Symptomatic therapies used in IBD include the antidiarrheals and antispasmodics, as well as pain medications. Loperamide use has not been associated with an increased rate of first trimester fetal malformations, spontaneous abortion, low birth weight or premature delivery [35], and is considered low risk. One has to keep in mind, however, that increased stool frequency may be a sign of increased activity and loperamide use should be monitored. Diphenoxylate with atropine is teratogenic in animals, and fetal malformations have been observed in infants exposed during the first trimester [36]. Antispasmodics and anticholinergics have been associated with nonlife threatening fetal malformations and are best avoided during pregnancy [37]. Codeine has been used for many years during pregnancy without reports of associated fetal abnormalities. Drug dependence and withdrawal in the newborn can occur, but fortunately is rare.

Aminosalicylates are first-line medications for mild to moderate IBD. While the exact mechanism is unknown, it is thought that the effects result from topical action on the gastrointestinal mucosa rather than through systemic action [38]. Sulfasalazine has been used for over 50 years in the treatment of IBD. Sulfasalazine is the combination of sulfapyridine and 5-ASA, which is eventually cleaved by enteric bacteria, allowing the main anti-inflammatory effects to take place in the colon. The majority of the side effects of sulfasalazine are due to the sulfapyridine moiety; it crosses the placenta with fetal serum levels equivalent to maternal levels [39]. Multiple studies have shown that despite this phenomenon, there appears to be no increased incidence of abnormal birth outcomes [40].

Sulfasalazine has been shown to inhibit folate acid metabolism, which can cause neural tube defects in the fetus [41]. While the risk of fetal abnormalities has not been shown to increase with sulfasalazine and its derivatives, pregnant female patients taking sulfasalazine should still supplement their diet with 2 mg of folate daily.

Mesalamine is a sulfa-free aminosalicylate that is also effective in mild-moderate IBD and is widely used for acute and long-term management of IBD. 5-Aminosalicylic acid (5-ASA) and its metabolite acetyl-5-aminosalicyclic acid are found in both maternal and fetal plasma in female patients taking mesalamine; however, a meta-analysis of published studies has demonstrated its lack of effect on the incidence of adverse birth outcomes [42]. There is a single case report of renal interstitial damage in a child born to a woman taking mesalamine [43]. This finding has not been confirmed in any other studies. Use of topical 5-ASA agents during pregnancy has not been associated with any increase in adverse birth outcomes related to its use during pregnancy.

Olsalazine, another 5 ASA compound, is rated as an FDA class C rating based on the paucity of prospective safety data with this agent. Recently, the FDA has changed the pregnancy rating of Asacol® (Warner-Chilcott) from B to C secondary to the association of phlalate, an entity in the coating of Asacol®. In rodent studies, administration is associated with urogenital abnormalities in male offspring [44]. However, the amount of phlalate required to replicate those animal studies in humans would equate to over 200 tablets of medication a day and thus not likely to be of clinical consequence.

Currently, the most frequently used antibiotics in IBD include metronidazole and ciprofloxacin. Animal studies have not shown any evidence of teratogenicity or increased fetal loss with metronidazole. Short courses of metronidazole during the first trimester of pregnancy for Trichomonas vaginalis has been shown to be well tolerated and low risk [45, 46]. The most recent study of 228 female patients exposed to metronidazole during pregnancy followed female patients prospectively through their pregnancy [45]. Eighty-six percent of female patients were exposed during the first trimester. The malformation rate was 1.6 % in the treatment group and 1.4 % in the control group. Female patients with IBD require the use of metronidazole for longer periods of time, and there are limited data regarding prolonged use of this medication.

In animal studies, no teratogenicity has been seen with ciprofloxacin, although musculoskeletal abnormalities have been identified in immature animals [47]. Moskovitz et al. found that in 27 patients who were receiving 1 g/day, even in the first trimester, its use appeared to be safe, while ciprofloxacin at doses of 1 g/day also appeared safe (18 patients) [48]. Another study investigated the effects of fluoroquinolones in the first trimester and did not show an increased risk of congenital malformations, prematurity, or low birth weight [49].

While these data are comforting, this information applies to the non-IBD population, where antibiotics are used short term. These agents are more commonly used for longer durations in IBD and the use of these two antibiotics during pregnancy should currently be restricted to short-term courses.

