Intentional infliximab use during pregnancy for induction or maintenance of remission in Crohn's disease

Authors


Dr U. Mahadevan, UCSF Center for Colitis and Crohn's Disease, 2330 Post Street No. 610, San Francisco, CA 94115, USA.
E-mail: umamah@itsa.ucsf.edu

Summary

Aim:  To study the effects of infliximab on pregnancy and foetal outcome.

Methods:  We conducted a retrospective chart review of women with Crohn's disease treated intentionally with infliximab during pregnancy. The primary outcome measure was the occurrence of congenital malformations. Secondary outcome measures were the rate of premature birth, low-birth weight, small for gestational age infants, intrauterine growth retardation and caesarean section.

Results:  Ten women were identified. Eight women received maintenance infliximab infusions throughout their pregnancy and two women received their initial infliximab infusions during pregnancy. All 10 pregnancies ended in live births. No infants had congenital malformations, intrauterine growth retardation or small for gestational age parameters. Three infants were premature and one had low-birth weight. Eight women had a caesarean section.

Conclusions:  This is the first reported series of intentional infliximab use throughout pregnancy. These data, combined with other studies of inadvertent use of infliximab during pregnancy, suggest that the benefits of infliximab in achieving response and maintaining remission in mothers with Crohn's disease may outweigh the risk to the foetus of exposure to the drug. Further prospective data collection will be helpful to confirm these findings.

Introduction

The highest age- and sex-adjusted incidence rates of inflammatory bowel disease (IBD; Crohn's disease, ulcerative colitis) occur during the peak reproductive years.1 Gastroenterologists frequently provide advice and medical care to women with Crohn's disease who desire conception. Optimal medical management of these women requires accurate information about the effects of IBD medications on fertility as well as on foetal outcome. Advances in medical therapy for Crohn's disease and the resultant improved health have allowed more women to consider conception, raising concerns regarding the use of IBD medications during conception, pregnancy and lactation. Unfortunately, data on the safety of medications in these clinical settings is limited. Infliximab (INF; Centocor, Malvern, PA, USA), a chimaeric monoclonal antibody to tumour necrosis factor (TNF)-α, is effective for induction2 of remission and for maintenance3 of response and remission when administered routinely every 8 weeks to patients with Crohn's disease. The effect of regular scheduled maintenance INF treatment during pregnancy on foetal outcome is not known. We report a series of women with Crohn's disease who were intentionally treated with INF for induction or maintenance of remission during pregnancy, and review the available literature.

Methods

We conducted a retrospective multicentre chart review of women with Crohn's disease who were intentionally treated with INF during pregnancy. Data was compiled from four referral centres in the United States (University of California San Francisco, University of Chicago, Mayo Clinic Rochester and Medical College of Wisconsin). This study was approved by each institutional review board (IRB). Doctors and nurse practitioners specializing in IBD identified women in their practice who had Crohn's disease and received or were receiving INF during their pregnancy.

A 55-question chart abstraction form was completed by the treating doctor or nurse practitioner to retrospectively collect maternal and newborn data. Patients were contacted to obtain missing data. Data collected included information on the mother's history of Crohn's disease (duration and extent of disease, surgery, medications, disease activity during pregnancy, etc.), on the pregnancy (marital status, tobacco and alcohol intake, prenatal visits, prior pregnancies, history of infertility, pregnancy and labour complications, etc.) and on the newborn (Apgar scores, foetal outcomes, illness, hospital stay, etc.). Crohn's disease activity was determined using a scale derived from published guidelines.4

The primary outcome measure was occurrence of congenital malformations. Secondary outcome measures were the rate of premature birth (defined as <37 weeks gestation), low-birth weight (defined as <2500 g), small for gestational age (defined as <10th percentile weight for gestational age),5 intrauterine growth retardation (if documented by the obstetrician) and caesarean section.

