Splanchnic Vein Thrombosis in Overt Myeloproliferative Neoplasms: The Mayo Clinic Experience with 107 Unselected Consecutive Cases

Background:

Splanchnic vein thrombosis (SVT) is a characteristic feature of JAK2 mutation-enriched myeloproliferative neoplasms (MPN). MPN constitutes the most frequent cause of Budd-Chiari syndrome (BCS) and non-malignant, non-cirrhotic portal vein thrombosis (Blood. 2012;120:4921). Reported risk factors for recurrent SVT included BCS, thrombosis history, splenomegaly, leukocytosis and absence of warfarin therapy (Blood Cancer J. 2016;6:e493). The main objective of the current study was to share the experience from a single tertiary center.

Methods:

Study patients were selected from institutional databases, with additional cases identified through Advanced Cohort Explorer Tool. Conventional diagnostic criteria were utilized (Blood 2016;127:2391) Diagnosis of SVT in all cases was confirmed by imaging studies. Cases with latent MPN (MPN-U) or evidence of other cancer were excluded. All analyses considered clinical and laboratory parameters obtained at time of SVT.

Results:

i) Presenting features

107 patients (median age 56 years; range 22-84; 62% females) were considered; specific MPN diagnosis was primary myelofibrosis (PMF) 36%, polycythemia vera (PV) 29%, essential thrombocythemia (ET) 29%, post-PV 6% and post-ET myelofibrosis (MF) 1%. Among evaluable cases, JAK2 mutation was detected in 85% and abnormal karyotype in 33%. At the time of SVT, 80% were not receiving aspirin (ASA) therapy and 4% had history of OCP use. Thrombosis history was recorded in 51% and included history of SVT in 38%. 39 (36%) patients had pre-SVT history of splenectomy, including 16 (15%) in the month before. Presentation included abdominal pain 62%, ascites 31% and GI bleed 15%.

ii) Type of SVT and clinical correlates

SVT types included BCS ± other vessels (n=9; 6 PV, 2 ET and 1MF), portal ± splenic (n=73; 38 MF, 20 PV, 15 ET), portal + mesenteric ± splenic (n=23; 14 ET, 5 MF and 4 PV). Comparison of these three SVT categories disclosed significant associations with MPN sub-category (p<0.01; BCS with PV and "portal + mesenteric" with ET), abdominal pain (p=0.02; with BCS and "portal + mesenteric") and ascites (p=0.007; with BCS).

iii) Initial management of SVT

Initial treatment included systemic anticoagulation (SA) only in 39% of the patients, SA + ASA 4%, SA + cytoreductive drug 21%, SA + ASA + cytoreductive drug 4%, and no SA 33%. In addition, trans-jugular intrahepatic portosystemic (TIPS) or other surgical shunting was undertaken in 13% and variceal banding in 14%.

iv) Post-SVT survival, causes of death and prevalence of liver failure

At median post-SVT follow up of 2.6 years (range .02-17.5), 35 (33%) deaths and 5 (5%) leukemic transformations were noted; the corresponding percentages for PV/ET vs MF were 18% vs 53% and 1.6% vs 9%. Causes of death were unrelated to SVT in 17 instances, unknown in 6, liver failure in 5 (all MF), upper gastrointestinal bleed in 4 (3 MF 1 PV), AML in 2 MF cases and thrombosis in one ET patient.

v) Risk factors for post-SVT survival and recurrent SVT

As expected, post-SVT survival was significantly shorter in patients with MF vs PV/ET (HR 4.6, 95% CI 2.2-9.6). In multivariable analysis, older age (p<0.01), MF diagnosis (vs PV/ET) (HR 4.0; 95% CI 1.9-8.4) and presentation with ascites (HR 2.7, 95% CI 1.3-5.8) were independently associated with shorter survival; the latter was also associated with post-SVT complications (p=0.01). When analysis was restricted to the 62 patients with PV/ET, median post-SVT survival was not reached and advanced age was the only predictor of shortened survival (p<0.01), while associations of BCS with PV and "portal + mesenteric" with ET were confirmed (p<0.01). Seven of 8 BCS cases occurred in patients younger than age 60 (p=0.07) and 7 of 8 BCS presented with ascites (p<0.01). The only predictor of SVT relapse was male sex (p<0.01). In general, post-SVT survival and recurrence rate were not affected by type of SVT or initial treatment strategy.

Conclusions:

Outcome in MPN-associated SVT mostly depends on the underlying MPN type, with limited influence from SVT-directed treatment strategies; importantly, there was no overt evidence of adverse effect, from SVT, on overall or leukemia-free survival in PV or ET; liver failure and GI bleed were relatively frequent causes of death in MF-associated SVT. Other noteworthy observations from the current study included association of BCS with PV, ascites and younger age; and recurrent SVT with male sex.