In moderate to severe disease, corticosteroids are a mainstay of treatment for IBD. As shown in studies for rheumatological conditions as well as for IBD, corticosteroids during pregnancy have largely been regarded as low risk [33]. Corticosteroids cross the placental barrier but the fetal:maternal serum concentration of the steroids varies between different corticosteroid preparations. Prednisolone and prednisone are more efficiently metabolized by the placenta than dexamethasone or betamethasone and fetal levels of this steroid are approximately eight- to tenfold lower than that of the maternal circulation [50]. Since corticosteroids are conjugated more rapidly to biologically less active sulfates in the fetus than the adult, a suppressive fetal blood concentration is not often reached with therapeutic doses used during pregnancy.

Among female patients with IBD, corticosteroids have not been found to be harmful to the fetus [20]. Mogadam et al. studied the effects of steroid use in 185 out of 531 pregnancies in female patients with IBD and did not find a statistically significant increased incidence of prematurity, spontaneous abortions, stillbirth, or development defects in the ulcerative colitis subgroup (4.6 % in the treated group vs. 2.2 % in the untreated group; P  >  0.10) [51]. In the Crohn disease subgroup, patients did significantly worse in the treated group compared to the untreated group (13.5 % vs. 1.9 %, P  <  0.1). Patients with CD may have more severe disease and require more medical intervention to control the activity of the disease and it is possible that the severity of the illness in CD itself may have caused these patients to not fare as well as the ulcerative colitis patients.

Clearly as more and more patients with IBD are treated with immunomodulators there is a growing need for information on their effects on the pregnant patient and growing fetus. There is a large body of literature on the use of these agents among pregnant transplant recipients and those patients with autoimmune diseases [52, 53]. It is generally believed by the most experienced IBD clinicians that immunomodulators such as 6-MP, azathioprine, and even cyclosporine can be used safely during pregnancy if the mother’s health mandates therapy, based on the evidence from these other conditions.

Thiopurines are used for steroid-sparing and steroid-dependent IBD. Azathioprine (AZA) is a prodrug of 6-mercaptopurine (6-MP) and does cross the placental barrier but the immunomodulatory effects of azathioprine do not affect the fetus due to the lack of inosinate pyrophosphorylase in the fetus, an enzyme which converts azathioprine into the active metabolites of 6-MP and S-methyl-4-nitro-5-thioimidazole [54]. Placental concentrations of azathioprine range from 64 to 93 % of the maternal blood level, while 6-MP and AZA levels in the fetal blood run about 1–5 % and 1.2 %, respectively [55]. Despite this, based on animal studies showing an increased risk of teratogenicity at doses significantly higher than what is used in clinical practice, thiopurines have been considered category D medications [56]. Several human studies have suggested that AZA and 6-MP are low risk during pregnancy when used for IBD [34, 57].

A large Danish cohort study used a population-based prescription registry to examine the risk of congenital malformation and perinatal mortality among patients taking azathioprine or mercaptopurine during pregnancy [58]. Nine pregnancies exposed to 6-MP in the first trimester and 10 pregnancies exposed throughout the pregnancy were compared to 19,418 pregnancies without 6-MP exposure. Fifty-five percent of the exposed female patients had a history of IBD and 45 % had other autoimmune diseases. The adjusted odds ratios for congenital malformations, perinatal mortality and preterm birth, and low birth weight were 6.7 % (95 % CI, 1.4–32.4), 20.0 (95 % CI, 2.5–161.4), 6.6 % (95 % CI, 1.7–25.9), and 3.8 (95 % CI, 0.4–33.3) in female patients treated with 6-MP or AZA. The conclusion of the investigators was that a causal relationship between drug use and outcome could not be established.

Francella et al., in a retrospective cohort study, investigated the possible toxicity of 6-MP from a review of records of 485 patients who had received the drug [57]. Of the 462 female patients who were contacted, 155 had conceived at least one pregnancy after developing IBD. Pregnancies were analyzed based on whether the patients had taken 6-MP before or at the time of conception compared with those IBD patients who had their pregnancies before taking 6-MP. There was no statistically significant increase in spontaneous abortion rates or major congenital malformations among patients taking 6-MP compared to control subjects [RR 0.85 (95 % CI, 0.47–1.55), P  =  0.59]. The authors concluded that the use of 6-MP or AZA and its beneficial effect on maternal health outweighed any risk to the fetus and that it was not unreasonable to continue its use throughout pregnancy. Recently, results regarding pregnancy in the prospective French cohort of Crohn patients have become available. There has been no association between thiopurine use compared to women with CD who used no medications during pregnancy [34].

