Introduction

Infant outcomes are favourable post in-utero exposure to thiopurines (TP), with no increase in infant infections or adverse growth and developmental outcomes^1-2. However, infant thrombocytosis, anaemia, lymphopenia and liver function derangement have been observed in small studies^1,3,4. We aimed to identify the frequency of and associations with these abnormalities in TP exposed infants and determine the clinical impact.

Aims & Methods

Infants born to participants in the multi-centre prospective PICCOLO-X study receiving TP were compared to infants exposed to biologic monotherapy (BM) in-utero. TP metabolites (6-thiogaunine (TGN) and 6-methylmercaptopurine (MMP)), liver function tests (LFT) and full blood examination (FBE) were taken from cord blood at delivery, and repeated at 6 weeks (6W). Abnormal parameters were repeated thereafter until normalization and infants reviewed by a paediatric gastroenterologist. Infants are followed for 2 years. TGN and MMP levels were assessed using high performance liquid chromatography. The lower limit for detection was 15 for TGN and 100pmol/8 x 10⁸ RBCs for MMP. Age specific biochemical and haematological reference ranges were used.

Results

78 TP and 52 BM infants (IFX 5, ADA 9, 23 UST and 15 VDZ) were included. Delivery outcomes were equivalent between the groups. TGN and MMP were undetectable at delivery in 6/55 and 47/55 respectively, and all tested cleared both at 6W (n=44). The median delivery infant:maternal TGN and MMP ratios were 0.38 (IQR 0.24-0.50, n=49) and 0.18 (0.05-0.45, n=8). Thrombocytosis was seen in 80% (62/77) of all infants at 6W and 77% (49/64) at 3 months (3M). Elevated ALT was seen in >50% of infants at 3M and 6 months (6M), and anaemia in 12% (8/67) at delivery. There was no increased risk of thrombocytosis or transaminitis in TP exposed infants vs controls (Table). Maternal TP shunting (RR 2.06, 95% CI 1.33-3.18, p<0.01) and MMP (coef. 0.03, 0.02) at delivery and preceding infection (RR 1.65, 1.01-2.70, 0.05) were associated with increasing ALT at 6W, as were preceding infection and elevated ALT (1.25, 1.14-1.37, <0.01) and ALP (1.89, 1.49-2.40, <0.01) at 3M. Vaccination prior to bloods increased the risk of elevated ALP at 6W (1.87, 1.11-3.13, 0.02). Neutropenia risk was increased in those with preceding infection (1.94, 1.05-3.59, 0.04) and vaccination (6.33, 1.09-36.83, <0.04) at 3M. Neutropenia at 6W was associated with an increased risk of infection at 3M (4.89, 1.64-14.57, <0.01). Maternal CRP in trimester 2 was associated with increasing ALT at 6W (coef. 0.7, 95% CI 0.4-1.0, p<0.01), 3M (1.6, 1.2-2.1, <0.01) and 6M (1.6, 1.2-2.0, <0.01), as was maternal calprotectin in trimester 3 and infant 6W platelets (0.08, 0.02-1.5, <0.01). There was no increased risk of infection, chronic disease, allergy or growth <10^th percentile in TP vs BM exposed infants at 6W, 3M and 6M.

Infant outcome	Timepoint	Thiopurines (%, n)	Biologic monotherapy (%,n)	Relative risk (95%) p, univariate binary regression	Adjusted coefficient (95%CI) p, Ordinary multivariate linear regression
Thrombocytosis/ Platelet count	6 weeks	80.7 (42/52)	80 (20/25)	1.01 (0.80 – 1.28) 0.93	38.3 (-330.2-406.7) 0.76
	6 months	59.5 (22/37)	62.5 (10/16)	0.95 (0.60-1.51) 0.83	360.8 (-370.8-1092.4) 0.24
Neutropaenia/neutrophil count	3 months	2.94 (1/33)	26.3 (5/19)	0.11 (0.01-0.89) 0.04	1.82 (-4.02-7.67) 0.39
Elevated ALT/ALT level	6 weeks	51.9 (28/54)	50 (13/26)	1.04 (0.65-1.65) 0.88	14.3 (-66.2-94.8) 0.67
	6 months	18.2 (6/33)	7.14 (1/14)	2.55 (0.37-19.24) 0.37	22.7 (-599.1-644.5) 0.89
Elevated ALP/ALP count	3 months	12.5 (5/40)	40 (8/20)	0.3 (0.11-0.83) 0.02	-71.9 (-317.1-173.3) 0.46
Anaemia/Haemaglobin	Delivery	11.7 (6/51)	12.5 (2/16)	0.94 (0.21-4.21) 0.94	-36.5 (-91.7-18.3 0.2

Conclusion

Abnormal biochemical and haematologic outcomes were seen in infants born to women with IBD, but not more commonly in those exposed to TP in-utero, and with favourable clinical outcomes. Abnormalities were associated antenatal maternal inflammation, altered maternal thiopurine metabolism and preceding infant infection and vaccination.

