Therapeutic challenges in the treatment of systemic inflammatory disease in pregnancy

Short abstract

Background

Granulomatosis with polyangiitis and Behçet’s disease can occur during pregnancy and may be treated by ophthalmologists, rheumatologists, and obstetricians. We hypothesized that specialty training would affect the way physicians selected therapy.

Methods

Using an online questionnaire, 209 uveitis specialists, 853 rheumatologists, and 2500 obstetricians were surveyed. Respondents were given clinical vignettes containing a female patient who was contemplating pregnancy or in the first trimester and was diagnosed with granulomatosis with polyangiitis or Behçet’s disease.

Results

In the patient with granulomatosis with polyangiitis, therapy choice between specialties for biologic versus non-biologic systemic immunosuppressive medications was significantly different for both the non-pregnant and pregnant patient (p < 0.00001, p < 0.00003). In the non-pregnant patient diagnosed with Behçet’s disease, the therapy choice between biologic versus non-biologic medications was also significantly different (p < 0.0003).

Conclusions

Specialty training affects how physicians manage granulomatosis with polyangiitis and Behçet’s disease. Development of inter-specialty guidelines and treatment plans may improve outcomes, communication, and patient care.

Pregnancy presents challenges in the treatment of many illnesses including systemic inflammatory diseases. While randomized controlled trials (RCTs) are the gold standard for developing evidence-based guidelines, often RCTs are difficult or impractical due to many factors, such as cost and patient recruitment. In pregnancy, ethical issues further compound the difficulty of performing RCTs and clinicians are left with clinical judgement and data from case reports to guide their management.

Granulomatosis with polyangiitis (GPA) and Behçet’s disease (BD) are two systemic inflammatory diseases that may have ocular involvement and can occur during pregnancy. Ocular GPA occurs in 14–60% of cases^1,2 and ocular BD occurs in 29–100% of cases.^3,4 It has been shown that various specialties manage diseases differently.^5–10 Due to the rarity of these diseases as well as the complexity of pregnancy, we hypothesized that obstetricians, ophthalmologists, and rheumatologists would manage pregnant women with GPA and BD differently. We developed an online survey to evaluate therapeutic strategies of these differing specialties.

Methods

An online survey was developed that included clinical vignettes of a woman diagnosed with either GPA or BD who was contemplating pregnancy or was currently in the first trimester of pregnancy. The survey included a total of four clinical vignettes with questions relating to the management of these diseases as well as counseling regarding pregnancy. The study was approved as exempt by the Colorado Multiple Institution Review Board (protocol 17–7778) and the requirement for consent was waived by the exempt approval.

An invitation to the online questionnaire was distributed to 209 members of the American Uveitis Society, 853 members of the American College of Rheumatology, and 2500 obstetricians via email and social media posts. The electronic invitations contained a link that led the respondents to the online questionnaire. Research Electronic Data Capture software was used to conduct the survey. All data were collected voluntarily and anonymously. The survey respondents were asked to identify whether they were an ophthalmologist, an obstetrician, or a rheumatologist.

The survey respondents were questioned on the following vignettes:

Scenario 1: “You are consulted about a 26-year-old woman, who wishes to get pregnant in the next year, and presents with progressive onset of bilateral lacrimal gland enlargement as well as painful necrotizing scleritis of the right eye. Sinus and chest CT confirm the presence of severe mucosal thickening and inflammation of the sinuses without pulmonary disease. Kidney function and urinalysis are normal. Lab testing reveals a positive C-ANCA and PR-3 antibody. A diagnosis of granulomatosis with polyangiitis (GPA; formerly Wegener's granulomatosis) is made.” The respondents were then asked to select the best choice for treatment. Choices included prednisone + methotrexate, prednisone + azathioprine, prednisone + mycophenolate, prednisone + cyclophosphamide, prednisone + adalimumab or other tumor necrosis factor (TNF) inhibitor, prednisone + rituximab, dexamethasone implant, fluocinolone acetonide implant, subtenon Kenalog® [triamcinolone] injection. The respondents were then asked how they would advise the woman regarding the planned pregnancy.

