that presented at the meeting.