Ker077 1.6

Rheumatology Advance Access published April 29, 2011
Pregnancy and fetal outcome in women withprimary Sjo¨gren’s syndrome compared withwomen in the general population: a nestedcase–control study Sarwen Z. Hussein1, Lennart T. H. Jacobsson1, Pelle G. Lindquist2 andElke Theander1 Objective. To study pregnancy and fetal outcome in women with primary SS (pSS) compared with womenin the general population.
Methods. In a nested case–control setting, variables related to pregnancy and fetal outcomes werecompared. Cases (n = 16) were identified through registry linkage (Malmo¨ pSS registry and a database entailing information of all pregnancies and deliveries in Malmo¨ from 1990 through 2006). For each preg-nancy with pSS, the following five deliveries from the same registry were chosen as controls (n = 80).
Results. All cases fulfilled the American European Consensus Criteria for pSS and were positive for ANAand anti-SSA antibodies. Date of diagnosis was before pregnancy in 10 women and after delivery in 6.
Mean age at delivery was significantly higher in women with pSS (33.6 years) vs controls (29.8 years).
While pregnancy duration did not differ, mean birthweight was significantly lower in offspring of pSSmothers (3010 g) vs babies of control mothers (3458 g). Normal partus in contrast to vacuum extractionor Caesarean section was significantly more frequent in healthy women than in pSS women (83 vs 56%).
Other pregnancy outcomes such as parity, miscarriages and Apgar score did not differ. There wereno significant differences between women with a pSS diagnosis before or after the index pregnancy.
Only one of the included pregnancies was complicated by intrauterine AV block.
Conclusion. Pregnancy occurs later in life in pSS women. Mothers with pSS give birth to offspring withlower birthweight and less commonly have normal partus.
Key words: Primary Sjo¨gren’s syndrome, Pregnancy, Fetal outcome, Nested case–control study.
and eyes, gastrointestinal and urogenital discomfort) [3],reduced health-related quality of life [4], disabling fatigue Primary SS (pSS) is the second most common autoim- [5] and increased risk of and death due to non-Hodgkin’s mune disease after RA, with a prevalence of $0.3–0.6% lymphoma [6, 7] make pSS a bothersome condition for [1], when defined strictly according to the American both patients and physicians. Secondary SS, in contrast, European Consensus Criteria (AECC) [2] and a female : co-exists with other autoimmune diseases such as RA male ratio of 9 : 1. Uncomfortable symptoms (dry mouth Although the incidence estimates are uncertain, it is well known that the disease often starts in the fourth or fifth Department of Rheumatology, Lund University, Ska˚ne University Hospital, Malmo¨ and 2Department of Obstetrics and Gynecology, decade of life; thus, the majority of pSS patients are Karolinska University Hospital, Huddinge, Sweden.
post-menopausal. In a cross-sectional analysis of our Submitted 7 September 2010; revised version accepted female pSS patients, only 13% were 445 years. At diag- nosis, only 29% were 445 years and 13% were 435 years Correspondence to: Elke Theander, Department of Rheumatology, of age (own unpublished registry data). However, pSS Lund University, Ska˚ne University Hospital, 20502 Malmo¨, Sweden.
E-mail: [email protected] ! The Author 2011. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: [email protected] Pregnancy complications due to the occurrence of anti-Ro/ delivery, numbers of parity and miscarriages, pregnancy SSA and anti-La/SSB autoantibodies in the maternal duration, birth-length, birthweight, expected birthweight, serum are well recognized as neonatal lupus and con- artery and venous pH, Apgar score at 5 min and way of genital heart block (CHB) [9]. The incidence of neonatal delivery (normal delivery, vacuum extraction delivery or lupus in an offspring of a mother with anti-Ro/SSA anti- Caesarean section and if any delivery induction). From bodies is estimated at 1–2% [10], but may be as high the Malmo¨ Sjo¨gren’s registry: autoantibody profile, dis- as >20% if the mother has given birth to a child with ease duration and age were collected for all pSS cases.