Scenario 2: Identical to scenario 1, except the woman was seven weeks pregnant. The respondents were asked again what, in their opinion was the best choice for treatment and how they would advise the woman regarding pregnancy.

Scenario 3: “A 26-year-old woman who wishes to get pregnant in the next year presents with her third episode of retinal vasculitis. She has a history of recurrent oral and genital ulcers that are consistent with Behçet’s Disease (BD). Vision is 20/30 in both eyes. She has been on prednisone 40mg for the past six weeks.” The respondents were then asked what, in their opinion, was the best choice for treatment. Choices for this case were the same as for the GPA case. The respondents were then asked how they would advise the woman regarding the planned pregnancy.

Scenario 4: Identical to scenario 3, except the woman was seven weeks pregnant and the respondents were asked what the best choice for treatment was, and how they would advise the woman regarding pregnancy.

For statistical analysis, TNF inhibitors and rituximab were grouped together (biologic medications), azathioprine, methotrexate, and mycophenolate were grouped together (non-biologic systemic immunosuppressive medications), dexamethasone implant and sub-tenon injections were grouped together (local therapy), and cyclophosphamide was left in its own group (alkylating therapy; cytotoxic medication). As most physicians chose to use biologic medications or a drug from the non-biologic systemic immunosuppressive group, chi-square analysis was performed on the resultant 2 × 3 contingency table of these choices for each clinical scenario.

Results

A total of 176 physicians responded to the questionnaire, 41/2500 obstetricians (1.6% response rate), 58/209 ophthalmologists (28% response rate), and 77/853 rheumatologists (9% response rate). Table 1 illustrates the mechanism of action and the potential fetal side-effects of the drugs studied in our survey. Figure 1 shows the data from scenario 1 (the non-pregnant GPA) and scenario 2 (the pregnant GPA). In scenario 1, the difference in therapy choice between specialties for biologic versus non-biologic systemic immunosuppressive medications was significantly different (p < 0.00001). With regard to pregnancy counseling, 71% of obstetricians, 52% of ophthalmologists, and 86% of rheumatologists stated that they would advise the woman to defer pregnancy, while 24% of obstetricians, 45% of ophthalmologists, and 9% of rheumatologists stated that rather than counseling the woman they would recommend that the woman discuss the issue with other medical providers.

Table 1.

Pregnancy-related toxicities of immunosuppressive agents.

Drug	Mechanism of action	Known pregnancy-related toxicity	Comments
Methotrexate	Inhibits DNA synthesis via folic acid reductase inhibition	Congenital malformations and fetal loss	Known to induce fetal death
Mycophenolate mofetil	Inhibits DNA synthesis via IMPDH inhibition	Congenital malformations and first trimester loss	Active metabolite is mycophenolic acid
Azathioprine	Inhibits DNA synthesis as a purine antagonist	Possible congenital malformations, fetal growth restriction	Fetus may not metabolize to active 6-MMP
Cyclophosphamide	Crosslinks guanine in DNA	Fetal loss	Known to induce fetal death
Adalimumab/infliximab	Monoclonal IgG antibody, inhibits tumor necrosis factor activity	No known toxicity	IgG molecules do not cross the placenta in the first trimester
Rituximab	Monoclonal IgG antibody, targets CD20 leading to B cell depletion	Neonatal cytopenia	IgG molecules do not cross the placenta in the first trimester

IMPDH: inosine monophosphate dehydrogenase; 6-MMP: 6-methylmercaptopurine.

Figure 1.

Percentage of physicians selecting therapy choices for a non-pregnant and pregnant patient diagnosed with granulomatosis with polyangiitis.

