ELECTIVE TERMINATION OF PREGNANCY IN AUTOIMMUNE RHEUMATIC DISEASES: EXPERIENCE FROM TWO DATABASES

Abstract

Objective

To provide reference data regarding frequency and safety of elective termination of pregnancy (ETOP) in women with autoimmune rheumatic diseases (ARD) followed in two referral databases.

Methods

Two large databases, a referral ARD clinical practice (BVC) with a known interest in pregnancy and an observational study of SLE and antiphospholipid antibody-associated pregnancies (PROMISSE), were interrogated for histories of prior ETOP and complications related to incident ETOP.

Results

Of women who had had prior pregnancies, 21.7% of 1,307 women in the BVC database and 25.3% of 297 patients in the PROMISSE database gave histories of 1 to 5 prior ETOPs; BVC patients reported no flare-up or hospitalizable complication due to ETOP. Of 674 incident pregnancies, termination for fetal or maternal reasons was recommended for 15 (2%); of these, 2 fetuses died before the procedure was carried out and one woman declined termination and though gravely ill successfully delivered. She died of cardiomyopathy three years later.

Conclusion

Many ARD patients undergo ETOP; few report complications. In medically indicated ETOP there are no adverse signals of unusual complications or disease flare.

INTRODUCTION

The American College of Rheumatology Reproductive Health Guidelines (RHG) (1) do not offer recommendations about elective termination of pregnancy (ETOP) in patients with autoimmune rheumatic diseases (ARD) because the RHG recommendations require published evidence and because systematic data on this topic are not available. The Society for Maternal-Fetal Medicine recently published expert opinion-based, not data-based, recommendations about pregnancy termination for women at high risk for maternal death; those recommendations do not separately consider patients with ARD (2). To provide data for future reference we offer historical and incident pregnancy termination experiences of two patient groups.

METHODS

BVC database

The database of the Barbara Volcker Center for Women and Rheumatic Disease (BVC, a tertiary care center with a focus on pregnancy and autoimmune disease at the Hospital for Special Surgery), systematically and prospectively recorded clinical data in a proprietary, Microsoft Access-based electronic medical record (EMR) converted to an Excel file. The database is a complete EMR that contains all charted outpatient visits between May, 2002, and January, 2016, and all laboratory data from January 2007 through January 2016 (some earlier laboratory data recorded in physician notes are available). The first visit outpatient charts include numbers of pregnancies, term deliveries, pre-term deliveries, fetal deaths, embryonic deaths, elective abortions, hospitalizations, and ARD flares of all prior pregnancies as stated by the patients, not independently verified. Detailed histories of incident pregnancies and their outcomes are available in the BVC database, supplemented when appropriate by review of hospital inpatient charts. All patients were seen by one author (MDL). Maternal diagnoses are those clinically assigned at the time of first visit or current pregnancy. A patient is counted once for pregnancy history but each time for incident pregnancy when there was more than one. All incident pregnancy patients were enrolled before pregnancy outcome occurred.

PROMISSE database

PROMISSE (Predictors of Pregnancy Outcome: Biomarkers in Antiphospholipid Antibody Syndrome and Systemic Lupus Erythematosus) is a multicenter, prospective observational study of 445 women with American College of Rheumatology criteria-defined systemic lupus erythematosus (SLE), antiphospholipid antibody/antiphospholipid syndrome (aPL/APS), or SLE and aPL/APS together and singlet pregnancies of 12 or fewer weeks’ duration (3–5). Seriously ill patients and those with multiple gestation pregnancies were excluded. PROMISSE recorded historical ETOP frequency (but not safety) for the patients who had had prior pregnancies. No patient is counted twice. Detailed histories of current pregnancies and their outcomes are available in the PROMISSE database. Maternal diagnoses are those clinically assigned by the treating physicians during the pregnancy. Since BVC contributed patients to PROMISSE, to avoid duplication of data, pregnancies reported from the Hospital for Special Surgery were removed from the PROMISSE database for this analysis. Patients with aPL and APS, often treated similarly during pregnancy, are combined.

This study addresses two questions: frequency of ETOP in historical pregnancies and prospective data regarding ETOP of incident pregnancies.

RESULTS

Historical pregnancies

The BVC database contains 10,947 visits of 2,358 women (two anatomically female who identify as male), of whom 83.4% are white. Diagnosis was an ARD in 83.6%. First visit obstetrical histories were available for 2,156 women; for 202 women first visits antedated the start of database and are not counted. In this database, 1,307 women (60.6%), including 14 of the 15 (6 APS; 6 other AID, 2SLE, and 1 UCTD) who were pregnant at first visit, had had one or more prior pregnancies. Of these, 284 (21.7%) had undergone 1–5 prior ETOPs (Table 1). The time of ETOP relative to onset of ARD is unknown. No patient in the BVC database reported complications, hospitalizations, or disease exacerbations associated with a termination.

