Prognostic Impact of Pregnancy on Long-Term Outcome in Acute Leukemias Patients

Abstract

BackgroundAcute leukemia (AL) during pregnancy - is a rare clinical condition that limits the possibility to conduct large prospective clinical studies. All publications present small retrospective data and case reports. Most of them conclude that pregnancy doesn't affect the prognosis of AL.AimTo assess the pregnancy, as independent prognostic factor, in AL-patients (pts), prospectively treated within Russian AL multicenter studies.MethodsFrom 1990 to 2017 the Russian AL-study group has treated 73 pregnant women with de novo AL 18-42 yrs (Me-27y): acute myeloid leukemia (AML) - 35, acute lymphoblastic leukemia (ALL)- 27 and acute promyelocytic leukemia (APL) - 11. AL was diagnosed at the different gestation age with 89% - in the IInd-IIIrd trimesters.In 8 pts at the 1st trimester medical abortion was conducted. 17 pts delivered at the 34-40 weeks before chemotherapy (CT) and 48 (66%) pregnant women received CT, that was started at 23 (14-32) weeks of gestation.Pregnancy was not an exclusion criteria for standard AL-treatment approach. APL-pts were treated by 7+3 (Dauno 60 mg/m2)+ATRA or AIDA-protocol. ALL-pts were included in the Multicenter ALL-2009 trial (NCT01193933). In AML pts standard 7+3 was applied: either with Dauno 45-60 mg/m2 (NCT01587430), or Mito10 mg/m2, or Ida 12 mg/m2 according to the on-going study-protocol.The control group consisted of the childbearing age women (15-51y) with the same type of AL, who were treated by the same protocols within 2011-2017 yy in NRCH (16 APL, 28 AML and 129 ALL).Results54,5% (n=6) of APL-pts were referred to the low/intermediate (L<10*109/l; Plt<>100*109/l) and 45,5% (n=5) to the high risk (HR) group (L>10*109/l). In the control group the distribution of the APL-pts according to the risk factors was similar: 75% and 25%, rsp.The cytogenetic risk group was estimated in 27/35 pregnant women with AML. 52% (n=14) were referred to the intermediate and 48% (n=13) to the HR group. The HR group comprised complex karyotype (n=5), -7/del7 (n=4), MLL gene translocations (n=2), inv3/-7 (n=1) and AML with normal karyotype, FLT-3+ (n=1). In AML non-pregnant women HR was detected in much less pts - 7% (p=0,001*).In ALL 37% of pregnant women (n=10) were referred to the standard and 63% (n=17) - to the HR group (BI, L>30,TI,II,IV, L>100, LDH>2N, t(4;11) and t(9;22)). There were no differences in comparison with non-pregnant ALL pts: 28,2% and 72,1%, respectively.Overall induction results in pregnant and non-pregnant women treated according the standard CT protocols were comparable in all AL subtypes (Tab.1). It should be noted that early death rate in APL pregnant women was slightly higher than in the control cohort. And in AML - despite the high portion of HR cytogenetics - CR rate was similar to the control.All differences appeared at the long-term follow-up (Tab.2). High relapse rate was noted in APL pregnant women (n=3), and the main reason for that were the refusals of 2 pts to continue CT due to postpartum psychological problems. 5-y OS for APL pregnant pts constituted 54,5%, DFS - 40,9%. APL-pts in the control group are alive in CR.The long term outcome in ALL pts represented the same results in both cohorts. 5-y OS and DFS were 70% and 55% in the pregnant ALL pts, and 60% and 65% - the control group. So we may assume that in ALL pregnancy does not affect the prognosis.On the contrary in AML we registered very unsatisfactory long term results. For AML pregnant pts 5-y OS was 23,5% and DFS -21,3% in comparison with 58,6% and 59% in the control group. The differences are highly significant though the same treatment protocols were applied. In order to evaluated the role of allo-HSCT in AML-pts, we performed a landmark analysis (land-mark = 6 mo of CR - median time to allo-HSCT). We note, that if allo-HSCT was performed in the 1st CR in AML pts, diagnosed while pregnancy, all those pts are alive; in contrast - 5-y OS without allo-HSCT constituted only 11,9% (p=0,03). In a multivariate analysis pregnancy and allo-HSCT in the 1st CR became the independent prognostic factor for AML, diagnosed in women during pregnancy.ConclusionThe long term results in APL-pregnant women in our study were somewhat worse than in non-pregnant women, mostly due to the poor adherence to the treatment after delivery. The survival of ALL-pregnant women did not differ from the general population. And AML diagnosed during pregnancy was characterized by extremely poor outcome, and these pts are the candidates for allo- HSCT in 1st CR. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.