Biologics During Pregnancy and Breastfeeding Among Women With Rheumatic Diseases: Safety Clinical Evidence on the Road

Abstract

Introduction

Autoimmune diseases affect around 3%–5% of the population. The prevalence is high for some diseases like rheumatoid arthritis (RA) representing ≈0.5%–1% and low for other diseases like systemic sclerosis (≈0.04%) (Jacobson et al., 1997; Cooper and Stroehla, 2003; Gabriel and Michaud, 2009; Schirmer et al., 2012). Systemic autoimmune rheumatic diseases (SARD) are generally more common in women during reproductive age, with female to male ratio up to 13:1 in diseases like systemic lupus erythematosus (SLE) and Sjogren syndrome, making antirheumatic drug exposure during pregnancy and lactation a frequent issue (Jacobson et al., 1997; Petri, 2002; Fava and Petri, 2019).

Pregnancy and SARD are reciprocally related; a flare in disease activity can occur in pregnancy and disease activity negatively affects pregnancy course and outcome (Andreoli et al., 2019; Giles et al., 2019). Therefore, it is crucial to reach and maintain total or near total remission before and during pregnancy for good pregnancy outcomes. Conventional drugs sometimes do not to achieve this therapeutic target, and safe drug choices are limited in pregnancy (Kuriya et al., 2011; Götestam Skorpen et al., 2016; Ngian et al., 2016).

The management of SARD has changed significantly with the revolutionary advent of biological disease-modifying antirheumatic drugs (bDMARD). Prescription of biologics to millions of patients with SARD has surged remarkably over the last two decades. Common SARD treated with bDMARDs include RA, SLE, spondyloarthritidis (SpA), juvenile idiopathic arthritis (JIA), and autoinflammatory syndromes. Upgrading treatment to biologics led to better control of disease activity that was considered resistant, improved quality of life, and prevented long-term functional disabilities of many patients with SARD (Shadick et al., 2019).

It is not uncommon to encounter a woman who desires a pregnancy or has unplanned pregnancy while on treatment with biologic therapies. Hence, several queries arise about issues related to the gestational safety and efficacy of this particular treatment. The overall experience with the use of biologics and the quality of evidence is not as strong as it should be since most of the available studies are observational with limited capability to conduct experimental trials in pregnancy (Sammaritano et al., 2020). More knowledge about long-term outcomes of children born to mothers treated with biologics in pregnancy is still needed, and it is expected to grow with ongoing studies. In 2016, two sets of recommendations were issued guiding the use of conventional or biological DMARDs in pregnant females with rheumatic diseases with not much discrepancy between them except for addressing paternal use and specifying the timing for stoppage of some drugs before conception by the British Society of Rheumatology and British Health Professionals in Rheumatology (BSR-BHPR) guidelines (Flint et al., 2016; Götestam Skorpen et al., 2016; Wu and Ying, 2019). Updated recommendations were published early 2020 with more consistency and evidence for safety (Sammaritano et al., 2020).

In this article, we review the literature available concerning what is considered safe during pregnancy and lactation, what is not, and the future prospects.

General Concepts

Maternal-Fetal Transport of Immunoglobulins

IgG and low levels of IgA are the only antibodies transferred from the mother to the fetus. Early in pregnancy, insignificant amounts of IgG are slowly transported by passive diffusion. Only 5–10% of the maternal level was measured at the end of the first trimester in blood samples obtained by cordocentesis (Jauniaux et al., 1995; Malek et al., 1996). A significant spike in maternal IgG transfer to the fetus starts at weeks 17–22 and increases thereafter resulting in higher, or at least similar, IgG levels in cord blood than maternal circulation at term, providing the newborn with a temporary passive immunity (Jauniaux et al., 1995; Malek et al., 1996; Saji et al., 1999; Simister, 2003; Lozano et al., 2018). The transfer is supported by placental structures such as neonatal Fc receptors (FcRn) as well as other placental Fcγ receptors (FcγR) (Wood et al., 1982).

