[Frontiers in Bioscience E2, 1514-1519, June 1, 2010]
Effects of epidural and systemic maternal analgesia in term infants: the NoPiL study
Carlo Dani1, Silvia Perugi1, Giulia Fontanelli1, Giovanna Bertini1, Simone Pratesi1, Giuseppe Buonocore2, Mariangela Longini2, Fabrizio Proietti2, Cosetta Felici2, Riccardo Ciuti3, Paola D'Onofrio4, Annamaria Melani Novelli4, Marco Pezzati5, Davide Gambi6, Gianfranco Scarselli7, Alessandro Frigiola8, Alessandro Giamberti8, Raul Abella8, Firmino F. Rubaltelli1
1
TABLE OF CONTENTS
1. ABSTRACT
The aim of the No Pain in Labour (NoPiL) study was evaluate the stress and clinical outcome of infants vaginally born without maternal analgesia and after maternal epidural or systemic analgesia. We studied 120 healthy term infants, 41 in the no analgesia group, 38 in the epidural analgesia group, and 41 in the systemic analgesia group. Cortisol, ß-endorphin, oxidative stress markers (ie: total hydroperoxide (TH) and advanced oxidation protein products (AOPP)), interleukin-1β (IL-1β), and interleukin-8 (IL-8) cytokines were measured in arterial cord blood samples. Infants in the 3 groups had similar Apgar score, cord blood pH and occurrence of hypoglycaemia, hyperbilirubinemia, and respiratory depression. Cortisol and endorphin plasma levels did not differ in the groups, nor did TH and AOPP values. IL-1β and IL-8 cytokine were higher in infants born after maternal epidural analgesia than in other groups. Short-term outcome and stress were similar in infants vaginally born without maternal analgesia and after epidural and systemic analgesia. The possible implications of the highest interleukin levels in the epidural analgesia group deserve further study.
2. INTRODUCTION
Obstetric analgesia is a widely diffused procedure although there is a paucity of data regarding the potential side-effect on term infants due to maternal drugs administration during delivery (1). In this regard, newborns' outcomes have been extensively investigated through several methods, such as the Apgar score, blood gas analysis, and tests evaluating neonatal behaviour (2). Results on potential side-effect on newborns whose mothers were subjected to obstetric analgesia are to date controversial and matter of debate.
Labor pain can cause a generalized neuro-endocrine stress response that has widespread physiological effects on the parturient, on fetus and finally on newborn (2-4). Of note, it is also known that labor and delivery mode may affect newborn oxidative balance (5) and cytokines plasma levels (6,7). These facts could be of relevance since free radicals are known to be involved in the pathophisiological cascade of events leading to damage neonatal organs, especially the brain (8). Furthermore cytokine profiles at birth might influence the neonatal immune response occurring in the transition from an almost sterile to a normal environment (6,7) and the development of allergic disorders (9).
On the other hand, side-effects of maternal pain such as hyperventilation, secretion of stress related hormones (i.e. catecholamines, cortisol), and increased oxygen consumption can be prevented and/or limited by local (epidural anesthesia) or systemic analgesia. Data on different analgesia methods relationship with neonatal outcomes such as Apgar scores, umbilical cord gas analyses, respiratory depression requiring naloxone use following epidural (10-12) or systemic maternal treatment support the notion on their safety (13-15). Thus, systemic analgesia is currently used in clinical practice when epidural analgesia is contraindicated (16,17) or refused by the parturient.
To date, data on potential side-effects on short term outcome in term infants due to vaginal delivery and maternal analgesia, are still lacking. Therefore, the No Pain in Labour (NoPiL) study group investigated by a prospective multicenter study whether there were any differences among clinical outcomes of infants vaginally delivered with or without maternal analgesia (epidural or systemic). Infants' stress was evaluated by a panel of biochemical markers, such as cortisol, ß-endorphin, oxidative stress markers, interleukin-1β (IL-1β), and interleukin-8 (IL-8) cytokines in arterial cord blood samples.
