Fernanda Garanhani de Castro Systemic erythematosus Surita Mary Ângela Parpinelli and : clinical evolution, Ema Yonehara Fabiana Krupa maternal and perinatal outcomes José Guilherme Cecatti and placental fi ndings Department of and Gynecology, Faculdade de Ciências Médicas,

Universidade Estadual de Campinas (Unicamp), Campinas, São Paulo, Brazil ORIGINAL ARTICLE

INTRODUCTION OBJECTIVE ABSTRACT Systemic lupus erythematosus (SLE) is The objective of the present study was to an autoimmune disease of unknown etiology evaluate the clinical evolution, perinatal out- CONTEXT AND OBJECTIVE: Systemic lupus erythematosus is a chronic disease that is more that can affect various organs and systems. comes and most frequently observed placental frequent in women of reproductive age. The Since it predominantly affects women (in alterations among pregnant lupus patients, relationship between is the proportions of 9 to 1) and since in the according to the presence or absence of disease problematic: maternal and fetal outcomes are worse than in the general population, and majority of cases it is diagnosed between fl are-ups. These women were receiving care the management of fl are-ups is diffi cult during the ages of 20 and 40, it is the connective tissue at a specialized prenatal clinic, Centro de this period. The aim here was to compare the disease that is most frequently associated with Atenção Integral à Saúde da Mulher (Women’s outcomes of 76 in 67 women with pregnancy and the puerperium. Remission Full Healthcare Clinic), Universidade Estadual lupus, according to the occurrence or absence of fl are-ups. of the disease around the time of conception de Campinas (CAISM/Unicamp), over an 1,2 DESIGN AND SETTING: An observational cohort is related to favorable pregnancy outcome. eight-year period, and they gave birth at the clinical study evaluating the evolution of pregnant On the other hand, a diagnosis of SLE during same institution. women with lupus who were receiving care at the pregnancy and fl are-up around the time of prenatal outpatient clinic, Centro de Atenção In- conception or during pregnancy are related METHODS tegral à Saúde da Mulher, Universidade Estadual de Campinas (CAISM/Unicamp), between 1995 to poor prognosis for both the pregnancy and This observational cohort study was and 2002. 1 the course of the disease. carried out at the specialized prenatal clinic METHODS: Data were collected on a precoded , intrauterine growth restric- CAISM/Unicamp, which is a tertiary clinic for form. The women were divided into two groups ac- tion, prematurity, and perinatal morbidity high-risk pregnancies. The patients included cording to the occurrence or absence of fl are-ups, and mortality are among the most common in the study were followed up according to as defi ned by the systemic lupus erythematosus disease activity index (SLEDAI). The presence or adverse perinatal outcomes in pregnant lupus a specifi c protocol for the care of pregnant absence of fl are-ups and renal involvement was patients. Flare-ups, nephritis and arterial women with lupus, including investigation of considered to be the independent variable and the are factors that increase the risk of the current clinical and laboratory conditions other results were dependent variables. perinatal complications,2,3 as is an association (cardiac, immunological, renal, hematological RESULTS: Flare-ups occurred in 85.3% of cases, with antiphospholipid antibody syndrome and hepatic status). They delivered their babies and were most signifi cant when there was renal involvement. This was related to greater numbers (APS), which is present in 30-40% of SLE at the same institution between 1995 and of women with preeclampsia and poor perinatal cases.4 Although rare, neonatal lupus is also a 2002. A total of 67 women and 76 pregnancies outcome. Intrauterine growth restriction was complication that may be serious and should were included in the study. They represent all more common in the women with active disease. Placental weight was signifi cantly lower in the always be considered in this population. of the cases with SLE that were managed at women with renal involvement. The adverse perinatal outcomes resul- this service during this period. There was no CONCLUSIONS: Flare-ups and renal involve- ting from SLE are believed to occur as a loss to follow-up. ment in lupus patients during pregnancy are consequence of immunological alterations in Data were collected on a precoded form associated with increased maternal and perinatal the . The histology of the placenta and data entry was performed using the Epi- complications. frequently reveals vascular abnormalities in Info software program, version 6.1. After KEY WORDS: Lupus. Pregnancy. Perinatology. the uteroplacenta or alterations in coagula- evaluating the general characteristics of the Placenta. Maternal welfare. tion. These lesions are generally similar to sample, the women were divided into two those found in preeclampsia, hypertension groups according to the presence or absence and diabetes mellitus.1,5,6 Improvements in of fl are-ups of the disease during pregnancy, the treatment and control of systemic lupus as defi ned by the systemic lupus erythemato- erythematosus have led to better quality of life sus disease activity index (SLEDAI).8 They for patients with this pathological condition were also divided into two groups according and a consequent increase in the number of to whether renal involvement with SLE was pregnancies in this population.7 detected or not.