At the maternal-fetal interface, FcRn are expressed on the placenta together with the three types of FcγR. However, it has been demonstrated that FcRn promotes the most IgG transport via a process of binding at the syncytiotrophoblast layer, endocytosis, and release into the fetal circulation (Figure 1) (Firan et al., 2001). IgG1 are preferentially transported, followed by IgG4, IgG3, and IgG2 (Leach et al., 1996; Simister et al., 1996; Kane and Acquah, 2009; Lozano et al., 2018). On the other side, FcRn on maternal endothelial cells and liver are responsible for maintaining IgG half-life including therapeutic mAbs via endosomal recycling and re-expression to the cell surface and unbinding (Junghans and Anderson, 1996; Pyzik et al., 2019).

Open in a separate window

FIGURE 1

Placental transfer of IgG mAbs. IgG is the only antibody class that crosses the placenta through the syncytiotrophoblast layer. FcRn internalizes the maternal IgG in the endosome and releases it on the other side of the syncytiotrophoblast layer. IgG transmission significantly increases during the third trimester of pregnancy.

Based on these concepts, whole mAbs with the Fc domains, particularly IgG1, will be transferred at high levels to the fetus through placental FcRn. Antibody formulations lacking the Fc portion or less readily transferable IgG subclass like IgG2, will keep the fetus relatively safe from the effects of the biologic therapy. The main concern about the transfer of bDMARD to the fetus is their theoretical potential to disrupt children`s immunity and response to infections and vaccinations.

Biological Drugs Used in Rheumatology

Denosumab

Denosumab is a fully human IgG2 which is the least IgG subclass to be transferred through the feto-maternal interface. Its use for the treatment of pregnancy and lactation-associated osteoporosis, which is a relatively underreported condition with no identifiable treatment consensus, was addressed in some case reports with good results (Sánchez et al., 2016; Ijuin et al., 2017). However, it is contraindicated in pregnancy and lactation following discouraging animal studies which showed high still-birth and newborn mortality rates (Bussiere et al., 2013; Okamatsu et al., 2017).

A summary of the characteristics of the common drugs used in SARD including the risk category according to Food and Drug Administration (FDA) and Therapeutics Good Administration (TGA) is reported in Tables 1, ​,22.

TABLE 1

Common Biological Drugs in SARDs and their indications.

Biologic name (Trade name)	Target	Type	IgG subclass	Structure	Risk category		FDA approved indication	Common off label and investigatory indications in rheumatology
		Schmid and Neri (2019)			FDA	TGA Ngian et al. (2016)
Infliximab (remicade)		Chimeric mAb	IgG1		B	C	-AS	BD Hatemi et al. (2018)
							-RA	Sarcoidosis Loza et al. (2011).
							-Pediatric and adult CD and UC	Refractory OAPS Alijotas-Reig et al. (2019)
							-PsA and plaque Psoriasis	Pyoderma gangrenosum. Maverakis et al. (2020)
Adalimumab (humira)	TNFα	Fully human mAb	IgG1		B	C	-As
							-RA
							-JIA
							-Pediatric and adult CD and UC
							-PsA and plaque psoriasis
							-Uveitis
Golimumab (simponi)		Fully human mAb	IgG1		B	C	-As
							-PsA
							-RA -
							UC
Etanercept (enbrel)		Recombinant TNFα receptor added to Fcγ	IgG1		B	B2	-As
							-RA
							-Polyarticular JIA
							-PsA and plaque psoriasis
Certolizumab (cimzia)		Humanized fab fragment and PEG	—		B	C	-RA
							-As
							-Non-radiographic axial SpA
							-CD
							-PsA
Abatacept (orencia)	CD80	Extracellular region of CTLA-4 added to Fcγ	IgG1		C	C	-RA	RA-ILD Fernández-Díaz et al. (2020)
	CD86						-JIA
Rituximab (rituxan)	CD20	Chimeric mAb	IgG1		C	C	-RA	Refractory LN Stolyar et al. (2020)
								Refractory NPSLE Pamfil et al. (2015)
								Refractory IHA and thrombo-cytopenia Fanouriakis et al. (2019)
								ANCA associated vasculitis Yates et al. (2016)
								DM Oddis et al. (2013)
								SS Ramos-Casals et al. (2020)
								RA-ILD Dai et al. (2020)
								IgG4-RD Brito-Zerón et al. (2016)
Belimumab (benlysta)	BAFF	Fully human mAb	IgG1		C	C	-Active, autoantibody positive SLE	Add-on therapy in LN Ward and Tektonidou (2020)
Secukinumab (cosentyx)		Humanized mAb	IgG1		B	C	-As	NA
							-PsA and plaque psoriasis
Ixekizumab (taltz)	IL-17	Fully human mAb	IgG4		NA	C	-As	NA
							-PsA and plaque psoriasis
Ustekinumab (stelara)	IL-12/IL-23	Fully human mAb	IgG1		B	B1	-PsA	NA
							-CD and UC
Tocilizumab (actemra)	C5	Humanized mAb	IgG1		C	C	-RA	Castleman’s disease
							-GCA	PMR
							-Polyarticular and systemic JIA	Uveitis Rubbert-Roth et al. (2018)
							CRS
Anakinra (kineret)	IL-1R	Recombinant non-glycosylated protein	—		B	B1	-RA	AOSD
								BD
							-CAPS	FMF
								Systemic JIA
Canakinumab (ILARIS)	IL-1	Fully human mAb	IgG1		C	D/X	-FCAS	RA Vitale et al. (2016)
							-MWS	Gout So et al. (2018)
Rilonacept (arcalyst)	IL1	IL-1R binding motifs coupled to the Fc domain	IgG1		C	NA	-CAPS	Gout So et al. (2018)
							-FCAS
							-MWS
Eculizumab (soliris)	IL-6	Humanized mAb	IgG2/4		C	B2	-aHUS	HELLP
							-gMG	C-APS Stefanovic (2019)
							-PNH
Denosumab (prolia/Xgeva)	RANKL	Fully human mAb	IgG2		D	NA	Postmeno-pausal osteoporosis	PLO Sánchez et al. (2016)

