Diagnosis of essential thrombocythemia at platele counts between 400 and 600x10(9)/L.

BACKGROUND AND OBJECTIVE: Diagnostic criteria for essential thrombocythemia (ET) remain essentially negative, that is, exclusion of other myeloproliferative diseases and causes of reactive thrombocytosis. A platelet count above 600x10(9)/L is still generally considered an absolute diagnostic criterion although new protocols for positive diagnostic criteria have recently been proposed, reducing the stringency of a definite platelet limit. This study demonstrates that a platelet count 600x10(9)/L is not a reliable diagnostic criterion for ET, especially in the early stages. DESIGN AND METHODS: An ongoing retrospective study by the GIMMC analyzed 2,316 ET patients diagnosed between 1986 and 1995. Of these 2,316 patients, diagnosed according to the PVSG criteria, 68 had a platelet count 600x10(9)/L and were analyzed separately; 37 out of 68 were excluded from this analysis because of a follow-up shorter than 2 years and/or because of treatment with myelosuppressive agents. The remaining 31 patients were the subjects of our study. RESULTS: After a median follow-up of 4.56 years (range 2-9.6 years) none of the 31 patients had a spontaneous decrease of platelets to the normal range. Transformation to a different chronic myeloproliferative disorders was never observed and no patient developed a condition known to produce reactive thrombocytosis. During follow-up, 23 patients (74%) were treated with anti-aggregating drugs, mainly aspirin. The disease did not evolve into acute leukemia in any patient, 1 had a thrombotic event and none presented hemorrhagic episodes. Median platelet count during follow-up was 534x10(9)/L (range 398-997x10(9)/L). INTERPRETATION AND CONCLUSIONS: Long term follow-up has documented that our 31 patients were correctly diagnosed as having ET, although platelet count was 600x10(9)/L. Our patients were probably in a early phase of their disease and following updated PVSG criteria would have been misdiagnosed leading to incomplete recognition of the natural history of the disease. Further, because an early diagnosis could also have a clinical relevance, our results outline the need for new criteria for the diagnosis of ET. The exclusion of patients with a platelet count between 400 and 600x10(9)/L may prevent patients, nevertheless at risk of vascular complications, from being treated.