3. MATERIALS AND METHODS
3.1. Approval
The NoPiL-study was approved by the local ethic committees and the parents of infants admitted into the study gave informed and signed consent.
3.2. Patients
The subject of the present study were 120 term infants of whom 41 were enrolled in the no analgesia group, 38 in the local analgesia group, and 41 in the systemic analgesia group. Maternal age (32.9 + 5.7 vs. 32.9 + 4.3 vs. 33.6 + 4.9 years) and parity (1.0 + 1.0 vs. 0.6 + 1.0 vs. 0.8 + 0.8) were similar in the 3 groups.
From May 2008 to April 2009 healthy singleton infants (18) with gestational age >37 weeks and birth-weight appropriate for gestational age, born after uncomplicated pregnancy by vaginal delivery without analgesia or after maternal epidural or systemic analgesia. were enrolled. Exclusion criteria were the lack of parental consent, the need of operative vaginal delivery, and incomplete data recording.
Women received local or systemic analgesia, regardless of parity and whether labor was spontaneous or induced, only after their specific request and after obtaining their written informed consent at 32 weeks of gestation. All women were in active labor (cervical dilation of 3-6 cm) with regular contractions every 3-4 min lasting for 1 min.
Epidural analgesia was performed with the following procedures: under asepsis the skin was infiltrated with 2 mL 2% lidocaine (Lidocaina® Angelini, Rome, Italy); the epidural space was identified at L2-L3 interspace using a midline approach and loss of resistance technique, with a 18G Tuohy needle. An epidural catheter (Perifix®, B Braun, Rome, Italy) was advanced cephalad 3-4 cm into the epidural space. After negative aspiration, analgesic boluses such as 15-20 mL of 0.0625% bupivacaine (Bupivacaina®, Angelini, Rome, Italy) or 10 ml of 0.125% bupivacaine, in relationship to the labor condition, plus fentanyl (Fentanest®, Pfizer, Milan, Italy) 50 μg until a maximum total dose of 100 μg. Boluses were administered to maintain a visual analogue pain scale (VAPS) <4, until the beginning of the delivery phase.
Systemic analgesia was obtained with intravenous continuous infusion of remifentanil (Ultiva®, GlaxoSmithKline, Verona, Italy) as previously reported (19). Briefly, remifentanil solution (25 μg�ml- 1) was given at the dose of 0.0125-0.15 μg�Kg-1�min-1 based on vital maternal and fetal signs (systemic arterial pressure, heart rate, respiratory rate, oxygen saturation, cardiotocography) and to maintain a visual analog pain scale (VAPS) <4.
Maternal age and parity, gestational age, birth weight, sex, type of delivery, umbilical cord artery pH, Apgar score at 1 and 5 minutes, respiratory depression necessitating naloxone use, and possible complications, such as jaundice or hypoglycemia occurring before discharge from the nursery were recorded for each newborn. Jaundice was recorded when the neonatologist on charge judged it necessary to measure bilirubin; hypoglycemia was recorded when glycemia was <45 mg/dL (20).
Blood samples of 5 mL were obtained at birth from the cord artery of all enrolled infants to study neonatal pH and perform the following assays.
3.3. Cortisol and endorphin assays
Plasma cortisol was measured by solid-phase chemiluminescent immunoassay (ADVIA Centaur® XP Immunoassay System, Bayer Health Care Diagnostics, Milan, Italy) with a lower limit of sensitivity of 1μg/dL. ί-endorphin was assayed in neonatal plasma by an enzyme immuno-assay (ß-Endorphin, EIA Kit, Burlingame, CA, USA) with a lower limit of sensitivity at 0.23 ng/ml.
3.4. Oxidative stress assay
Total hydroperoxide (TH) concentration is a measure of overall oxidative stress, given that it is the intermediate oxidative product of lipids, peptides, and amino acids. Its production was measured using a spectrophotometric method (d-ROMs Kit®, Diacron srl, Grosseto, Italy) as described by Buonocore and coll. (21). This method makes it possible to estimate the TH. The results were expressed in conventional units, (Carr units: the value of 1 Carr unit is equal to a concentration of 0.08 mg/dL of hydrogen peroxide).
