ORIGINAL E n d o c r i n e

ARTICLE R e s e a r c h

Study of Anti-Müllerian Hormone and Its Relation to the Subsequent Probability of Pregnancy in 112 Patients With Systemic Lupus Erythematosus, Exposed or Not to Cyclophosphamide Nathalie Morel, Anne Bachelot, Zeina Chakhtoura, Pascale Ghillani-Dalbin, Zahir Amoura, Lionel Galicier, Olivier Aumaitre, Jean-Charles Piette, Jacques Pourrat, Du Boutin, Karim Sacre, Jean-Emmanuel Kahn, Pierre Duhaut, Dominique Farge, Camille Francès, Gaëlle Guettrot-Imbert, Jean-Robert Harle´, Olivier Lambotte, Véronique Le Guern, Damien Sène, Salim Trad, Elisabeth Vidal, Francoise Sarrot-Reynauld, Anne Gompel, Marie-Laure Tanguy, Philippe Touraine, Jean-Marc Lacorte, and Nathalie Costedoat-Chalumeau, on behalf of the PLUS group* Context: Cyclophosphamide is used for renal and major extrarenal involvement in systemic lupus erythematosus (SLE) and is associated with a risk of premature ovarian failure. There are no data available about the relation between anti-Müllerian hormone (AMH) serum levels and the probability of subsequent pregnancy in SLE patients. Objective: We analyzed AMH levels and the probability of pregnancy in SLE women exposed to cyclophosphamide. Design and Setting: We conducted a matched cohort study in referral centers for SLE. Patients: Fifty-six cyclophosphamide-exposed SLE women younger than 40 years of age and 56 control SLE women matched for age within 6 months participated in the study. Main Outcome Measures: AMH was measured in samples from the PLUS study (ClinicalTrials.gov no. NCT00413361). All patients were interviewed in May 2012 regarding their obstetric status. Results: The mean age ⫾ SD of the 112 patients was 31.6 ⫾ 5.8 years. The mean AMH level was low (1.21 ⫾ 1.01 ng/mL) and was significantly lower in patients exposed to cyclophosphamide (P ⫽ .03) and in patients older than 30 years (P ⫽ .02). During a median follow-up (interval between sampling and the interview) period of 4.2 (range, 2.5– 4.8) years, 38 patients sought to become pregnant, and 32 (84.2%) succeeded. In the univariate analysis, the risk of failure was associated with cumulative cyclophosphamide dose (P ⫽ .007) and older age (P ⫽ .02), but not with AMH. Conclusion: We confirmed that AMH levels are low in SLE patients and decrease significantly with age and cyclophosphamide exposure. Nonetheless, the risk of failure to conceive was low and was predicted by cyclophosphamide exposure and age, but not by AMH levels. (J Clin Endocrinol Metab 98: 3785–3792, 2013)

yclophosphamide (CYC) is widely used to treat systemic lupus erythematosus (SLE) with renal and major extrarenal involvement. Its use, however, is associated with a risk of ovarian failure resulting in irreversible amen-

C

orrhea and infertility (1–5). This risk is associated mainly with age (2– 4, 6) and cumulative dose of CYC (1–3, 5–7). Most studies in this area estimate infertility by measuring the percentage of women with sustained amenorrhea, a

ISSN Print 0021-972X ISSN Online 1945-7197 Printed in U.S.A. Copyright © 2013 by The Endocrine Society Received January 27, 2013. Accepted July 1, 2013. First Published Online July 5, 2013

* Author affiliations are shown at the bottom of the next page. Abbreviations: AMH, anti-Müllerian hormone; BMI, body mass index; CYC, cyclophosphamide; SLEDAI, SLE disease activity index; SLE, systemic lupus erythematosus.

doi: 10.1210/jc.2013-1235

J Clin Endocrinol Metab, September 2013, 98(9):3785–3792

jcem.endojournals.org

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 January 2017. at 19:17 For personal use only. No other uses without permission. . All rights reserved.

3785

3786

Morel et al

AMH in SLE and Probability of Pregnancy

definition that does not quantify the full impact of CYC on ovarian function in young women who may not experience cytotoxic damage severe enough to cause a definitive full cessation of ovarian function. Anti-Müllerian hormone (AMH) is produced by the granulosa cells of growing follicles that shelter oocytes through maturity; its level thus reflects ovarian reserve (ie, the number of eggs remaining in the ovaries) (8 –11). In a recent study of 3260 healthy premenopausal females, Kelsey et al (12) generated the first model of AMH concentration from conception through menopause: AMH peaks at age 24.5 years at a mean serum level of 5 ng/mL and then declines until menopause, reflecting the progressive loss of the nongrowing follicle pool. At age 40, the mean serum AMH level is close to 2 ng/mL (12). AMH is increasingly considered a useful biomarker of CYC-induced ovarian damage and has been proposed for guiding treatment choices and family planning decisions, including in SLE patients (13). Five studies have examined AMH serum levels in women with SLE (14 –18). Their mixed results indicate a relation between CYC exposure and AMH and also suggest that lower AMH is associated with SLE, regardless of CYC exposure. To our knowledge, there are no data available about the relation between AMH serum levels and the probability of subsequent pregnancy in patients with autoimmune diseases including SLE. The aims of our study were to assess AMH serum levels in a large cohort of SLE patients, to compare these levels between patients who were and were not exposed to CYC treatment, and to analyze the subsequent probability of pregnancy.

