Differential Expression of SHP-1 Levels in Chronic Phase and Advanced Disease CML and in AML Patients

Abstract 1449

Despite unprecedented success of tyrosine kinase inhibitors (TKI) in treating chronic myeloid leukaemia (CML) patients, optimum clinical management is still impeded by lack of reliable predictive and prognostic markers to identify those at risk of progression from chronic phase (CP) to advanced disease (AD), i.e. accelerated phase (AP) and blast crisis (BC). Significantly, SHP-1 (Src homology 2 domain-containing phosphates-1) is reported to bind to BCR-ABL1. Furthermore, SHP-1 knockout mice fail to thrive and develop a myeloproliferative-like disease, while point mutations have been detected in patients with acute myeloid leukaemia (AML) disorder. These observations are consistent with SHP-1 function as a negative regulator of PI3K-Akt pathway and a tumour suppressor. More recently data implied SHP-1 expression levels were prognostic and predictive of TKI response. Therefore we quantified SHP-1 mRNA levels in CML patients to assess these findings. In addition, we included acute myeloid leukaemia (AML) and adult blood donor samples controls.

SHP-1, BCR-ABL1 and GUSβ (endogenous control gene) and MSI2 mRNA levels were retrospectively assessed by Taqman quantitative real time polymerase chain reaction in 78 highly heterogeneous CML patients [median age of 50 years (20–76); M: 44; F: 34]; 54 in CP; 24 in AD (6 in AP and 18 in BC). Among the 78 patients 59 were treated with two or more agents. But 19, all in CP, were prescribed single drug only; Imatinib (n=11); nilotinib (n=6); dasatinib (n=1) and interferon + cytarabine (n=1). Sanger sequencing of the BCR-ABL1 kinase domain in 40 (CP: 27; AD: 13) of the 78 patients identified 18 (CP: 9; AD: 10) with one or more mutations. In 8 patients the mutations mapped to the P-loop, 3 had T315I in isolation or in combination with another KD variant and 7 had non-P-loop variants. Seventyseven diagnostic samples from AML patients (M: 42; F: 35) with a median age of 63 years (8–85) were included and 18 normal control (NC) samples from blood donors with median age 44 years (35–61). Target gene expression levels were reported as ratio to GUSβ. Samples with <5500 GUSβ copies were excluded from the study.

SHP-1 expression was significantly higher (p=<0.0001) in CML patients (median: 31.54; range: 0.82–675.1) when compared with NC (median: 3.66; range: 1.40–6.36). Among the CML patients SHP-1 mRNA copies were significantly lower (p=<0.0002) in AD patients [median 14.0 (0.8–211.9)], compared with CP (median 37.7; range: 5.2–675.1). However, there was no significant difference between 9 patients failing to achieve a major molecular response (MMR) within 18 months and 7 patients who did. Furthermore, among the CP patients we observed no significant difference in SHP-1 mRNA levels between those patients prescribed 1 (n=18), 2 (n=7), or ≥3 TKI (n=8), which generally correlates with optimal, sub-optimal and/or failed response. Similarly, we found no significant difference in SHP-1 expression between mutated KD and wild type alleles in CP. The number of patients in BC was limited to assess this. Importantly, we observed no significant difference between AML and NC samples (p=0.801). But there was a significant difference between CML patients in AD and normal, p=<0.0001. We and others previously reported a significant increase in expression of MSI2, a stem cell renewal regulator, between CP and AD, but we found no correlation among 51 CML patients assessed for MSI2 and SHP-1 mRNA levels.

We report here differential expression of the putative tumour suppressor SHP-1 in CML patients in CP and AD, with lower levels in the latter. However, SHP-1 expression in these two groups was higher than that detected in AML and NC samples. In contrast to earlier reports we did not observe a significant difference between those achieving and failing MMR within 18 months nor between patients with different degrees of response to TKI therapy. This variance may have been affected by the different time points of sample acquisition during the course of the treatment. Our data imply SHP-1 regulates or is regulated by BCR-ABL1. Moreover, it is reported that SHP-1 may counteract oncogenic effect of BCR-ABL1, however, we observed no correlation in mRNA levels between them in our 78 CML patients. The data presented here warrant a prospective study to assess whether SHP-1 expression levels is able to identify patients at risk of progression prior to other markers, e.g. BCR-ABL1 copy numbers.