Pregnancy Outcomes in Patients with Myeloproliferative Neoplasms: A Systematic Review and Meta-Analysis

Introduction

Essential thrombocythemia (ET), polycythemia vera (PV) and myelofibrosis (MF) are a group of hematopoietic stem cell-derived clonal disorders collectively known as the classical, Philadelphia chromosome-negative myeloproliferative neoplasms (MPNs). Although MPN diagnoses are typically made in the sixth or seventh decade of life, approximately 20% of ET patients and 15% of PV patients are under the age of 40 years. Pregnancy with an MPN is associated with maternal thrombosis, hemorrhage and placental dysfunction leading to fetal growth restriction or loss. The aims of this systematic review were to determine the risk of maternal and fetal complications in pregnancy complicated by MPN and to quantify the benefit of commonly used interventions.

Methods

A systematic search was conducted using MEDLINE (1946 to July 2018), EMBASE (1947 to July 2018), Cochrane Database of Systematic Reviews (2005 to July 2018), CCRCT (inception to June 2018), and Epub Ahead of Print & Other Non-Indexed Citations (inception to July 2018). Bibliographic references were reviewed for additional studies. There were no date or language restrictions and the review was registered with PROSPERO (CRD42018090680). Studies were included if they reported on maternal or fetal outcomes of pregnancy in patients with an MPN. Case reports and series including fewer than 10 patients were excluded. Duplicate reports were excluded, with only the most recent or most informative included. Data were extracted independently by two investigators (D.M. and S.K.).

The primary outcome was live birth. Secondary outcomes included fetal complications (e.g. intrauterine growth restriction) and maternal complications (e.g. pre-eclampsia, thrombosis, hemorrhage). Study quality was assessed using the Newcastle-Ottawa Scale. Analyses were performed using Open Meta-Analyst version 10.12 and Review Manager version 5.3.

Results

The search strategy resulted in 4191 records of which 22 met the inclusion criteria (Figure 1). The studies included a total of 779 women and 1226 pregnancies. Fifteen of the studies included patients with ET, 3 included patients with PV, and 4 included patients with any classical MPN. There were no studies of MF patients that met our inclusion criteria. Most of the studies were retrospective cohort studies (19) and 3 were prospective cohort studies.

Among 1226 pregnancies, the live birth rate was 70%. Of 821 pregnancies with ET and 171 with PV, the live birth rate was 71% and 51%, respectively. The live birth rate was higher in patients with ET who received low dose aspirin during pregnancy than those managed with observation alone (OR 5.0; 95% CI 2.2 - 11.3; I²26%; Figure 2). There was no further improvement with the addition of low molecular weight heparin (LMWH) to aspirin (OR 2.8; 95% CI 0.67 - 11.7; I²0%). Interferon alpha (INF) during pregnancy was also associated with a higher live birth rate (OR 3.9; 95% CI 1.5 - 10.4; I²2.5%; Figure 3). The presence of the JAK2 mutation resulted in a lower live birth rate (OR 0.58; 95% CI 0.4 - 0.9; I²6%; Figure 4). Maternal events were unaffected by the addition of aspirin (OR 0.48; 95% CI 0.16 - 1.4; I²0%) or INF (OR 0.1; 95% CI 0.01 - 1.5; I²0%).

Risk of bias assessment was performed using the Newcastle-Ottawa scale. Most studies included a representative cohort and follow up was generally adequate to at least 6 weeks postpartum. Only 3 studies included a control group.

Conclusions

The preliminary results of our systematic review determined that the live birth rate is lower in pregnant patients with MPN as compared to the general population. The chance of a successful pregnancy was higher in patients with ET than PV and the presence of the JAK2 mutation increased the risk of fetal loss. Patients who received low-dose aspirin during pregnancy had a 5-fold higher odds of a successful pregnancy than those managed with usual care and observation alone. Cytoreduction with INF improved the odds of live birth by nearly 4-fold. Maternal events were uncommon overall and no intervention was found to be beneficial. Our review is limited by generally small sample sizes and the retrospective nature of most studies. Additionally, given that the studies span up to 4 decades, our results may be affected by changing standards of care. Pregnancy in patients with MPN is uncommon and this is the only systematic review to determine the chance of successful pregnancy and quantify the benefit of commonly used interventions.

View largeDownload slide

View largeDownload slide

Close modal