Objective To assess relapses, disability progression and the role of disease modifying drugs (DMDs) in the year after delivery in women with multiple sclerosis (MS).
Methods We prospectively followed-up pregnancies occurring between 2002 and 2008 in women with MS, recruited from 21 Italian MS centres. The risk of relapses and disability progression in the year after delivery was assessed using time-dependent Cox regression analysis.
Results 350 out of 423 pregnancies were assessed (pregnancies not resulting in live birth and with a postpartum follow-up period shorter than 1 year were excluded from the analysis). 148 patients (42.3%) had at least one relapse in the year after delivery. An Expanded Disability Status Scale (EDSS) score at conception ≥2.0 (HR=1.4; 95% CI 1.1 to 2.0; p=0.046) and a higher number of relapses before (HR=1.5; 95% CI 1.2 to 1.8; p<0.001) and during pregnancy (HR=2.3; 95% CI 1.6 to 3.4; p<0.001) were related to a higher risk of postpartum relapses. On the contrary, early DMD resumption after delivery marginally reduced the risk of postpartum relapses (HR=0.7, 95% CI 0.4 to 1.0; p=0.079). Moreover, 44/338 women progressed by at least one point on the EDSS. Disability progression was associated with a higher number of relapses before (HR=1.4, 95% CI 1.1 to 1.9; p=0.047) and after delivery (HR=2.7, 95% CI 1.4 to 5.2; p=0.002).
Conclusions Our findings show an increased risk of postpartum relapses and disability accrual in women with higher disease activity before and during pregnancy. Since it may reduce the risk of postpartum relapses, early DMD resumption should be encouraged, particularly in patients with more active disease.
Statistics from Altmetric.com
Over the past few years, pregnancy-related issues in multiple sclerosis (MS) patients have gained growing attention. The postpartum period is a critical phase during which a recrudescence of disease activity is expected, particularly in patients with more active disease course before and during pregnancy. Moreover, there is limited and controversial information on the overall impact of pregnancy on MS course and disability.1–7
The reference information on the influence of pregnancy on MS course comes from the seminal Pregnancy in MS (PRIMS) study,8 an international, multi-centric, prospective study involving 269 pregnancies in 254 women with MS. This study clearly showed a decrease of relapse rate during pregnancy in comparison with the prepregnancy year, followed by an increase early after the delivery, with disability progression remaining in general unaffected.
However, the PRIMS study was carried out before the introduction of disease modifying drugs (DMDs) that are currently used starting from the earliest stages of the disease. Therefore, we have no data on the possible impact of pregnancy on MS course in the scenario of DMD treatment.
Using the dataset collected (the Italian pregnancy dataset), in this prospective, collaborative study, we assessed the risk of clinical relapses and disability accrual in the year after delivery and the relevant clinical predictors, taking into account the role of DMD treatment in the period preceding pregnancy and after delivery.
The Italian Pregnancy dataset has been described in detail elsewhere.2 ,9 In brief, all pregnancies occurring in MS patients10 referred to the participating centres between 2002 and 2008 were identified and tracked over the whole gestational period. The network of the Italian Pregnancy dataset includes 21 sites representing the main Italian MS centres located throughout the entire country. In the present study, we included all pregnancies resulting in live births and having a postpartum follow-up duration of at least 1 year. In each centre, all the patients were prospectively followed-up every 6 months and in the case of relapse before, during and after the pregnancy. Data were gathered by the neurologist using a standardised information form. After delivery, the neurologist administered a semistructured interview to each patient dealing with pregnancy outcomes, breast feeding and potential confounders (see below). For comparison with previous studies,3 ,4 patients were divided into two groups: those who breast fed their infants for at least 2 months consecutively and exclusively (breastfeeding group) and those who breast fed their infants for less than 2 months or did not breast feed at all (not breastfeeding group). The cut-off of 2 months was chosen since it was reported to be associated with a protective role of breast feeding on postpartum relapses.3 The introduction of DMDs after delivery was arbitrarily defined as early (early DMDs) when started within 3 months after delivery (and before a relapse in the first postpartum trimester) or delayed (delayed DMDs) when started in the first postpartum trimester but after the occurrence of a relapse, after the first postpartum trimester or not started at all. As for disease activity, the date of onset and number of relapses in the year preceding pregnancy, during pregnancy and in the year after delivery were recorded. A relapse was defined as the appearance or reappearance of one or more symptoms attributable to MS accompanied by objective deterioration, as shown by neurological examination, lasting at least 24 h, in the absence of fever and preceded by neurological stability for at least 30 days. Moreover, at each visit disability was measured through the functional systems and Expanded Disability Status Scale (EDSS).11 Baseline EDSS and prepregnancy information were recorded within 1 month from conception (determined as 14 days after the mother's last menstrual period) and outside relapses. Disability progression was defined as a worsening of at least 1.5 points on the EDSS for patients with baseline EDSS=0, at least 1 point for patients with baseline EDSS between 1 and 5.5, and 0.5 point for patients with baseline EDSS=6.0, evaluated outside relapse periods and confirmed without any significant recovery at 6 and 12 months.