Results

Ten women with Crohn's disease receiving INF during pregnancy were identified. Patient characteristics are listed in Table 1. None of the patients had a prior history of infertility. The total number of pregnancies for the group, including the index case, was 18, with a mean of 1.8 per patient (range: 1–4). One patient had a prior history of one therapeutic abortion and one spontaneous abortion while another patient had a previous stillbirth; all other prior pregnancies were live births. All of the women had a history of Crohn's disease prior to the index pregnancy during which INF was administered.

Table 1.  Patient characteristics
  1. LMP, last menstrual period.

N10
Mean age (range)31.2 years (24–48)
Marital status
 Married8
 Single2
Ethnicity 
 Caucasian7
 Black2
 Hispanic1
Smoking history
 Non-smokers7
 Former smokers2
 Current smoker1
Alcohol use
 Non-drinkers6
 Former drinkers (last drink prior to LMP)4
 Recreational drug use (former user)1
 Mean number of pregnancies(including index case, range)1.8 (1–4)
Mean duration of disease (range)11.1 years (1–24)
Disease location
 Small bowel7
 Colon7
 Perianal7
 Gastroduodenal2
Prior surgery (eight of 10 patients)
 Perianal4
 Small bowel resection4
 Total colectomy, ileorectal anastomosis1
 Colostomy and sigmoid resection1
 Exploratory laparotomy1

Luminal and perianal disease activity during pregnancy and the medications used during each trimester are outlined in Table 2. Four women had no change in disease activity from conception to the postpartum period: two were in remission and two had moderate disease activity throughout their pregnancy. Of the latter two, one received INF in the first trimester but was non-compliant with therapy and received no further infusions. Two women had improved disease activity by the conclusion of pregnancy and four women worsened during pregnancy. Of the latter group, two patients in remission had a mild flare during the postpartum period (one was on every 8 week maintenance and one had a 14 week interval between doses), another had a mild flare in the second trimester that progressed to severe by the third trimester (INF was only initiated in third trimester) and a fourth patient had a severe flare in the third trimester despite maintenance INF (corticosteroids were initiated).

Table 2.  Crohn's disease activity and medical therapy during pregnancy
 Conception (n)Trimester 1 (n)Trimester 2 (n)Trimester 3 (n)Postpartum (n)
Disease activity
 Inactive55643
 Mild13314
 Moderate42132
 Severe   21
Perianal disease activity (seven of 10 patients)
 Active55444
 Inactive22333
Medication use
 Mesalazine2211
 AZA/MP4443
 Steroids4453
 Steroid enema2222
 Ciprofloxacin1 11
 Metronidazole1  1
 6-Tioguanine1   
Infliximab use9 887

There were five women receiving concomitant purine analogue treatment for their Crohn's disease. One patient receiving 6-tioguanine (6-thioguanine) stopped the drug in the first trimester. All four women on mercaptopurine (6-mercaptopurine)/azathioprine continued the drug throughout pregnancy, although one stopped in the postpartum to breastfeed.

There were eight women who received maintenance INF infusions throughout their pregnancy. One patient chose to stop INF once pregnancy was known, but then restarted due to recurrence of Crohn's disease symptoms. She received INF in all three trimesters. Of the INF-naive patients, one woman began INF in the third trimester for a severe flare and another woman began INF in the first trimester for steroid-dependent disease but then was non-compliant with therapy and did not receive further INF during her pregnancy. All patients were on 5 mg/kg of INF per infusion. Table 3 provides further details of INF use in these women. Infections were noted in three women – one upper respiratory tract infection, one urinary tract infection and one patient with both vaginal yeast infection and urine positive for group B Streptococcus.

Table 3.  Infliximab (INF) use before and after pregnancy
History of prior INF use
 Yes8
 No2
Maintenance INF use prior to pregnancy
 Yes8
 No2
Mean interval between last INF and pregnancy start5.1 weeks (2–10)
INF stopped at onset of pregnancy1/8 patients
Mean number of infusions during pregnancy4.2 (2–6)
Mean dose of infliximab5 mg/kg (all patients)
Mean interval between doses6.4 weeks (2–8)
Indication for INF
 Maintenance7
 Flare2
 Steroid dependence1
Breast feeding while on INF5 patients
Infusion reactionsNone