Open in a separate window

NA, no available data; ACR, american college of rheumatology; aHUS, atypical hemolytic uremic syndrome; ANCA, anti-neutrophil cytoplasmic antibodies; AOSD, adult onset still’s disease; AS, ankylosing spondylitis; BAFF, B-cell activating factor; BD, behcet’s disease; C, complement; C-APS, catastrophic antiphospholipid syndrome; CAPS, cryopyrin-associated periodic syndromes; CD, crohn’s disease; CRS, cytokine release syndrome; DM, dermatomyositis; EULAR, european league against rheumatism; Fab, fragment antigen-binding; Fc, fragment crystallizable; FCAS, familial cold autoinflammatory syndrome; FDA, US food and drug administration; FMF, familial mediterranean fever; GCA, giant cell arteritis; gMG, generalized myasthenia gravis; IgG, immunoglobulin G; IHA, immune hemolytic anemia; IL, interleukin; ILD, interstitial lung disease; IgG4-RD, immunoglobutlin G4 related disease; JIA, juvenile idiopathic arthritis; LN, lupus nephritis; mAb, monoclonal antibody; MWS, muckle-wells syndrome; NPSLE, neuropsychiatric systemic lupus erythematosus; OAPS, obstetric anti-phospholipid syndrome; PEG, poly-ethylene glycol; PMR, poly-myalgia rheumatica; PNH, paroxysmal nocturnal hemoglobinuria; PLO, pregnancy and lactation associated osteoporosis; PsA, psoriatic arthritis; RA, rheumatoid arthritis; RANKL, receptor activator of nuclear factor kappa-Β ligand; SARD, systemic autoimmune rheumatic diseases; SS, sjögren syndrome; TGA, therapeutic goods administration; TNFα, tumor necrosis factor α; UC, ulcerative colitis.

TABLE 2

Risk categorization by FDA and TGA for use of drugs in pregnancy.