The study was approved by the ethics committee of the University of Florence, and written consent was obtained from all patients.
Baseline characteristics were reported as frequency (percentage), mean±SD, median and IQR, and compared with Pearson's χ2, Student t and Mann–Whitney U tests when appropriate.
Possible predictors of relapses and disability progression in the year after delivery were assessed through a multivariable backward stepwise survival analysis (time-dependent Cox regression model). Proportion hazards assumption has been tested graphically using log-log survival curves for categorical variables and partial residual plots for continuous variables. The following covariates were entered into the model: current age, disease duration and EDSS at conception, DMDs before pregnancy (yes vs no), number of relapses in the year before pregnancy and during pregnancy, smoking, alcohol and toxin exposure during pregnancy (yes vs no), breast feeding (yes vs no) and DMD introduction after delivery (early vs delayed). In order to account for possible immortal bias due to the classification into early and delayed DMD, DMD status was included as a time-dependent variable. For early DMD, the value of the time-dependent variable is 0 until drug introduction and changes to 1 thereafter. For delayed DMD, the value remains 0 in the follow-up.12 ,13 Current age, disease duration and number of relapses in the year before and during pregnancy were entered as continuous variables. EDSS at conception was categorised as EDSS <2.0 or ≥2.0, according to the median value. Consistent with previous papers,2 ,9 substance exposure status was defined as ‘exposed’ (any exposure to any substance in any trimester) and ‘never exposed’ (no exposure). Maternal smoking status was defined as ‘never smoked’ (no smoking in any trimester) and ‘smoked’ (smoking in any trimester). Alcohol exposure status was defined as ‘exposed’ (drinking more than one unit per day) and ‘not exposed’ (drinking less than one unit per day). In addressing disability progression, the number of relapses in the year after delivery was also included (occurred before disability progression). Since the number of multiple births to the same mother was low (see Study population section), a clustered analysis was not applied.
All analyses were performed using SPSS V.18.0 software running on Windows (SPSS, Chicago, Illinois, USA).
During the study period, a total of 423 pregnancies were tracked in 415 women. The last pregnancy included in the study took place in January 2008. Among these pregnancies, 350 (in 345 women) resulted in live births, with a postpartum follow-up of at least 1 year (figure 1). A total of 37 pregnancies not resulting in live birth and 36 pregnancies with a postpartum follow-up period shorter than 1 year were excluded from the analysis. Table 1 shows the main demographic and clinical characteristics of the study cohort. Overall, 74 (21.1%) patients started DMDs within 3 months (39 in the first month, 23 in the second month, 12 in the third month) after delivery (24 with intramuscular interferon β-1a, nine with subcutaneous interferon β-1b, 27 with subcutaneous interferon β-1a, 12 with glatiramer acetate, two with experimental drugs). All of these patients were still on the same treatment at the 1-year follow-up. There were no differences between patients with early and delayed DMD treatment after delivery in terms of the main demographic and clinical characteristics, with the exception of the proportion of DMDs before pregnancy, which was higher in patients with early DMD introduction (82.4% vs 58.5% in delayed DMD and 37.7% in untreated subjects; p<0.001) (see online supplementary e-table 1).