J Clin Endocrinol Metab, September 2013, 98(9):3785–3792

ticenter trial conducted from 2007 through 2010 at 37 French centers (19) (Plaquenil LUpus Systemic; ClinicalTrials.gov no. NCT0041336). Briefly, relevant inclusion criteria were: adult SLE patients according to American College of Rheumatology criteria (20) treated with hydroxychloroquine for at least 6 months and with a stable SLE. The study included 573 patients, all of whom had serum samples taken at inclusion and stored at ⫺80°C. All participants provided written informed consent. The ethics committee at Saint Louis Hospital, Paris, approved the study protocol. This ancillary study includes all the women younger than 40 years at serum sampling who had been exposed to CYC and compares them with CYC-unexposed women from the study, matched for age within 6 months. We were unable to collect accurate information about GnRH-a for all CYC-exposed patients. Extrapolation indicates, however, that this rate was less than 10% in our population. The cumulative CYC dose was the total dose received from the first administration to the last and was calculated from medical records. Because no standard protocol was recommended at that time, the doses varied among patients, as did the mode of administration (9 women received oral CYC), and the number of courses (11 women had more than 1 course of CYC). Accordingly, we used cumulative doses, determined from the medical record. The last CYC dose was given before inclusion in all cases.

AMH levels AMH concentrations were determined by ELISA with an Immunotech kit (Beckman Coulter). AMH concentrations were determined once from the serum banked at inclusion in the PLUS study. All AMH samples were measured in the same laboratory by the same laboratory-developed test methods, and all values can therefore be compared uniformly. The assay range was 0.4 –21 ng/mL. The intra- and interassay coefficients of variation were 5.3 and 8.7%, respectively. The sensitivity of the AMH assay was 0.05 ng/mL. AMH levels had been assessed in the same laboratory for 335 women with normal menstrual cycles, and the normal range for days 3–5 of the cycle was established at 2.2– 6.8 ng/mL.

Patients and Methods Interviews Patients This was an ancillary study of women included in the PLUS Study, a randomized, double-blinded, placebo-controlled, mul-

All patients were contacted by telephone in May 2012 and provided information regarding their desire to become pregnant, the results of attempts to become pregnant, and their pregnancies

University Pierre and Marie Curie (UPMC) (N.M., Z.A., J.-C.P., D.B., D.S., N.C.-C.), Université Paris 6, Paris, France; Assistance Publique-Hôpitaux de Paris (AP-HP) (N.M., Z.A., J.-C.P., D.B., D.S., N.C.-C.), Hôpital Pitié-Salpêtrière, Centre de référence national pour le Lupus Systémique et le syndrome des Antiphospholipides, service de médecine interne, 75651 Paris Cedex 13, France; UPMC (A.B., Z.C., P.T.), Université Paris 6, Paris, France; AP-HP (A.B., Z.C., P.T.), Hôpital Pitié-Salpêtrière, service d’endocrinologie et médecine de la reproduction, Centre de Référence des Maladies Gynécologiques Médico-Chirurgicales Rares and Centre de Référence des Maladies Endocriniennes Rares de la Croissance, 75651 Paris Cedex 13, France; UPMC (P.G.-D.), Université Paris 6, Paris, France; AP-HP (P.G.-D.), Hôpital Pitié-Salpêtrière, service d’immunologie, 75651 Paris Cedex 13, France; Université Paris Diderot (L.G.), Sorbonne Paris Cité, Paris, France; AP-HP (L.G.), Hôpital St Louis, service d’immunologie clinique, 75010 Paris, France; Université de Clermont-Ferrand (O.A., G.G.-I.), 63003 Clermont-Ferrand, France; Centre Hospitalier Universitaire (CHU) Clermont-Ferrand (O.A., G.G.-I.), Hôpital Gabriel Montpied, service de médecine interne, 63003 Clermont-Ferrand Cedex 1, France; Université Paul-Sabatier (J.P.), Toulouse, France; CHU Toulouse (J.P.), Hôpital Rangueil, service de néphrologie, TSA 50032, 31000 Toulouse, France; Université Paris-Diderot (K.S.), Sorbonne Paris-Cité, F-75205 Paris, France; AP-HP (K.S.), Hôpital Bichat Claude-Bernard, service de médecine interne, 75018 Paris, France; Hôpital Foch (J.-E.K.), service de médecine interne, 92150 Suresnes, France; CHU Amiens (P.D.), Hôpital Nord, service de médecine interne, 80000 Amiens, France; Université Paris Diderot (D.F.), Sorbonne Paris Cité, Paris, France; AP-HP (D.F.), Hôpital St Louis, service de médecine interne, 75010 Paris, France; UPMC (C.F.), Université Paris 6, Paris, France; AP-HP (C.F.), Hôpital Tenon, service de dermatologie allergologie, 75020 Paris, France; Hôpital de la Conception (J.-R.H.), Assistance Publique-Hôpitaux de Marseille, Service de médecine interne, 13385 Marseille cedex 5, France; Hôpital Bicêtre (O.L.), service de médecine interne, 94270 Le Kremlin Bicêtre, France; Université Paris-Descartes (V.L.G.), Paris, France; AP-HP (V.L.G.), Hôpital Cochin, service de médecine interne, 75014 Paris, France; AP-HP (S.T.), Hôpital Ambroise Paré, service de médecine interne, 92104 Boulogne-Billancourt cedex, France; Centre Hospitalier Régional Universitaire Dupuytren (E.V.), service de médecine interne, 92104 Boulogne-Billancourt cedex, France; Site Nord Hôpital A. Michallon (F.S.-R.), 38700 La Tronche, France; Université Paris-Descartes (A.G.), Paris, France; AP-PH (A.G.), Hôpital Cochin, service de gynécologie, 75014 Paris, France; UPMC (M.L.T.), Université Paris 6, Paris, France; AP-HP (M.L.T.), Hôpital Pitié-Salpêtrière, unité de recherche clinique, 75651 Paris Cedex 13, France; UPMC (J.-M.L.), Université Paris 6, Paris, France; and AP-HP (J.-M.L.), Hôpital Pitié-Salpêtrière, service de biochimie, 75651 Paris Cedex 13, France