All 10 pregnancies ended in live births (Table 4). All patients underwent ultrasounds during their pregnancy. Eight of 10 women underwent caesarean sections, four for active perianal disease, two for active Crohn's disease, one for preterm birth and one for prior history of caesarean section. No infants were diagnosed with congenital malformations, intrauterine growth retardation or small for gestational age at a mean follow-up of 6 months. There were three premature infants and one low-birth weight infant. There were two neonatal illnesses. One child, whose mother received maintenance INF throughout pregnancy, had neonatal jaundice, which resolved. It was felt to be unrelated to INF. A second infant, born at 39-weeks gestation to a mother on maintenance INF, had respiratory distress and desaturation and was admitted to the neonatal intensive care unit for 1 day. The baby was also diagnosed with a gastric ulcer at 5 days of age. At 6-month follow-up, the infant was doing well with no medical problems. The relationship to maternal INF use is not known.

Table 4.  Pregnancy outcomes
Live births10
Caesarean section8
Gender
 Male6
 Female4
Mean gestational age (week)37.1 (32–39)
Mean weight (g)3211.2 (2436–4215)
Preterm birth (<37 weeks)3
Low-birth weight (<2500 g)1
Small for gestational ageNone
Intrauterine growth retardationNone
Congenital malformationsNone
Newborn illnesses
 Jaundice1
 Respiratory distress, gastriculcer with 1 day intensivecare unit (ICU) stay (term delivery)1

Discussion

We report clinical outcomes in 10 women who received intentional INF infusions for induction and maintenance of remission of Crohn's disease during pregnancy. Previous reports have described inadvertent exposure or rescue treatment of severely ill women with this agent in the setting of pregnancy.

The INF is classified by the FDA as a pregnancy class B agent (no documented human toxicity). There is no published information regarding intentional use, specifically continuation of maintenance infusions, in women who have required this agent for effective treatment of their chronic gastrointestinal inflammation. Theoretically, the chimaeric structure of the INF molecule, which contains a human immunoglobulin G1 (IgG1) constant region, allows little placental transfer of the molecule during the first trimester.6 However, IgG subclasses are readily passed into the foetus during the second and third trimesters, which specifically raises questions regarding safety of INF administration beyond the month 3 of pregnancy. Fortunately, our data, although limited, suggests that maternal treatment with INF is not associated with adverse outcome to the pregnancy or the child.

Biological therapy with INF for Crohn's disease has changed practice patterns by improving medical treatment outcomes for refractory patients and by raising awareness of the need for maintenance of remission therapy. Data supporting the effectiveness of INF in the maintenance of remission in patients with luminal3 and fistulizing Crohn's disease7 has led to an increasing number of patients on regularly scheduled infusions. This practice is further strengthened by the concern for prevention of antibodies to INF with its attendant risk of infusion reactions and loss of response to drug.8 While this relatively new medication and aggressive treatment approach to medical management may have put more women with Crohn's disease in the position of being healthy enough to consider pregnancy, it also raises difficult issues about the safety of these medications to the foetus.

Limited data exists on the safety on INF in pregnancy, which has resulted in its FDA pregnancy category B designation. All available published data is summarized in Table 5. There are multiple case reports of INF use in pregnancy. There are four reports in patients with Crohn's disease. In one case,9 the mother received INF during the conception period and first trimester, had active disease throughout and was also on azathioprine, metronidazole and mesalazine (mesalamine). The pregnancy ended in premature birth at 24 weeks and death of the infant 3 days later of intracerebral and intrapulmonary bleeding. In the three other cases, the pregnancy ended in a live birth – two full term, one preterm at 36 weeks – and the infants were healthy at last follow-up.10–12 There is one report of a patient with psoriatic arthritis, but the outcome of the pregnancy was not known.13 In the rheumatoid arthritis literature, three patients have been reported, of which one pregnancy ended in live birth, one had outcome unknown14 and one ended in miscarriage at 6 weeks.15 The latter patient was on methotrexate and diclofenac.

