Serum Tryptase Is a Strong Predictive Biomarker That Improves Prognostication in Ph+ Chronic Myeloid Leukemia.

Abstract 2783

Although the prognosis of patients with chronic myeloid leukemia (CML) has improved significantly with the advent and use of novel BCR/ABL tyrosine kinase inhibitors (TKI) in recent years, still a number of patients fail to achieve a durable molecular remission or they relapse. It is of particular importance to define the prognosis in individual patients as early as possible, as timely intervention with novel TKI may improve long term outcome. Basophilia is one of the most predictive and best documented biomarkers in CML at diagnosis. However, in high risk patients, the basophil-committed compartment of the clone may exceed the morphologically identifyable fraction of mature granulated basophils detected in MGG or Giemsa stains. In addition, even granulated basophils may be immature cells and may therefore escape microscopic examination. Tryptase is a serine protease primarily expressed by mast cells but also by immature basophils. In healthy controls, the serum tryptase level ranges between <1 and 15 ng/mL. In freshly diagnosed CML, the serum tryptase level has been reported to be either normal or elevated. In those with elevated serum tryptase, the enzyme is considered to be expressed and released by immature CML basophils. In the current project we asked whether tryptase levels at diagnosis correlate with the phase of disease and/or represent a prognostic marker in CML. To address this question, serum samples from a total number of 69 chronic phase (CP) CML patients (median age: 55, range: 21–84; f:m ratio = 1:1.55) and 10 patients with advanced disease (i.e. 9 patients in accelerated phase = AP and one in blast phase = BP) seen at the Medical University of Vienna between September 1998 and June 2011, were analyzed retrospectively. The median serum tryptase level at diagnosis was significantly lower in CP patients (11.7 ng/mL, range 1.4–65.5 ng/mL) compared to patients in advanced disease, i.e. AP/BP (29.9 ng/mL, range 8.1–67.7 ng/mL; p<0.05). We also found that serum tryptase levels correlate with basophil counts (p<0.05). The median observation time in our CP patients was 3.5 years. All patients received therapy with TKI (imatinib n=66; dasatinib n=2; nilotinib n=1). The median progression-free survival (PFS) in our CP patients was 2.7 years. Overall, 12 CP patients (17%) showed a progression. Eight of these patients (75%) were still alive at the end of the observation period. In the cohort of CP patients with elevated tryptase (>15 ng/mL), 31% showed a progression, whereas in the cohort with normal tryptase (≤15 ng/ml), 9% of the patients were found to progress (p<0.05). Of all peripheral blood parameters analyzed, including white blood cell counts, platelets, eosinophils, basophils, and peripheral blast cells, only basophils and tryptase were found to be significant prognostic variables in univariate analysis. As assessed by multivariate analysis performed with the same paramters as above except basophils, tryptase was found to be an independent prognostic variable. Since basophilia is a key variable used in the recently established EUTOS score, we asked whether basophils can be replaced by serum tryptase levels. Indeed, with regard to PFS, this novel EUTOS-T score was found to be of prognostic significance. In particular, the EUTOS-T score was found to predict the PFS of our patients with high confidence and a p-value of 0.004 compared to a p-value of 0.019 in the original EUTOS score. In addition, the EUTOS-T score was found to be a superior score for predicting event-free survival compared. Together, these data suggest that the serum tryptase level is a useful prognostic biomarker in CML. Moreover, the substitution of basophils by tryptase in established scoring systems may improve the predictive value of these scores.