Prognostic Factors for Thrombosis-Free Survival and Overall Survival in Polycytemia Vera: A Retrospective Analysis of 623 Patients Series with Long Follow-up

Polycythemia Vera (PV) is a myeloproliferative neoplasm characterized by trilinear marrow expansion and an increased susceptibility to thrombo-embolic complications. Data of 623 patients (pts) followed in 11 Hematological centers of our region from 1978 to December 2010 were collected in our database. The diagnosis was made according to PVSG criteria, WHO 2001 and 2008 criteria, respectively, based on the year of diagnosis. The main epidemiological and clinical features of all pts are reported in table.

Of 623 pts, 161( 25,8%) died, 87 (13,9%) were lost to follow up and 375 (73,1%) were alive at the time of evaluation. The median follow up was 8.5 years.

The thrombotic events during follow-up were 107 (17,2% of 622 evaluable pts): the arterious events were 67 (10.8%), the venous were 40 (6,4 %). The rate of thrombosis (patients/year) was 1,71 %.

At the univariate analysis, the risk factors for thrombosis-free survival (TFS) at diagnosis that resulted statistically significant were: age (> 60 yrs, p= 0,036), WBC (> 10.2 x 10⁹/L, p= 0,034), previous thrombosis (p< 0,0001). The presence of cardiovascular risk factors, Hb (>18.2 g/dL), PLT count (either > 457 x 10⁹/L or >1000 x 10⁹/L), JAK2V617F allele burden > 59.15% and spleen enlargement, did not reach the cut-off value of significance. At multivariate analysis with the Cox proportional hazards model method, age (> 60 yrs, p= 0,049) previous thrombotic events (p< 0.0001) and platelet count < 457 x 10⁶/L (p= 0.019) maintained an independent prognostic value (p< 0,05), while WBC count (> 10.2 x 10⁹/L, p =0,065) did not. The risk factors significant for overall survival (OS) at univariate analysis were: age > 60 yrs (p <0,0001), WBC> 10.2 x 10⁹/L (p< 0,0001), previous thrombosis (p< 0,0001). diabetes (p= 0,0008), platelet count< 457 x 10⁹/L (p= 0.013) and spleen enlargement (p= 0.02); Hb level < 18,2 gr% showed only a trend of significance (p= 0.056), while allele burden of JAK2 > 59.15% and the presence of at least 1 CV risk factor did not reach the cut-off value of significance. At multivariate analysis, age (p< 0,0001), WBC count > 10.2 x 10⁹/L (p< 0,0001), previous thrombosis (p= 0,0004), diabetes (p= 0.0035) and Hb level < 18,2 gr% (p= 0.0078) maintained their independent prognostic value on OS; in contrast platelet count < 457 x 10⁹/L and spleen enlargement lost their prognostic significance.

In conclusion our retrospective analysis of a large series of PV confirm the prognostic value of age and previous thrombosis on TFS and OS, while seems to exclude any impact of spleen enlargement and high Hb level. Interesting, PLT count and Hb below median value resulted as independent risk factors for TFS and OS, respectively.