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 January 2017. at 19:17 For personal use only. No other uses without permission. . All rights reserved.

doi: 10.1210/jc.2013-1235

jcem.endojournals.org

3787

significance was set at 0.05. All analyses were performed with SAS software version 9.2 (SAS Institute).

Results Study population The study population included 112 patients (56 exposed to CYC and 56 unexposed). Table 1 summarizes patient characteristics. Mean age was 31.6 ⫾ 5.8 years, and median disease duration was 9.2 years (range, 0.6 –27.5). Antiphospholipid syndrome was present in 17% of the patients. The mean estimated creatFigure 1. Design of the study. *, Pregnancies during the interval between the date of the first exposure to CYC for exposed patients or the matched date for each unexposed control and the inine clearance calculated by the phone interview (median follow-up ⫽ 12.4 y). **, Pregnancies during the interval between the Cockroft-Gault equation was 113 ⫾ date of the AMH samples (inclusion) and the phone interview (median follow-up ⫽ 4.2 y). 35 mL/min. The median SLE disease activity index (SELENA-SLEDAI) (in their whole life, after CYC exposure, and after the sampling score was 2 (range, 0 –11). All women were being treated date) (Figure 1). The same practitioner conducted all interviews, which were standardized, were conducted blinded to CYC exwith hydroxychloroquine at inclusion (ie, at sampling). posure status, and lasted as long as necessary. Other current SLE medications included corticosteroids (77.7% with a mean dose of 8.7 ⫾ 4.3 mg/d) and immuStatistical analysis nosuppressants (28.6%). Predictive factors of AMH serum levels ⱕ1 ng/mL were idenThe 2 groups did not differ significantly for SLE activity tified by univariate analysis with the ␹2 test for qualitative vari(SELENA-SLEDAI score), renal insufficiency, compleables or the Wilcoxon test for quantitative variables. Variables ment level, proportion of patients positive for anti-DNA with a univariate P value ⬍.2 were included in multivariate stepwise logistic regression. Predictive factors of the risk of failure to antibodies, body mass index (BMI), or antiphospholipid become pregnant, for women who wanted to, were identified in syndrome. Women exposed to CYC had all received imthe univariate analysis with Fisher’s exact test for qualitative munosuppressive drugs, compared with 19.6% (n ⫽ 11) variables and the Wilcoxon rank sum test for quantitative variof the unexposed women (P ⬍ .0001). This group also had ables. No multivariate analysis was performed for this question a longer median disease duration (10.5 vs 6.6 y; P ⫽ .008). given the low number of women who failed to achieve planned pregnancy (n ⫽ 6 of 38). All tests were 2-sided. The level of During the PLUS study, a higher proportion of the exTable 1.

Characteristics of the Study Patients

Characteristics

Total

Nonexposed

Exposed

P

n Age, mean ⫾ SD, y Disease duration, median [range], y Associated APS, n (%) Renal disorder, n (%) Ever use of prednisone, n (%) Ever use of immunosuppressive drugs, n (%) Current prednisone use, n (%) Prednisone dose, mean ⫾ SD, mg/d Current use of immunosuppressive drugs, n (%) Creatinine clearance, mean ⫾ SD, mL/min Active disease (SELENA-SLEDAI ⬎ 4), n (%) SELENA-SLEDAI score, median [range] Low levels of complement, n (%) Increased anti-double stranded DNA, n (%) BMI, mean ⫾ SD, kg/m2 Active smokers, n (%) AMH levels, mean ⫾ SD, ng/mL

112 31.6 ⫾ 5.8 9.2 [0.6 –27.5] 19 (17) 57 (50.9) 103 (92) 67 (59.8) 87 (77.7) 8.7 ⫾ 4.3 32 (28.6) 113 ⫾ 35 18 (16.1) 2 [0 –11] 9 (8.1) 47 (42) 23.9 ⫾ 5.7 25 (22.3) 1.21 ⫾ 1.01

56 31.6 ⫾ 5.8 6.6 [0.6 –27.5] 7 (12.5) 11 (19.6) 47 (83.9) 11 (19.6) 39 (69.6) 8.6 ⫾ 4.3 9 (16.1) 111 ⫾ 29 9 (16.1) 2 [0 –11] 4 (7.1) 22 (39.3) 22.9 ⫾ 5.2 17 (30.4) 1.41 ⫾ 1.01

56 31.6 ⫾ 5.8 10.5 [0.8 –20.5] 12 (21.5) 46 (82.1) 56 (100) 56 (100) 48 (85.7) 8.7 ⫾ 4.3 23 (41.1) 116 ⫾ 40 9 (16.1) 2 [0 –10] 5 (8.9) 25 (44.6) 24.8 ⫾ 6.1 8 (14.3) 1.02 ⫾ 0.97

.99 .008 .21 ⬍.0001 .002 ⬍.0001 .04 .84 .003 .85 1.00 .55 .73 .57 .08 .04 .03

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 January 2017. at 19:17 For personal use only. No other uses without permission. . All rights reserved.