Simultaneous determination of the advanced oxidation protein products (AOPP) provides information regarding another aspect of protein involvement in free radical reactions, namely oxidized proteins that have lost their oxidant properties (22). We measured AOPP by the method of Witko-Sarsat and coll. (22) using a spectrophotometric method. AOPP concentrations were expressed as m
mol/L chloramine-T equivalents.
3.5. Cytokine assays
To measure IL-8 we used a commercial immunoassay kit (IL-8 Human Elisa Kit, Invitrogen Corporation, Camarillo, CA, USA) following the prescribed procedures. The minimum detectable dose for IL-8 is <0.1 pg/ml. To measure IL-1β we used another commercial immunoassay kit (IL-1β Human Elisa Kit, Invitrogen Corporation, Camarillo, CA, USA) which follows the same protocol as IL-8. The minimum detectable dose for IL-1β is 0.1 pg/ml. All samples were analyzed in duplicate by personnel unaware of the patients' assigned group.
We decided to measure these cytokines because: 1) IL-1β is an important mediator in the early inflammatory response by recruiting and activating inflammatory cells and causes the release of other cytokines (including itself); 2) IL-8 is probably the most important chemotactic factor (23) in the human immune system.
3.6. Data analysis and statistical comparisons
In planning our study, we calculated that a sample size of at least 37 infants for each group (infants born without maternal analgesia, infants born after maternal local analgesia, infants born after maternal systemic analgesia) was required to detect a statistically significant change of 20% in cortisol plasma value with 80% power at 0.05 level.
The clinical characteristics of the two groups were described by median (and ranges) or mean (+ SDs) values, while laboratory results were reported as mean values (+SDs). Statistical analysis was performed using Student's "t" test for parametric continuous variables (since studied variables showed a normal distribution) and Fisher's exact test for categorical variables. A p<0.05 was considered statistically significant. Data analysis was performed by D.C..
4. RESULTS
Neonatal characteristics are given in Table 1. Infants in the 3 groups had similar gestational age, birth weight and sex distribution, as well as Apgar score and cord blood pH. Moreover, the occurrence of respiratory depression, hypoglycemia and hyperbilirubinemia did not vary between the groups.
Cortisol and ß-endorphin plasma levels did not differ between the groups, nor did TH and AOPP values. On the other hand, IL-1β and IL-8 cytokine blood concentrations were similar in infants born without maternal analgesia or after maternal systemic analgesia, but lower than in infants born after maternal epidural analgesia (Table 2). In addition, IL-1β was not detectable in 61% (25/41) of infants born without maternal analgesia, in 27% (11/41) of infants born after epidural analgesia, and in 66% (25/38) of infants born after maternal systemic analgesia; IL-8 was not detectable in 8% (3/41) of infants born without maternal analgesia, in 12% (5/41) of infants born after epidural analgesia, and in 13% (5/38) of infants born after maternal systemic analgesia.
5. DISCUSSION
In our study we evaluated the effects of local and systemic maternal analgesia on term infants' stress at birth and outcome at nursery discharge. We did not find any clinical differences between infants born without or after maternal labor analgesia. In addition, we observed that the oxidative stress was similar in the studied groups, while IL-1β and IL-8 concentrations were higher in infants in the epidural analgesia group than in other groups.
Our data confirm that the outcome of infants born after epidural maternal analgesia is similar to that of infants born with no maternal pain relief, particularly in relationship to the post-delivery adaptation (24). In fact a recent meta-analysis detailed that these infants have the same risk (relative risk 1.56, C.l. 955 0.54-4.52) of having an Apgar score less than 7 at 5 minutes of life (25). In addition we observed that the most common neonatal complication occurring in healthy infants in the nursery, namely jaundice and hypoglycemia, have the same rate in infants born without maternal analgesia or after local maternal analgesia.
