Management of Chronic Myeloid Leukemia (CML) during Pregnancy Among Patients (pts) Treated with a Tyrosine Kinase Inhibitor (TKI): A Single-Center Experience

Background: Exposure to TKI has teratogenic effects in animal studies and humans. CML is being identified in younger patients who need to continue treatment for long periods of time, making management of CML during pregnancy an increasingly common medical issue, one for which there is very limited information.

Methods: We reviewed the records of all CML pts treated and identified 42 pregnancies involving 34 patients (27 females and 7 males) who were either diagnosed with CML during pregnancy or were receiving TKI for CML at the time they became pregnant.

Results: Thirteen women were diagnosed with CML during pregnancy. Of these, 4 had no complications; 8 had leukocytosis managed with short pulses of hydroxyurea (n=5), short-term interferon-alfa (n=1), leukapheresis (n=1) or observation (n=2). Pts were followed with weekly blood counts; PCR for BCR-ABL was checked at a median interval of 6 weeks (range, 4-8). These pregnancies resulted in healthy babies on term (n=12, including 2 twins), spontaneous miscarriages (n=2; at weeks 4 and 27 of gestation), and 1 elective abortion. One baby had 1 episode of seizure at birth attributed to delivery complications, later growing and developing normally. Within 1 month (mo) of delivery or abortion, all women started imatinib (n=2), dasatinib (n=6) or nilotinib (n=5). Best responses were MR4.5 (n=5) within 3-12 mos, and MMR (n=7) within 6-48 mos. One pt had no response to dasatinib and transformed to blast phase within 3 mo. Four of these women later conceived 5 other pregnancies during the course of their disease.

Twenty-five patients (7 men and 18 women) conceived while on TKI, resulting in 29 pregnancies (20 unexpected and 9 planned). After a median time of exposure to TKI of 17 mos (range, 11-56), all male patients receiving nilotinib (n=5) or dasatinib (n=2) fathered healthy babies. Upon recognition of unplanned pregnancy, all female patients discontinued TKI immediately: dasatinib (n=9), imatinib (n=4), nilotinib (n=1), and bosutinib (n=1). At the time of pregnancy, all women had been receiving TKI for a median of 35.5 mos (range, 5-116). The median time of exposure to TKI from conception to discontinuation was 3 weeks (range, 2-6). Two women had miscarriages and 2 electively terminated pregnancies (including 1 therapeutic abortion for molar pregnancy following bosutinib). Of all 12 born babies, 2 had minor abnormalities (hypospadias and rotation of small intestine, respectively, both successfully surgically repaired). Two uncomplicated pregnancies are ongoing. At the time any pregnancy was recognized, clinical response was MR4.5 (n=6), MMR (n=6), CyR (1 major; 1 complete) and CHR (n=2). The median time pts stayed off-TKI was 10 mos (range, 7-21) and 6/15 women (40%) lost their responses during pregnancy. At the end of pregnancies, all women but 2 resumed TKI. At a median follow-up of 42 mos (range, 13-160), the 6 pts who lost responses have achieved MR4.5, MMR and CCyR (2 each).

Of the 9 planned pregnancies (involving 5 women), all were receiving dasatinib and had undetectable BCR-ABL. All 5 pts stopped dasatinib prior to conception for a median of 4 mos (range, 1-20). During pregnancy, 2 pts had gradual increase in PCR that became undetectable 3-6 mos after resuming dasatinib post birth. All pts resumed therapy a median of 1 mo (range, 0.5-6) after delivery. Of the 5 women, 3 gave birth to healthy babies carried to term, 1 premature and 1 miscarriage.

Conclusion: Conception among CML pts chronically exposed to TKI can lead to uncomplicated pregnancies. Although 40% of pts may lose response following treatment interruption, pts will typically regain response upon resuming therapy and rarely is therapy of any kind needed during pregnancy. Adequate contraception while on TKI and planning pregnancy while in deep molecular response with close monitoring remain the best advice.