Pregnancy outcomes in interferon-beta-exposed patients with multiple sclerosis: results from the European Interferon-beta Pregnancy Registry

Family planning is an important consideration for female patients with multiple sclerosis (MS) who are receiving treatment with disease modifying drugs (DMDs) that may be contraindicated in pregnancy. The introduction of DMDs for the treatment of MS over 20 years ago has improved clinical outcomes for many patients, yet studies at this time were limited and did not address concerns regarding the potential adverse effects of DMDs on pregnancy and infant outcomes [1]. Patients are, therefore, advised to discuss treatment options with their healthcare professional (HCP) prior to family planning, and in many instances, it may be recommended, or even required, for a woman to discontinue her treatment with DMDs during pregnancy planning or once pregnancy is confirmed.

There is currently no consensus in the literature regarding MS treatment up to and during pregnancy. Clinical treatment guidelines recommend that injectables such as interferon-beta (IFN-beta) are continued until pregnancy occurs; however, the current evidence to support this is considered weak based on the quality of evidence and risk–benefit balance [2‐4]. There are some data, suggesting that exposure to IFN-beta during pregnancy may lead to an increased risk of lower mean birth weight, shorter mean birth length, and preterm birth (generally considered as a birth before 37 weeks of gestation) [5‐8]. However, in recent years, several systematic reviews [5, 9, 10] and observational studies [6‐8, 11‐15] have shown no increase in adverse pregnancy outcomes in patients exposed to IFN-beta before or during pregnancy.

Data on the risks of IFN-beta use during pregnancy were limited at the time which the DMDs were launched by their respective pharmaceutical companies (Betaseron^®, Bayer AG; Avonex^® and Plegridy^®, Biogen; Rebif^®, Merck KGaA; and Extavia^®, Novartis Pharma AG). The prospective European IFN-beta Pregnancy Registry was initiated in 2009 at the request of the Committee for Medicinal Products for Human Use (CHMP), to address the lack of evidence on the safety of IFN-beta treatment before conception and during pregnancy. The registry was comprised of spontaneous and solicited reports from patient support programmes and pharmacovigilance databases conducted by the Market Authorisation Holders (MAHs), Bayer AG, Biogen, Merck KGaA, and Novartis Pharma AG.

Here, we present the final cumulative pregnancy exposure data and prevalence of pregnancy and infant outcomes in IFN-beta-exposed pregnant women from the European IFN-beta Pregnancy Registry.

Between 2009 and 2017, a total of 948 pregnancy cases with known outcomes were identified in which the prevalence of spontaneous abortions, congenital anomalies in live births and live births, still births, and elective terminations remained within the ranges reported in the general population (10.7% vs. up to 10–21% [18], 2.1% vs. 2.1–4.1% [19, 21] and 2.8% vs. 2.6–6.9% [22], respectively) (Fig. 2). This similarity was also observed when considering only the HCP-confirmed, spontaneously reported cases with a confirmed diagnosis of MS. A number of systematic literature reviews and observational studies have been published on the effect of IFN-beta exposure during pregnancy and the related pregnancy outcomes. Overall, the rates of congenital anomalies and spontaneous abortions reported in these studies are comparable with the rates reported in the general population [5, 9]. The estimated prevalence of pregnancy outcomes among IFN-beta-exposed patients, collected by the European IFN-beta Pregnancy Registry, in all cases including those that were spontaneously reported cases of HCP-confirmed pregnancies with a confirmed diagnosis of MS, remained within the ranges reported for the general population. Based on these results, there is no evidence of a signal, suggesting that exposure to IFN-beta before conception and/or during the first trimester of pregnancy adversely increases the rate of congenital anomalies or spontaneous abortions. Sensitivity analyses showed that the pregnancy outcomes of the cases included after 2015 were comparable between spontaneous and solicited cases (Fig. 2).

The results presented in Table 1 show that in the majority of pregnancy cases, IFN-beta treatment was stopped during the first trimester, a time at which most women become aware that they are pregnant, and that there is no evidence of a signal suggesting exposure to IFN-beta during this period adversely affects pregnancy or infant outcomes. Similar results have been shown in other observational studies, where exposure in the early stages of pregnancy has not been shown to worsen pregnancy outcomes [7, 8, 12, 14]. The European IFN-beta Pregnancy Registry collected a limited number of pregnancies exposed to IFN-beta during the second and third trimesters (3.3%; 31/948), so it is not possible to determine the impact of IFN-beta exposure during the latter stages of pregnancy; however, exposure occurred at a critical time in organ development [24]. In previous studies, the timing of exposure to IFN-beta was also relatively narrow, with a very low number of pregnancies being exposed to IFN-beta beyond the first trimester. Mean duration of foetal exposure time to IFN-beta ranged from 4.0 to 9.1 weeks [5‐8, 12, 14]. It should be noted that Table 1 presents one spontaneous abortion that occurred during the third trimester. This may be due to an error in which a pregnancy case was incorrectly reported to the pharmacovigilance database, and was not possible to be clarified during follow-up. This particular case is not in accordance with the EMA definition of spontaneous abortion and should, therefore, be disregarded.