3788

Morel et al

Table 2.

AMH in SLE and Probability of Pregnancy

J Clin Endocrinol Metab, September 2013, 98(9):3785–3792

Characteristics of Women Who Failed to Become Pregnant

Patient No.

Age, y

CYC Exposure

Age at CYC, y

AMH Level, ng/mL

CYC Cumulative Dose, g

1 2 3 4 5 6 7

29.4 38.9 39.9 37 29.5 38.7 37.5

N Y Y Y Y Y Y

NA 27 27 21 26 23 25

1.58 0.23 0.11 1.42 0.45 0.04 0

0 55 (PO) 18 (iv) 46 (PO) 6 (iv) 16 (iv) 42 (iv and PO)

No. of Pregnancies Before CYC

No. of Pregnancies Between CYC and Inclusion

0 2 0 1 0 1 0a

0 0 1 0 0 1 0

No. of Pregnancies After Inclusion 0 0 0 0 0 0 b

Etiology of Infertility Endometriosis Premature ovarian insufficiency Premature ovarian insufficiency Premature ovarian insufficiency Male infertility Tubal factor Premature ovarian insufficiency

Abbreviations: Y, yes; N, no; PO, per os; NA, not applicable. a

Two pregnancies obtained with egg donation were excluded.

b

Case 7 was diagnosed with infertility between CYC and inclusion and did not wish or try to become pregnant afterward.

posed women were undergoing immunosuppressive (41.1 vs 16.1%; P ⫽ .003) and steroid (85.7 vs 69.6%; P ⫽ .04) treatment. A higher proportion also had had renal involvement (82.1 vs 19.6%; P ⬍ .0001) and steroid treatment (100 vs 83.9%; P ⫽ .002), but fewer were active smokers (14.3 vs 30.4%; P ⫽ .04). Of these 112 patients, 107 women answered the questions regarding their desire of pregnancy with a precision of 1 month, and only 5 were less precise. Eighty patients had been pregnant at least once in their lives and had had a total of 175 pregnancies with 100 live births at their inclusion in the PLUS study. AMH levels The mean AMH serum level was 1.21 ⫾ 1.01 ng/mL. AMH levels were ⱕ2 ng/mL for 82% (n ⫽ 92), ⱕ1 ng/mL for 50% (n ⫽ 56), and ⱕ0.5 ng/mL for 34% (n ⫽ 38). This level was significantly lower in patients exposed to CYC than in unexposed patients (1.02 ⫾ 0.97 vs 1.41 ⫾ 1.01 ng/mL; P ⫽ .03) and in patients older than 30 years (1.02 ⫾ 0.97 vs 1.43 ⫾ 1.02 ng/mL; P ⫽ .02). In the univariate analysis, higher SLE activity (SLEDAI ⱖ 4), CYC exposure, longer disease duration, higher BMI, and higher age were associated with an AMH level ⱕ 1 ng/mL with a P value ⬍.2 and were subsequently included in a multivariate stepwise logistic regression model. Of these variables, only age (odds ratio ⫽ 1.1, 95% CI ⫽ [1.04 –1.19]; P ⫽ .004) and CYC exposure (odds ratio ⫽ 2.6, 95% CI ⫽ [1.16 –5.71]; P ⫽ .02) were associated with an AMH level ⱕ1 ng/mL. AMH levels and the probability of pregnancy We then analyzed the relation between AMH levels and the probability of pregnancy. The median follow-up (the interval between sampling and the telephone interview) was 4.2 (range, 2.5– 4.8) years. During this period, 37 patients became pregnant at least once: unplanned pregnancies in 5 women, and planned pregnancies in 32.

