WASHINGTON – Prenatal exposure to tumor necrosis factor–inhibiting drugs does not significantly increase the risk of a serious antenatal infection in infants born to women taking the drugs for rheumatoid arthritis, according to a large database study.
Researchers from McGill University, Montreal, and the University of Alabama at Birmingham who conducted the study did find a higher rate of serious infections among infants born to users of a tumor necrosis factor inhibitor (TNFi), especially among those exposed to infliximab, but after adjustment for maternal age and other antirheumatic drugs, the risk was not statistically significant.
Infliximab is unique among the TNFi drugs in that it concentrates in cord blood, reaching levels that can exceed 150% of the maternal blood level, Dr. Vinet noted. Adalimumab concentrates similarly, although the current study did not find any significantly increased infection risk associated with that medication.
Dr. Vinet and her colleaguesin 2,455 infants born to mothers with rheumatoid arthritis (RA), who were included in the PregnAncies in RA mothers and Outcomes in offspring in the United States cohort (PAROUS) registry. This cohort is drawn from data in the national MarketScan commercial database. The infants were age- and gender-matched with more than 11,000 matched controls born to women without RA, and with no prenatal TNFi exposure. Among these drugs, she looked for exposure to adalimumab, certolizumab pegol, etanercept, golimumab, and infliximab, as well as corticosteroids and other biologic and nonbiologic disease-modifying antirheumatic drugs (DMARDs).
Two exposures were considered: drugs taken during pregnancy and drugs taken before conception but not during pregnancy. These were compared with infants of mothers with RA who didn’t take TNFi drugs, and to the control infants. Serious infections were those that required a hospitalization during the first 12 months of life; only the index incident was counted.
Among the RA cohort, 290 (12%) were exposed to a TNFi during pregnancy and 109 (4%) were born to women who had taken a TNFi before conception. The remainder of the cohort was unexposed to those medications.
The mean maternal age was 32 years and similar in all RA categories and controls.
Corticosteroid use was common in women with RA, whether they took a TNFi during pregnancy (55%), before pregnancy (44%), or not at all (26%). Nonbiologic DMARDs were given to 19% of the TNFi cohort during pregnancy and 16% before pregnancy, as well as to 15% of those who didn’t take a TNFi.
The rate of serious neonatal infection was 2% among both infants born to RA mothers who didn’t take a TNFi and those born to RA mothers who took a TNFi before conception. Control infants born to women without RA had a serious infection rate of 0.2%.
Among infants exposed to a TNFi during pregnancy, the serious infection rate was 3%; it was also 3% among those exposed only in the third trimester.
A multivariate analysis that controlled for maternal age, prepregnancy diabetes, gestational diabetes, preterm birth, and exposure to the other drug categories determined that TNFi drugs did not significantly increase the risk of a serious infection in neonates with gestational exposure (odds ratio, 1.4) or whose mothers took the drugs before conception (OR, 0.9), compared with controls. The findings were similar when the analysis was restricted to TNFi exposure in the third trimester only.
When Dr. Vinet examined each drug independently, she found numerical differences in infection rates: golimumab and certolizumab pegol, 0%; adalimumab, 2.4%; etanercept 2.7%; and infliximab, 8.3%.
Because of the relatively small number of events, this portion of the regression analysis could not control for preterm birth and gestational diabetes. But after adjusting for maternal age and in utero corticosteroid exposure, Dr. Vinet found no significant associations with serious neonatal infection and any of the TNFi drugs, including infliximab (OR, 3.5; 95% CI, 0.8-15.0).
She and her colleagues had no financial disclosures.