Premature birth was considered as birth before gestation- Reports on pregnancy outcomes beyond neonatal al Week 37 + 0. Low birthweight (LBW) was defined as lupus and congenital AV block are rare in pSS in contrast birthweight <2500 g. Birthweight deviation was birth- to the situation in SLE and APS. Pregnancy outcome in weight in relation to expected weight for the gesta- pSS has not been extensively studied, but has in general tional age [(expected weight–newborn weight)/expected not been considered to be associated with impaired fetal weight], presented as percentage. Severe fetal outcome outcome [12], although two studies have reported an is defined as a sum of the following: death (before or after increased rate of spontaneous abortion and fetal loss in delivery, Apgar score 0, 5 min after delivery), severe fetal pregnancies before SS diagnosis [13, 14]. The only study distress (umbilical cord blood pH < 7, Apgar score < 4).
up to now applying the most recent and widely acceptedclassification criteria for pSS [2] could not reveal signifi- cant differences in pregnancies in pSS before diagnosiscompared with controls [15]. The aim of the present study Mann–Whitney U-test and chi-squared test were used to was to study the impact of pSS on fetal and pregnancy compare pSS patients with a clinically diagnosed disease outcome compared with pregnancy outcomes in the gen- and their offspring with those in a pre-diagnostic state and eral population by linking two registries covering the broad their offspring. Conditional logistic regression analyses majority of pSS patients and all deliveries in Malmo¨ since were used to compare cases and controls taking the matched design of the study into account. Statistical sig-nificance was set at P < 0.05.
The present study is designed as a nested case–control registry linkage study. The Malmo¨ SS registry was estab- Mean age at delivery was significantly (P = 0.004) higher lished in 1984 as an approved research registry covering for the pSS patients than for controls [33.6 (range 29–37) the majority of clinically symptomatic pSS cases in Malmo¨ and 29.8 (range 24–35) years, respectively] (Table 1). Ten and additionally a number of cases from the surrounding deliveries occurred in women with pSS diagnosis before regions. All patients, independently of disease severity, pregnancy (mean disease duration 2.9 years), whereas in were followed prospectively at intervals of 6 months to six deliveries pSS was not yet diagnosed (mean delay of 2 years. Clinical, laboratory and histological data were diagnosis 5.7 years). ANA and Ro/SSA antibodies were collected. After 2002, all SS diagnoses were re-evaluated present in all the cases at the time point of SS diagnosis.
according to the AECC criteria for pSS [2] and only pa- Anti-La/SSB and RF were present in 70 and 60% of those tients fulfilling the AECC criteria are included in this study.
with pSS before pregnancy and 67 and 83% of those with All deliveries since 1990 at the only delivery department pSS after pregnancy. Among the patients with pregnancy in Malmo¨ have been continuously included in a prospect- preceding diagnosis of pSS, retrospective analysis of ive database and research registry. The Malmo¨ SS regis- saved serum samples revealed the presence of autoanti- try and the Malmo¨ delivery registry were linked using the bodies to Ro and La in all but two mothers, who were unique Swedish personal identification number and all seronegative for Ro and La during pregnancy. These pregnancies in patients with pSS between 1990 and two pregnancies were completely uncomplicated. The 2006 were included, as long as delivery occurred at number of parity and miscarriage was not significantly dif- Malmo¨ University Hospital irrespective of whether they ferent either between cases and controls or between the received their pSS diagnosis before or after pregnancy.
Since this study was considered as a normal follow-upof health-care production and quality, there was no reason for applying to the Swedish research ethicscommittee.
Mean pregnancy duration was 277 days for controls vs Sixteen deliveries occurred in 14 pSS patients. For each 272 days for Sjo¨gren’s patients (P = 0.4). Normal delivery delivery in a pSS patient, the next five deliveries from the in contrast to instrumental and Caesarean section was general population in the registry were selected as control significantly more common in controls than pSS cases deliveries (n = 80 deliveries). The study period was from (P = 0.02). Generally, pregnancy outcome did not differ January 1990 through December 2006. The following in- between patients with a pSS diagnosis before or after formation was extracted from the delivery register: date of Pregnancy and fetal outcome in primary SS TABLE 1 Demographic data for cases and controls ( n = number of deliveries) Disease duration at delivery, mean (S.D.), years TABLE 2 Pregnancy outcome for cases and controls ( n = number of deliveries) diagnosed before pregnancy. The Apgar score betweenthe cases and controls and between the case groups did Seventy-two (90%) deliveries ended at term in the con- not differ significantly. Only one case of intrauterine AV trols vs 14 (87%) in the cases (NS). Premature birth block II occurred and was treated with high-dose dexa- occurred in 8 (10%) of the controls vs 2 (13%) of the methasone (published in detail elsewhere [16]).