To determine the success rate, we excluded the unplanned pregnancies; no percentage could be calculated because the denominator was unknown. We then focused on the probability of planned pregnancies. In our interviews, we found that 38 of the 112 patients had sought to become pregnant at least once after inclusion and 32 (84.2%) succeeded. The mean time to pregnancy was 7 ⫾ 10.5 months (ie, from the date of the first attempt to become pregnant after stopping contraception until pregnancy). This time exceeded 12 months for only 4 women (12.5%), including 1 whose infertility was explained by male infertility. The AMH levels of these 4 patients were 0.78, 1.84, 2.44, and 3.83 ng/mL. Of the 6 women who tried and failed to become pregnant after inclusion, 1 was in the unexposed group, and 5 were in the exposed group (their characteristics are listed in Table 2). Their mean age at inclusion was 35.5 years. Infertility was attributed to CYC exposure in 3 cases and to other reasons in 3 others. More specifically, the unexposed patient apparently failed to become pregnant because of endometriosis. Among the 5 exposed patients, 1 was 29.5 years old at inclusion, had received only 6 g of CYC, and failed to become pregnant after 24 months, presumably because of male infertility. Three had premature ovarian insufficiency after CYC. They had received high doses of CYC (18, 46, and 55 g) and were ages 37, 38.9, and 39.9 years at inclusion (ie, before the study period during which they attempted to become pregnant). The last patient was 38.7 years old at inclusion and failed to become pregnant after 30 months, presumably because of a tubal factor. In the univariate analysis, the risk of failure was associated with cumulative CYC dose (17 [range, 0 –55] g in the failure group vs 0 [range, 0 –22] g in those who succeeded; P ⫽ .007) and with older age (38 [range, 29 – 40] y at sampling vs 29 [range, 23–39] y, respectively; P ⫽ .02) (Table 3). They did not differ for antiphospholipid syndrome, SLE activity (SLEDAI ⬎ 6), or high occupational

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 January 2017. at 19:17 For personal use only. No other uses without permission. . All rights reserved.

doi: 10.1210/jc.2013-1235

jcem.endojournals.org

Table 3. Univariate Analyses of Probability of Pregnancies After Inclusion in the 38 Patients Who Wished to Become Pregnant Characteristics

Failure

Success

P

n CYC exposure, n (%) CYC cumulative dose, median [range], g Age at inclusion, median [range], y AMH level, median [range], ng/mL

6 5 (83.3) 17 [0 –55]

32 12 (37.5) 0 [0 –22]

.07 .007

38 [29 – 40]

29 [23–39]

.02

0.34 [0.04 –1.58] 1.27 [0 – 4.14] .055

status. The risk of failure to achieve planned pregnancy with lower serum AMH level did not meet statistical significance (P ⫽ .055). Interestingly, pregnancies occurred in 6 of 10 women with very low AMH serum levels (ⱕ0.5 ng/mL) and in 11 of 15 with levels ⱕ1 ng/mL. CYC exposure and the probability of pregnancy We then analyzed the link between CYC exposure and the probability of pregnancy. The study period was defined as the interval between the date of the first exposure to CYC for exposed patients or the matched date for each unexposed control and the telephone interview. The median follow-up was 12.4 years (range, 3.3–22.9). During this follow-up, 61 patients became pregnant at least once. Another patient (case 7) had 2 pregnancies obtained with egg donation, which we treated as failure

3789

rather than pregnancies. Of these 61 pregnancies, 9 were unplanned pregnancies and 52 planned. We then focused on the probability of planned pregnancies. Overall, 56 of the 112 women (50%) sought to become pregnant at least once after starting CYC treatment, and 51 succeeded (91%). The median time to pregnancy was 3 months (range, 1– 48) (Figure 2). The number of failures (n ⫽ 5; cases 1, 2, 4, 5, and 7) differs from the number after inclusion (n ⫽ 6) because 2 women (cases 3 and 6) had pregnancies between CYC and inclusion but not after, and 1 patient (case 7) was diagnosed with infertility between CYC and inclusion and did not wish or try to become pregnant afterward (Table 2). In the univariate analysis, risk of failure was associated with cumulative CYC dose (P ⫽ .01; Table 4).

Discussion This study is, to our knowledge, the first to examine AMH levels and the probability of subsequent pregnancy in SLE patients. First, our results confirm prior reports that AMH levels are unexpectedly low in female SLE patients, even those not exposed to CYC (16). The mean AMH level was 1.21 ⫾ 1.01 ng/mL in our 112 patients, with a mean age of 31.6 ⫾ 5.8 years; it was 1.41 ⫾ 1.01 ng/mL in the 56 unexposed SLE patients. These levels correspond to the mean levels found in women 39 and 38 years old, respectively, in a large cohort of 17 120 women attending a fertility clinic (21) and were lower than the mean found in

Figure 2. Time to pregnancies for women who wished to become pregnant according to CYC exposure. Dotted line, women who were not exposed to CYC; solid line, women who were exposed to CYC.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 January 2017. at 19:17 For personal use only. No other uses without permission. . All rights reserved.

3790

Morel et al

AMH in SLE and Probability of Pregnancy

J Clin Endocrinol Metab, September 2013, 98(9):3785–3792

Table 4. Univariate Analyses of Probability of Pregnancies After CYC (or the Matched Date for Patients Unexposed to CYC) in the 56 Patients Who Wished to Become Pregnant Characteristics

Failure

Success

P

n Exposure to CYC, n (%) CYC cumulative dose, median [range], g Age at CYC or matched date, median [range], y Age at desire to become pregnant, median [range], y

5 4 (80) 42 [0 –55]

51 21 (40.4) 0 [0 –23]

.15 .01

25 [19 –26] 24 [13–34]