References

1. Kanis SL, de Lima-Karagiannis A, de Boer NKH, van der Woude CJ. Use of Thiopurines During Conception and Pregnancy Is Not Associated With Adverse Pregnancy Outcomes or Health of Infants at One Year in a Prospective Study. Clin Gastroenterol Hepatol. 2017;15(8):1232-1241.e1231.
2. Mahadevan U, Long MD, Kane SV, Roy A, Dubinsky MC, Sands BE, Cohen RD, Chambers CD, Sandborn WJ. Pregnancy and Neonatal Outcomes After Fetal Exposure to Biologics and Thiopurines Among Women With Inflammatory Bowel Disease. Gastroenterology. 2021;160(4):1131-1139.
3. Thomas C, Monteil-Ganiere C, Mirallié S, Hémont C, Dert C, Léger A, Joyau C, Caldari D, Audrain M. A Severe Neonatal Lymphopenia Associated With Administration of Azathioprine to the Mother in a Context of Crohn’s Disease. Journal of Crohn's and Colitis. 2018;12(2):258-261.
4. Flanagan E, Wright EK, Hardikar W, Sparrow MP, Connell WR, Kamm MA, De Cruz P, Brown SJ, Thompson A, Greenway A, Westley I, Barclay M, Ross AL, Kiburg KV, Bell SJ, Group PS. Maternal thiopurine metabolism during pregnancy in inflammatory bowel disease and clearance of thiopurine metabolites and outcomes in exposed neonates. Aliment Pharmacol Ther. 2021.

Disclosure

Nil

Introduction

The maternal safety of thiopurines (TP), including azathioprine (AZA) and mercaptopurine (6MP), in pregnancy for IBD management is well established. However, intrahepatic cholestasis of pregnancy (ICP) occurs more commonly in TP exposed IBD pregnancies^1,2,. TP ‘shunting’ with a 6-methylmecrcaptopurine (MMP) to 6-thiogaunine (TGN) ratio of >11, is associated with hepatotoxicity, and increasingly occurs in later pregnancy³. We aimed to explore the mechanisms behind TP exposure and ICP, and identify risk minimisation strategies.

Aims & Methods

Pregnant or preconception participants with IBD receiving 6MP, AZA or biologic monotherapy (BM) were recruited to this prospective multi-centre cohort study. Demographic and disease activity data, as well as TP metabolites, biochemical and haemotologic parameters were obtained prior to conception (PC), in each trimester of pregnancy (T1, T2, T3), at delivery (D), and 6 weeks post-partum (PP). Bile acids (BA) were collected in T2 and T3 in TP exposed pregnancies. ICP was defined by BA >10 with clinical symptoms and LFT derangement. Obstetric outcomes were compared between TP and BM exposed controls. TP metabolites were assessed with high performance liquid chromatography and reported in pmol/8 x 10⁸ RBCs.

Results

131 TP and 147 BM exposed pregnancies were included. 63% of TP pregnancy (82) received AZA and 60% (78) had CD. 12% receiving TP (13/109) had a pre-existing shunting phenotype. 59% (77) received concomitant biologics. 60% (75) had a prior pregnancy, with 5.3% (4) of those having had ICP. TP pregnancies were more likely to be exposed to aminosalycylates (39 vs 14%, p<0.01), and infliximab (32 vs 10, <0.01) and less likely exposed to ustekinumab (8 vs 44, <0.01) and vedolizumab (3 vs 29, <0.01). The median AZA dose was 100mg (IQR 75-150) and 6MP 50mg (IQR 50-75).
Rates of biochemically active disease were equivalent between TP and BM pregnancies. The median faecal calprotectin (FC) for TP exposed pregnancies was 42mg/ml (95% CI 11.9-126.5, n=72) PC, 45mg/ml (9.3-149, 89) in TI, 27mg/ml (12-74, 109) in T2, 21mg/ml (9.8-76, 86) in T3 and 19.7mg/ml (7.9-82.6, 72) PP. For BM pregnancies, the FC was 58mg/ml (95% CI 18.2-251, n=51) PC, 39mg/ml (18.6-150, 75) in T1, 38mg/ml (14.8-106, 80) in T2, 49.9mg/ml (17-189, 68) in T3 and 60mg/ml (13.9-156, 33) PP. FC did not change over the course of pregnancy in TP (p=0.63, Skilling’s Mack) and BM exposed (p=0.73) participants. In TP pregnancies, FC and TGN weakly negatively correlated prior to conception (-0.36, p=0.02, Spearman coefficient), but not at any other timepoint (first trimester; r=-0.07 p=0.59, second trimester; -0.09 p=0.39, third trimester; r=-0.09 p=0.44, postpartum; -0.14 p=0.29).