Category	FDA	TGA
A	Adequate and well-controlled studies have failed to demonstrate a risk to the fetus in the first trimester of pregnancy (and there is no evidence of risk in later trimesters).	Drugs which have been taken by a large number of pregnant women and women of childbearing age without any proven increase in the frequency of malformations or other direct or indirect harmful effects on the fetus having been observed.
B	Animal reproduction studies have failed to demonstrate a risk to the fetus and there are no adequate and well-controlled studies in pregnant women.	B1:
		Drugs which have been taken by only a limited number of pregnant women and women of childbearing age, without an increase in the frequency of malformation or other direct or indirect harmful effects on the human fetus having been observed.
		Studies in animals have not shown evidence of an increased occurrence of fetal damage.
		B2:
		Drugs which have been taken by only a limited number of pregnant women and women of childbearing age, without an increase in the frequency of malformation or other direct or indirect harmful effects on the human fetus having been observed.
		Studies in animals are inadequate or may be lacking, but available data show no evidence of an increased occurrence of fetal damage.
		B3:
		Drugs which have been taken by only a limited number of pregnant women and women of childbearing age, without an increase in the frequency of malformation or other direct or indirect harmful effects on the human fetus having been observed.
		Studies in animals have shown evidence of an increased occurrence of fetal damage, the significance of which is considered uncertain in humans.
C	Animal reproduction studies have shown an adverse effect on the fetus and there are no adequate and well-controlled studies in humans, but potential benefits may warrant use of the drug in pregnant women despite potential risks.	Drugs which, owing to their pharmacological effects, have caused or may be suspected of causing, harmful effects on the human fetus or neonate without causing malformations. These effects may be reversible.
D	Animal reproduction studies have shown an adverse effect on the fetus and there are no adequate and well-controlled studies in humans, but potential benefits may warrant use of the drug in pregnant women despite potential risks.	Drugs which have caused, are suspected to have caused or may be expected to cause, an increased incidence of human fetal malformations or irreversible damage. These drugs may also have adverse pharmacological effects.
X	Studies in animals or humans have demonstrated fetal abnormalities and/or there is positive evidence of human fetal risk based on adverse reaction data from investigational or marketing experience, and the risks involved in use of the drug in pregnant women clearly outweigh potential benefits.	Drugs which have such a high risk of causing permanent damage to the fetus that they should not be used in pregnancy or when there is a possibility of pregnancy.

Open in a separate window

Biologic Drug-Induced Maternal and Infant Immunological Changes

Complex immunological changes naturally occur in pregnancy. Pregnant women are generally considered immunocompromized with more susceptibility to various infections (Mor and Cardenas, 2010). Administration of biological drugs surely induces additional variations in the maternal immune responses.

An insufficiently studied aspect is how much the drug level and efficacy are affected by pregnancy status. By measuring maternal infliximab and adalimumab levels each trimester in 25 pregnancies with IBD, an increase in infliximab with the progress of pregnancy was unexpectedly found. Adalimumab showed steady drug levels, although infliximab was discontinued earlier than adalimumab (Seow et al., 2017). Such findings were considered as a possible explanation for the higher cord blood levels of infliximab, which inversely correlated with the interval between the last dose and birth date, and the more delayed infliximab clearance from infant blood in former studies (Mahadevan et al., 2013; Julsgaard et al., 2016). Therefore, the authors concluded that monitoring anti-TNFα drug level, particularly in the second trimester, may offer a guide for third trimester dosing to ensure safety of the baby and protection to the mother from intrapartum or postpartum disease activity.

The 2014 meta-analysis by Narula et al. concluded that the benefit from anti-TNFα therapy overweighs any possible poor outcomes to the fetus or the mother (Narula et al., 2014). However, as their use is becoming more common, there are reports of a higher incidence of maternal infections that might be related to the use of anti-TNFα treatment in pregnancy. A study reported the records of twenty-five children exposed during gestation to infliximab/adalimumab; 80% of them were also breastfed till the median age of 6 months. Seven children had a mild decrease in immunoglobulin levels (IgA and IgG) without relevant clinical signs, while lymphocytic subpopulation counts were normal. Infections, mostly mild upper respiratory tract infections, were recorded in 80% of the babies. However, the reported infections do not seem to be therapy-related since they occurred late after the drug clearance (Bortlik et al., 2014). In the case series of women treated with certolizumab throughout pregnancy, maternal infections developed in three out of 13 included women; however, two of them were on concomitant low dose steroid therapy (Förger et al., 2016). A large cohort of French IBD patients was retrospectively investigated; 1,457 anti-TNFα (mostly infliximab and adalimumab) exposed patients were compared with 9,818 non exposed patients. Maternal infection rate in the first group was significantly higher than in non-exposed patients (Luu et al., 2018).