Sao Paulo Med J. 2007;125(2):91-5. 92

Statistical analysis was carried out after can College of Rheumatology (ACR), while Of the 67 women enrolled in this study, performing consistency tests. Qualitative 17 (22.4%) had probable lupus (three positive 20 (26.3%) suffered some form of hyper- variables were analyzed using the χ² test criteria) and 12 (15.8%) had possible lupus tensive syndrome during pregnancy. Nine and, when applicable, Fisher’s exact test. (two positive criteria). The ACR criteria most women (11.8%) had preeclampsia (defi ned Student’s t test was used for comparison frequently found in this population were: im- as the raising of blood pressure after the of the means of continuous quantitative munological disorders, blood abnormalities and twentieth week of pregnancy plus variables. Statistical significance was esta- renal involvement, serositis, butterfl y-shaped above 300 mg over a 24-hour period), and all blished as p < 0.05. The study obtained rash and a positive antinuclear antibody (ANA) of these patients had some degree of renal in- prior approval from the institution’s Re- test (Table 2). volvement. Another three patients (4%) had search Ethics Committee. Although the SLEDAI criteria do not and eight (10.5%) recommend any specifi c test for evalua- had chronic hypertension. Some degree of RESULTS ting the immunological profi le, the ANA lupus nephropathy was present in 32 women During the study period, there were test was positive in 73% of cases during (45.1%) and there was no statistically signifi - 23,676 deliveries at the institution, of which prenatal follow-up. There were positive cant difference in gestational loss between 76 occurred among lupus patients. There- fi ndings of anti-DNA antibodies in 19%, these patients and the women who had no fore, the rate of deliveries to lupus patients anti-Ro in 37% and anti-La in 11% of the renal alterations. However, the incidence of at this institution was 3.21:1000. The mean cases. Anticardiolipin antibodies (via im- preeclampsia was greater, and the newborn age of the lupus patients was 25.9 years munoglobulin G and immunoglobulin M ’s weight and placental weight were (range: 18-39 years), and these patients had determination) and/or lower in this group of women. had between one and seven pregnancies, in- (via Russell viper venom time and kaolin One occurred during cluding the current one. Around 22.4% of clotting time) were detected in 36% of the puerperium. This was a patient who had the women had previously miscarried. The the patients. Immunosuppressor treatment secondary pulmonary hypertension associated women in this study had attended a mean of for SLE was required in 93.4% of the with lupus and who had had a fl are-up during 8.5 prenatal consultations. The mean time pregnant women. The immunosuppressors a twin pregnancy. She went into premature elapsed between diagnosis of lupus and com- used included prednisone (at doses of 5 to labor and underwent cesarean section at mencement of pregnancy was 48.9 months, 80 mg/day) and . Antihyperten- 28 weeks due to breech presentation of the i.e. four to fi ve years (Table 1). sive drugs were used by 26.7% and aspirin fi rst twin and premature labor. She developed Of the 76 cases studied, 47 (61.8%) fulfi lled and/or heparin by 14% of the women (data central nervous system vasculitis and died the diagnostic criteria established by the Ameri- not presented in table). from reentrant convulsions on the third day of the puerperium. Table 1. General characteristics and history of pregnant lupus patients. Campinas, With regard to the type of delivery, 68.6% 1995-2002 of the patients underwent cesarean section. Maternal variables Mean (n = 76) SD The principal indication for this was fetal Age (years) 25.9 5.64 distress (32 cases). Of the 76 cases studied, Number of prenatal visits 8.5 4.02 14 led to gestational loss. Of these, six were Disease duration (months) 48.9 39.0 (7.8%), three were cases of fetal death (3.9%) and fi ve were neonatal deaths n% (6.5%) (Table 3). In all the cases in which With history of abortion 17 22.4 the woman was not a primigravida, she had First pregnancy 21 27.6 a history of at least one previous . Nulliparous 28 36.8 Among the cases of fetal death, all the patients Without living children 31 40.8 had some degree of renal involvement. With history of cesarean sections 30 39.5 There was one twin pregnancy; the rate SD = standard deviation. of prematurity, as defi ned by the Capurro method, was 50.0% of all live births. The Table 2. Frequency of diagnostic criteria (American College of Rheumatology) among percentages of with Apgar score < 76 pregnant lupus patients. Campinas, 1995-2002 7 at the fi rst and fi fth minutes were 13.4% Criterion n % and 3.0%, respectively. The mean weight Immunological disorder 56 73.7 of the newborns was 2,324.35 grams. The Hematological disorder 49 64.5 majority were considered to be adequate for Renal disorder 47 61.8 (AGA), while around 22.3% Antinuclear antibody 46 60.5 were considered small for gestational age Serositis 38 50.0 (Table 3). Among the adverse neonatal out- Butterfl y wing rash 33 43.4 comes occurring in the surviving infants, the Discoid lupus 23 30.3 most frequent were intraventricular hemor- Photosensitivity 15 19.7 rhage (four cases), patent ductus arteriosus 14 18.4 (fi ve cases) and congenital dislocation of the Mouth ulcer 6 7.9 hip (three cases). There were also two cases Neurological disorder 5 5.7 of thrombocytopenia, in which a diagnosis