.89

31 [29 –36] 29 [18 –39] .07

a healthy population of 42-year-old women (12). In 416 healthy women with a median age of 34 (interquartile range, 27– 40 y; range, 18 –51 y), the median serum AMH level was 3.45 ng/mL (interquartile range, 1.64 –5.64 ng/ mL; range, 0 –19.34) (22). The multivariate analysis showed that only age and previous CYC exposure predicted low AMH levels. Specifically, low AMH levels were not associated with SLE activity (measured by the SELENA-SLEDAI), anti-DNA antibodies, mean BMI, use of oral contraceptives, or progestin-only contraceptives. The low AMH levels might be explained by a true reduction in ovarian reserve. It has been hypothesized that such a reduction could be due to anti-corpus luteum antibodies and anti-endometrial antibodies, which are significantly correlated with increased FSH levels (23, 24). Another explanation, discussed below, could be that AMH levels are diminished for other reasons in SLE patients and that these levels do not accurately reflect ovarian reserve in this specific population. AMH levels were significantly lower in patients exposed to CYC, an agent that has been implicated in dosedependent damage to oocytes and granulosa cells. Although Aikawa et al (14) found no significant difference in AMH levels between 11 SLE patients exposed to CYC and 16 unexposed patients, Marder et al (17) found results similar to ours with significantly lower mean AMH levels in 11 premenopausal SLE patients who received CYC alone than in 27 unexposed SLE patients (0.18 ng/ml ⫾ 0.20 vs 1.33 ng/ml ⫾ 1.59; P ⫽ .015) or in 10 CYCexposed SLE patients who also received GnRH-a (0.86 ng/ml ⫾ 1.06; P ⫽ .018). Similarly, Mok et al (18) found significantly lower levels of AMH in past CYC users than in nonusers; after adjustment for age, AMH level was inversely correlated with age and the cumulative dose of previous CYC exposure. Similar results have been found in other populations exposed to CYC (3). The determination of AMH level has many advantages. Blood can be stored for later analysis, there is minimal dependence upon operators, and levels do not vary sig-

nificantly throughout the human menstrual cycle or after the administration of oral contraceptives (25, 26). We know that women exposed to CYC have a significant risk of ovarian failure, essentially related to age at CYC onset, to duration of treatment, and to cumulative dose (1– 4). Because AMH levels are lower in women exposed to CYC, AMH has— quite logically— been proposed as a biomarker of CYC-induced ovarian damage, the levels of which should guide treatment choices and family-planning decisions, including in SLE patients (13). To evaluate the appropriateness of this proposition, we conducted what appears to be the first study of AMH as a prognostic marker for pregnancy. We showed that even in our population characterized by low AMH levels, the risk of infertility was quite low because 84.2% of the 38 patients who wanted a child became pregnant at least once during a mean follow-up period of 3.9 ⫾ 0.6 years. Time to pregnancy was usually relatively short. Success was observed even in the women with the lowest AMH levels. In the univariate analysis, only cumulative CYC dose and mean age were associated with failure to become pregnant. This result suggests that low AMH levels may not predict poor prognosis for pregnancy in women with SLE and that physicians should be cautious before using AMH levels for counseling their patients about family-planning decisions. Women with SLE and very low AMH levels may still become pregnant easily. Similar discrepancies between AMH levels and “true” ovarian reserve have been reported for obese women, who also have reduced levels of AMH (27) that suggest obesity may be associated with impaired ovarian reserve. A study assessing AMH levels and ultrasound-derived markers of ovarian reserve in obese women found, however, that their antral follicle count was similar to that in age-matched women of normal weight, despite lower serum AMH levels (28). It has also been shown that AMH levels do not predict spontaneous pregnancy in some populations with poor ovarian reserve (29, 30). We are planning a similar prospective study comparing AMH levels and antral follicle counts in women with SLE to further confirm that their low AMH levels do not accurately reflect altered ovarian reserve. Finally, our results raise the question of why AMH levels are low in women with SLE. One hypothesis is that some antibodies interfere with the laboratory test. This is not supported by the lack of correlation between antiDNA antibodies and AMH levels, but other antibodies might be involved. Other physiological or SLE-related reasons may be implicated, and further research is needed. We also analyzed the probability of pregnancy after the onset of CYC exposure (or the matched date for patients not exposed to CYC). During a median follow-up period

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 January 2017. at 19:17 For personal use only. No other uses without permission. . All rights reserved.

doi: 10.1210/jc.2013-1235

of 12.4 years, 91% of the 56 women who wanted to become pregnant did so at least once. Among the 5 patients who failed to become pregnant, infertility was attributed to CYC exposure in 3 cases and to other reasons in 2 cases (endometriosis and male infertility). The main limitation of our study is that we did not prospectively determine our patients’ desires regarding pregnancy. It is therefore possible that some women did not recall this perfectly, although their precision was excellent. A prospective study design with a predefined protocol for interviews and monitoring of desire to become pregnant and pregnancy status is strongly recommended. Another limitation is that we did not analyze unplanned pregnancies. The number of pregnancies would have been higher had we included them, but it was not possible to calculate a percentage because the denominator was unknown. This calculation of the pregnancy rate in SLE patients, for the first time, was another strength of our study. In conclusion, we confirmed that AMH levels are low in many SLE patients and that this level decreases significantly with age and with CYC exposure. Nonetheless, we also showed that the likelihood of becoming pregnant is high: 84.2% of the patients who wanted to become pregnant did so. Success was observed even in patients with the lowest AMH levels. Failure was associated with CYC exposure but not with AMH levels. We believe that these results are reassuring for SLE patients. They should be helpful for physicians counseling women with SLE.