In those with no dose changes, MMP/TGN increased from PC to T3 (Skilling's Mack p<0.01) (Table 1).

Table: Thiopurine metabolites over the course of pregnancy (median, IQR)

24% (18/76), 25% (26/103), 26% (24/92) and 17% (12/72) were shunting in T1, T2, T3 and at D respectively. 2.6% (3/116) were receiving split dose TP PC, and of these with bloods taken, 50% (3/6) were shunting in T1, 75% (2/8) in T2, 43% (3/7) in T3 & 20% (1/5) at D. Of PC shunters, 38.5% (5/13) had their dose split over pregnancy; 7.7% (1/13) PC, 23% (3/13) in T1 and 8% (1/13) in T2. Overall, 4% had their dose split (4/100) in T1, 7% in T2 (8/111) and 2% in T3 (2/105). In those who had their dose split in T2 for shunting management, the median MMP/TGN ratio decreased from 18 (IQR 6-57) in T2 to 7 (IQR 3-8) in T3 (p<0.01 Wilcoxon sign-rank) and 3 (IQR 2-3.5) at D (p=0.04).
Elevated ALT was observed in 8.1% (5/62) of TP pregnancies in T1, 5.7% (5/89) in T2, 18.3% (15/82) in T3 and 34.5% (20/38) at D. Increasing MMP was associated with an increasing ALT in T1 (0.001, 0.03), T3 (coeff. 0.002, p=0.001), PP (0.003, p=0.02) and at D (0.002, p<0.01). 60% (3/5), 50% (6/12) and 21.1% (4/19) with elevated ALT in T2, T3 and at D were diagnosed with ICP. The risk of ICP was increased in TP (6.7%, 7/105) vs BM (0%, 0/112) pregnancies (p<0.01). With multivariate binary linear regression, ICP risk was increased with shunting in T3 (RR 8.9, 95% CI 1.2-65.6 p=0.03) and at D (11.4,1.18-110.3, 0.04), TP dose increase antenatally (5.3, 1.4-23.5, <0.01) and prior ICP (21.1, 1.8-249.0, 0.02). Increasing MMP in T2 (0.011, 0.03), T3 (0.001, <0.01) and at D (0.006, <0.01) were associated with increasing bile acids. Of the 7 TP pregnancies with ICP, 1 had a history of shunting PC, and 1 had a prior history of ICP. All were shunting in pregnancy, with 2/7 having an MMP/TGN >75 in T3. 4 cases occurred following a TP dose increase. There was no increased risk of adverse obstetric outcomes, including preterm delivery (RR 0.40, 95% CI 0.05-2.79, p=0.3) and low birth weight (0.61, 0.14-2.62, 0.5) in TP vs BM pregnancies.

Conclusion

TP exposure was associated with an increased risk of ICP. In particular, TP shunting in late pregnancy, TP dose increases and prior ICP were associated with increased ICP risk. Shunting can be effectively managed with TP split dosing. TP dose increases should be avoided in pregnancy, and TP metabolites routinely monitored.

References

1.Jharap B, de Boer NK, Stokkers P, Hommes DW, Oldenburg B, Dijkstra G, van der Woude CJ, de Jong DJ, Mulder CJ, van Elburg RM, van Bodegraven AA, Dutch Initiative on C, Colitis. Intrauterine exposure and pharmacology of conventional thiopurine therapy in pregnant patients with inflammatory bowel disease. Gut. 2014;63(3):451-457.
2. Kanis SL, Modderman S, Escher JC, Erler N, Beukers R, de Boer N, Bodelier A, Depla ACTM, Dijkstra G, van Dijk A-BRM, Gilissen L, Hoentjen F, Jansen JM, Kuyvenhoven J, Mahmmod N, Mallant-Hent RC, van der Meulen-de Jong AE, Noruzi A, Oldenburg B, Oostenbrug LE, Ter Borg PCJ, Pierik M, Romberg- Camps M, Thijs W, West R, de Lima A, van der Woude CJ. Health outcomes of 1000 children born to mothers with inflammatory bowel disease in their first 5 years of life. Gut. 2021;70(7):1266.
3. Flanagan E, Wright EK, Hardikar W, Sparrow MP, Connell WR, Kamm MA, De Cruz P, Brown SJ, Thompson A, Greenway A, Westley I, Barclay M, Ross AL, Kiburg KV, Bell SJ, Group PS. Maternal thiopurine metabolism during pregnancy in inflammatory bowel disease and clearance of thiopurine metabolites and outcomes in exposed neonates. Aliment Pharmacol Ther. 2021.

