Outcomes of Abdominal Thrombosis in Patients with Myeloproliferative Neoplasms in a Single Center Experience

Background

Abdominal thrombosis (AT) is a concerning complication of myeloproliferative neoplasms (MPNs), leading to significant morbidity and mortality. While the epidemiology of AT in MPNs has previously been described, outcomes based on timing of AT relative to MPN diagnosis are unknown. Additionally, it is unclear how the treatment of the MPN affects outcomes including esophageal variceal bleeding (EVB), development of ascites, and additional thrombosis.

Methods

We conducted a retrospective review of patients at a single tertiary care institution. Inclusion criteria included: 18 years or older, a diagnosis of an MPN, including polycythemia vera (PV), myelofibrosis (MF), essential thrombocythemia (ET), and AT, including portal vein thrombosis (PVT), Budd Chiari Syndrome (BCS), or other splanchnic vein thrombosis. Primary outcome measures included EVB, additional thrombosis, development of ascites, hepatic encephalopathy, and death due to any cause. Years to outcome events were calculated by Kaplan Meier analysis.

Results

Baseline disease characteristics are summarized in Table 1. Sixty-four eligible patients were identified, 46 (72%) were female. The median age at time of AT was 45 years (range, 18-89). PV was the most common MPN, followed by ET then MF. Sixty patients (95%) harbored a mutation in JAK2.

Characteristics of the AT are summarized in Table 2. Twenty-nine patients (45%) were diagnosed with AT after the MPN, with a median of 44 months (1-288) between diagnoses. Nineteen patients (30%) were diagnosed with AT before MPN, a median of 4 months (1-90). Sixteen patients (25%) were diagnosed concurrently (within 1 month). There was no difference in age at diagnosis of MPN among the three groups, however, patients diagnosed with AT before MPN were significantly younger (37 [18-89]) than those diagnosed with an AT after MPN (52 [31-85]) or concurrently (48 [20-70]) (p=0.0045). There was no significant difference among these three groups with respect to other AT characteristics.

The median overall survival (OS) of the cohort was not reached. Five-year OS probability was 98%. No significant difference in overall survival (OS) was observed among those diagnosed with an AT before, concurrent, or after being diagnosed with an MPN. Treatment of the AT was primarily with warfarin (39%), although 19% of patients were treated with a direct oral anticoagulant. Additionally, 20% of patients received a transjugular intrahepatic portosystemic shunt (TIPS). Seven patients (11%) received no treatment for their AT.

Of the entire cohort, 16 patients (25%) experienced an EVB. For patients who were on MPN directed therapy at time of AT, the hazard ratio (HR) for years to bleeding event was 1.24 (0.28-5.57) as compared to those who were not treated (p=0.7798). Ten patients (15.6%) experienced a non-abdominal thrombosis, predominantly deep vein thrombosis or pulmonary embolism. Thirty patients went on to develop ascites. HR for years to ascites from thrombosis was 1.94 (0.17-21.64) in the MPN treated patients (p=0.5823). Six patients (9%) developed hepatic encephalopathy. There was no difference between the MPN treated patients and non-MPN treated patients in a composite outcome of EV bleed, additional AT, ascites, or hepatic encephalopathy.

Conclusions

In this single center experience, a significant portion of patients were diagnosed with an AT before their MPN diagnosis. Treatment of the MPN at time of AT diagnosis did not appear to affect outcomes including EVB, additional thrombosis, development of ascites, or hepatic encephalopathy. These results suggest that once developed, treatment of the underlying MPN may not decrease complication rates of an AT. Further analyses are underway to clarify whether this finding is true in each AT subtype.

View largeDownload slide

View largeDownload slide

Close modal