Acknowledgments The authors are grateful to the patients; to the Association France Lupus for a grant for this study; to the Clinical Research Unit of Pitié-Salpêtrière Hospital, which dealt with the methodological aspects, data management, and monitoring; to the sponsor of the study, Assistance Publique-Hôpitaux de Paris; to the team of the “laboratoire de biochimie” of Pitié-Salpêtrière Hospital for technical help; and to Sanofi, which provided the hydroxychloroquine and placebo tablets. Address all correspondence and requests for reprints to: Prof Nathalie Costedoat-Chalumeau, Faculté Paris 6, Hôpital Pitié-Salpêtrière, service de médecine interne, 47-83 Boulevard de l’Hôpital, 75013 Paris, France. E-mail: nathalie. [email protected]. This study was funded by a grant from the French Programme Hospitalier de Recherche Clinique 2005 Ministère de la santé. The “Direction de la Recherche Clinique et du Développement” provided logistical and administrative support. Sanofi had no role in the initiation, planning, conduct, data assembly, analysis, or interpretation of the study. N.M. and N.C.-C. wrote the initial draft of the manuscript. All the authors contributed to subsequent drafts and had unre-

jcem.endojournals.org

3791

stricted access to the data during this process. All the authors made the decision to submit the manuscript for publication and assume responsibility for the accuracy and completeness of the data. All the authors attest that the study was performed in accordance with the protocol and the statistical analysis plan.

Collaborator Group Félix Ackermann, Bouchra Asli, Leonardo Astudillo, Cristina Belizna, Nadia Belmatoug, Olivier Benveniste, Audrey Benyamine, Holly Bezanahary, Patrick Blanco, Olivier Bletry, Bahram Bodaghi, Pierre Bourgeois, Benoît Brihaye, Patrice Cacoub, Emmanuel Chatelus, Judith Cohen-Bittan, Richard Damade, Eric Daugas, Christian De-Gennes, Jean-François Delfraissy, Céline Delluc, Aurélien Delluc, Hélène Desmurs-Clavel, Alain Dupuy, Isabelle Durieu, Hang-Korng Ea, Olivier Fain, Christian Funck-Brentano, Frédérique Gandjbakhch, Justine Gellen-Dautremer, Bertrand Godeau, Cécile Goujard, Catherine Grandpeix, Claire Grange, Loïc Guillevin, Eric Hachulla, Julien Haroche, Pierre Hausfater, Jean Sébastien Hulot, Moëz Jallouli, Jean Jouquan, Gilles Kaplanski, Homa Keshtmand, Mehdi Khellaf, David Launay, Philippe Lechat, Gaëlle Leroux, Hervé Levesque, Olivier Lidove, Nicolas Limal, Frédéric Lioté, Eric Liozon, Kim Ly, Matthieu Mahevas, Kubéraka Mariampillai, Xavier Mariette, Alexis Mathian, Karin Mazodier, Marc Michel, Luc Mouthon, Lucile Musset, Rokiya Ngack, Jacques Ninet, Eric Oksenhendler, Thomas Papo, Jean-Luc Pellegrin, Laurent Perard, Olivier Peyr, Anne-Marie Piette, Vincent Poindron, Fabienne Roux, David Saadoun, Sabrinel Sahali, Laurent Sailler, Bernadette Saint-Marcoux, Yoland Schoindre, Jérémie Sellam, Jacques Serratrice, Pascal Seve, Jean Sibilia, Claude Simon, Amar Smail, Christelle Sordet, Jérôme Stirnemann, Benjamin Terrier, Jean-François Viallard, Bertrand Wechsler, Pierre-Jean Weiller, and Noël Zahr. Disclosure Summary: The authors have nothing to disclose.

References 1. Appenzeller S, Blatyta PF, Costallat LT. Ovarian failure in SLE patients using pulse cyclophosphamide: comparison of different regimes. Rheumatol Int. 2008;28:567–571. 2. Boumpas DT, Austin HA 3rd, Vaughan EM, Yarboro CH, Klippel JH, Balow JE. Risk for sustained amenorrhea in patients with systemic lupus erythematosus receiving intermittent pulse cyclophosphamide therapy. Ann Intern Med. 1993;119:366 –369. 3. Clowse ME, Copland SC, Hsieh TC, et al. Ovarian reserve diminished by oral cyclophosphamide therapy for granulomatosis with polyangiitis (Wegener’s). Arthritis Care Res (Hoboken). 2011;63: 1777–1781. 4. Huong DL, Amoura Z, Duhaut P, et al. Risk of ovarian failure and fertility after intravenous cyclophosphamide. A study in 84 patients. J Rheumatol. 2002;29:2571–2576. 5. Katsifis GE, Tzioufas AG. Ovarian failure in systemic lupus erythematosus patients treated with pulsed intravenous cyclophosphamide. Lupus. 2004;13:673– 678. 6. Park MC, Park YB, Jung SY, Chung IH, Choi KH, Lee SK. Risk of ovarian failure and pregnancy outcome in patients with lupus ne-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 January 2017. at 19:17 For personal use only. No other uses without permission. . All rights reserved.