Another example of confusing maternal responses to biologic therapy during pregnancy is the need to increase doses and frequencies of eculizumab and increase the need for red blood cell transfusions in a subset of pregnant patients with PNH which might be explained by increased lysosomal metabolism of eculizumab and physiologic dilutional pregnancy changes (Kelly et al., 2015). Following treatment of a pregnant APS patient with eculizumab infusions, complement regained normal activity soon after infusions earlier than expected (Gustavsen et al., 2017). For the infant, the cord blood levels of various biologic drugs at birth display different values. Cord blood levels were high for infliximab and adalimumab (≈2.6 and 1.5% fetal to maternal ratio, respectively) (Mahadevan et al., 2013; Kanis et al., 2018), and not detected or negligible in the case of certolizumab, etanercept, and eculizumab (Murashima et al., 2009; Mariette et al., 2018). Most drugs with high levels at birth persisted in fetal blood for 4–9 months (Julsgaard et al., 2016).

Reversible change in immune cell counts is a well-recognized side effect particularly with the anti B-cell therapy. Neutropenia with serious skin infections was reported after treatment with infliximab in some reports (Guiddir et al., 2014). Apparently harmless reversible low infant total and B cell lymphocyte count without a significant rise in infection rates or abnormal vaccination course was reported after exposure to rituximab (Azim et al., 2010; Chakravarty et al., 2011; Das et al., 2018). A similar effect was seen in one case report with belimumab exposure (Bitter et al., 2018). These cellular effects were mostly transient and reverted to normal within months. Following eculizumab exposure, assessment of infant complement revealed normal activity (Hallstensen et al., 2015).

Since patients on anti-TNFα therapy display more susceptibility to mycobacterial infections due to interference with the IL-12/IFN-γ pathway, this pathway was investigated in newborns exposed to infliximab/adalimumab throughout pregnancy. Seven exposed and eight non-exposed children were followed for 12 months. Apart from one child who had recurrent infections after the 6th month of age, all showed no increase in infection rates with good developmental milestones. Leucocyte and lymphocyte counts were normal. T-regulatory cell values were lower than average in exposed children and defects in IL-12/IFN-γ pathway led to subnormal response to mycobacterial challenge. B-cell development and maturity are also thought to be affected by TNFα inhibition. Although this study showed that the B-cell population was shifted more to the naïve cells in exposed children, immunoglobulin levels were normal and antibody responses to diphtheria, tetanus, and pneumococcal vaccines were average (Esteve-Solé et al., 2017).

Breastfeeding

The risk of transfer of biologics in breast milk is generally lower than that of placental transfer. Unlike IgA, overall secretion of IgG is scarce in breast milk. The preferential secretion of IgG subclasses is different from the placental one in which IgG3 and IgG4 are secreted more than IgG1 (Gasparoni et al., 1992). Transfer of the Fab fragment is even much lower than full IgG (Nesbitt et al., 2006). Moreover, mAb oral bioavailability is limited due to their large molecular size and the fact that degradation by proteolytic enzymes (Zelikin et al., 2016; Ryman and Meibohm, 2017). Nevertheless, FcRn are expressed at the intestinal mucosal surface of the newborn and can mediate the transfer of antibodies which escape proteolysis (Clowse et al., 2017; Pyzik et al., 2019).

Following these concepts, breast milk levels of almost all anti-TNFα and other mAb biologics are either very low or undetectable and mostly not be delivered to the nursing child’s circulation (Murashima et al., 2009; Ben-Horin et al., 2011; Clowse et al., 2017; Matro et al., 2018). Nevertheless, the reassurance of physicians for the use in clinical practice needs more studies reporting actual outcomes with breastfeeding.