Table 5.  Infliximab use during pregnancy and the conception period
AuthorNDisease typeNumber of infusionsPregnancy stagePregnancy outcome
  1. A summary of published data on infliximab in pregnancy.

  2. T, Trimester; CD, Crohn's disease; P, psoriasis; RA, rheumatoid arthritis; UC, ulcerative colitis.

Srinivasan91CD2Conception/T1Premature at 24 weeks,death at 3 days
James101CD1T2Live term birth
Bank111CD, PMaintenanceThroughoutLive term birth, healthy at 2 years
Burt et al.121CD1ConceptionPremature at 36 weeks,healthy at 20 months
Lichtenstein et al.16 TREAT registry36CDAny exposure11.1% miscarriage (P = 0.53),8.3% neonatalcomplications (P = 0.78)
Dechant and Wendler131P, arthritisMaintenanceConception/T1Unknown
Chakravarty et al.142RAMaintenanceUnknownLive term birth, unknown
Kinder et al.151RAMaintenanceConception/T1Miscarriage, 6 weeks
Katz et al.17 InfliximabSafety Database96 (100 births)CD, UC, RAVarious7 >3 months priorto conception,53 conception,30 T1, 6 unknown68 Live birth
 Preterm death 24 weeks
 Full-term tetralogy of fallot(corrected, doing well)
 Perinatal sepsis (doing well)
 Intestinal malrotation (doing well)
 Developmental delay
14 miscarriage
 1 intrauterine death 19.5 weeks
 1 Stillbirth 27 weeks
18 therapeutic terminations
 No abnormalities
 Elective

The two largest studies are from the TREAT registry16 and the Infliximab Safety Database17 maintained by Centocor. The TREAT registry is a prospective registry of patients with Crohn's disease. Patients may or may not be treated with INF. Of the 5807 patients enrolled, 66 pregnancies were reported, 36 with prior INF exposure. Foetal malformations have not occurred in any of the pregnancies. The rates of miscarriage (11.1% vs. 7.1%, P = 0.53) and neonatal complications (8.3% vs. 7.1%, P = 0.78) are not significantly different between INF-treated and INF-naive patients, respectively.

The Infliximab Safety Database is a retrospective data collection instrument. Limitations include reliance on reporting by treating doctors and incomplete data. Katz et al.17 queried this database for pregnancies and 146 were identified: direct exposure in 131 women and indirect by 15 male partners. The majority of the women had Crohn's disease (82), while 10 had rheumatoid arthritis, one had ulcerative colitis and three were unknown. Pregnancy outcome data is available for 96 women with direct exposure to INF. The 96 pregnancies ended in 100 births, the outcomes of which are summarized in Table 5. Of note, the case that ended in death at 24 weeks is the same as that reported by Srinivasan.9 The expected vs. observed outcomes among women exposed to INF were not different from those of the general population.

This is the first series of intentional INF use throughout pregnancy. All 10 pregnancies ended in live births, with no reported congenital malformations. The three preterm births and one low-birth weight infant are not unexpected in a population of women with Crohn's disease significant enough to require INF. Five of the mothers breastfed. It is not known whether INF is excreted in human milk or absorbed systemically after ingestion. In a single case, INF levels were either not present or were too low to be detected in breast milk.18 Also four further assays on breast milk have not detected INF levels (S. Kane, personal communication).

In conclusion, of the 151 progeny with known outcomes of mothers exposed to INF at some time during their pregnancy, there have been 19 miscarriages, 18 therapeutic terminations and 114 live births of whom, one was born 24 weeks preterm and died, one had tetralogy of fallot and one had intestinal malrotation. Based on available data, the benefits of INF use in keeping the mother's disease under control may outweigh the unknown risk to the foetus of exposure to drug. This is a very complex decision for the mother and for the treating doctor, and further studies are needed to confirm these findings.

Disclosure

U. Mahadevan, S. Kane, R. D. Cohen, W. J. Sandborn, J. P. Terdiman and D. G. Binion have received research support, honoraria, consulting fees or are members of the Centocor Speaker's Bureau.

Acknowledgement

No financial support was required for this study.

Ancillary