Cytometry Part B (Clinical Cytometry) 74B:9–16 (2008)

Positive Selection for CD90 as a Purging Option in Acute Myeloid Leukemia Stem Cell Transplants Nicole Feller, Ange`le Kelder, Guus Westra, Gert J. Ossenkoppele, and Gerrit J. Schuurhuis* Department of Hematology, VU University Medical Center, Amsterdam, The Netherlands

Background: Several studies showed the benefit of purging of acute myeloid leukemia (AML) stem cell transplants. We reported previously that purging by positive selection of CD34+ and CD133+ cells resulted in a 3–4 log tumor cell reduction (TCR) in CD342 and/or CD1332 AML, but has been shown to be potentially applicable in only about 50% of cases. Similar to CD34 and CD133, CD90 marks the hematopoietic CD34 positive stem cells capable of full hematopoietic recovery after myeloablative chemotherapy, and therefore, in the present study, we explored whether a similar purging approach is possible using CD90. Methods: CD90 expression was established by flowcytometry in diagnosis AML on the clonogenic AML CD34+ blast population by flow cytometry. Positivity was defined as >3% CD90 (CD34+) expression on blasts. For the calculation of the efficacy of TCR by positive selection, AML blasts were recognized by either prelabeling diagnosis blasts with CD45-FITC in spiking model experiments or using expression of leukemia associated marker combinations both in spiking experiments and in real transplants. Results: In 119 patients with AML and myelodysplastic syndrome, we found coexpression of CD34 and CD90 (>3%) in 42 cases (35%). In AML patients 60 years or younger, representing the patients who are eligible for transplantation, only 23% (16/69) of the patients showed CD90 expression. Positive selection for CD90 in transplants containing CD90 negative AML resulted in a 2.8–4 log TCR in the models used. Conclusions: Purging by positive selection using CD90 can potentially be applied effectively in the majority of AML patients 60 years or younger. q 2007 Clinical Cytometry Society Key terms: AML; purging; THY-1; stem cell transplants

How to cite this article: Feller N, Kelder A, Westra G, Ossenkoppele GJ, Schuurhuis GJ. Positive selection for CD90 as a purging option in acute myeloid leukemia stem cell transplants. Cytometry Part B 2008; 74B: 9–16.

There is still debate about the benefits for patients with acute myeloid leukemia (AML) to be treated with myeloablative therapy, followed by autologous peripheral blood stem cell (PBSC) transplantation. Although this treatment is potentially beneficial, clinical outcome of these AML patients is adversely affected by therapyrelated mortality/morbidity and putatively also by increased frequency of relapse due to contamination of stem cell products with minimal residual disease (MRD) cells. In AML several studies showed an actual advantage for survival after purging of these transplants with maphosphamide (1,2), but no randomized studies have been performed to unequivocally prove the principle. The reluctance to incorporate purging with maphosphamide, the most commonly used purging agent for AML transplants, as a logical component in stem cell processing, is partly caused by its relatively high toxicity

q 2007 Clinical Cytometry Society

towards normal progenitor/stem cells. Recoveries of colony forming unit ability down to 5% of controls have been allowed in this setting (1). Due to the delayed hematopoietic recovery this may result in a relatively high morbidity and/or mortality. In a previous study, we reported an immunologic purging strategy based on expression of the stem/progenitor cell markers CD34 and CD133 (3). Positive selection using CD34 or CD133 *Correspondence to: G. J. Schuurhuis, Department of Hematology, VU University Medical Center, CCA Building Rm. 4.24, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands. E-mail: [email protected] Received 9 March 2007; Revision 10 July 2007; Accepted 3 August 2007 Published online 30 November 2007 in Wiley InterScience (www. interscience.wiley.com). DOI: 10.1002/cyto.b.20375

10

FELLER ET AL.