The best evidence available is for certolizumab. It was not detected in the breast milk of nursing mothers following treatment with certolizumab (Mahadevan et al., 2013; Förger et al., 2016). A post-marketing multicenter study confirmed minimal or no excretion of certolizumab or PEG in breast milk and negligible relative infant dose of certolizumab (Clowse et al., 2017). In parallel, insignificant amounts of etanercept were detected in breast milk following treatment during breastfeeding and the drug was not detected in infant circulation (Murashima et al., 2009; Berthelsen et al., 2010; Keeling and Wolbink, 2010). Breast milk levels of other TNFα blockers and rituximab follow the same pattern as etanercept with low levels excreted in breast milk (Matro et al., 2018). The 2020 ACR guidelines considered all the TNFα inhibitors and rituximab compatible with breastfeeding with a strong recommendation although based on a low level of evidence. No data were available to support recommendations for its use during lactation by the EULAR and the BSR-BHPR guidelines. It is advised against nursing for 6 months after exposure to rituximab by FDA (US Food and Drug Administration, 2018).

There are no or very few data regarding other IgG1 mAbs (belimumab, tocilizumab, secukinumab, ustekinumab) and anakinra. They are conditionally recommended for use in pregnancy putting into consideration patient benefits vs. risks to the child (Sammaritano et al., 2020). Breast milk levels of belimumab were recently reported to be low at 1/200 to 1/500 of maternal values (Saito et al., 2020). Anakinra is profusely excreted in breast milk but infant drug levels have not yet been clearly measured. Since IL1 receptor antagonists are naturally present in breast milk, it was presumed that anakinra would not have a detrimental effect on the infant. That was confirmed by several studies that showed average developmental progress in children breastfed by mothers receiving anakinra (Chang et al., 2014; Youngstein et al., 2017; Smith and Chambers, 2018). Similarly, the IgG2/4κ eculizumab was not detected in breast milk samples and no developmental problems were published (Kelly et al., 2015; Miyasaka et al., 2016).

Table 3 summarizes the guidelines for using biological drugs during pregnancy, lactation, and for male partners according to ACR, BSR/BHPR and EULAR committees.

The data from animal studies are also summarized in Table 3.

Vaccine Responses and Long-Term Outcomes

More informative long-term data regarding outcomes of children after exposure to biological drugs in-utero or during nursing are needed. In addition, precise assessment of responses to vaccines, especially for the first-year doses, is definitely needed to ensure successful immunization after inactivated vaccines and exclude the risk of serious side effects with live-attenuated vaccines. Anticipated persistence of mAb biologics in infant blood and B-lymphocyte depletion following exposure to anti B-cell therapy requires more research to determine the time of clearance of different biologics and proper planning of an effective and safe vaccination schedule. Reports about responses to inactivated vaccines are more available than live-attenuated since it is a common practice to delay live-attenuated vaccines till after the first year.

Although data from drug registries and observational studies report minimal or no side effects in infants exposed to infliximab in-utero, early reports revealed a serious disarray in infants` immune responses to Bacille Calmette–Guérin (BCG) vaccines (Cheent et al., 2010).

A study showed that, postnatally, infliximab was cleared from infant blood over a longer duration than adalimumab (mean 7.3 vs. 4 months), raising concerns about infant`s response to live-attenuated vaccines, particularly BCG, which is still obligatory in many countries (Julsgaard et al., 2016). In this study, authors found that 4/80 (5%) of infants suffered bacterial infections and 16/80 (20%) had viral infections, but most of them were exposed to anti-TNFα in-utero and also to thiopurines.

During a 2-years follow-up period, 34 children exposed in-utero to anti-TNFα agents (8 after the first trimester) showed no difference in growth and developmental parameters nor in the incidence of congenital malformations in comparison to controls. Post-vaccination responses and adverse events were also similar except for one etanercept exposed child who contracted chickenpox infection after insufficient response to the relevant vaccine (Dall'ara et al., 2016). The World Congress of Gastroenterology statements in 2011 recommended giving special attention to the timing of biological therapy during pregnancy and postponing delivery of live-attenuated vaccines to exposed children until the drug is no longer detected in their circulation (Mahadevan et al., 2011).