cases (NS) (Table 3). Birthweight differed significantly be-tween cases and controls (P = 0.025). Newborn of womenwith pSS were characterized with greater birthweight Medication and disease activity during pregnancy deviation as compared with controls (À11.4 vs À1.3%, We anticipate that patients without pSS diagnosis when P = 0.007), and 25% of these babies were small-for- pregnant did not have active systemic disease and no gestational age (SGA) as compared with 7.5% in controls immunosuppressive treatment. Three out of 10 patients (P = 0.04). Of the four SGA fetuses there was one who became pregnant after pSS was diagnosed were intrauterine death (weight deviation À47%), one had pH on low-dose prednisolone (maximum 5 mg/day), whereas umbilical artery 6.96 and was delivered by vacum extrac- one patient had been treated with high-dose dexametha- tion, and two were delivered by Caesarean section. Mean sone during pregnancy due to intrauterine fetal AV block II birthweight for pSS patients’ babies was 3010 vs 3458 g in [16]. This patient was on ciclosporin as well due to severe controls. All babies born prematurely and with LBW interstitial lung disease. This case is one of the two with among the cases were found in the group of pSS patients LBW and preterm delivery. Another patient on low-dose TABLE 3 Fetal outcome for cases and controls (n = number of deliveries) SGA = small-for-gestational-age. aApgar score: determined at 5 min.
prednisolone was also on AZA due to chronic active hepa- difference between RA and controls [21]. It may partly titis. Pregnancy was terminated at term by Caesarean be explained by physical or psychological preconditions section. The other 10 patients had neither prednisolone (pain, fatigue and sicca problems) or maternal or fetal fac- nor any other DMARDs during pregnancy. One stopped tors not formally registered in the birth registry. Also higher anti-malarials when pregnant. The other patient with LBW/ age at first delivery compared with controls was demon- stillbirth had had severe RP, but no other systemic dis- strated in the RA study similar to our results in pSS [21].
ease symptoms before/during pregnancy and neither According to an older study, women with RA experience a during the following uncomplicated pregnancy. In general, longer time to conception [22] than healthy women, when disease activity is difficult to assess in pSS and only very trying to become pregnant at the same age. This might be recently have assessment instruments become available similar in pSS. The mechanism is unclear. However, the [17, 18]. However, young-onset pSS is usually associated higher age of the mother was most pronounced in those with more systemic disease complications and high con- having been diagnosed with pSS before pregnancy.
centrations of autoantibodies, which was the case in sev- Young women with pSS often have systemic and severe eral of the patients in the present study [19].
disease compared with pSS patients with later onset. In these young women, in concordance with the situation in SLE, reduction of disease activity is necessary beforepregnancy. In some of the patients in the present study, Beyond intrauterine AV block and neonatal lupus syn- this was certainly the reason for delayed pregnancy. Parity drome, it is usually anticipated that pSS is not associated is not different and Skopouli et al. [20] could not detect with unfavourable pregnancy outcome. The few studies statistically significant deviations from normal sexual be- investigating the topic are with one exception >10 years haviour in their study. Despite normal average gestational old [13–15, 20] and rely on older classification criteria for length, birthweight in offspring to pSS patients was sig- case definition. Therefore, we performed a registry linkage nificantly lower due to 13% (2 out of 16) of babies being study including all available pSS pregnancies (n = 16)in our region between 1990 and 2006 matched with born with birthweight <2500 g. These two children were population-derived controls (n = 80) avoiding any selection born preterm. Thus, 2 (20%) out of 10 children of mothers with a diagnosis of pSS before pregnancy were born pre- Our results confirm normal fertility expressed by parity term, one with CHB and one as stillbirth. No abnormalities and lack of excess numbers of fetal losses or preterm were found in pSS pregnancies before diagnosis. Niewold deliveries. However, maternal age at delivery is higher in et al. [23] documented recently that Type I IFN activity was pSS patients, birthweight in pSS offspring lower and de- high in symptomatic in contrast to asymptomatic individ- livery by Caesarean section or vacuum extraction more uals with pSS or SLE having given birth to children with frequent than in the background population.
neonatal lupus. Possibly, the presence of severe preg- This increased risk of operative delivery in our study was supposedly due to an increased risk of fetal growth women with pSS might be associated with Type 1 IFN restriction in the pSS pregnancies resulting in increased up-regulation. The retrospective nested case–control risk of severe fetal outcome. It was not detected in the design of our study does not allow comparison of IFN recent Norwegian study [15]. A recent publication on RA levels at the time of pregnancy. Intrauterine growth retard- and birth outcomes demonstrated exactly the same ation is described in SLE pregnancies. SLE and pSS are Pregnancy and fetal outcome in primary SS partly overlapping diseases and all of our patients are In conclusion, this study indicates that pregnancy seropositive for SSA/SSB. Mechanisms resulting in pla- occurs later in patients with pSS, they give birth to cental insufficiency may be similar to SLE in some sero- babies with lower birthweight and have a lower frequency positive pSS patients. Also in the study by Julkunen et al.
of normal deliveries compared with controls. Larger stu- [14], birthweight of pSS offspring was lower than that of dies are needed to investigate the mechanisms behind controls, but not as low as that of SLE patients.
these findings in pSS pregnancies in more detail.
We could not confirm a higher number of abortions (either spontaneous or medical) or fetal losses as seen in the Greek study from 1988 and in the publication fromFinland 1995. This difference most probably is to be . Pregnancy occurs later in life in pSS.
ascribed to the small numbers of pregnancies in all the . Mothers with pSS give birth to offspring of lower studies. However, spontaneous or medical abortions occurring without following pregnancy resulting in delivery . Normal partus is less common in pSS patients than or stillbirth are not included due to the design of our studyderiving information from the birth registry.
The main drawback of our study is the low number of deliveries. Despite linkage of the Malmo¨ SS registry with the Malmo¨ delivery registry, the number of available SS deliveries is as low as 16. As controls, five We want to thank biostatistician Jan A˚ke Nilsson for his population-derived consecutive deliveries following the expert advice when performing the statistical analysis.
index delivery in a pSS patient were chosen. The usually Funding: This study was supported by the Swedish late onset of pSS results in low numbers of pSS pregnan- Rheumatism Association (E.T. and L.T.H.J.), Anna-Greta cies, mirrored also by the lack of reliable epidemiological Crafoord Foundation (E.T.) and the Swedish Sjo¨gren’s data regarding maternal/fetal outcome of pregnant pa- tients with this autoimmune disease in the literature. Thefact that pSS diagnosis is often made with several years of Disclosure statement: The authors have declared no delay from symptom onset explains the fact that 6 out of 16 pregnancies had occurred before the patients werediagnosed with pSS. There was no statistically significantdifference between pregnancy outcomes in SS patients with delivery before or after the diagnosis was made.
Explanations for this could be both lack of statistical Bowman SJ, Ibrahim GH, Holmes G, Hamburger J, power to detect small differences or that the disease in Ainsworth JR. Estimating the prevalence amongCaucasian women of primary Sjo¨gren’s syndrome in fact in most cases does not influence pregnancy outcome two general practices in Birmingham, UK. Scand One strength of our study despite its limited size is Vitali C, Bombardieri S, Jonsson R et al. Classification the population-based approach with cases as well as criteria for Sjo¨gren’s syndrome: a revised version of the controls representing the underlying population in the European criteria proposed by the American-European catchment area, which minimizes selection bias of cases Consensus Group. Ann Rheum Dis 2002;61:554–8.
and ensures representativeness of the controls. None of Jonsson R, Bowman S, Gordon T. Sjo¨gren’s syndrome.
the previous studies has been truly population based.
In: Koopman W, Moreland L, eds. Arthritis and allied Another strength, with regard to the outcome variables conditions - a textbook of rheumatology. 15th edition.
related to pregnancy and delivery is the prospective regis- Philadelphia: Lippincott Williams & Wilkins, tration in identical fashion in both cases and controls.
A few studies that have evaluated the pregnancy and Stro¨mbeck B, Ekdahl C, Manthorpe R, Wikstro¨m I, fetal outcome in patients with pSS (the early studies Jacobsson LT. Health-related quality of life in primary from Yoannina/Greece from 1988 [13] and 1994 [20]) Sjo¨gren’s syndrome, rheumatoid arthritis and fibromyalgia used questionnaire and interview techniques, while Haga compared to normal population data using SF-36. ScandJ Rheumatol 2000;29:20–8.
et al. [15] combined questionnaire and birth registry datain their study.
