Higher Rates of Perioperative Thrombosis and Distinct Management Practices in Secondary Erythrocytosis Versus Polycythemia Vera

Background: Polycythemia vera (PV) is a myeloproliferative neoplasm (MPN) characterized by prothrombotic JAK2V617F mutations. Elevated baseline risk of thrombosis, plus that of paradoxical hemorrhage, is further compounded perioperatively (Blood, 2008). Thrombosis is controversial in non-clonal yet more prevalent secondary erythrocytosis (SE), though recent studies suggest appreciable thrombotic risk (Blood Cancer J, 2021). Scarce perioperative outcomes data exist for either cohort, accounting for lack of evidence-based guidelines. The objective of this study was to comparatively assess perioperative complication rates (thromboembolic, hemorrhagic, and survival) and management practices in SE and PV.

Methods: Consecutive patients with i) World Health Organization (WHO) defined PV and ii) SE, defined as sustained hemoglobin (Hb)/hematocrit (Hct) above WHO PV thresholds, not meeting PV criteria (JAK2-unmutated/normal-high serum erythropoietin) were recruited from 3 Quebec centers. Endpoint (within 90 days post-surgery) and surgical definitions were according to convention. Standard statistical methods were used (JMP® Pro 14.1.0 software; SAS Institute, Cary, NC, USA).

Results: A total of 117 procedures were undertaken in 78 patients (24 SE, 31%; 54 PV, 69%). SE vs PV patients were predominantly male (75 vs 41%; p=0.005), with similar median age at diagnosis (62 vs 64 years; p=0.74) (Table 1). Cohorts were stratified as having 0-1 or 2+ cardiovascular risk factors; these were proportional among SE and PV (p=0.5), as well as rates of previous thrombosis (p=0.57-0.92).

SE patients, compared to PV, were younger at time of surgery (median 65 vs 70 years; p=0.03) (Table 2). Nature of interventions (major/minor, subtypes: general/orthopedic/cardiovascular, urgent/elective) was balanced (p=0.13-0.9). Blood counts at time of surgery however, differed substantially between groups. SE compared to PV cases exhibited significantly higher median Hb (167 vs 137 g/L; p<0.0001) and Hct (49.9 vs 41.5%; p<0.0001), as well as lower platelet counts (193 vs 277 x10⁹/L; p=0.002); no differences in leukocytes (p=0.9). Regarding therapy, while 90% of PV patients were on antiplatelet drugs pre-surgery, this was the case for only 42% of SE patients (p<0.0001). Few were on anticoagulation preoperatively (n=7/cohort; p=0.15). Therapeutic phlebotomy was ongoing in 41% of SE patients at the time of surgery. Perioperative therapy modifications were documented: antiplatelet agents were stopped in 25 and 19% of SE and PV patients respectively (p=0.71); cytoreduction was held in 16% of PV patients. Antithrombotic prophylaxis was proportional in both cohorts (p=0.08). Few patients in either group were the subject of either pre- (SE vs PV, 9 vs 17%; p=0.3) or postoperative (SE vs PV, 8 vs 24%; p=0.08) hematology consultations, though these were more frequent in PV.

Outcomes data disclosed similar rates of surgical bleeding (4 vs 5%; p=0.7), and postoperative hemorrhage (3 vs 6%; n=0.5) in SE vs PV patients, respectively. Conversely, however, statistically significantly higher total rates of arterial and venous thrombosis were found in SE vs PV (13 vs 3%; p=0.04). No perioperative deaths were recorded.

Analysis of strictly SE patients revealed those having at least one perioperative complication (versus none), had significantly lower leukocyte counts (5.3 vs 7.7 x10⁹/L; p=0.03) and less frequent exposure to antiplatelets (p=0.05). Analysis in PV showed correlation of overall perioperative risk with higher leukocyte counts (p=0.007) and Hb/Hct levels (p=0.003), as well as cytoreduction discontinuation (p=0.01); adverse trends were seen for absence of antiplatelets (p=0.08) and ruxolitinib vs hydroxyurea therapy (p=0.06).

Conclusions: SE compared to PV patients, exhibit unexpectedly higher rates of perioperative thrombosis, conceivably reflecting distinct disease biology but also, inadequate disease knowledge and underappreciation of risk, driving conservative perioperative practices (underutilization of phlebotomy, tolerance of uncontrolled Hb/Hct, limited antiplatelets). Larger studies will be required to validate findings, though this data serves as a preliminary call to vigilance for greater awareness of risk and advocacy for education/prevention, more stringent preoperative workup, and proactive management strategies in SE, as well as PV patients.

Szuber:Novartis: Honoraria. Busque:Novartis: Consultancy.