Introduction: BP of PV presents with dyserythropoiesis, including ring sideroblasts, and bilineage dysplasia in most cases. Clonal evolution or acquisition of new cytogenetic clone(s) plays a critical role in progression to BP of PV. SF3B1 mutation is associated with 75% of MDS with ring sideroblasts (RS). We hypothesized that an increase in ring sideroblast counts is an early event in the BP evolution. We postulate that genomic lesions other than in splicing factors genes are implicated in RS development in PV during BP evolution. Methods: We identified all cases in the last 14 years (2004 - 2018) with a diagnosis of PV in the MDACC files. We collected demographic, clinical, morphological, cytogenetic, and NGS information from 61 patients who developed BP progression from PV. We did an assessment and scoring of all bone marrow (BM) samples available at BP evolution. Results: A total of 61/477 (13%) patients with diagnosis of BP of PV were identified. Median age at BP diagnosis was 67 yrs. (32-82). This included 34 (56%) men and 27 (44%) women. At BP presentation, median BM blast percentage was 28%; 51 (84%) patients had BM dysplasia; 35 patients (69%) with dyserythropoiesis, 22 (43%) cases showed bilineage dysplasia. At BP, 43 pts had complex CG, 9 were were normal karyotype, 3 had double and 5 had single abnormalities. Prussian blue stain for Iron evaluation was available in 54 patients (89%); 24 patients (44%) had 0% RS; 19 patients (35%) had 1-14% RS; and 11 pts (20%) had >15% RS. Mutation analysis for splicing factor was performed on 13 patients. Of these, 2 patients with >15% RS had SRSF2 mutation and 1 patient with 6% RS had SF3B1 mutation. The other patients were wild type for SF3B1, SRSF2, U2AF1 (5 patients: >15% RS, 2 patients: 1-14% RS and 3 patients: 0% RS). Twenty-four (39%) patients had BM available before the BP diagnosis; median time between first BM at MDACC and BP BM diagnosis was 35 months (range, 1-135). At Baseline BM assessment, median BM blast was 3%; 13 (54%) patients had BM dysplasia. Prussian blue stain for Iron evaluation was available in 13 pts (55%): 10 pts (77%) had 0% RS and 3 pts (23%) had 1, 3, and 60% RS, respectively. To evaluate the level of RS during disease evolution we identified 10/26 pts who had at least 1 BM sample available between the first BM and the BP. 5/10 pts were in spent phase with increased blast, multilineage dysplasia and complex CG (RS: 6%, 25%, 0%, 1% and 50%) median time between samples 5 months. 4/10 pts had 1-4% blast, isolated dyserythropoiesis and normal diploid CG (RS: 12%, 5%, 2% and 0%). The median time between samples was 56 months. The last pt. had 0% blast, no dysplasia, del(20q) on CG and no RS. Pt. developed BP of PV 41 months after this follow up sample. Ten pts with PV chronic phase and were used as control group, median follow up of 14,5 yr. None of the patients had RS at baseline or at last follow up BM. Median time between samples 6 yrs. RS increase during PV evolution were statistically significant more frequent in pts who develop BP than in pts who remain in chronic phase of PV (7/10 vs. 0/10. P=0.003) Conclusion: Splicing factor gene mutation is infrequent in BP of PV despite the presence of dyserythropoiesis and frequent RS in 55% of patients. When present , RS may identify disease evolution and correlate with the disease phase of PV before BP. Dyserythropoiesis and the acquisition of RS is an early event BP of PV that precede other markers of disease progression like CG. Disclosures No relevant conflicts of interest to declare.