Goodchild CE, Treharne GJ, Booth DA, Kitas GD,Bowman SJ. Measuring fatigue among women with The present study is too small to investigate the pos- Sjogren’s syndrome or rheumatoid arthritis: a comparison sible mechanism leading to the lower number of normal of the Profile of Fatigue (ProF) and the Multidimensional delivery and lower birthweight in the cases. It is likely that Fatigue Inventory (MFI). Musculoskeletal Care 2008;6: it may be a result of immunological abnormalities, such as autoantibodies, B-cell, complement or IFN activation and Theander E, Henriksson G, Ljungberg O, Mandl T, inflammation. A multicentre prospective approach for de- Manthorpe R, Jacobsson LT. Lymphoma and other tecting markers of unfavourable pregnancy outcome in malignancies in primary Sjo¨gren’s syndrome: a cohort pSS would probably be necessary for sufficient statistical study on cancer incidence and lymphoma predictors.
Theander E, Manthorpe R, Jacobsson LTH. Mortality and syndrome-treatment of fetal incomplete atrioventricular causes of death in primary Sjo¨gren’s syndrome. Arthritis block with dexamethasone. J Rheumatol 2001;28: Theander E, Jacobsson LT. Relationship of Sjogren’s 17 Seror R, Ravaud P, Bowman S et al. EULAR Sjogren’s syndrome to other connective tissue and autoimmune syndrome disease activity index (ESSDAI): development disorders. Rheum Dis Clin North Am 2008;34:935–47.
of a consensus systemic disease activity index in Lee LA. Transient autoimmunity related to maternal primary Sjogren’s syndrome. Ann Rheum Dis 2010;69: autoantibodies: neonatal lupus. Autoimmun Rev 2005;4: 18 Seror R, Ravaud P, Mariette X et al. EULAR Sjo¨gren’s 10 Brucato A, Frassi M, Franceschini F et al. Risk of syndrome patient reported index (ESSPRI): development congenital complete heart block in newborns of of a consensus patient index for primary Sjo¨gren’s mothers with anti-Ro/SSA antibodies detected by syndrome. Ann Rheum Dis 2010;69:1103–9.
counterelectrophoresis. Arthritis Rheum 2001;44:1832–5.
19 Ramos-Casals M, Solans R, Rosas J et al. Primary 11 Buyon JP. Neonatal lupus: bedside to bench and back.
Sjogren syndrome in Spain: clinical and immunologic expression in 1010 patients. Medicine 2008;87:210–9.
12 Mecacci F, Pieralli A, Bianchi B, Paidas MJ. The impact of 20 Skopouli FN, Papanikolaou S, Malamou-Mitsi V, autoimmune disorders and adverse pregnancy outcome.
Papanikolaou N, Moutsopoulos HM. Obstetric and gynaecological profile in patients with primary Sjogren’s 13 Siamopoulou-Mavridou A, Manoussakis MN, Mavridis AK, syndrome. Ann Rheum Dis 1994;53:569–73.
Moutsopoulos HM. Outcome of pregnancy in patients with 21 Norgaard M, Larsson H, Pedersen L et al. Rheumatoid autoimmune rheumatic disease before the disease onset.
arthritis and birth outcomes: a Danish and Swedish nationwide prevalence study. J Intern Med 2010;268: 14 Julkunen H, Kaaja R, Kurki P et al. Outcome in women with primary Sjogren’s syndrome. A retrospective 22 Nelson JL, Koepsell TD, Dugowson CE, Voigt LF, case-control study. Clin Exp Rheumatol 1995;13:65–71.
Daling JR, Hansen JA. Fecundity before disease 15 Haga HJ, Gjesdal CG, Koksvik HS, Skomsvoll JF, onset in women with rheumatoid arthritis. Arthritis Rheum Irgens LM, Ostensen M. Pregnancy outcome in patients with primary Sjogren’s syndrome. A case-control study.
23 Niewold TB, Rivera TL, Buyon JP, Crow MK. Serum type I interferon activity is dependent on maternal diagnosis in 16 Theander E, Brucato A, Gudmundsson S, Salomonsson S, anti-SSA/Ro-positive mothers of children with neonatal Wahren-Herlenius M, Manthorpe R. Primary Sjo¨gren’s lupus. Arthritis Rheum 2008;58:541–6.

Source: http://www.gynae4u.gr/articles/Rheumatology-2011-Hussein-rheumatology_ker077.pdf