Abstract Background: MPD is found in up to 50% of BCS and 30% of PVT, but its diagnosis is often difficult, bone marrow biopsy (BMB) and endogenous erythroid colony (EEC) formation being the most reliable tests in our experience (Br J Haematol, 2005;129:553) and that of others. We assessed the diagnostic and prognostic value of V617F JAK2 detection in this situation. Methods: Underlying prothrombotic factors were studied in 241 pts (104 BCS, 137 PVT) from 3 centers (Paris, Barcelona, Rotterdam). Based on BMB results (blindly reviewed), 31% of patients had MPD, 67% no-MPD, and 2% were uncertain. Based on combined BMB+EEC results, 29% had MPD, 41% no-MPD and 31% had discordant features. V617F JAK2 mutation was studied in all patients by quantitative RT-PCR. Results: V617FJAK2 was detected in 45% of BCS and 34% of PVT (p=.08). Mean % circulating V617F allele was 37% and 28% in BCS and PVT pts, respectively (p<.04). In pts diagnosed with MPD using conventional criteria, V617F mutation was found in 84% pts with positive BMB, and in 97% pts with concordant BMB+EEC results (table). In pts previously diagnosed as no-MPD, V617F was present in 19% pts with negative BMB and in 7% pts with both negative BMB and EEC. This discrepancy was greater in BCS patients, where V617F was found in 25% pts previously not diagnosed with MPD. V617F was also found in 58% of pts with discordant BMB and EEC results, allowing to increase the rate of definite MPD diagnosis. On the other hand, 16% pts with wild type JAK2 had MPD features on BMB and 20% had positive EEC. Taking into account V617F JAK2 in addition to BMB and EEC, definite MPD incidence increased from 30 to 49% overall; from 32% to 60% in BCS patients; and from 29% to 49% in PVT patients. V617F JAK2 was significantly associated with higher Hb, WBC, and platelet counts, larger spleen size, lower serum EPO, and higher red cell mass. Other prothrombotic states were found in 19/92 V617F patients and 29/144 wtJAK2 patients (NS). Baseline BCS prognostic scores (Child Pugh, Clichy and Rotterdam scores) did not significantly differ between V617F and wtJAK2 patients. BCS and PVT survival was not affected by the presence of V617F JAK2. Conclusion: Detection of V617FJAK2 (1) is almost as common in PVT as in BCS but the % of mutant allele is higher in BCS; (2) is not fully redundant with EEC or BMB findings and increases the diagnosis of underlying MPD by 28% in BCS and by 20% in PVT; (3) has no discernible impact on medium term outcome of BCS or PVT. (Supported by EU: QLG1-CT-2002-01686). % of V617F JAK2 positive All patients BCS PVT BMB / EEC + / + 97 100 94 − / − 7 25 0 Discordant 58 53 62 BMB + 84 75 92 − 19 27 12 EEC + 77 84 72 − 25 38 18 inconclusive 33 36 30