Table 1.

Historical pregnancies: Demographic and clinical information regarding prior pregnancies and pregnancy terminations from first visit obstetrical histories. Age is that at first visit in both populations (age was 33.4 ± 6.0 years for the 15 BVC patients who were pregnant at first visit.) The timing of ETOP relative to onset of ARD is unknown. Abbreviations: ETOP, elective termination of pregnancy. BVC, Barbara Volcker Center; PRO, PROMISSE study.

Diagnosis	Site	PRIOR PREGNANCY			ETOP n (%)	NO PRIOR PREGNANCY n
		n	Age Yr + SD	White n (%)
SLE only	BVC	161	44.6 ± 12.2	120 (74.5)	43 (27.6)	191 A
	PRO	188	31.0 ± 4.9	169 (54.9)	56 (29.8)	120
aPL/APS only	BVC	183	45.5 ± 13.3	167 (90.8)	28 (15.7)	77
	PRO	73	31.4 ± 4.5	72 (86.7)	10 (13.7)	10
SLE+aPL/APS	BVC	57	43.1 ± 12.3	46 (80.7)	12 (21.1)	37
	PRO	36	31.2 ± 4.8	37 (68.5)	9 (25.0)	18
Vasculitis	BVC	55	53.3 ± 14.8	48 (87.2)	14 (25.5)	33
UCTD	BVC	92	45.8 ± 13.4	81 (88.0)	21 (22.8)	65 A
RA	BVC	99	53.4 ± 15.4	76 (76.8)	16 (15.7)	54
Other AIDB	BVC	306	47.8 ± 13.2	267 (87.3)	71 (23.2)	184
Non AID	BVC	354	51.4 ± 13.8	318 (89.8)	79 (22.3)	208

^Aone anatomically female patient identifies as male

^Btotal with and without prior pregnancy includes 84 Sjogren’s, 92 seropositive without diagnosis, 27 scleroderma, and 287 miscellaneous (dermatomyositis, Raynaud’s, palindromic rheumatism, etc.) patients.

In the PROMISSE database, 75 of 297 (25.3%) patients who had had a prior pregnancy had undergone one or more prior ETOPs; safety data of ETOP were not available for this group.

In the two databases, 27.6 of BVC and 29.8% of PROMISSE SLE patients, 13.7 and 15.7% of aPL/APS patients, and 21.1 and 25.0% of patients with both SLE and aPL/APS had had a prior ETOP.

Incident pregnancies

Maternal and/ or fetal health concerns led to a recommendation to terminate 10 of 229 (4.8%) incident pregnancies of 205 women in the BVC database (one of whom was pregnant at first visit and seeking a second opinion) and 5 of 445 (1.1%) pregnancies of 445 women in the PROMISSE database (Table 2). In 8 cases termination was recommended for maternal indications, 4 also having fetal indications (of which 2 led to embryonic or fetal death before the procedure took place). In one patient with severe pulmonary hypertension an attempt to terminate her pregnancy with methotrexate failed but surgical termination then took place without incident.

Table 2.

Incident pregnancies: Terminations advised, with maternal and fetal indications and outcomes.