Sao Paulo Med J. 2007;125(2):91-5. 93

of neonatal lupus was made. There were no cases of atrioventricular blockage. Histological examinations were carried Constitutional 2.4% out on 45 and 44.4% of these were found to have no abnormalities. The mean Serositis 8.4% placental weight was 403.36 ± 193.3 grams. The most common findings were infarct, Osteoarticular 19% and intervillous thrombosis. The occurrence of disease fl are-ups during Cutaneous 43.5% pregnancy was defi ned as the presence of any of the SLEDAI criteria: hematological, im- Renal 42.2% munological, renal, cutaneous, constitutional or osteoarticular alterations, or serositis. These Immunological alterations were analyzed in 68 of the 76 women 43.6% in the study; 58 of them (85.3%) had some Hematological form of alteration (Figure 1). When the whole 19.5% sample was dichotomized into those who did and those who did not have fl are-ups according to the SLEDAI criteria, it was observed that Figure 1. Disorder rates due to lupus activity among pregnant women. Campinas, preeclampsia only developed in those with 1995-2002. some degree of active disease. The gestational loss rate, calculated by adding the percentages of miscarriages, fetal deaths and neonatal deaths, was not signifi cantly different between the Table 3. Type of delivery, indication for cesarean section and perinatal outcome among two groups. The women with disease fl are-ups pregnant lupus patients. Campinas, 1995-2002 had a greater number of low-weight infants, Type of delivery n (70)* % although there was no difference in relation to Vaginal 22 31.4 prematurity or adequate weight for gestational Cesarean section 48 68.6 age. There were no signifi cant differences in the 32 66.6 placental analysis, although a greater number Breech presentation 3 6.3 of placentas with abnormalities were found in Repeated cesarean section 2 4.2 patients who had fl are-ups (Table 4). Other indications 11 22.9 On the other hand, when the group was dichotomized into patients who had and pa- Perinatal morbidity-mortality (n = 77) tients who did not have lupus nephropathy, Gestational loss rate (%) 14 18.2 statistically signifi cant differences were seen Abortion 6 7.8 in perinatal outcome. There was a greater Fetal death 3 3.9 number of low-weight infants, a greater rate of Neonatal death 5 6.5 prematurity and fewer healthy infants at dis- Living 63 81.8 charge from hospital among the patients with Weight: AGA 49 77.7 lupus nephropathy. Preeclampsia occurred Weight: SGA 14 22.3 exclusively in this group. The mean placental Prematurity rate 34 50.0 weight was also lower (Table 4). Women with mean SD antiphospholipid antibody syndrome were not Weight of newborn (g) 2324.35 866.57 signifi cantly different with respect to gesta- AGA = adequate for gestational age; SGA = small for gestational age; SD = standard deviation. tional loss or placental abnormalities. * 6 cases of abortion were excluded.