According to Bortlik et al., serologic antibody responses to first year scheduled inactivated vaccines were found satisfactory except for Hemophilus influenza B (HiB) vaccine which produced below threshold protective antibody levels in 6/15 examined children (Bortlik et al., 2014). In a recent evaluation of the long-term impact of in-utero exposure to anti-TNFα therapy, Duricova et al. followed 72 exposed children older than 12 months of age and compared them with 69 non-exposed children for a duration of follow up that ranged from 3 to 4 years. Most of the mothers had CD and were treated with infliximab, which was used until 17–30 weeks of gestation. There was no significant variation between groups of children regarding growth, psychomotor development, allergy status, and infection rate during the first year of life and the period of follow-up (Duricova et al., 2019). Vaccines caused more adverse events in the exposed group, but that difference was not statistically significant, and none of the adverse events was critical (23 vs. 11.6% p = 0.06). Of the children who received scheduled vaccines, serologic response was adequate in >95% of exposed children except for HiB and mumps vaccines, for which fewer children in exposed and control groups had an adequate serologic response (65.3 vs 12.5%) and (75.7 vs. 81.3%) of tested children (Duricova et al., 2019). In another study, children exposed to biologics (including anti-TNFα and ustekinumab) during breastfeeding in the first year of life showed average developmental milestones and infection rates similar to controls (Matro et al., 2018). Another set of extracted data from the Pregnancy in Inflammatory Bowel Disease and Neonatal Outcomes registry (PIANO) registry was published in 2018 reporting infant serologic responses to HiB and tetanus vaccines after antenatal exposure to biologics including anti-TNFα and ustekinumab. The percentage of infants who had protective serological titers after HiB and tetanus vaccines did not differ between the exposed and non-exposed children, but the comparison group was small and included infants exposed to other immunosuppressive and immunomodulatory agents. Moreover, the response to vaccines was not related to the different titers of the biologics in cord blood at birth. Most of the mothers avoided the live-attenuated Rotavirus vaccine based on medical advice. Sequelae, including diarrhea and fever, were reported in 17% (7/40) of infants receiving the Rotavirus vaccine after exposure to anti-TNFα drugs (Beaulieu et al., 2018).

In an effort to determine the ability to achieve an adequate serological response after receiving an inactivated vaccine, de Lima et al. measured anti-hepatitis B antibodies one year after vaccination and proved that effective vaccination could be achieved in anti-TNFα exposed children similar to non-exposed children. Growth and developmental outcomes were also similar between the two groups at one year of age (de Lima et al., 2018). Recently a Korean study examined 12 children who were exposed to infliximab/adalimumab in-utero with a mean age of 28.3 ± 16.6 months. All showed average growth and development: 33% needed a booster dose after hepatitis B virus (HBV) vaccination for seroconversion which was similar to other non-exposed Korean children. Seven out of 12 children received live-attenuated vaccines before 6 months of age without complications (Lee et al., 2019).

Following the early discouraging results, BCG vaccine was given at different ages during the first year of life in various observational studies with generally few adverse events, but the relationship to the timing of the last dose of biologic therapy was sometimes poorly reported (Bortlik et al., 2014; Luu et al., 2018). Proper timing for BCG vaccination was investigated by retrospectively analyzing records of offspring of 74 women treated by anti-TNFα drugs during pregnancy, including those who received third trimester doses. Adverse events were very few and self-limited in infants who received BCG vaccine at a range of 0.25–11 months and totally absent in infants receiving it at or after the 6th month of age. The authors eventually recommended 6 months after birth as the optimal timing for delivering the BCG vaccine and the same recommendation may be applied to other live-attenuated vaccines (Park et al., 2020).

Few more results about some biologic drugs other than anti-TNFα are continuously published showing supporting evidence of long-term safety. Sixteen children were followed up to 1 year (6–48 weeks) after abatacept exposure in one study. No abnormal increase in infection rate or immunodeficiencies was observed (Kumar et al., 2015). Follow up of breastfed children exposed to anakinra since fetal stage and up to 10 years postnatal in some cases was mostly uneventful with no serious infections or developmental delays (Youngstein et al., 2017; Smith and Chambers, 2018). Response to vaccination after rituximab has been discussed in case reports. All showed adequate antibody responses to vaccines, including Tetanus, Diphtheria, HBV, HiB, MMR, and polio (Decker et al., 2006; Friedrichs et al., 2006; Mandal et al., 2014; Ling and Koren, 2016), except in one case report with an insufficient response after Diphtheria (Kimby et al., 2004).