Predictors of postpartum relapses
In all, 126 patients (36%) had a relapse and 22 (6.3%) had two or more relapses in the postpartum year. Relapses were treated with intravenous methylprednisolone 1 g for 3–5 consecutive days. Patients with relapses had higher EDSS scores at conception (median and IQR 1.5, 1.0–2.0 vs 1.5, 1.0–1.5; p=0.007) and experienced a higher number of relapses in the year before pregnancy and during pregnancy (table 1). Proportional hazards assumptions for variables included in the Cox regression model were met. The multivariable analysis confirmed EDSS scores ≥2.0 at conception (HR=1.4; 95% CI 1.1 to 2.0; p=0.046) and higher number of relapses in the year before pregnancy (HR per unit increase=1.5; 95% CI 1.2 to 1.8; p<0.001) and during pregnancy (HR per unit increase=2.3; 95% CI 1.6 to 3.4; p<0.001) as significant predictors of postpartum relapses (table 2). Early introduction of DMDs after delivery was marginally associated with a lower risk of postpartum relapses (HR=0.7, 95% CI 0.4 to 1.0; p=0.079; mean time to first relapse 10.32 (95% CI 9.72 to 11.04) months vs 8.88 (95% CI 8.28 to 9.36) months; Log-Rank p=0.006) (table 2).
Predictors of postpartum disability
Full data on disability progression in the first year after delivery were available in 338 patients. A total of 44 patients (12.6%) progressed on the EDSS in the year after delivery (table 3). Patients with disability progression had a higher number of relapses in the year before pregnancy and after delivery (table 3). Proportional hazards assumptions for variables included in the Cox regression model were met. In the multivariable analysis, significant predictors of disability progression in the year after delivery were confirmed to be a higher number of relapses in the year before pregnancy (HR per unit increase=1.4, 95% CI 1.1 to 1.9; p=0.047) and after delivery (HR per unit increase=2.7, 95% CI 1.4 to 5.2; p=0.002) (table 4). There was no significant relationship between early DMDs and disability progression in the year after delivery.
In the last decade, the issues of conception, pregnancy and delivery in MS patients have received renewed interest following the introduction of DMDs and a paradigm shift towards early treatment. There is consistent evidence that, although pregnancy reduces the risk of relapses, particularly during the third trimester, the 12-month period after delivery, and especially the first trimester, is characterised by a significant increase in relapse rate and represents a critical phase for patient counselling and therapeutic decision-making.6 In a meta-analysis of 13 published studies, the risk of relapse during pregnancy decreased on average from a mean value of 0.44±0.02 to 0.18±0.01 (p<0.0001) and increased up to 0.70±0.02 in the year after delivery.6
Different strategies have been suggested to prevent postpartum relapses, including treatment with intravenous immunoglobulins14 or the use of progestin and oestradiol.15 The role of breast feeding, in this context, remains controversial, with a recent study suggesting a protective effect.3 The effect of breast feeding, however, seems to be neutral in most published studies, including the large, prospective, collaborative European and Italian studies.2 ,4 Breastfeeding behaviour may simply reflect different patient choice depending on disease activity, so that women with less active disease are usually more likely to breast feed.
In this collaborative, prospective study based on the Italian pregnancy dataset, we focused on the predictors of postpartum relapses taking into account breastfeeding behaviour and DMD treatment. We confirmed the well-known profile of relapse rate during pregnancy and after delivery. Consistent with the PRIMS study,8 ,16 ,17 patients with higher EDSS at conception and higher relapse rate prepregnancy and during pregnancy were at higher risk of postpartum relapses. In particular, taking into account a series of possible confounders, the risk of relapse in the first year after delivery was higher in patients who experienced relapses in the prepregnancy year and during pregnancy. However, 30% of patients experienced postpartum relapses despite the absence of previous clinical activity, highlighting the postpartum period as a ‘critical’ phase for possible recrudescence of disease activity.
While the risk of relapses after delivery and possible clinical predictors have been previously described, our study adds to the knowledge in the field. On the one hand, the occurrence of relapses in the year after delivery appears not to be neutral in terms of disability progression. Indeed, postpartum relapses were associated with an approximately twofold increased risk of disability accrual in the year after delivery. This observation is in line with previous data obtained from randomised trials that highlight the role of relapses in disability accumulation, at least in the short-term period.18–20 On the other hand, the early introduction or resumption of DMDs within the first trimester after delivery marginally reduced the risk of postpartum relapses. A protective role of DMD exposure of at least 8 weeks during pregnancy on the risk of postpartum relapses has been already reported.21 In our study, 82.4% of patients starting DMD early after delivery were already on DMD before pregnancy. This suggests that, at least in our cohort, decision on treatment in the postpartum period was steered by DMD history. However, the protective role of DMDs did not apply to disability accumulation. In this regard, it is possible that the sample size, low disability levels at conception and duration of follow-up did not provide adequate power to reveal DMD efficacy on disability progression. In the PRIMS study, disability progression in the follow-up period was estimated to corresponded to what is expected in the general population of MS patients.8 ,16 ,17 In the absence of a concurrent control group, this remains a controversial point. Recently, a population-based study from the British Columbia MS Clinic found that term pregnancies had no effect on the time to reach EDSS 6.0, which was predicted only by a progressive course and a greater age at MS onset.22 However, in this study a ‘conception behaviour bias’ cannot be excluded. Moreover, the analysis also took into account pregnancies occurring before MS onset. It is also possible to speculate that, despite relapse-related disability increase in the year after delivery, long-term outcome remains largely independent of postpartum relapses and therefore not significantly affected by pregnancy. Indeed, studies on the long-term prognosis of MS have revealed that early features of the disease do not fully account for disease course in the long-term.23 The role of breast feeding remains controversial. Current data obtained on a larger sample of patients confirmed our previous observation that, when analysed together with other confounders, breast feeding did not seem to protect against the risk of postpartum relapses.2
A few possible limitations of our study are worth being discussed. The study sample included 350 out of the 386 pregnancies in the original cohort. However, the 36 (9.3%) women lost to follow-up did not significantly differ in terms of the main clinical and demographic characteristics from patients assessed in the present study, with the exception of the proportion of DMD treatment before pregnancy, which was higher in dropped patients (75% vs 49.7%; p=0.004) (see online supplementary e-table 2). Moreover, although the participating MS centres cover the entire national territory, the study population may not be entirely representative of the general population of MS patients. The assessment of DMD efficacy after delivery was performed using an arbitrary definition of early DMD introduction on non-randomised groups, without information on disease activity according to MRI. Moreover, DMD efficacy did not reach statistical significance, possibly due to a lack of statistical power (sample size). It has to be noted, however, that the trend towards a protective role of early DMD treatment was confirmed using an earlier cut-off for the classification of patients (ie, 2 months; see online supplementary e-table 3), whereas it was lost using a later cut-off (ie, 4 months; online supplementary e-table 3). This could be related to the occurrence of the majority of postpartum relapses during the first trimester after delivery, and points to a possible advantage only for ‘very’ early introduction of DMD in the postpartum period. Despite these considerations, the study findings are relevant to MS patient counselling and management. In particular, our study highlights that: (a) even with a large proportion of patients treated before pregnancy (49.7%), 43% are expected to relapse in the postpartum year; (b) the more disabled patients and those with a history of relapses before and during pregnancy have an increased risk of postpartum relapses; (c) patients who relapse after delivery have an approximately twofold increase in the risk of disability accumulation during the postpartum year; (d) the early introduction or resumption of DMDs after delivery may be a suitable strategy to reduce the risk of postpartum relapses.
Taken as whole, our results foster early DMD (re)-introduction after delivery, particularly in women with higher disability levels and history of relapses before and during pregnancy. To date, these variables may be overlooked in the therapeutic decision-making process since DMDs are started mainly in patients who were already treated before pregnancy, independently of disease activity.
Finally, it is noteworthy that our findings apply to disease course in the short-term period.
Over the long-term period, available evidence does not point to any major deleterious effect of pregnancy on disease course and a few studies do suggest a possible protective role24–26 so that, overall, patients can be reassured as to the impact of pregnancy on the long-term prognosis of the disease.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Collaborators The MS Study Group of the Italian Neurological Society: University of Florence: MP Amato, MD, E Portaccio, MD, B Hakiki, MD, A Sturchio, MD, L Pastò, MD, M Giannini, MD, L Razzolini, MD E Piscolla, MD, G Siracusa, MD. Hospital of Gallarate: A Ghezzi, MD, A Rizzo, MD, M Zaffaroni, MD. San Raffaele Institute, University of Milan: V Martinelli, MD, M Radaelli, MD, L Moiola, MD, G Comi, MD. Niguarda Hospital, Milan: A Protti, MD, C Spreafico, MD, R Marazzi, MD. University of Torino: P Cavalla, MD; S Masera, MD. Neurological Institute Mondino: R Bergamaschi, MD. University of Genova: G Mancardi, MD, E Capello; MD. Department of Neurology, Genova: C Solaro, MD. University of Ferrara: MR Tola, MD, L Caniatti, MD. University of Parma: F Granella, MD, P Immovilli, MD. University of Siena, Clinical Neuroimmunology Unit: P Annunziata, MD; L De Santi, MD. Hospital of Grosseto: K Plewnia, MD. Hospital of Empoli: L Guidi, MD, ML Bartolozzi, MD. Hospital of Lucca: M. Mazzoni, MD. University of Rome: C Pozzilli, MD, L De Giglio, MD. University of L'Aquila: R Totaro, MD, A Carolei, MD, M Rossi, MD. University of Chieti: A Lugaresi, MD, G. De Luca, MD, V Di Tommaso, MD. University of Bari: M Trojano, MD, D Paolicelli, MD, A Carrozzo, MD, M D'Onghia, MD. University of Cagliari: MG Marrosu, MD, L Musu, MD. University of Catania: F Patti, MD, L Carmela, MD, S Lo Fermo, MD, PhD.
Contributors EP: Study concept and design, analysis and interpretation, critical revision of the manuscript. AG, VM, FP, GLM, CS, MRT, RT and MGM: Acquisition of data, critical revision of the manuscript. BH, AS, LM, LDG, GDL and DP: Acquisition of data. CP, AL, GC and MT: Analysis and interpretation, critical revision of the manuscript. MPA: Study concept and design, analysis and interpretation, critical revision of the manuscript, study supervision.
Competing interests EP serves on a scientific advisory board for Biogen-Idec, Merck Serono and Bayer, received honoraria for speaking from Biogen-Idec and Teva, and receives research support from Merck Serono, Biogen-Idec, Bayer Schering and Sanofi-Aventis. AG received honoraria for speaking from Bayer Schering, Merck Serono and research grants from Sanofi-Aventis, Biogen Idec and Merck Serono. BH received research support from Merck Serono, Novartis and Teva. VM has received speaker honoraria and funding for travel from Biogen-Dompè SG, Merck Serono, Bayer Schering Pharma, Novartis and Sanofi-Aventis and has served as a consultant to Bayer Schering Pharma, Sanofi-Aventis and Teva Pharmaceutical Industries. FP served on a scientific advisory board for Bayer, Biogen-Idec, Merck Serono, Novartis and Sanofi-Aventis, received honoraria for speaking from Bayer, Biogen-Idec, Merck Serono, Novartis and Sanofi-Aventis and funding for travel from Merck Serono and TEVA. GLM has received honoraria for lecturing, travel expenses for attending meetings and financial support for research from Bayer Schering, Biogen Idec, Sanofi-Aventis and Merck Serono Pharmaceuticals. CP received honoraria for consultancy or speaking from Sanofi-Aventis, Biogen, Bayer Schering, Novartis, Teva, Merck Serono and research grants from Merck Serono, Sanofi-Aventis, Novartis and Biogen. RT received honoraria for consultancy or speaking from Bayer-Schering Pharma, Biogen Idec, Merck Serono, Novartis Pharma, Sanofi-Aventis and Teva. AL serves on a scientific advisory board for Bayer Schering, Biogen Idec, Genzyme and Merck Serono, received travel grants and honoraria from Bayer Schering, Biogen Idec, Merck Serono, Novartis, Sanofi-Aventis and Teva and research grants from Bayer Schering, Biogen Idec, Merck Serono, Novartis, Sanofi-Aventis and Teva, received travel and research grants from the Associazione Italiana Sclerosi Multipla and is a Consultant of “Fondazione Cesare Serono”. MGM serves on a scientific advisory board for Merck Serono and Bayer Schering, and receives research support and honoraria for speaking from Biogen-Idec, Merck Serono, Bayer Schering and Sanofi-Aventis. GC received honoraria for consultancy or speaking from Teva Neuroscience Plough Corporation and Merck Serono. MT received honoraria for consultancy or speaking from Sanofi-Aventis, Biogen and Bayer Schering, and research grants from Merck Serono. MPA serves on scientific advisory boards for Biogen-Idec, Merck Serono, Bayer Schering and Sanofi-Aventis and receives research support and honoraria for speaking from Biogen-Idec, Merck Serono, Bayer Schering and Sanofi-Aventis.
Patient consent Obtained.
Ethics approval Ethics Committee of the University of Florence.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.