Pregnancy Outcomes, Risk Factors, and Gestational Cell Count Trends in Pregnant Women with Essential Thrombocythemia and Polycythemia Vera

BACKGROUND: Myeloproliferative neoplasms (MPNs) including essential thrombocythemia (ET) and polycythemia vera (PV) are commonly diagnosed in the sixth decade, but as many as 20% of patients are younger than 40. This introduces the possibility of pregnancy and need for specialized management. ET/PV patients have a higher risk of pregnancy complications, with a reported 68.5% live birth rate (LBR), and 1.8% and 2.4% risk of major thrombotic and bleeding events respectively. Risk factors for pregnancy complications are conflicting, and further contemporary studies are needed.

METHODS: We retrospectively analyzed 130 pregnancies in 58 ET/PV women seen at 3 affiliate hospitals from 1995-2019. Patients were identified using ICD-9/10 codes in the electronic medical record. We analyzed 12 input variables as risk factors associated with first pregnancy complications after MPN diagnosis by univariate logistic regression (α=0.1, type I error rate; Table 1a). Variables with p<0.10 were included in multivariate logistic regression modeling. Cell counts were averaged at baseline, each trimester, delivery, and the postpartum period.

RESULTS: The analysis included 53 ET patients and 5 PV patients. Driver mutation status for JAK2, CALR, or MPL was available in 79% of all patients: 60.3% were JAK2+ and 12.1% were CALR+. 2 patients were high-risk by International Prognostic Score of thrombosis (IPSET). Median maternal age was 34, median year of delivery was 2010, and median time between MPN diagnosis and index pregnancy was 3 years. There were no complications in 46% of pregnancies. The LBR was 70%. Aspirin (ASA) and interferon were used in 58.5% and 1.5% of pregnancies, respectively. In the pregnancies that resulted in live birth, postpartum low-molecular-weight heparin (LMWH) was used in 23.5% of cases. Antepartum LMWH was used in 6.9% of cases. PV patients were more likely to be treated with ASA (p=0.039) or postpartum LMWH (p=0.013) compared to ET patients. The most common complication was spontaneous abortion (SAB), defined as fetal loss <20 weeks, which occurred in 24.6% of all cases. Thrombosis and hemorrhage occurred in 2.3% and 6.9% of pregnancies respectively.

On univariate analysis, only history of miscarriage prior to MPN diagnosis (13.2% of women) and ASA use during pregnancy were significantly associated with outcomes. On multivariate analysis, history of prior miscarriage remained statistically significant (OR 8.82, p=0.023), while ASA use during pregnancy trended toward improved outcomes (OR 0.33, p=0.12). All other variables were not significantly associated with pregnancy outcomes (Table 1a).

Cell counts throughout pregnancy were available in 53 ET patients (Figure 1). There was a significant decline in platelets (p<0.001) from baseline to delivery. The average percent platelet, hemoglobin (Hg), and white blood cell (WBC) change from baseline to delivery was -42.98%, -9.25%, and +35.65%, respectively. The median nadir was at delivery. In 12 women with postpartum labs drawn within one month of delivery, 7 had recovered their platelet counts to at least 75% of their baseline.

CONCLUSIONS: In this contemporary analysis of pregnancy outcomes in 130 ET/PV women, we demonstrate that overall prognosis is good, with a LBR of 70%. However, pregnant ET/PV patients remain at high risk, with SAB and complication rates of 24.6% and 54%, and maternal thrombosis and hemorrhage occurring in 2.3% and 6.9% of pregnancies. We show for the first time that prior miscarriage is a significant predictor of pregnancy complication in MPN patients (OR 8.82).

Cell counts have never been systematically evaluated in MPN patients throughout pregnancy. We found that platelet counts decrease dramatically, often to normal, in ET. Platelet recovery occurs quickly in the postpartum period, with most patients returning to 75% of their baseline platelet count by one month postpartum. The observed 43% platelet nadir from starting counts to the day of delivery is greater than the expected 20% platelet decrease in the normal pregnant population. This decrease is unlikely to be explained by hemodilution, as the observed Hg decrease in ET patients approximates that seen in normal women. While we found no effect of JAK2 or CALR mutation status on pregnancy complications, it is possible that changes in allele burden influence platelet trends and outcomes. Measuring these changes in available women is the next step in our investigation.

View largeDownload slide

View largeDownload slide

Close modal