Changes in disease responses after discontinution of tyrosine kinase inhibitors (TKI) for pregnancy in established chronic myeloid leukaemia (CML): Evidence for the basis of advice for women of child-bearing age

Abstract The ability of women of child-bearing age to have successful pregnancies despite their diagnosis of chronic myeloid leukaemia (CML) is now well-documented. All are advised to discontinue their tyrosine kinase inhibitor (TKI) before, or immediately on discovery of, conception. What is less well-known is their risk of losing their pre-existing disease responses during their pregnancy, which in turn limits our ability to advise best practice. We collected the disease responses before, during and after pregnancy from women with established CML, to better understand the kinetics of their disease whilst off TKI, and after re-starting treatment. Our study was restricted to pregnancies resulting in live births. 152 pregnancies were observed in 107 women (median age 33 yrs, range 18-43), after a median treatment duration of 6 yrs (0.3-19.4). 28 infants were premature (<37 wks gestation) and fetal abnormalities including 5 cardiac, were seen in 10 (6.4%). 2 patients with poorly controlled disease and poor compliance prior to unplanned conceptions progressed to blast phase: both are alive after allografting. We initially focussed on 107 first pregnancies to observe disease responses unaffected by previous periods of treatment discontinuation. The median duration of TKI therapy was 4.7 yrs (0.3-15), 67 (63%) of pregnancies were planned and 57 women (53%) were on 1st line therapy at, or prior to, conception. Molecular responses at time of stopping TKI were molecularly undetectable disease, MR4.5, MR4, MMR, CCyR or less than CCyR in 13%, 22%, 18%, 31%, 9% and 7% respectively (84% in at least MMR). Of 38 women in at least MR4.5 at stopping, 19 (50%) lost MMR with 13 losing CCyR. Of 57 women in at least MR4 at stopping, 32 (56%) lost MMR with 23/32 also losing CCyR. Duration of TKI therapy impacted the risk of losing MMR but only after 3 years duration of at least MR4: 73% of women in MR4 for <3 yrs lost MMR compared to 40% in MR4 for >3 yrs. Achievement of MMR only prior to conception did not protect against loss of MMR whilst off TKI: 26 of 33 women (78%) whose best response was MMR, lost MMR with 18 losing CCyR and 2 losing CHR. There was no clear impact of duration of MMR on the risk of loss of MMR with 24% and 13% of women treated for more or less than 2 years retaining MMR. 32 women received interferon during their pregnancy of whom 30 were In at least CCyR at the time of stopping TKI: 24, 3 and 3 lost CCyR, MMR or MR4 off TKI: 4 of the 24 regained CCyR after starting interferon. 2 patients remained in at least MMR on interferon and stayed on interferon after delivery without re-starting TKI. 3 women received TKI after the first trimester for loss of prior responses. The median duration off TKI for all 152 pregnancies was 10.2 mths (2.3-72.4). 25 women remained in at least MMR and did not re-start a TKI, 2 remained on interferon only, 2 are yet to re-start and 1 was treated for disease progression, Of 120 women who have been re-treated with a TKI for at least 6 months, 97%, 94% and 68% achieved or remained in CCyR, MMR and MR4 respectively. The median times to CCyR, MMR or MR4 were 3.1, 3.9 and 6 mths respectively.In conclusion patients in at least MR4 behave as candidates for TKI discontinuation with 44% remaining in at least MMR at the time of delivery. The high number of these patients who also lost CCyR is most likely attributable to the inability to re-start TKI at the time of MMR loss. In contrast stopping TKI for pregnancy when in MMR only is associated with a high risk of loss of MMR with no obvious protection for longer durations of response. Despite this the incidence of disease progression was <2% and probably explicable by unplanned pregnancies in poorly controlled disease. The role of interferon in maintaining or deepening prior responses remains unclear. We cannot exclude the possibility that the occurrence of congenital abnormalities at higher than background levels (3%) may be attributable in some to delay in discontinuing TKI. The vast majority of women have regained or improved on prior responses on re-starting TKI. We demonstrate the feasibility of managing pregnancy in CML, highlighting the importance of individualised planning and vigilant monitoring to balance disease control and pregnancy outcomes.