Expression of TGF-β Superfamily Receptor ALK-5 Is Associated with Decreased Overall Survival of AML Patients Receiving Standard AML Chemotherapy.

Bone marrow neoangiogenesis plays an important role in acute myeloid leukaemia. TGF-β has been shown to be involved in angiogenesis. TGF-β receptors are comprised of class I, class II and accessory receptors. Class I TGF-β receptor ALK-1 is mainly expressed on endothelial cells whereas ALK-5 is found on endothelial cells and various cell types including tumor cells.

m-RNA and protein expression of ALK-1 and ALK-5 were studied in several leukemic cell lines by PCR and FACS analysis. Expression of both receptors and their respective ligands, bone morphogenic proteins (BMP) -9 and -10 and TGF-β1, were analyzed by real-time quantitative PCR in pretherapeutic samples from 93 acute myeloid leukaemia (AML) patients enrolled into the AMLSG 07-04 study of the Austrio-German AMLSG study group. To evaluate ALK-1 and ALK-5 receptor activation, expression of downstream target genes ID-1 and PAI-1 were also determined by quantitative PCR. Uni- and multivariate analyses were performed using clinical and gene expression variables.

ALK-1 and ALK-5 were expressed by several leukemic cells lines including KG-1 and UKE-1 where endothelial cells (HUVECs and OECs) served as positive controls. Expression of ALK-1 and ALK-5 were found in the majority of AML patients (82 % and 88 %, respectively). All patients expressed TGF-β1, whereas the percentage of BMP-expressing patients was low (8 % for BMP-9, 13 % for BMP-10). Expression of downstream targets ID-1 and PAI-1 was detected in almost all patients (99 % and 100 %, respectively).

The impact of gene expression on achievement of CR and event-free (EFS), relapse-free (RFS) and overall survival (OS), was analyzed. Due to low numbers of patients with BMP-9 and -10 expression and inconclusive PCR results for ALK-1 and PAI-1 in several patients, these variables were only studied by univariate analysis. No effect of ALK-1, PAI-1, BMP-9 or -10 expression on clinical outcome was detected.

Patients with low vs. high ALK-5 expression achieved a CR rate 63 % vs. 27 %, respectively. Multivariate analysis included ALK-5, TGF-β1 and ID-1 gene expression as well as the baseline variables sex, age, karyotype and FLT3 mutation status. During the stepwise procedure of the Cox regression, several variables were eliminated. Only ALK-5 and the karyotype showed a significant impact on event-free, relapse-free and overall survival (ALK-5: p=0,001 for EFS, p<0,001 for RFS, p=0,008 for OS; karyotype: p<0,001 for EFS, p<0,001 for RFS, p=0,008 for OS).

ALK-1 and ALK-5 expression is found in bone marrow and peripheral blood samples from the majority of AML patients. Both receptors are found on endothelial cells and on leukemic blasts. Both receptors are in an activated state as demonstrated by expression of downstream target genes. High ALK-5 expression results in a low CR rate which translates into reduced EFS, RFS and OS of AML patients. Therefore ALK-5 expression represents a new prognostic marker in AML. Recently published data (Shao, Blood. 2009 Jun 8. [Epub ahead of print]) demonstrate involvement of both ALK-1 and ALK-5 in the regulation of vascular endothelial growth factor which is believed to be the central growth factor in angiogenesis. Hence, the ALK-1/ALK-5 pathway might represent a novel therapeutic target in AML.

No relevant conflicts of interest to declare.