Introduction: Recent evidence supports the clinical significance of leukemia stem cells (LSCs) in acute myeloid leukemia (AML). However, the identification of LSCs in acute lymphocytic leukemia (ALL) has proved challenging, as transplantation studies in immunocompromised mice have yielded conflicting results. The distinction between Philadelphia chromosome-positive (Ph+) ALL and lymphoid blast crisis (LBC) chronic myeloid leukemia (CML) is also controversial. We previously identified a clinically relevant CD34+CD38- population of LSCs with intermediate (int) levels of aldehyde dehydrogenase (ALDH) activity (CD34+CD38-ALDHint) in AML [Gerber, et al. Blood, 2012]. This population was not present in healthy controls and could be distinguished from normal hematopoietic stem cells (HSCs), which had higher levels of ALDH activity (CD34+CD38-ALDHhigh). We hypothesized that the same approach could be used to identify a putative LSC population in ALL. Furthermore, in contrast to most cases of AML, the chronic phase CML stem cell was found to reside in the same CD34+CD38-ALDHhigh population as normal HSCs [Gerber, et al. Am J Hematol, 2011]. We therefore also hypothesized that the presence of BCR/ABL mutations in the CD34+CD38-ALDHhigh population might help distinguish LBC CML from Ph+ ALL.

Methods: Bone marrow and/or peripheral blood specimens were collected at diagnosis from patients with B cell ALL or LBC CML on an IRB-approved protocol. A total of 7 patients were evaluated: 2 Ph- ALL, 2 Ph+ ALL, and 3 LBC CML patients. CD34+ cells were isolated by magnetic bead and column selection, then analyzed by flow cytometry with respect to CD38 expression and ALDH activity. Sorted cell populations were analyzed by fluorescence in situ hybridization (FISH) for leukemia-specific abnormalities. Polymerase chain reaction was performed on clinical samples to determine the presence of a p190 vs. p210 transcript.

Results: All patients harbored an aberrant CD34+CD38-ALDHint population, similar to that previously seen in AML. This population was ≥95% positive for BCR/ABL by FISH in all Ph+ ALL and LBC CML cases. It was similarly positive (≥75%) for other leukemia-specific FISH abnormalities (including trisomy 4, 8, 10, 12, and/or 21) in all four ALL cases, as well as one LBC CML case. Conversely, the CD34+CD38-ALDHhigh population (which typically contains the normal HSCs) lacked any of the other cytogenetic abnormalities in all of the cases, irrespective of Ph status or a diagnosis of ALL vs. CML. Notably, the CD34+CD38-ALDHhigh population was negative for BCR/ABL in the Ph+ ALL cases but was >95% positive for BCR/ABL by FISH in the LBC CML cases. The B cell differentiation marker, CD19, was expressed on the CD34+CD38-ALDHint but not the CD34+CD38-ALDHhigh population in all ALL cases, both Ph- and Ph+. In contrast, CD19 expression was variable in the LBC CML cases. Both Ph+ ALL cases possessed a p190 BCR/ABL transcript, whereas all of the LBC CML cases contained a p210 transcript. Also of note, the CD34+CD38-ALDHint population was persistently detectable in one of the LBC CML patients while in complete remission after induction therapy; that patient subsequently relapsed.

Conclusions: An abnormal CD34+CD38-ALDHint population was identified in all cases of B cell ALL and LBC CML. This population is analogous to a previously identified, clinically relevant LSC population in AML and may represent a putative LSC population in ALL. The CD34+CD38-ALDHhigh population was normal by FISH in the ALL cases but contained the BCR/ABL mutation in the LBC CML cases, thus permitting distinction between Ph+ ALL and LBC CML (which also differed based on the presence of p190 vs. p210 transcripts, respectively). Additionally, clonal evolution from chronic phase to lymphoid blast crisis CML was apparent, based on the acquisition of additional cytogenetic abnormalities unique to the CD34+CD38-ALDHint population as compared to the CD34+CD38-ALDHhigh population. The presence of CD19 on the putative LSCs in the four cases of ALL suggest that CD19-directed therapies may target the LSCs and thus may have curative potential in those cases. This assay may serve as a means to evaluate other possible therapeutic targets. Lastly, the detection of the abnormal CD34+CD38-ALDHint population may have utility as a minimal residual disease assay for monitoring response to treatment. These findings warrant validation in a larger patient cohort.


Gerber:Janssen: Research Funding; Alexion: Membership on an entity's Board of Directors or advisory committees; Spectrum: Membership on an entity's Board of Directors or advisory committees; Seattle Genetics: Membership on an entity's Board of Directors or advisory committees. Grunwald:Alexion: Membership on an entity's Board of Directors or advisory committees; Amgen: Research Funding; Incyte Corporation: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Medtronic: Equity Ownership; Janssen: Research Funding; Ariad: Membership on an entity's Board of Directors or advisory committees; Forma Therapeutics: Research Funding. Avalos:Seattle Genetics: Membership on an entity's Board of Directors or advisory committees.

Author notes


Asterisk with author names denotes non-ASH members.