Disclosure

Nil

Introduction

Thiopurines are vital therapeutic agents for inflammatory bowel diseases but often cause severe myelosuppression. The NUDT15 polymorphism, particularly p.Arg139Cys (NUDT15R139C), is strongly linked to severe leukocytopenia and is prevalent in Asian populations. Despite thiopurine use during pregnancy, it remains unclear if NUDT15R139C affects fetal thiopurine sensitivity.

Aims & Methods

The human and murine NUDT15 proteins share 89% homology with the preservation of arginine at codon 139 of human NUDT15 in codon 138 of murine Nudt15. We established a knock-in mouse model with a point mutation (c.412C>T) generating p.Arg138Cys (Nudt15^R138C), mimicking the human phenotype. In the current study, we investigated offspring generated from Nudt15^+/R138C or Nudt15^+/+ female mice that were mated with Nudt15^+/R138C male mice and simultaneously administered with mercaptopurine (MP).To evaluate the effects of thiopurine on the fetus, we evaluated the number of hematopoietic stem cells in the fetal liver, which is the center of fetal hematopoiesis.

Results

In administration with the tolerable MP dose (0.5 mg/kg) for adult Nudt15^+/R138C mice, genotypes of the genotypes of neonatal mice born from breeding between Nudt15^+/R138 male mice and Nudt15^+/R138 female mice were 35 (37.2%) Nudt15^+/+, 59 (62.8%) Nudt15^+/R138C, and 0 Nudt15 ^R138C/R138C. On embryonic day 14.5, Nudt15^R138C/R138C fetal mice showed shorter body lengths and fewer hematopoietic stem cells in the fetal liver than Nudt15^+/+ fetal mice. The number of fetal HSCs was not altered in any of the Nudt15genotypes without MP treatment. However, it was significantly reduced by MP treatment in Nudt15^+/R138C fetal mice (p< 0.01, mean number 1104 in DMSO and 577 in MP) and in Nudt15^R138C/R138C fetal mice (p< 0.01, mean number 927 in DMSO and 402 in MP), but not in Nudt15^+/+ fetal mice (p= 0.39, mean number 1004 in DMSO and 820 in MP). The number of HSCs in Nudt15^R138C/R138C fetal livers was significantly reduced to less than 50% of that in Nudt15^+/+ fetal livers by MP treatment during gestation (p= 0.0089). In administration with the tolerable MP dose (1.0 mg/kg) for Nudt15^+/+ adult mice, genotypes of neonatal mice born from breeding between Nudt15^+/R138 male mice and Nudt15^{++ /}female mice were 23 Nudt15^+/+ and 0 Nudt15^+/R138C. Without MP administration, neonatal mice were born in a Mendelian fashion.

Conclusion

Fetal toxicity may arise if the fetus has more NUDT15 R139C alleles than the mother, despite safe thiopurine doses for the mother. Analysis of the father's NUDT15 genotype may enable safer thiopurine use during pregnancy, emphasizing the importance of considering parental genotypes in therapeutic decisions.

Introduction

Women with inflammatory bowel disease (IBD) are highly recommended to maintain treatment during pregnancy because discontinuation has been associated with flare-ups, which are linked to a worse prognosis for the mother and the fetus. It is widely known that biologic drugs cross the placenta, but most studies focused on the impact on the risk of infection. Data regarding the impact on psychomotor development are very limited, particularly on non anti-TNF biologics.

Aims & Methods

Our aim was to evaluate the impact of the exposure of biologics in utero on the psychomotor development of children during the first year of life.
Data from children included in the DUMBO registry with complete AQS-3 available up to 12 months of age were analysed. DUMBO is a prospective, observational, and multicentre registry endorsed by GETECCU, which enrols pregnant women with IBD throughout 5 years in 60 centres in Spain. This subanalysis to evaluate the neurodevelopment of the babies, follows up mothers and babies since the pregnancy diagnosis until delivery during each trimester and then, every 3 months during the first 12 months of babies´ life. Normal psychomotor development was defined by ASQ-3 scores above the lower limit of normality (referral zone) in all domains. Serious adverse events (SAE) were defined in accordance with the ICH Topic E 2 A Clinical Safety Data Management.