Indirect Exposure to Biologic Therapy via the Male Partner

A limited number of studies addressed paternal exposure to biological drugs, infertility, and pregnancy outcomes in the current literature. Cohort studies have reported no relevant difference between peri-conceptional paternal exposure to different anti-TNFα and undesirable pregnancy outcomes or congenital malformations with rates comparable to population rates (Viktil et al., 2012; Wallenius et al., 2015; Larsen et al., 2016).

The majority of papers reported no impact of TNFα inhibitors on male fertility (Saougou et al., 2013; Micu et al., 2014; Mouyis et al., 2019). Only a few reported decreased spermatozoa number or questionable abnormality of motility and morphology in relation to infliximab (Mahadevan et al., 2005; Montagna et al., 2005).

With other drugs, outcomes of 10, 54, and 13 pregnancies after paternal exposure to abatacept, secukinumab, and tocilizumab were satisfactory with the majority achieving healthy live births free of congenital anomalies (Kumar et al., 2015; Hoeltzenbein et al., 2016; Warren et al., 2018).

Followed pregnancies after rituximab treatment in the male partner were few. Only 9 of 22 pregnancies could be followed in one study with seven live births, four premature infants, and two spontaneous miscarriages (Chakravarty et al., 2011).

Unlike the EULAR recommendations in 2016 (Götestam Skorpen et al., 2016), the BSR-BHPR and the recent ACR guidelines devoted statements for prescribing antirheumatic drugs in male partners with SARD planning to conceive a pregnancy. Both guidelines recommended the continuation of infliximab, adalimumab, etanercept, and rituximab in men attempting to conceive. The BSR-BHPR did not make recommendations for certolizumab, and anakinra use in breastfeeding due to lack of supporting data, while the ACR guidelines strongly recommended certolizumab and conditionally recommended anakinra continuation (Flint et al., 2016; Sammaritano et al., 2020). Relevant recommendations for other biologics were not made due to limited data.

Table 4 summarizes the data on the transfer of the different biologicals and their effect on infants.

Discussion

Ensuring safety of biological therapies during preconception, pregnancy, and breastfeeding offers an important treatment option in disorders in which disease control is crucial. Several reports and observational studies are reassuring especially in dealing with anti-TNFα therapies. Biologics may offer a safer choice than non-biologic drugs in regard to neonatal infections and congenital anomalies. Heterogeneity in study design, disease populations, target biologics studied, the timing of exposure to biologics, outcome measures, inadequate comparison groups and difficulty in adjustment for confounders such as disease activity do not provide very strong evidence for recommendations.

The largest experience is from IBD cases with exposure to anti-TNFα in pregnancy. TNFα inhibitors seem to be safe and effective without a significant maternal or fetal risk despite previous worries about the risk of an increase in congenital malformation. Except for certolizumab, a PEGylated monovalent Fab fragment, all other available anti-TNFα drugs cross the placental and can be detected in fetal blood. Eculizumab is another biologic which recently showed limited placental passage owing to the unique engineered structure of IgG2/4. A comparable low placental transfer was described for etanercept as well because of its molecular structure. While some unfavorable pregnancy outcomes like low birth weight and preterm labor were reported with the use of biologics in pregnancy, it is arguable that they may be a consequence of the autoimmune disease itself. Evidence about the safety of other biologics has not yet been sufficient. Exposure through fathers has shown safe outcomes.

Some studies have reported alterations in maternal and infant immune systems; however, the whole spectrum of these changes and their effects need to be further evaluated. First-year tracking of exposed children is comforting. Live-attenuated vaccines are still avoided in clinical practice at least for the first 6 months of life due to the anticipated persistence of some mAbs. Proper timing of biological dosing in pregnancy is yet to be more clearly determined. Longer follow-up periods are still needed to monitor for possible late effects like the risk of malignancy and delayed immunological effects in the exposed children. Assessment of available biosimilar drugs’ safety is also crucial since they provide more affordable therapy than biologic drugs. Only one paper published information about outcomes of a small number of pregnancies after peri-conceptional treatment by biosimilar infliximab with results not different from the original (Kolar et al., 2018).

Overall, the available data are encouraging. With the emergence of more studies every year, recommendations are stronger for considering treatment by certain biological drugs throughout gestation and breastfeeding.

References