can be applied in autologous PBSC transplants only when AML blasts at diagnosis show no or dim expression of CD34 and/or CD133 (3). This procedure resulted in a consistent 3–4 log tumor cell reduction (TCR). Importantly, this was accompanied by a high recovery of normal CD34þ and CD133þ progenitor/stem cells which is typical for positive selections with MACS-based technology (4,5). However, this approach applies only for half of the AML patients. Other targets for positive selection are therefore needed to extend the applicability for purging using positive selection. CD90, also known as THY-1, is a good candidate. Firstly, CD90 is a stem cell marker, expressed on 20–60% of normal CD34þ cells in mobilized PBSC transplants (6). It has been proven that CD34þCD90þ cells are able to provide hematopoietic reconstitution in vivo in animal models (7–9). In addition, purified CD34þCD90þ cells were able to cause engraftment in both multiple myeloma patients either treated with high-dose melphalan or with total body irradiation (10) and in patients with metastatic breast cancer treated with high dose chemotherapy including carmustine, cisplatin, and cyclophosphamide (11). Overall, transplantation with CD34þCD90þ cells thus seems as effective as reinfusion of the total CD34þ population. Secondly, part of AML patients does not show expression of CD90 in blasts at diagnosis. Conflicting data on the percentage of patients presenting with CD90 have been published: expression varied between 5 and 66% (12–14). Recently, these discrepancies have been clarified: CD90 turned out to be expressed predominantly in myelodysplasia syndrome (MDS) patients, patients with secondary AML, AML patients >60 years, and patients presenting with poor-risk karyotypes (15,16). Autologous transplantation, however, is predominantly performed in patients 60 years FAB M0 M1 M2 M3 M4 M5 M6 BAL RAEB(-t) CMML AML from MDS Unclassified AML Cytogenetic abnormalitya Favorable Intermediate Adverse No metaphases

Percent

61 58

51 49

63 56

71.7 28.3

2 11 22 8 14 26 4 0 20 0 4 8

1.7 9.2 18.5 6.7 11.8 21.8 3.4 0 16.8 0 3.4 6.7

14 62 10 33

11.8 52.1 8.4 27.7

a Cytogenetic aberrations were according to Grimwade et al. (18).

frozen in liquid nitrogen until use. Patients with >20% blasts in BM at diagnosis were referred as AML, patients with 5% blasts in peripheral blood (PB) or Auer rods in peripheral blood in BM/PB, were referred as refractory anemia with excess blasts in transformation (RAEB-t). Patients with 5–20% blasts in BM with 3% on blast cells) in 42 out of 119 Table 2 Total CD90 and CD34CD90 Expression on Blasts at Diagnosis in Different Patient Groups

AML
60 years AML >60 years Secondary AML
60 years Secondary AML >60 years MDS
60 years MDS >60 years Total

CD90þ on total blast component Number of cases Percent

CD34þCD90þ on total blast component Number of cases Percent

31/68 15/29

16/68 12/29

45 52

0/1 1/1 8/11 8/9 63/119

23 41

0/1 73 89 54

0/1 7/11 7/9 42/119

64 78 35

A patient is defined as CD90þ or CD34þCD90þ if expressed >3% on CD45dim blasts. All groups are divided by age since patients
60 years are eligible for autologous transplantation.

12

FELLER ET AL.

Table 3 CD34CD90 and Total CD90 Expression on Blasts at Diagnosis in Different Patient Groups

Cytogenetic aberrations Favorable Intermediate Adverse Unknown Total

CD90 on total blast component Number of cases Percent 4/13 31/63 6/10 22/33 63/119

31 49 60 67 53

CD34þCD90þ on total blast component Number of cases Percent 3/13 18/63 6/10 15/33 42/119

23 29 60 45 35

A patient is defined as CD90þ or CD34þCD90þ if expressed >3% on CD45dim blasts. Cytogenetic aberrations were according to Grimwade et al. (18). The presence of cytogenetic aberrations in the evaluable group of 86 patients was weakly associated with the presence of CD34þCD90þ expression (P 5 0.13) and single CD90 expression (P 5 0.17).

patients (35%). Next, we subdivided the patient group on age, the presence of dysplastic features, and secondary disease. The age of 60 years was chosen since this age determines the eligibility for autologous transplantation in our HOVON/SAKK clinical protocols. CD34/ CD90 expression was found predominantly in patients with MDS (70%, 14/20). This percentage was significantly higher compared to non-MDS AML patients (28%, 28/99) (P 5 0.006) and does not seem to be affected by differences in age: 7/11 in MDS patients
60 years and 7/9 in MDS patients >60 years. In agreement with the studies of Buccisano et al. (15,16), we found a tendency towards more positivity in elderly patients (AMLþMDS): 49% (19/39), versus 29% (23/80) in younger patients (P 5 0.02). For AML alone these percentages were 40% (12/30) versus 23% (16/69). For secondary AML the number of patients was too small to draw separate conclusions. Besides the association with age, in accordance with Buccisano et al. (15,16), the presence of CD34þCD90þ cells was associated, although weakly because of small numbers of patients, with adverse cytogenetics (see Table 3). Since the stem cell in MDS has been reported to be CD90 positive (21), we restricted the further analyses to AML (including secondary AML). Summarized, 23% of diagnosis AML patients