Results

352 children born to 343 mothers were included (9 pair of twins).134 children (38%) had been exposed to biologics in utero; from them, 50 (37%) had been exposed to adalimumab, 44 (32%) to infliximab, 3 (2.2%) to certolizumab, 1 (0.7%) to golimumab, 28 (20%) to ustekinumab, and 10 (7.5%) to vedolizumab. To assess psychomotor development, 2073 ASQ-3 questionnaires were analyzed. No differences were observed in any of the evaluations or domains throughout the first year, except at 12 months, where exposed babies were more likely to achieve normal results in the domain of gross motor skills.
In the univariate analysis between babies with normal development at 12 months and those who did not we only found differences in: Exposure to biologic drugs, which was higher in those who presented a normal ASQ-3. The type of maternal IBD, that babies from mothers with UC had lower likelihood of normal psychomotor development, and prematurity, serious adverse events, and hospitalizations, all of which were more frequent in those who presented abnormal ASQ-3 results. In the multivariate analysis, to have been born to a mother with CD (vs. UC) was associated with higher likelihood (OR=2, 95%CI=1.1-3.9), while to be premature was associated with lower likelihood (OR=0.3, 95%CI=0.1-0.6) of having ASQ-3 scores above the limit of normality in all domains at 12 months of age.

Conclusion

In the multicenter, prospective DUMBO registry, the exposure to biologics for the treatment of IBD in utero (including anti-TNF and non-anti-TNF agents) did not impair the psychomotor development of the children. Our findings support the safety of these drugs and reinforce the recommendation to continue them during pregnancy.

Introduction

The evidence on efficacy of vaccination given to children exposed to ustekinumab (UST) or vedolizumab (VDZ) in utero is missing.

Aims & Methods

The aim of our study was to assess serologic response to vaccination in children exposed prenatally to this medication due to inflammatory bowel disease (IBD) of their mothers. Children ≥ 1 year of age exposed to UST or VDZ in utero who had completed at least 1 year mandatory vaccination were included. Serological response to the following vaccination was assessed: non-live vaccines: tetanus, diphtheria; live-vaccines: mumps, rubella, measles. Data on perinatal period were collected using predefined questionnaire filled by mother and/or treating physician and supplied from patient medical files. Information on cord blood levels of biologics were retrieved from patients' medical files.

Results

Twenty-one children were included, 10 exposed to VDZ and 11 to UST. Median age at blood sampling was 26 months (range: 13-49). Except one child exposed to VDZ, all children were born at term with median birth weight of 3250 g (range: 650 - 4000) with 3 (14%) infants having low birth weight. Median gestational week of the last drug administration was week 31 and week 30 in case of VDZ and UST, respectively. Cord levels were measured in 7 VDZ and 10 UST- exposed infants with median (range) cord blood levels of 5.3 mg/l (1.1-21.3) and 2.4 mg/l (0.4-30), respectively. Results on serologic response to vaccination are outlined in Table 1. Tetanus, diphtheria, rubella and measles vaccines demonstrated adequate serologic response in ≥ 80% of children exposed either to VDZ or UST. Regarding mumps lower serologic response was found in children exposed to UST (62.5%), with all infants exposed to VDZ responding adequately.

Table 1. Adequate serologic response to vaccination

	UST	VDZ
Diphtheria	11/11 (100%)	8/10 (80%)
Tetanus	10/11 (91%)	9/10 (90%)
Mumps	5/8 (62.5%)	8/8 (100%)
Rubella	8/8 (100%)	8/8 (100%)
Measles	8/8 (100%)	7/8 (87.5%)

Conclusion

Despite the limited sample size, most of the infants exposed to VDZ or UST in utero were successfully immunised to non-live (tetanus, diphtheria) and live vaccines. Somewhat lower rate of adequate serologic response to mumps in UST group is in line with previous report of anti-TNFa exposed infants and their controls (1). There is a need for a control group of unexposed children to make comparison.

References

Duricova D. et al. Safety of Anti-TNF-Alpha Therapy During Pregnancy on Long-term Outcome of Exposed Children: A Controlled, Multicenter Observation. Inflamm Bowel Dis 2019; 14;25(4):789-796.

Disclosure

The study was supported by IBD-Comfort foundation.