The majority of obstetricians and ophthalmologists did not change their management strategy if the woman was pregnant. However, rheumatologists selected a non-biologic systemic immunosuppressive medication more often than a biologic medication in this setting. Physician specialty training significantly affected choice of therapy between the non-biologic systemic immunosuppressive and biologic medication groups (p = 0.00003). Five percent of obstetricians, 9% of ophthalmologists, and 1% of rheumatologists stated they would advise the woman to terminate the pregnancy. Fifty-one percent of obstetricians, 72% of ophthalmologists, and 36% of rheumatologists would recommend that the woman discuss this issue with her other medical providers.

The data collected from the BD clinical vignettes are shown in Figure 2. In the non-pregnant woman (scenario 3), a large portion of the obstetricians chose to treat the patient with a drug from the non-biologic systemic immunosuppressive medication group, while most ophthalmologists chose a biologic medication, and rheumatologists were split between the two. These differences were significant (p = 0.0003). Fifty-six percent of obstetricians, 33% of ophthalmologists, and 73% of rheumatologists stated they would advise the woman to defer pregnancy. Thirty-two percent of obstetricians, 57% of ophthalmologists and 14% of rheumatologists stated they would recommend that the woman discuss this issue with her other medical providers.

Figure 2.

Percentage of physicians selecting therapy choices for a non-pregnant and pregnant patient diagnosed with Behçet’s disease.

In a pregnant woman diagnosed with BD (scenario 4), obstetricians and rheumatologists were split between medications from the biologic and non-biologic systemic immunosuppressive classes, but the majority of ophthalmologists continued to select biologic medications. No obstetricians, 3% of ophthalmologists, and 3% of rheumatologists would recommend terminating pregnancy; 44% of obstetricians, 78% of ophthalmologists, and 29% of rheumatologists would recommend that the patient discuss this issue with her other medical providers.

Of the drugs in the non-biologic systemic immunosuppressive group, azathioprine was the clear majority in every clinical scenario; chosen 80–95% of the time. Figures 3 and 4 show the breakdown of biologic medication chosen by each specialty in GPA and BD, respectively. In GPA, pregnancy status significantly affected therapy choice by ophthalmologists as the majority switched their preference from rituximab to an anti-TNF agent (p = 0.0008).

Figure 3.

Percent of physicians selecting biologic medications for therapy in a non-pregnant and pregnant patient with granulomatosis with polyangiitis.

Figure 4.

Percent of physicians selecting biologic medications for therapy in a non-pregnant and pregnant patient with Behcet’s disease.

Discussion

The results of this study show that there is a specialty dependent difference in management of GPA and BD in the pre-pregnancy patient between obstetricians, ophthalmologists, and rheumatologists. The results also show that there is a difference between specialties when it comes to managing these diseases in a pregnant woman. Furthermore, the data suggest that the physician surveyed felt more comfortable providing pre-pregnancy counseling, compared to counseling in the pregnant state, as the percentage of respondents electing to defer counseling to another provider rose in all three physician groups in this scenario. These conclusions have potential implications for women and for healthcare as a whole.

In a non-pregnant woman with GPA, most obstetricians chose a non-biologic systemic immunosuppressive medication, whereas most ophthalmologists and rheumatologists chose a biologic medication. The decision by the ophthalmologists and rheumatologists may have been influenced by recent studies demonstrating the effectiveness of rituximab for induction therapy (non-inferiority compared to the alkylating-drug cyclophosphamide) as well as maintenance therapy (superiority over azathioprine).^11,12 While our survey did not assess the reason behind each selection, it is possible that training and familiarity with certain drugs influenced this decision.