We also found that the neonatal outcome and the complication rate of infants born after systemic analgesia are similar to those of infants born with no maternal pain relief or after epidural maternal analgesia. This depended certainly on the favorable pharmacological characteristics of remifentanil which is considered an ideal drug in obstetrics because it has the most rapid onset of peak effect, shortest half-life, and greatest clearance of the commonly used opioids (26). Its concentration decreases by 50% within 3 to 5 minutes of stopping drug administration, regardless of the duration of the infusion (26). The drug rapidly crosses the placenta and is quickly metabolized by fetal esterases (27). Therefore, as expected, our data are in agreement with previous studies which did not report lower Apgar scores, unacceptable umbilical cord pH, or neonatal respiratory depression necessitating naloxone use after maternal remifentanil treatment (14,15,19,27,28).
It is interesting that our study is the first to show the occurrence of neonatal hyperbilirubinemia, and hypoglycaemia in term infants born after maternal labor analgesia, which were evaluated because they are practical markers of infant extra-uterine well-adaptation. Therefore, we cannot compare these findings with others studies and they deserve further confirmation.
In agreement with previous studies (29,30) we found that the cortisol blood level was similar in infants born without maternal analgesia and after epidural analgesia studied groups, as well as the ß-endorphin blood level. This latter result disagrees with findings of Vogl and coll. who found that infants born after epidural analgesia have a higher ß-endorphin level than those born with no maternal pain relief (30). This discrepancy might be due to different inclusion criteria and study design, since Vogl's study included only nulliparous women at nearly the same length of gestation, and/or epidural analgesia is not detailed by the authors (30). It is noteworthy that there are no other data in the literature on cortisol and ß-endorphin levels in infants born following systemic maternal analgesia.
The issue of fetal/neonatal oxidative stress has not been thoroughly investigated and is still subject to debate. Although a number of studies have compared the oxidative stress of infants born by vaginal delivery or cesarean section with contradictory and inconclusive results (31-33), we could not find any study comparing the oxidative stress of infants born by vaginal delivery with or without maternal analgesia. However, our study demonstrated that maternal analgesia does not affect the oxidative stress of infants born by vaginal delivery. To explain this finding we considered that oxidative stress takes place as a consequence of inadequate inactivation of accumulating oxygen free radicals by the antioxidant defense system (34). The effects of this imbalance are difficult to predict especially in the second half of pregnancy. The antioxidant capacity of rat embryo was reported to gradually increase with gestational age (35) and this increase might be the basis for better fetal coping with oxidative stress in the second half of pregnancy. In addition, term labor was found to induce an up-regulation of fetal antioxidant reserve that was thought to protect against the increased production of reactive oxygen species which occurs during labor secondary to the repeated ischemic reperfusion periods associated with uterine contractions and against the relative hyperoxia that is experienced by the newborn infant at birth (36). Thus, it is likely that maternal analgesia did not affect the oxidative stress of our patients because it cannot influence the aforesaid events.
We found that infants born after maternal epidural analgesia show higher levels of IL-1β and IL-8 than infants born without maternal analgesia or after systemic analgesia. The role of labor as an important trigger of neonatal pro-inflammatory cytokine production has been previously demonstrated (6), but the effect of different modes of maternal analgesia on their production has never been investigated. Previous studies demonstrated that IL-1β is higher in infants born by vaginal delivery than in infants born by cesarean section (6,37) while IL-8 is similar (37). In addition, Bessler and coll. demonstrated that the capacity of peripheral blood mononuclear cells to produce IL-1β is higher in newborns of mothers treated with epidural analgesia than in newborns of untreated mothers or those treated with systemic analgesia (38). To explain our results, we considered that some factors--including an increased risk of neurologic injury from needle- or catheter-induced mechanical trauma, local anesthetic toxicity, and neural ischemia secondary to local anesthetic additives (39) might help to induce maternal pro-inflammatory events which in turn also increase IL-1β and IL-8 production in infants born after epidural analgesia. Moreover, this hypothesis could explain the still unknown mechanism by which epidural labor analgesia induces an increase in maternal and neonatal temperature (40), since cytokine level increase is associated with fever development (41). However, to date the physiological role of pro-inflammatory cytokines in the neonatal period has not been clarified and the possible long-term effect of their changes, particularly on the development of allergic disorders, remains to be elucidated.