3792

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

Morel et al

AMH in SLE and Probability of Pregnancy

phritis treated with intravenous cyclophosphamide pulse therapy. Lupus. 2004;13:569 –574. Medeiros MM, Silveira VA, Menezes AP, Carvalho RC. Risk factors for ovarian failure in patients with systemic lupus erythematosus. Braz J Med Biol Res. 2001;34:1561–1568. Baird DD, Steiner AZ. Anti-Mullerian hormone: a potential new tool in epidemiologic studies of female fecundability. Am J Epidemiol. 2012;175:245–249. Hansen KR, Hodnett GM, Knowlton N, Craig LB. Correlation of ovarian reserve tests with histologically determined primordial follicle number. Fertil Steril. 2011;95:170 –175. Meduri G, Massin N, Guibourdenche J, et al. Serum anti-Mullerian hormone expression in women with premature ovarian failure. Hum Reprod. 2007;22:117–123. Visser JA, Schipper I, Laven JS, Themmen AP. Anti-Mullerian hormone: an ovarian reserve marker in primary ovarian insufficiency. Nat Rev Endocrinol. 2012;8:331–341. Kelsey TW, Wright P, Nelson SM, Anderson RA, Wallace WH. A validated model of serum anti-mullerian hormone from conception to menopause. PLoS One. 2011;6:e22024. Clowse ME, Harward L, Criscione-Schreiber L, Pisetsky D, Copland S. Anti-Mullerian hormone: a better marker of ovarian damage from cyclophosphamide. Arthritis Rheum. 2012;64:1305–1310. Aikawa NE, Sallum AM, Pereira RM, et al. Subclinical impairment of ovarian reserve in juvenile systemic lupus erythematosus after cyclophosphamide therapy. Clin Exp Rheumatol. 2012;30:445– 449. Browne H, Armstrong A, Decherney A, et al. Assessment of ovarian function with anti-Mullerian hormone in systemic lupus erythematosus patients undergoing hematopoietic stem cell transplant. Fertil Steril. 2009;91(suppl 4):1529 –1532. Lawrenz B, Henes J, Henes M, et al. Impact of systemic lupus erythematosus on ovarian reserve in premenopausal women: evaluation by using anti-Muellerian hormone. Lupus. 2011;20:1193– 1197. Marder W, McCune WJ, Wang L, et al. Adjunctive GnRH-a treatment attenuates depletion of ovarian reserve associated with cyclophosphamide therapy in premenopausal SLE patients. Gynecol Endocrinol. 2012;28:624 – 627. Mok CC, Chan PT, To CH. Anti-mullerian hormone and ovarian reserve in systemic lupus erythematosus. Arthritis Rheum. 2013;65: 206 –210.

J Clin Endocrinol Metab, September 2013, 98(9):3785–3792

19. Costedoat-Chalumeau N, Galicier L, Aumaitre O, et al. Hydroxychloroquine in systemic lupus erythematosus: results of a French multicentre controlled trial (PLUS Study) [published online ahead of print January 7, 2013]. Ann Rheum Dis. doi:10.1136/annrheumdis2012-202322. 20. Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum. 1997;40:1725. 21. Seifer DB, Baker VL, Leader B. Age-specific serum anti-Mullerian hormone values for 17,120 women presenting to fertility centers within the United States. Fertil Steril. 2011;95:747–750. 22. La Marca A, Spada E, Grisendi V, et al. Normal serum anti-Mullerian hormone levels in the general female population and the relationship with reproductive history. Eur J Obstet Gynecol Reprod Biol. 2012;163:180 –184. 23. Moncayo R, Moncayo HE. A new endocrinological and immunological syndrome in SLE: elevation of human chorionic gonadotropin and of antibodies directed against ovary and endometrium antigens. Lupus. 1995;4:39 – 45. 24. Pasoto SG, Viana VS, Mendonca BB, Yoshinari NH, Bonfa E. Anticorpus luteum antibody: a novel serological marker for ovarian dysfunction in systemic lupus erythematosus? J Rheumatol. 1999;26: 1087–1093. 25. Streuli I, Fraisse T, Pillet C, Ibecheole V, Bischof P, de Ziegler D. Serum antimullerian hormone levels remain stable throughout the menstrual cycle and after oral or vaginal administration of synthetic sex steroids. Fertil Steril. 2008;90:395– 400. 26. La Marca A, Stabile G, Artenisio AC, Volpe A. Serum anti-Mullerian hormone throughout the human menstrual cycle. Hum Reprod. 2006;21:3103–3107. 27. Freeman EW, Gracia CR, Sammel MD, Lin H, Lim LC, Strauss JF 3rd. Association of anti-mullerian hormone levels with obesity in late reproductive-age women. Fertil Steril. 2007;87:101–106. 28. Su HI, Sammel MD, Freeman EW, Lin H, DeBlasis T, Gracia CR. Body size affects measures of ovarian reserve in late reproductive age women. Menopause. 2008;15:857– 861. 29. Janse F, Donnez J, Anckaert E, de Jong FH, Fauser BC, Dolmans MM. Limited value of ovarian function markers following orthotopic transplantation of ovarian tissue after gonadotoxic treatment. J Clin Endocrinol Metab. 2011;96:1136 –1144. 30. Bidet M, Bachelot A, Bissauge E, et al. Resumption of ovarian function and pregnancies in 358 patients with premature ovarian failure. J Clin Endocrinol Metab. 2011;96:3864 –3872.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 14 January 2017. at 19:17 For personal use only. No other uses without permission. . All rights reserved.