Pregnancy status did not seem to affect the choices made by obstetricians or ophthalmologists, but it did affect choices made by rheumatologists. In pregnancy, rheumatologists showed a preference for non-biologic immunosuppressive medications. The specific reasons behind this preference were not elicited by this study, but it is plausible that the historical use of azathioprine in pregnant women with conditions such as inflammatory bowel disease, rheumatic disease, or post-solid organ transplantation, and its relative safety in these settings, played a role.^13–15 A study of umbilical cord blood from neonates whose mothers were treated with azathioprine demonstrated absence of the active 6-methylmercaptopurine metabolite suggesting that azathioprine is not fully metabolized in the fetus.¹⁶ Data on the relative safety of anti-TNF medications and/or rituximab in pregnancy are more limited in comparison. While it appears that rituximab administration may be safe among pregnant patients, data are mainly limited to case reports.^17,18 In a recent systematic review of the literature, the European League Against Rheumatism concluded that strong evidence is lacking on the safety of rituximab in pregnancy, although no clear evidence of fetal malformation has been documented.¹⁴ This group recommended the use of rituximab in pregnancy in “exceptional cases,” with preference for use earlier in gestation, given the potential for cytopenia in the newborn when the medication is administered late in pregnancy. Placental transfer of immunoglobulin (including biologic medications such as anti-TNF medications and rituximab) does not occur, to a large degree in the first trimester.¹⁹ As such, use of anti-TNF medications is generally accepted prior to conception, as well as during the first trimester. Use of these medications after this time point is left to the discretion of the patient and treating provider. Current data have not shown evidence of increased risk of fetal malformation among the anti-TNF medications; and despite placental transfer of these medications to the fetus across the second and third trimester, no signal of increased risk in these time periods has been shown.²⁰

A clear majority of ophthalmologists chose to treat BD with a biologic medication, whereas the other specialties were divided. The American Uveitis Society recommends the use of an anti-TNF alpha inhibitor as first line in the treatment of ocular BD.²¹ The data in the literature on the use of anti-TNF drugs in pregnancy, as noted above, mostly come from observational studies of women with chronic arthritis or inflammatory bowel disease.²² Since IgG molecules do not cross the placenta in the first trimester, many experts consider them safe in early pregnancy.^19,23 However, there may be differences in placental transfer among different anti-TNF drugs since adalimumab and infliximab are IgG molecules, whereas certolizumab is a Fab fragment and etanercept is a fusion protein of an IgG molecule with the TNF receptor.

There are a few limitations to our study. The small sample size and low response rates may have introduced bias into our study. Also, most general obstetricians seldom treat these diseases and so the responses may not accurately reflect the complete obstetric opinion. Our questionnaire was largely distributed to physicians in academia and so the results may be less generalizable.

In the pregnant woman diagnosed with GPA or BD, most obstetricians and ophthalmologists, and a third of rheumatologists, recommended that the woman discuss the issue of potential pregnancy termination with her other medical providers. This has the potential to cause confusion, especially if a single provider fails to take primary responsibility on this important issue. As we have also shown, obstetricians, ophthalmologists, and rheumatologists often recommend different treatments for the same disease and mirrors our studies in ocular infectious disease.²⁴ It is possible that a woman may see three different specialists and get three differing opinions about treatment as well as pregnancy counseling. While a randomized controlled trial will likely never be possible in these diseases, coordinated inter-specialty guidelines and treatment plans may be worthwhile to improve outcomes, communication, and ultimately, patient care.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research to Prevent Blindness, Inc., New York, NY provided a challenge grant to the University of Colorado, Department of Ophthalmology. The sponsor had no role in the design or analysis of this study.

Ethical approval

The study was approved as exempt by the Colorado Multiple Institution Review Board (protocol 17–7778) and the requirement for consent was waived by the exempt approval.

Guarantor

AGP

Contributorship

JSF, JRK and AGP researched literature and conceived the study. JSF, JRK, JLH and AGP were involved in protocol development and gaining ethical approval, JSF, JLL, JNC, AML and AGP were involved in study recruitment and data analysis. JSF wrote the first draft of the article. All authors reviewed and edited the article and approved the final version of the article.