Indication	Pt	Maternal diagnosis	ETOP week	Maternal status pre ETOP	Fetal status	Maternal status 1 month post delivery	Other
Maternal	BVC1	aPL/APS	7	Prior CVA, ARF, dialysis; SCr 1.9, BP 108/70		Unchanged	Lost to follow-up
	BVC2	SLE	< 8	Pulmonary hypertension stable on bosentan, valsartan, amlodipine, prednisone, hydroxychloroquine		MTX ETOP failed; surgical ETOP successful; stable	Mother recovered, uncomplicated pregnancy 4 years later
	PRO1	SLE	21	Preeclampsia; baseline persistent urine protein 30 mg/dL, BP 160/95; during >300; BP 184/110		No postpartum flare	Urine protein neg, LFTs normal, platelets normal, complement normal 17 wk postpartum
	BVC3	SLE	35	Anasarca, LVEF 25%, urinary protein 1.5 g, SCr 0.7	Baby normal	Declined termination; anasarca resolved with medical management	Mother died of cardiac and renal failure 3 years later
Maternal and fetal	BVC4	SLE	7	Polyarthritis, renal disease	Teratogenic medication	Unchanged	Lost to follow-up
	BVC5	aPL/APS	< 8	Severe mitral regurgitation, renal failure; BP 178/104	Failure to grow, baby died before procedure	Unchanged; BP better controlled	Mother died of cardiac and renal failure 1.5 y later
	BVC6	UCTD	20	Severe active arthritis	Severe IUGR, baby died before procedure	Arthritis improved with TNF inhibitor	Lost to follow-up
	PRO2	aPL/APS	22	HELLP, urinary protein 0.8g	Severe IUGR	No information	Normal LFTs, urine protein, platelets 36 wk postpartum; complement stable
Fetal	PRO3	SLE	14		Trisomy 21	No information	Lost to follow-up
	BVC7	aPL/APS	18		Parasitic twin	Stable	Successful pregnancies 2 and 4 years later
	BVC8	aPL/APS	<20		One fetus hydrocephaly/ anencephaly, remaining twin delivered successfully	Stable with recovery from late HELLP with small CVA	CAPS 6 years later
	BVC9	?UCTD	<20		Reduction, triplet to twin	Stable	Transient fever at 24 wk, resolved
	PRO4	SLE/ aPL/APS	21		Severe IUGR	No postpartum flare	Normal LFTs, urine protein, platelets 49 wk postpartum; complement stable
	BVC10	?aPL/APS	21		Severe IUGR	Stable	aPL/APS not confirmed
	PRO5	SLE	24		Fetal CCHB	No postpartum flare	Normal LFTs, urine protein, platelets 33 wk postpartum; complement stable

Abbreviations: aPL/APS, antiphospholipid antibody/antiphospholipid syndrome; CVA, cerebrovascular accident; ARF, acute renal failure; SCr, serum creatinine; BP, blood pressure; MTX, methotrexate; SLE, systemic lupus erythematosus; LVEF, left ventricular ejection fraction; UCTD, undifferentiated connective tissue disease; HELLP, hypertension, elevated liver enzymes, low platelets syndrome; IUGR, intrauterine growth restriction; CAPS, catastrophic antiphospholipid syndrome; CCHB, complete congenital heart block

One patient, who had severe cardiac failure and renal insufficiency but a viable fetus, declined termination and (in the medical intensive care unit) vaginally delivered a healthy child; she died three years later. Another patient, also with severe cardiac and renal failure, died 1.5 years after her pregnancy ended. In all cases of maternal health indications the patient stabilized or improved after delivery.

Termination was recommended for fetal reasons in 11 pregnancies (7 for fetal reasons only and 4 combined with maternal reasons). Multiple gestation pregnancies accounted for 3 of the 7 women with fetal indications. Reduction of a triplet to twin pregnancy resulted in 2 live births in a woman who had a history of several prior fetal losses. Reduction of a twin pregnancy in which one twin had a severe malformation incompatible with post-delivery life resulted in the successful delivery of the healthy second child. In another, a parasitic twin pregnancy was terminated for both twins.

Two patients—one with the parasitic twin pregnancy and the other with severe pulmonary hypertension that later resolved—had subsequent uncomplicated pregnancies.

DISCUSSION

This is a report of observational but systematically collected data from two databases, one of unselected ARD patients with all levels of disease activity and damage known to a referral clinic, the other a pregnancy outcome study of relatively well patients with SLE and/or aPL antibodies. As expected, because they were currently pregnant, SLE and/or aPL/APS patients in the latter study were younger and healthier but otherwise similar. Also expected (since fear of pregnancy loss is a reason for consultation), historical ETOP rates for prior pregnancies were lower in APS patients. The ETOP rates were similar to the national average of 188–233 per 1000 live births in 2006–2015 (6). Regarding circumstances in which physicians advised termination, serious maternal or fetal illness was present in all cases. Maternal illness did not worsen for any patient; most of those in whom maternal illness was the determining factor improved after termination; none worsened.

Limitations to this study are that data are descriptive and observational; the databases were not designed to study risks of ETOP; we did not verify ETOP histories; nor did we relate ETOP to disease status or duration at the time of the procedure. Nonetheless, no patient reported an associated hospitalization or complication associated with ETOP. That the patients derive from a referral clinic and from a focused study is both a weakness, in that they may be unrepresentative, and a strength, in that both databases are detailed and large and, for incident pregnancies, prospective, thus provide information that is otherwise unavailable. Accepting that observational data are not of the “high quality” required for inclusion in official guidelines, we believe that, because termination decisions are uncommon, circumstance-driven, and dictated by personal, moral, political, religious, and legal concerns, more rigorous data will not be seen soon. Limited as it is, this study shows no adverse signals for either complication or disease flare following ETOP or medically-advised termination in patients with ARD.