Table 4. Results according to systemic lupus erythematosus (SLE) fl are-up and SLE renal involvement (nephropathy) among pregnant women. Campinas, 1995-2002 Occurrence of Results No fl are-ups p Renal involvement No renal involvement p fl are-ups Preeclampsia 9/58 0/10 NS 9/32 0 < 0.05 Gestational loss rate 9/58 2/10 NS 2/32 2/39 NS Low 32/52 2/9 < 0.05 20/27 14/36 < 0.05 Prematurity 29/52 4/9 NS 7/32 5/39 NS Small for gestational age 14/52 0/9 NS 20/27 12/36 < 0.05 Placental characteristics n = 38 n = 6 n = 9 n = 6 NS Mean weight (g) 376.6 498 NS 330 459 < 0.05 NS = not signifi cant.

Sao Paulo Med J. 2007;125(2):91-5. 94

DISCUSSION renal involvement. The incidence of prema- pus renal involvement. These results are in Among woman with lupus, the interaction turity was higher in this group of patients, agreement with reports in the literature, thus between the physiological changes to the im- which leads us to conclude that patients with suggesting that is predictive of mune system brought about by pregnancy and maternal complications related to lupus fl are- poor prognosis for pregnancy.13 the pathological changes caused by the disease ups present higher perinatal risk. Differentia- Pregnant SLE patients are normally alter the normal course of the reproductive tion between lupus fl are-ups and preeclampsia referred for normal labor and delivery. How- process.4 A higher rate of poor perinatal out- is diffi cult in clinical practice. Preeclampsia ever, the repercussions of their condition on comes among women with systemic lupus occurs in approximately 13% of patients the placenta and , particularly if anti- erythematosus has been well-documented in with lupus and is frequently confused with cardiolipin antibody is present, increase the the literature, particularly in women who have lupus nephritis.4 Some clinical and laboratory incidence of chronic or acute fetal distress and disease fl are-ups during pregnancy, as was the abnormalities are similar in these two condi- consequently increase the cesarean section rate case in the present series. In the literature, the tions, such as edema, arterial hypertension, among these women.11 fl are-up rates during pregnancy range from 13 thrombocytopenia and other alterations in The prematurity rate was high, reaching to 60%. This wide range may be explained blood coagulation, and proteinuria. In ad- 74.1% among those with renal involvement, by the absence of a universal standard for dition, growth disorders and fetal vitality and this resulted in poor gestational prognosis, diagnosing lupus fl are-ups, which is done disorders may occur. even among the patients whose pregnancy subjectively in clinical practice. The laboratory test findings that may resulted in a live conceptus. In addition, pre- Various indices have been created in an help in reaching a differential diagnosis maturity in itself defi nes a greater degree of attempt to standardize the criteria that defi ne are: abnormal urinary sedimentation with severity of neonatal morbidity, thus making fl are-ups of this disease.8 These indices do not the presence of erythrocyte dysmorphism the perinatal prognosis for such pregnancies focus specifi cally on pregnant lupus patients and cell casts, which are present in lupus even poorer. and this makes their application in pregnancy nephritis, whereas preeclampsia is defined The small-for-gestational-age rate (22.3%) diffi cult, since the clinical and laboratory by the presence of isolated proteinuria; and was also higher than in normal populations, standards of normality vary during this phase. hypocomplementemia and increased anti- and it is important to stress the fact that in all However, in the absence of an adequate index DNA antibody titers, which are present in these cases the mothers had disease fl are-ups for assessing pregnant women, and since the lupus nephritis and not in preeclampsia.9,10 during pregnancy. In the literature, reports study demanded a non-subjective standard Nevertheless, both conditions are seri- describe better prognosis for pregnancies in for defi ning disease fl are-ups, we opted to ous and, while the treatment should be which no fl are-ups occur, and recommend use one of these existing indexes: the systemic different in each case with respect to chang- pregnancy only after a one-year period without lupus erythematosus disease activity index ing the immunosuppressors or increasing disease fl are-ups.1,2 In such cases, the aggres- (SLEDAI).8 On the basis of this index, we their doses if lupus flare-up is the first sion to the placenta is expected to be less, thus defi ned groups of patients in whom the disease hypothesis, various complications such allowing adequate intrauterine development. fl ared up or did not, in order to compare vari- as changes in fetal vitality and maternal With regard to histology, although not ables. In this study sample, fl are-ups occurred arterial hypertension should be treated all the placentas were analyzed, and despite in 85.3% of cases. This large difference in the similarly in both conditions. the fact that the analyses were not specifi cally numbers of patients with and without fl are- The gestational loss rate, calculated carried out for the purposes of this study, a ups, in comparison with other studies may be from the sum of the percentages of miscar- greater rate of abnormalities was found in the indicative of a lack of adequate contraceptive riages, fetal deaths and neonatal deaths, placentas of this study population, particularly counseling in this population. In view of the was 18.2%. In a previous study carried out concerning the proportion of infarcted areas importance of not having disease fl are-ups in our center, in which 40 cases were ana- (more than 10%). Nevertheless, a comparative around the time of a planned conception, this lyzed over a period of nine years, this rate study with normal placentas is required, to may have been a factor associated with the was 30%. This possibly demonstrates that evaluate the importance of these abnormali- poor maternal and perinatal outcomes. some improvement has occurred in the care ties. Immunohistochemistry would provide In the present study, not all the patients provided for pregnant lupus patients over further important data for analyzing the fulfi lled at least four of the ACR criteria, which recent years.11 However, this current rate physiopathology of the disease. is the minimum requirement for reaching a is still much higher than what is found in confi rmed diagnosis of lupus. However, these normal pregnancies, albeit compatible with CONCLUSIONS patients were managed in the same way as data in the literature from studies carried There was a statistically signifi cant dif- those with confi rmed diagnosis, because of out among lupus patients.12 ference with regard to the rate of low-weight the strongly suggestive signs and symptoms In the literature, a greater number of cases newborn infants, between patients who had presented and because SLE is an evolutive of miscarriage have been reported associated lupus fl are-ups during pregnancy and those disease in which there may be a limited range with antiphospholipid antibody syndrome. who did not. SLE patients with renal involve- of symptoms in its initial phase. On the other However, in our study, there was no statisti- ment during pregnancy had significantly hand, careful for these women cally signifi cant difference in the number of greater rates of preeclampsia, prematurity and may reduce the gestational risks and poor miscarriages between the two groups, perhaps low-weight newborn infants. They also had perinatal outcomes. due to the restricted sample size. The presence smaller placentas than did the other patients With regard to maternal complications, of APS was also not associated with gestational with lupus. The presence of antiphospholipid preeclampsia occurred only in patients who loss. It should be emphasized that, in all cases antibodies was not related to poorer perinatal had fl are-ups, and particularly in those with of fetal death, there was some degree of lu- outcomes in this study sample.