In IBD, methotrexate is used in the management of steroid-dependent or steroid-resistant Crohn disease as an alternative to azathioprine and 6-MP. Methotrexate is a known abortifacient, showing increased risk of spontaneous abortions in various studies, currently used therapeutically at high doses in tubal pregnancies [59]. As a folic acid antagonist, it is a mutagenic agent known to cause neural tube defects and palatal defects. Therefore, methotrexate is a Category X medication. Patients who are started on methotrexate should be strongly advised to use reliable contraception. If termination of a pregnancy is not possible, high doses of folic acid therapy are recommended to prevent CNS abnormalities, including anencephaly, meningomyelocele, and hydrocephaly. The optimum management includes careful counseling and effective contraception prior to any initiation with methotrexate therapy.

Infliximab was the first biologic agent FDA approved for the treatment of inflammatory and fistulizing Crohn disease. The early safety literature with infliximab includes one study by Katz et al. suggesting that infliximab exposure for CD or rheumatoid arthritis during pregnancy does not lead to a statistically significant increase in adverse outcomes compared to that of the general population, using the National Center for Health Statistics database between 1976 and 1996 for comparison [60]. Of 96 female patients who were studied, live births (67 %, 95 % CI: 56.3–76.0 vs. 67 %), miscarriages (17 %, 95 % CI: 8.2–23.2 vs. 15 %), and therapeutic terminations (16 %, 95 % CI: 11.5–28.0 vs. 19 %) were not statistically different from that of the general population. In this review, 8 of 14 miscarriages in female patients who were exposed to infliximab occurred at or before 10 weeks. It is thought that these miscarriages early in pregnancy were related more to disease activity than infliximab use. Maternal IgG is transported across the placenta as early as the late first trimester [61], but efficiency of transport is poor so that total fetal IgG levels are low until the late second or early third trimester, suggesting that it is not the infliximab exposure that would be contributing to this early miscarriage rate observation.

The second anti-TNF agent on the market is an injectable agent, adalimumab. There are no commercially available assays to measure serum drug levels so there are no data for pregnancy at this time. However, because it is also a full antibody the same pharmacokinetics are thought to hold true for adalimumab as infliximab.

The third anti-TNF to get FDA approved is certolizumab pegol. This is a pegylated Fab fragment, and early data suggest that it does not cross the placenta [62]. However, at this time there is no recommendation to switch a patient from one agent to another if she is in remission solely because of these properties. The choice of an agent still depends on multiple factors.

Natalizumab, a monoclonal antibody to alpha 4 integrin, is approved to treat Crohn disease as well as multiple sclerosis. Its properties are not well characterized, but the available data suggest that it is low risk during pregnancy and carries a B safety rating [63].

Cyclosporine is used particularly for refractory ulcerative colitis and pyoderma gangrenosum. Cyclosporine’s inhibition of interleukin-2 and its effect on cell proliferation and cell turnover makes its use in pregnancy concerning. While cyclosporine is known to cross the placenta, the concentration of the drug in the newborn falls rapidly within days after birth [64]. At a dose of 10 mg/kg/day, it has not been shown to cause any fetal toxicity in rats [65]. In the renal transplant literature, cyclosporine has been associated with premature births and growth retardation, but overall, cyclosporine has not been associated with poor pregnancy outcomes or severe adverse effects to the fetus [66]. One case report in 1995 describes the successful use of cyclosporine to avoid the risk of colectomy in a pregnant patient in her 29th week of pregnancy. This woman was effectively treated with cyclosporine resulting in the delivery of a healthy baby boy at 34 weeks [67]. A retrospective review of pregnant female patients hospitalized with severe ulcerative colitis found that female patients treated with cyclosporine in addition to corticosteroids had no more of an increased risk for adverse birth outcomes than those female patients treated with steroids alone [68]. While these are encouraging, more long-term studies are required to investigate the effects of cyclosporine in the pregnant IBD population.