6. CONCLUSIONS
In summary, we studied the clinical outcome and stress of infants vaginally born without maternal analgesia or after maternal epidural or systemic analgesia. We found that the short-term outcome and the occurrence of the main neonatal complications did not differ between the groups. Moreover, the stress of the newborns, measured in terms of concentration of cortisol, ß-endorphin, and oxidation markers in arterial cord blood samples, were similar among the three groups. We also found that IL-1β and IL-8 levels were higher in infants born after maternal epidural analgesia than in the other two groups but the possible implications of this finding deserve further study.
7. ACKNOWLEDGMENT
The authors declare that they have no conflicts of interest.
8. REFERENCES
1. Mattingly JE, D'Alessio J, and Ramanath J: Effects of obstetric analgesics and anesthetics on the neonate. Paediatr Drugs 5:615-627 (2003) $mAuthor11. Johnson RF, Cahana A, Olenick M, Herman N, Paschall RL, Mintzer B, Ramasubramanian R, Gonzalez H, and Downing JW: A comparison of the placental transfer of ropivacaine versus bupivacaine. Anesth Analg 89:703-708 (1999) 21. Buonocore G, Perrone S, Longini M, Terzuoli L, and Bracci R: Total hydroperoxide and advanced oxidation protein products in preterm hypoxic babies. Pediatr Res 47:221-4 (2000) Key Words: Epidural analgesia, Systemic Analgesia, Stress, Oxidative Stress, Interleukin, Newborn
Send correspondence to: Carlo Dani, Division of Neonatology, Careggi University Hospital, Viale Morgagni, 85, 50141 Florence, Italy, Tel: 39-055-7947428, Fax: 39-055-7947428, E-mail:cdani@unifi.It
doi:10.2165/00148581-200305090-00004
PMid:12956618
2. Littleford J: Effects on the fetus and newborn of maternal analgesia and anesthesia: a review. Can J Anaesth 51:586-609 (2004).
doi:10.1007/BF03018403
3. Brownridge P: The nature and consequences of childbirth pain. Eur J Obstet Gynecol Reprod Biol 59:S9-15 (1995)
doi:10.1016/0028-2243(95)02058-Z
4. Neumark J, Hammerle AF, and Biegelmayer C: Effects of epidural analgesia on plasma catecholamines and cortisol in parturition. Acta Anaesthesiol Scand 29:555-559 (1985).
doi:10.1111/j.1399-6576.1985.tb02253.x
PMid:4060996
5. Lurie S, Matas Z, Boaz M, Fux A, Golan A, and Sadan O: Different degrees of fetal oxidative stress in elective and emergent cesarean section. Neonatology 92:111-115 (2007)
doi:10.1159/000100965
PMid:17377411
6. Buonocore G, De Filippo M, Gioia D, Picciolini E, Luzzi E, Bocci V, and Bracci R: Maternal and neonatal plasma cytokine levels in relation to mode of delivery. Biol Neonate 68:104-110 (1995).