Sao Paulo Med J. 2007;125(2):91-5. 95

REFERENCES

1. Barros VIPVL. Correlações anátomo-clínicas de placentas de 7. Buyon JP, Tamerius J, Ordorica S, Young B, Abramson SB. 13. Julkunen H. Renal lupus in pregnancy. Scand J Rheumatol pacientes com lúpus eritematoso sistêmico. [Dissertation]. São Activation of the alternative complement pathway accompanies Suppl. 1998;107:80-3. Paulo: Universidade de São Paulo; 1994. disease fl ares in systemic lupus erythematosus during pregnancy. 2. Khamashta MA, Ruiz-Irastorza G, Hughes GR. Systemic lupus Arthritis Rheum. 1992;35(1):55-61. erythematosus fl ares during pregnancy. Rheum Dis Clin North 8. Bombardier C, Gladman DD, Urowitz MB, Caron D, Chang Am. 1997;23(1):15-30. CH. Derivation of the SLEDAI. A disease activity index for 3. Petri M. Hopkins Lupus Pregnancy Center: 1987 to 1996. lupus patients. The Committee on Prognosis Studies in SLE. Rheum Dis Clin North Am. 1997;23(1):1-13. Arthritis Rheum. 1992;35(6):630-40. 4. Kiss E, Bhattoa HP, Bettembuk P, Balogh A, Szegedi G. 9. Lockshin MD. Pregnancy does not cause systemic lupus ery- Pregnancy in women with systemic lupus erythematosus. Eur thematosus to worsen. Arthritis Rheum. 1989;32(6):665-70. J Obstet Gynecol Reprod Biol. 2002;101(2):129-34. 10. Repke JT. Hypertensive disorders of pregnancy. Differentiating 5. Nayar R, Lage JM. Placental changes in a fi rst trimester missed preeclampsia from active systemic lupus erythematosus. J Reprod abortion in maternal systemic lupus erythematosus with Med. 1998;43(4):350-4. antiphospholipid syndrome; a case report and review of the 11. Surita FGC, Cecatti JG, Barini R, Parpinelli MA, Silva JLCP. Sources of funding: Not declared literature. Hum Pathol. 1996;27(2):201-6. Lúpus e gravidez. Rev Bras Ginecol Obstet. 1997;19(6):413-7. Confl ict of interest: None 6. Benirschke K, Kauffman P. Lupus erythematosus and lupus anticoagu- 12. Dhar JP, Essenmacher LM, Ager JW, Sokol RJ. Pregnancy Date of fi rst submission: January 4, 2006 lant. In: Benirschke K, Kauffman P, editors. Pathology of the human outcomes before and after a diagnosis of systemic lupus ery- Last received: February 28, 2007 rd placenta. 3 ed. New York: Springer-Verlag; 1995. p. 512-20. thematosus. Am J Obstet Gynecol. 2005;193(4):1444-55. Accepted: March 6, 2007