Recently, a variety of other therapies have been used to treat refractory IBD. Thalidomide has been found to have some efficacy in the treatment of Crohn disease [69, 70]. Thalidomide, however, is a potent teratogen and should not be given to female patients of child-bearing age except those who have undergone careful counseling and strict adherence to contraceptive use. Its availability is restricted and only available to those female patients with negative serum pregnancy tests done on a monthly basis.

Some of the newer immunomodulators used in refractory Crohn disease include tacrolimus [71] and mycophenolate mofetil [72]. Again, the safety data during pregnancy comes from the transplant populations. The experience with tacrolimus has been overall favorable, with no evident increase in adverse outcomes based on case reports and National Transplant Registry data [73]. National Registry data followed 100 female patients status post solid organ transplant and reported 4 neonatal malformations in 68 live births.

The use of mycophenolate in CD has been recent, and to date there are no data regarding any pregnancies occurring with its use.

Given the poor nutrition that often accompanies IBD, it stands to reason that both total parenteral (TPN) and enteral supplementation have been used to support the pregnant IBD patient [74, 75]. Despite the theoretical concern about fat embolization to the placenta, pregnant patients receiving TPN with intravenous lipids have done well. The infants born to those mothers have been healthy and examination of the placenta has failed to show any signs of fat emboli. Elemental diets have also been used safely during pregnancy both as primary therapy for active CD as well as a source of supplemental nutrition. Because the nutritional needs of the pregnant IBD patient are quite different from the nonpregnant IBD patient, close monitoring along with a nutritional expert is necessary.

The advantages of breastfeeding in pregnancy are well known but the effects of IBD medications on breastfeeding still remain unclear. The breastfeeding initiation rates among adolescent mothers are approximately 35–40 % [76] and are significantly less than the national rate, which is 60 %. In a study by Kane et al., only 44 % (54/122) of female patients with IBD had breastfed their infants, the majority of whom had UC [77]. In a more recent study done in Canada, showed the opposite, that women with IBD nursed more frequently than the background population [78]. In neither study was there an increased risk for disease activity associated with the act of nursing itself, disease activity was related to cessation of therapies to treat disease.

Table 47.2 summarizes the safety data regarding medications and their use during breastfeeding.

Salazopyrine and other forms of 5-ASA are excreted into the breast milk with milk concentrations that are about 40–50 % of the maternal serum levels with outcomes study suggesting its safety during breastfeeding [79]. There is one case report of diarrhea in a nursing infant of a mother who used mesalamine suppositories 6 week after childbirth with four additional challenges of breastfeeding following suppository administration leading to similar results [80].

In a case series, low concentrations of mesalamine and its metabolite were found in the breast milk of two female patients taking mesalamine at 1 g orally three times daily. Milk to plasma ratios at day 7 and 11 postpartum were 0.17 and 0.09, respectively. The mesalamine metabolite ratios were 16.5 and 6.8, respectively. The estimated intake of mesalamine and metabolites per day by the infant was 0.065 mg (0.015 mg/kg) and 10 mg (2.3 mg/kg), respectively, which are considered to be negligible amounts [81].

Data pertaining to thiopurines comes from the prospective study done by Christensen [82]. Milk concentrations were highest within the first 4 h of maternal ingestion and only equated to 0.0075 mg/kg bodyweight. Thus, it seems reasonable to suggest that drug exposure is minimal, and to minimize it further that a mother pump and dump any milk produced within the first 4 h of thiopurine ingestion and nurse with milk produced the remainder of the day.

Cyclosporine has also been relatively contraindicated due to theoretical immunosuppression in the infant. However, a study by Nyberg et al. of transplant patients showed that breast fed infants of mothers on cyclosporine received <300 mcg/day of cyclosporine [83]. Blood cyclosporine levels ranged from 55 to 130 ng/ml in mothers (12-h trough), with a range of 50 to 227 ng/ml in breast milk, and were below the detection limit of 30 ng/ml in all infants. A case report of a 35-year-old woman who exclusively breast-fed her infant during the first 10.5 months of life while she was being treated with cyclosporine revealed mean breast milk/maternal blood level ratio of 84 %, with undetectable levels in the infant [50].

Approximately 5–25 % of the maternal serum concentration of corticosteroids reaches breast milk and the amount received by the infant is considered minimal [84]. The commonly used corticosteroids, prednisone and prednisolone, result in low breast milk concentrations with doses of <20 mg of corticosteroids deemed to be safe to use during nursing [85]. Some suggest that breastfeeding is safer if delayed for 4 h after ingestion of steroids [86]. Other medications, such as metronidazole, ciprofloxacin, and methotrexate should be discontinued in nursing mothers given their high concentrations in breast milk.

There is a small case series of women who nursed while receiving infusions of infliximab, and there was no evidence that infliximab is excreted into milk [87] and thus it is not mandated that women who receive infliximab stop nursing. There are no data for the other biologic agents.

The mode of delivery should most often be an obstetrical decision and not solely based on the presence of IBD. Adolescents have lower Cesarean section rates than adult women [88]. The indications for Cesarean section for obstetrical reasons are not different in female patients with IBD. The presence of UC does not have a significant impact on the method of delivery, nor is it an indication for a section per se. However, active perianal disease in Crohn disease may worsen after a vaginal delivery. One retrospective study of female patients with CD found that 18 % of those without previous perianal disease developed such disease after delivery, usually involving an extensive episiotomy [89]. General guidelines include a planned C-section for any woman with known perianal or rectal CD, or if the birth appears to be more complicated than initially presumed.

In the pregnant IBD patient, elective surgical procedures are uncommon, but those that are performed in the second trimester do not appear to carry a significant increase in perinatal morality in female patients without IBD [90].

The indications for surgery during pregnancy are identical to that of nonpregnant patients. These include obstruction, perforation, abscess, and hemorrhage. The approach of continuing medical therapy may only further increase the risk to both mother and fetus. In the ill pregnant IBD patient, the greater risk to the child is continued maternal illness rather than surgical intervention [91]. In general, doing what is best for the mother results in what is ultimately best for the fetus.

Numerous case reports but only one small case series make up the literature documenting successful surgical intervention for treatment of severe colitis in the pregnant patient. Anderson reported on the outcomes of four pregnant female patients who experience disease activity between the 28th and 37th week of gestation [92]. Three patients were treated medically and allowed to progress to labor. Two babies were stillborn and one child was healthy. All three mothers required colectomy in the weeks after delivery. The 4th patient relapsed at 28 weeks’ gestation, had surgery for toxic megacolon at 31 weeks, and deliver at 34 weeks.

In patients with Crohn disease, Hill and colleagues described three pregnant patients with intraperitoneal sepsis, requiring surgery [93]. All three female patients recovered and delivered healthy infants. Most reports suggest proceeding to surgery when indicated. A variety of procedures have been performed, including proctocolectomy, subtotal colectomy with ileostomy, hemicolectomy or segmental resection, and combined subtotal colectomy and cesarean section. Two general points should be made (1) primary anastomosis carries a greater risk of postoperative complication rate, and thus a temporary ileostomy is generally preferred, and (2) if the fetus is significantly mature, then cesarean section along with bowel resection is indicated.

In female patients who have a total proctocolectomy with IPAA prior to pregnancy, there is controversy regarding postoperative fertility and sexual function. An early study suggests that these are maintained [94] but most recent studies [95, 96] suggest a significant decrease in fertility following this type of surgery. There has been debate whether female patients who have had IPAA should be allowed to deliver vaginally, or whether cesarean section should be planned. In one study of 43 pregnancies in female patients status post IPAA, pregnancy was well tolerated, with a complication rate lower than in female patients who had had an ileostomy [96]. Although more cesarean sections were performed in female patients with IPAA, the explanation was likely due to the uncertainty about the pouch function. An extended follow-up of female patients with an IPAA who delivered vaginally showed no adverse long-term effects on pouch function. During actual pregnancy, however, female patients with IPAA did note an increase in stool frequency, incontinence, and pad usage, with symptoms resolving after delivery. The authors suggest that the type of delivery in patients with an IPAA be dictated by obstetrical considerations. Other surgeons feel that the risk to permanent pouch failure is higher with a vaginal delivery, and recommend any patient with surgery for UC undergo cesarean section.