Imatinib treatment has radically changed the prognosis of patients with CML. However, around 23-32% of patients discontinue this therapy due to lack of efficacy. Second generation TKI are available, which exhibit greater potency, so there is scope to further improve the strategy of selection of the appropriate TKI in the first line setting. Measurement of PTCH1 expression at diagnosis has been proposed as a useful strategy to tailor first line therapy as patients with low PTCH1 expression showed a worse outcome. Signalling via SMO is inhibited by non-Hedgehog ligated PTCH1 in Hedgehog pathway. SMO and PTCH1/SMO expression ratio has also been related to response to imatinib. Our aim was to corroborate imatinib outcome prediction in a different cohort and compare the prognostic power of PTCH1, SMO and PTCH1/SMO.

We have retrospectively studied 101 pre-treatment samples of patients who received first-line imatinib from 14 Spanish centres. Clinical data were recorded in the Spanish CML Registry (RELMC). Informed consent was signed by every patient. Predesigned assays for PTCH1, SMO and GUSB (control gene) were used in single qPCR reactions in duplicates and run in an ABI 7900. Receiver operating characteristic (ROC) curves were plotted for PTCH1, SMO and PTCH1/SMO expression ratio and the area under curve (AUC) was used to compare its capacity to predict imatinib failure free survival (IFFS). For the measurement with higher AUC a threshold was set to divide patients with high and low expression. TKI failure was defined as loss of CCyR, progression to advanced phase disease, death or change in treatment from imatinib due to lack of efficacy. Secondary endpoints were: probability of achieving <10% BCRABL/ABL at 3 months, probability of achieving CCyR; probability of achieving MMR, progression free survival (PFS) and overall survival related to CML (CML OS). TFFS, CCyR, MMR and CML OS were analyzed by Kaplan-Meier analysis and log-rank test. Fishers exact test was employed to analyze the relationship with <10% BCRABL/ABL at 3 months and PFS. All analysis were carried out in an intention-to-treat basis. Age, Sokal, and EUTOS scores were introduced with categorised PTCH1 expression in a forward stepping Cox regression analysis for prediction of IFFS. Sensitivity, specificity and negative predictive values for prediction of IFFS were calculated.

Patient median follow-up was 33 months (2-151). 13 patients (12.9%) showed imatinib failure. The AUC of PTCH1, SMO and PTCH1/SMO expression ratio were 0.72, 0.55 and 0.71. A PTCH1 expression of 0.026 was used as cut-off. Low and high PTCH1 expression groups had a 10 year rates of IFFS of 64% vs 95% (p=0.01), CCyR at 1 year of 91% vs 93% (p=0.261) and MMR at 12 months of 53% vs 81% (p=0.022). Median time of the entire cohort of achievement of CCyR was 6 months. Fishers exact test for achievement of <10% BCRABL/ABL at 3 months was significant (p=0.021). Three patients who progressed to accelerated or blastic phase and two of them who died from CML were included in the low expression group but no significant results were obtained due to the low number of events. PTCH1 expression was the unique independent predictor of IFFS in the multivariate analysis (p=0.023, HR=5.8(1.3-26)). Sensitivity, specificity and negative predictive values were 84.6%, 55.7 and 96.1%.

We have confirmed PTCH1 expression prognostic power and found a greater predictive capacity than SMO and PTCH1/SMO expression ratio. Compared to previous PTCH1 studies this is a more real-to-life cohort, extracted from tertiary and secondary hospitals and the results confirm PTCH1 expression can be applied to very different clinical settings (previous studies had been performed in a cohort from a national CML reference hospital). Maybe the greater prognostic power of PTCH1 expression reflects its biological role in CML expands further than controlling SMO activity. Therefore PTCH1 could be used as a therapeutic target instead of SMO inhibitors which have shown poor results and high toxicity in early phase clinical trials. A reference standard, similarly as made for BCRABL/ABL1 measurement, could be developed with a level of PTCH1 expression equivalent to the cut-off established in this study. In this way PTCH1 expression could be implemented in the clinical setting.

Disclosures

García-Gutierrez:Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; BMS: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Pfizer: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Ariad: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding. Martinez-Lopez:Janssen: Honoraria; Bristol-Meyer Squibb: Honoraria; Novartis: Honoraria, Research Funding; Celgene: Honoraria.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution