Introduction

Rheumatoid Arthritis (RA) and Psoriatic Arthritis (PsA) are two most frequent types of chronic systemic inflammatory arthritides (IA) that can cause irreversible joint damage, disability and increased mortality if sub-optimally treated1,2,3. Many studies have shown that early aggressive treatment of IA with disease modifying anti rheumatic drugs (DMARDs), which can be conventional synthetic DMARDs (csDMARDs), targeted synthetic DMARDs (tsDMARDs) and biologic DMARDs (bDMARDs), directed at achieving disease remission leads to better outcomes including reduction in pain, joint destruction, improved function, quality of life, and radiographic progression4,5,6,7.

For women of childbearing age with a diagnosis of RA/PsA, there are multiple issues surrounding the peri-partum period including medication use, fertility, risk of disease flare and potential impact on neonatal and maternal outcomes. We define peripartum period in this study as the immediate 3–6 month period preceding conception, time period during pregnancy and up to 24 months post partum. Furthermore, decision-making regarding medications before, during and after pregnancy are challenging, and can vary significantly between rheumatologists leading to great differences in the patient experience8,9.

Evidence about the peri-partum period in RA/PsA is limited to cohort studies with small numbers. In general, RA is seen to improve or remit in about 60–80% of patients during pregnancy; however, a significant number of patients still experience active disease, which require treatment decisions by the rheumatologist10,11. Therefore, some RA/PsA patients will choose to have smaller families (fecundity) to avoid the risk of flare during conception attempts or an unsuccessful pregnancy10,12. While certain DMARDs must be clearly discontinued due to teratogenic potential (eg. methotrexate), recommendations for the use of biologics during pregnancy emphasize that the decision should be individualized, although there is increasing evidence indicating safety of biologics during pregnancy13. Furthermore, the recommendation to discontinue certain TNF inhibitors (TNFi) at 20–32 weeks of pregnancy also lacks sufficient data8,14.

Considering all the challenging issues surrounding potential pregnancies in RA/PsA patients, the objectives of this study were: (1) To gain insight into fertility and fecundity in RA/PsA patients in northern Alberta participating in the Rheumatoid Arthritis Pharmacovigilance Program and Outcomes Research in Therapeutics (RAPPORT) prospective cohort (2) To describe the peripartum period for RA/PsA patients in northern Alberta interms of preconception counselling, unplanned/planned pregnancy, time to pregnancy, requirement for infertility treatment, pregnancy outcomes, number of pregnancies prior to and after IA diagnosis, disease activity during and after pregnancy, type of delivery and breast feeding (3) To understand the choices made by females with RA/PsA in childbearing age regarding arthritis medication usage (csDMARDs, tsDMARDs, bDMARDs, steroids, non-steroidal anti-inflammatories) (4) To understand the fetal and maternal outcomes in the peri-partum period for RA/PsA patients in northern Alberta in terms of neonatal complications, labour and delivery complications. We hypothesized that the presence of a RA/PsA diagnosis in women of childbearing age had important consequence on pregnancy, including decreased fecundity in patients after the diagnosis due to fear of adverse pregnancy outcomes as a result of the disease itself and medications, increased pregnancy complications in patients with active RA/PsA disease and treatment discontinuation.

Materials and Methods

We conducted a survey-based retrospective evaluation of the peri-partum period in RA/PsA patients participating in the RAPPORT (Rheumatoid Arthritis Pharmacovigilance Program and Outcomes Research in Therapeutics) prospective cohort in northern Alberta, Canada. The RAPPORT program is an ongoing prospective inception cohort of IA patients on biologics in the province of Alberta, Canada since 2004. Total number of patients seen at the time of the survey was 2659. Following ethics approval from the University of Alberta Research Ethics Board (Pro00064218), an anonymous electronic-based survey was sent by email link to all female patients of child-bearing age (<50 years) in the RAPPORT database (440 patients in total) who have previously provided informed consented to participate in the RAPPORT program and agreed to receive communication regarding future studies and questionnaires related to RAPPORT. Implied consent was granted by the action of completing the survey sent by email link. All methods were performed in accordance with relevant guidelines and regulations. No paper-based version of the questionnaire was utilized. No authors had competing interests related to this study.

This survey collected the following information from the patients: (1) the impact of the disease on fertility and fecundity, pregnancy and post-partum period, (2) course of disease during pregnancy assessed based on whether joint pain, swelling or tenderness was limited during pregnancy, (3) complications of pregnancy including Preterm labour (delivery <37 weeks), Premature rupture of membrane, Prolonged labour (labour lasting for >20 hours (1st pregnancy) and >14 hours (subsequent pregnancies), Haemorrhage (bleeding), Infection, Umbilical cord prolapse (umbilical cord comes out of the uterus with or before the presenting part of the fetus), Stillbirth (baby born with no signs of life at or after 24 weeks gestation) (4) medication use during pregnancy including medications taken when the patient realized she was pregnant, medications discontinued and continued during pregnancy and biologic therapies, (5) post-pregnancy issues for the mother and baby including disease flare, breastfeeding, and medication changes, birth defects, neonatal ICU admission as well as (6) choices women make around pregnancy in general (Supplemental Appendix 1: Survey).

All information was collected online through RedCap, a web-based survey tool, which is supported through EPICORE (Epidemiology Coordinating and Research Centre) at the University of Alberta. All responses were obtained and collected over a 6-week period with email reminders sent twice over the period for those who did not respond. Information obtained through the survey was evaluated for pregnancy-related outcomes using descriptive statistics, Chi-Square and Fisher’s exact data. Chi-square and Fisher’s exact tests were used to analyze pregnancy variables such as pregnancy complications, pregnancy planning, labour and delivery complications in pre and post RA/PsA pregnancies.

Results

During the 6-week collection period, 162 patients completed the survey (response rate of 37%), with 133 patients having a diagnosis of RA and 29 patients having a diagnosis of PsA, the majority of patients ranging in age from 41–50 (Table 1). One hundred and three (103) patients had 234 pregnancies with 164 (out of 234) pregnancies occurring prior to the RA/PsA diagnosis and 70 (out of 234) pregnancies occurring after the diagnosis. Seventeen patients had a pregnancy, developed RA/PsA and then had at least one more pregnancy after their diagnosis. Pregnancy outcomes from 103 patients included: 96% live births, 1.9% stillbirths, 23% miscarriages, and 15% therapeutic abortions (Table 1).

Table 1 Characteristics of RAPPORT Survey Patients.
Full size table

Thirty-six percent (36.4%) (59/162) of patients were never pregnant due to social reasons such as financial, relationship status or personal choice (18/59), RA/PsA medications (15/59), infertility (10/59) and RA/PsA disease activity (9/59) (7/59 patients provided no or unclear reason). Of the patients who had pregnancies, 33% of patients had fewer children than desired due to RA/PsA medications (15/34), RA/PsA disease activity (14/34), infertility (5/34), social reasons (5/34) and other co-morbidities (5/34) (Table 2). In comparison to patients who had their desired number of children, patients with fewer children than desired had fewer total pregnancies, live pregnancies, therapeutic abortions and more stillbirths and miscarriages (Table 3).

Table 2 Reasons for limiting family size and for not having children.
Full size table
Table 3 Detailed Pregnancy Outcomes.
Full size table

In the 63 pregnancies that occurred after their RA/PsA diagnosis (excluding those with therapeutic abortions): (1) pre-conception counseling was obtained in 49% of pregnancies; (2) no statistically significant difference was noted between those with pregnancies who received pre-conception counseling (31/63) and those that did not (32/63) in terms of disease control during pregnancy, discontinuing medication and pregnancy complications including pregnancy induced hypertension, pre-eclampsia/eclampsia, gestational diabetes, multiple pregnancies, IUGR, placental abruption, flare and hospitalizations; (3) most described good disease activity control during pregnancy (41/63 pregnancies, 65%) but flared (28/38, 73.7%) in the first 3 months postpartum (19/28 patients, 67.9%); (4) RA/PsA medication (including NSAIDs, steroids, s/bDMARDs) was discontinued during 79% of pregnancies (Table 4); (6) no statistically significant difference noted in disease control at the time of pregnancy and disease flare during pregnancy between pregnancies with medication discontinuation and those pregnancies who continued medication (7) 35% of pregnancies (22/63) occurred while on biologic therapy at the time of or prior to conception and continued for all or part of the pregnancies in 5 cases (3/5 with complications including ectopic pregnancy, IUGR, multiple birth, flare).

Table 4 Peri-partum outcomes in the 63 pregnancies that occurred after their RA/PsA diagnosis (excluding TAs).
Full size table

For patients planning conception (46/63 post-arthritis, 107/153 pre-arthritis diagnosis pregnancies) time to pregnancy was 0–2 months in 41% post-arthritis vs 46% pre-arthritis diagnosis pregnancies (Table 5). Gestational age at time of delivery was 37–40 weeks in 58% (33/57) post-arthritis vs 66% (83/126) pre-arthritis diagnosis pregnancies. There were no statistically significant differences in pregnancies that occurred before or after diagnosis regarding pregnancy planning, use of fertility treatment (Clomiphene, hCG/FSH/hMG/GnRH agonist/GnRH antagonist hormone injections, IVF and IUI), pregnancy complications (with and without disease flare), labor and delivery complications, delivery methods, breastfeeding, birth defect frequency or neonatal complications. Pregnancy complications (with and without disease flare), labor/delivery complications and C-section deliveries were lower in pregnancies prior to diagnosis; however, this was not statistically significant. Neonatal ICU admissions were significantly lower in pre-arthritis diagnosis pregnancies compared to post-arthritis diagnosis pregnancies. No pregnancy complications were noted in 24/54 pregnancies on medications compared to 6/9 pregnancies not on medications (not statistically different) (Table 5).

Table 5 Pregnancy variables in pre-and post-RA/PsA pregnancies.
Full size table

When comparing the 1st pregnancy of patients who had all pregnancies prior to arthritis diagnosis to 1st pregnancy of patients who had all pregnancies after the diagnosis, the patients with 1st pregnancy after arthritis diagnosis were older and had a longer time to achieve pregnancy, although this was not statistically significant. There was also no significant difference in terms of the number of patients who used infertility treatment such as in vitro fertilization, intrauterine insemination, and hormone injections as well as pregnancy complications between these 2 groups.

Discussion

This survey of the Northern Alberta RAPPORT patient participants provides real-world insight into pregnancy outcomes for RA/PsA patients and the careful considerations women must make in contemplating this journey in the context of these diseases. We identified that a third of women with RA/PsA had fewer children than desired due to factors including disease activity and medications. In women with a diagnosis of RA/PsA, flares were very common post-partum. Disease-related medications were discontinued in nearly eighty percent of women during or prior to pregnancy. No difference was found in peripartum outcomes for women who had pregnancies before compared to after their arthritis diagnosis except for neonatal ICU admissions.

This data is important in light of evolving clinical trial and observational data on newer agents in treating RA/PsA and changing opinions of safety of various agents (eg. TNFi). Continued heterogeneity exists in the peri-partum experience amongst women as evidenced by our results, which reflects a complex interplay between patient and rheumatologists and managing RA/PsA during childbearing years.

Our study confirms that patients with RA/PsA who had fewer children than desired had fewer total pregnancies, live pregnancies, therapeutic abortions and more stillbirths, and miscarriages compared to those patients who had expected number of children12. This likely represents the true nature of this population as the anonymity of our survey-based study likely lead to limitation of the social desirability bias, as the patient answers were not shared with care providers.

Our study demonstrates that approximately 36% of all the survey respondents were never pregnant mainly due to their arthritis related concerns and infertility. Furthermore, 33% of patients with pregnancies had less children than desired due to their arthritis related concerns and infertility. These results parallel previous studies indicating that arthritis related concerns in addition to infertility lead to smaller family sizes in a substantial number of patients who have not finished having children at the time of the RA/PsA diagnosis10,12. Given the availability of pregnancy safe medications for RA/PsA patients in the peripartum period10,15,16, the perception of the negative effects of RA/PsA medications seems to be an unnecessary reason for decreased fecundity. Provision of accurate and appropriate education regarding these medications and the impact of pregnancy and disease on each other is essential to ensure patients make informed decisions about reproduction. In contrast to Clowse et al.12, we found a lower infertility rate among patients who had fewer children than desired, possibly reflecting survey bias whereby patients with difficult reproductive histories may not have completed the survey.