The European IFN-beta Pregnancy Registry reported a total of 216 adverse events in 98 infant/foetus cases, yet no patterns of congenital anomalies were observed. These findings align with previous studies which found that in cases in which congenital anomalies were present, they did not centre on a specific system organ class [6, 11, 13, 14]. The cases of adverse events were reported retrospectively to the pharmacovigilance databases, and as such, 11 cases could not be matched to the mother’s case report. It should also be noted that some adverse events in infant/foetus cases may not have been reported to the pharmacovigilance databases due to the case being lost to follow-up.

The number of pregnancy reports from patients who met the initial inclusion criteria did not reach the estimated number of cases required to detect a doubling in outcome prevalence. Therefore, after June 2015, patients with an unconfirmed diagnosis of MS, solicited reports from patient support programmes and pregnancy cases not confirmed by an HCP were permitted to be included in the analysis. Through the inclusion of these cases, the required number of pregnancy reports, calculated to be 827, was exceeded by 121 pregnancy cases. Sensitivity analyses indicated that the inclusion of the additional non-HCP-confirmed and solicited cases were comparable to the pregnancy outcomes of the cases meeting the original inclusion criteria, and were, therefore, deemed adequate as a measure to reach the required sample size for the analysis. A limitation of this analysis is that 16.1% of patients with known pregnancy outcome did not have a confirmed diagnosis of MS; however, in Europe, treatment with IFN-beta automatically means that the diagnosis of MS is confirmed by an HCP.

At the end of the study period, approximately 61.3% of the total numbers of reported cases were either pending outcomes (25.3%) or the patients were lost to follow-up (35.9%); as such, there may be selection bias for pregnancies with documented outcomes versus those that were not documented. This selection bias may be either towards the presence of adverse pregnancy outcomes or towards a particular subgroup of women who have higher healthcare seeking behaviour. It is also thought that the seriousness of an adverse event contributes to whether or not it is reported, with serious events more likely to be reported than non-serious events [25]. However, it may also be possible that the mothers participating in the pregnancy registry may be healthier, and therefore, the rate of adverse events would be expected to be lower.

The European IFN-beta Pregnancy Registry did not separate the pregnancy outcomes by treatment type (i.e., IFN-beta-1a and beta-1b) to determine the adverse pregnancy outcomes associated with each treatment. However, if the results were separated for analysis, then the statistical power within each treatment group would be limited and meaningful conclusions could not be drawn from the data. Due to the design of the study, it was not possible to determine the dose or duration of IFN-beta exposure prior to conception and pregnancy. Therefore, it cannot be determined if pregnancies with congenital anomalies were associated with higher doses of IFN-beta, longer term DMD use, a combination of these factors, or whether adverse pregnancy outcomes occur regardless of the dosage and length of exposure. It should also be noted that the European IFN-Beta Pregnancy Registry did not adjust for potential confounders including family history, gestational age, parity, previous obstetrical history, smoking and alcohol consumption, comorbidities, or exposure to co-medications, which may have influenced the prevalence results. Nor did the study collect any substantial data on birth weight, birth length, or gestational age at birth to allow for a comprehensive analysis. Furthermore, 375 (54%) of the 694 cases reported in this study were obtained from patients in Germany; therefore, interpretations should take into consideration the differences in clinical practices between Germany and the rest of the countries participating in this study. Nevertheless, the available data do not suggest any concern regarding the ability to extrapolate the results to the intended target population.

The European IFN-beta Pregnancy Registry was initiated at the request of the CHMP, and was justified by the need to provide HCPs and women of childbearing age with MS additional information on IFN-beta exposure during pregnancy, so that informed treatment decisions can be made. Overall, the results from the European IFN-beta Pregnancy Registry represent the largest prospective cohort providing safety data of women with MS of childbearing age exposed to IFN-beta. Results from the registry showed no evidence that IFN-beta exposure before conception and/or during pregnancy increased the rate of congenital anomalies or spontaneous abortions, thus indicating that treatment with IFN-beta may be continued until pregnancy is confirmed.