AUTHOR INFORMATION RESUMO

Fernanda Garanhani de Castro Surita, MD, PhD. As- Lúpus eritematoso sistêmico e gravidez: evolução clínica, resultados maternos e perinatais e achados sistant professor, Department of Obstetrics and Gynecology, placentários Faculdade de Ciências Médicas, Universidade Estadual de CONTEXTO E OBJETIVO: O lúpus eritematoso sistêmico é uma doença crônica que acomete preferencial- Campinas, Campinas, São Paulo, Brazil. mente mulheres em idade reprodutiva. A associação entre lúpus e gravidez é problemática e os resultados Mary Ângela Parpinelli, MD, PhD. Assistant professor, maternos e perinatais são piores que na população geral. O objetivo foi determinar os resultados de 76 Department of Obstetrics and Gynecology, Faculdade de gestações de 67 mulheres lúpicas segundo a atividade da doença. Ciências Médicas, Universidade Estadual de Campinas, Campinas, São Paulo, Brazil. TIPO DE ESTUDO E LOCAL: Estudo clínico descritivo avaliando a evolução de gestantes lúpicas seguidas no Ambulatório de Pré-Natal Especializado do Centro de Atenção Integral à Saúde da Mulher, Universidade Ema Yonehara, MD. Former student at the Department Estadual de Campinas (CAISM/Unicamp), no período de 1995 a 2002. of Obstetrics and Gynecology, Faculdade de Ciências Médicas, Universidade Estadual de Campinas, Campinas, MÉTODOS: Os dados foram coletados a partir de uma fi cha pré-codifi cada. As mulheres foram divididas São Paulo, Brazil. em dois grupos segundo atividade do lúpus eritematoso sistêmico (LES) na gestação, conforme o índice Fabiana Krupa, MD, MSc. Obstetrician, Department de atividade de doença lúpica SLEDAI (Systemic Lupus Erythematosus Disease Activity Index). A presença of Obstetrics and Gynecology, Faculdade de Ciências ou não de atividade de doença e de envolvimento renal foram consideradas variáveis independentes e Médicas, Universidade Estadual de Campinas, Campinas, os demais resultados as variáveis dependentes. São Paulo, Brazil. RESULTADOS: A doença em atividade durante a gestação ocorreu em 85,3% dos casos, sendo o acometi- José Guilherme Cecatti, MD, PhD, Full professor of mento renal o mais importante, relacionando-se a um maior número de mulheres que tiveram pré-eclâmpsia Obstetrics, Department of Obstetrics and Gynecology, e pior evolução perinatal. Restrição do crescimento intra-uterino foi mais freqüente nas mulheres com doença Faculdade de Ciências Médicas, Universidade Estadual ativa. O peso da placenta também foi signifi cativamente menor nas mulheres com envolvimento renal. de Campinas, Campinas, São Paulo, Brazil. CONCLUSÕES: A atividade da doença e o envolvimento renal do LES na gestação associam-se com o aumento de complicações maternas e perinatais. Address for correspondence:

Fernanda Garanhani de Castro Surita PALAVRAS-CHAVE: Lupus. Gestação. Perinatologia. Placenta. Saúde materna. Rua Alexander Fleming, 101 Campinas (SP) — Brasil — CEP 13083-970 Tel./Fax. (+55 19) 3788-9304 E-mail: [email protected]

Copyright © 2007, Associação Paulista de Medicina

Sao Paulo Med J. 2007;125(2):91-5.