Similar to other studies, we found that the majority of RA/PsA patients reported low disease activity during pregnancy but flare post-partum. Previous studies have shown disease improvement rates between 48%–95%, which is consistent with the low disease activity state of 65% in our study. Furthermore, our postpartum flare rate of 74% was similar to rates of 39–90% demonstrated in previous studies10,17,18. The rates observed in our study are on the higher end of the spectrum, likely because subjective measurements were used in our study in keeping with the higher disease activity scores reported in patient reported outcomes compared to physician scores based on objective measurements used in other studies17. It is important to note that the rate of remission/improvement during pregnancy is less than what is commonly perceived and given the moderate- high rates of disease flare post-partum18,19,20, continuation of appropriate medication during and after pregnancy is imperative to prevent RA/PsA disease flares.

Our survey identified a statistically increased neonatal ICU admission rate in pregnancies that occurred after the diagnosis of RA/PsA, similar to the findings of Barnabe et al.21. We speculate that possible contributions to this adverse outcome include use of medications such as corticosteroids, reproductive technologies, and active disease. Although not statistically significant, pregnancies occurring after the diagnosis of IA had more pregnancy complications, labour and delivery complications and C-section deliveries compared to pregnancies occurring before the RA/PsA diagnosis10. Age may have been a contributing factor for increased complications as patients with all pregnancies after RA/PsA diagnosis were older compared to patients with all pregnancies prior to diagnosis, although this was not statistically significant in our study, likely due to the low and imbalanced sample size between the two groups. In addition to lower numbers, peri-natal complications may not have achieved statistical significance as the majority (65%) of pregnancies occurred in periods of low disease activity which is associated with better peri-partum outcomes and because of the demographics of the study participants18,19,20,21.

As expected, due to concerns of teratogenicity, the majority of patients discontinued RA/PsA medications during pregnancy in our study15. Our study found no statistically significant difference in disease control during pregnancy between pregnancies with medication discontinuation and those pregnancies who continued medication likely due to the fact that not all medications (i.e s/bDMARDs) were discontinued. Furthermore, if just one medication was discontinued (eg. Steroids, or 1 DMARD if on multiple), these patients were counted under “discontinued medication group” even if the patients may have still been maintained on certain s/bDMARDs. It is also possible that pregnancy may have kept the disease under better control. While the use of certain DMARDs such as methotrexate and leflunomide is restricted during pregnancy and breast-feeding, increasingly more treatment options (eg. hydroxychloroquine, sulfasalazine, TNFi) are available to maintain remission or low disease activity state than ever before10,15,16. Despite the data showing low risk of tumor necrosis factor inhibitors (TNFi) use during pregnancy8,14,16,20,22, only 5 pregnancies in the study continued biologic therapy and three cases had notable complications, the details of which are not available due to the nature of the survey. Possible explanations include incorrect fear of medication teratogenicity, pregnancies that occurred in the earlier part of the RAPPORT cohort where experience in using TNFis was limited and patient reluctance. Of the 5 pregnancies, only 1 continued biologic therapy into the 3rd trimester.

Several limitations must be noted in this study. Patients included both RA and PsA patients even though these are two distinct disease entities that may confer different risks. Furthermore, the smaller sample size may have led to the lack of significant associations between perinatal events and RA/PsA. This was a retrospective, self-reported survey where quality would be downgraded due to selection and recall bias. Issues surrounding childbearing can be extremely sensitive and our response rate of thirty-seven percent may reflect a reluctance of some women to share their experience but was considered reasonable in comparison to other pregnancy-survey studies23. We created responses requiring dichotomous or drop-down list answers to reduce free text and ensure more homogenous data collection.

In conclusion, our study demonstrates that peri-partum complications in RA/PsA patients do not significantly increase in patients with good disease control; however, neonatal ICU admissions are increased in women with RA/PsA revealing the importance of decision-making before, during and after pregnancy. Most patients discontinued RA/PsA medication during pregnancy and despite the data showing low risk of TNFi use during pregnancy, only a minority of patients continued biologics during pregnancy, highlighting the lack of consensus on use and the appropriate time to discontinue TNFi. Therefore, providing patients with medication and disease information is essential to ensure that patients make informed and educated reproductive decisions.