doi:10.1159/000244225
7. Ly NP, Ruiz-Perez B, Onderdonk AB, Tzianabos AO, Litonjua AA, Liang C, Laskey D, Delaney ML, Dubois AM, Levy H, Gold DR, Ryan LM, Weiss ST, and Celedon JC: Mode of delivery and cord blood cytokines: a birth cohort study. Clin Mol Allergy 4:13 (2006)
doi:10.1186/1476-7961-4-13
PMid:17002791 PMCid:1592116
8. Buonocore G, Perrone S, Longini M, Paffetti P, Vezzosi P, Gatti MG, and Bracci R: Non protein bound iron as early predictive marker of neonatal brain damage. Brain 126:1224-1230 (2003)
PMid:12690060
9. Sugiyama M, Arakawa H, Ozawa K, Mizuno T, Mochizuki H, Tokuyama K, and Morikawa A: Early-life risk factors for occurrence of atopic dermatitis during the first year. Pediatrics 119:e716-723 (2007)
doi:10.1542/peds.2006-0893
PMid:17332188
10. Santos A, Karpel B, and Noble G: The placental transfer and fetal effects of levobupivacaine, racemic bupivacaine, and ropivacaine. Anesthesiology 90:1698-1703 (1999)
doi:10.1097/00000542-199906000-00027
PMid:10360869$Title
doi:10.1097/00000539-199909000-00032
12. Bolukbasi D, Sener EB, Sarihasan B, Kocamanoglu S, and Tur A: Comparison of maternal and neonatal outcomes with epidural bupivacaine plus fentanyl and ropivacaine plus fentanyl for labor analgesia. Int J Obstet Anesth 14:288-293 (2005)
doi:10.1016/j.ijoa.2005.04.007
PMid:16140519
13. Kan RE, Hughes SC, Rosen MA, Kessin C, Preston PG, and Lobo EP: Intravenous remifentanil. Placental transfer, maternal and neonatal effects. Anesthesiology 88:1467-1474 (1998)
doi:10.1097/00000542-199806000-00008
PMid:9637638
14. Saunders TA and Glass PS: A trial of labor for remifentanil. Anesth Analg 94:771-774 (2002)
doi:10.1097/00000539-200204000-00001
15. Blair JM, Hill DA, and Fee JP: Patient-controlled analgesia analgesia for labor using remifentanil: a feasibility study. Br J Anaesth 87(3):415-420 (2001)
doi:10.1093/bja/87.3.415
PMid:11517125
16. Hebl JR, Horlocker TT, and Schroeder DR: Neuraxial anesthesia and analgesia in patients with preexisting central nervous system disorders. Anesth Analg 103:223-228 (2006)
doi:10.1213/01.ane.0000220896.56427.53
17. Stamer UM, Stuber F, Wiese R, Wulf H, and Meuser T: Contraindications to regional anaesthesia in obstetrics: a survey of German practice. Int J Obstet Anesth 16:328-335 (2007)
doi:10.1016/j.ijoa.2007.05.011
PMid:17698339
18. American Academy of Pediatrics. Committee on Fetus and Newborn: Hospital stay for healthy term newborns. Pediatrics 125:405-9 (2010)
doi:10.1542/peds.2009-3119
PMid:20100744
19. D'Onofrio P, Melani Novelli AM, Mecacci F, and Scarselli G: The Efficacy and Safety of Continuous Intravenous Administration of Remifentanil for Birth Pain Relief: An Open Study of 205 Parturients. Anesth Analg 109:1922-1924 (2009)
doi:10.1213/ane.0b013e3181acc6fc
20. Cornblath M, Hawdon JM, Williams AF, Aynsley-Green A, Ward-Platt MP, Schwartz R, and Kalhan SC: Controversies Regarding Definition of Neonatal Hypoglycemia: Suggested Operational Thresholds. Pediatrics 105:1141-1145 (2000)
doi:10.1542/peds.105.5.1141
PMid:10790476$mAuthor
doi:10.1203/00006450-200002000-00012
PMid:10674350
22. Witko-Sarsat V, Freidlander M, Capeillere-Blandin C, Nguyen-Khoa T, Thu Nguyen A, Zingraff J, Jungers P, and Descamps-Latsca B: Advanced oxidation protein product as a novel marker of oxidative stress in uremia. Kidney Int 49:1304-1313 (1996)
doi:10.1038/ki.1996.186
PMid:8731095
23. De Dooy JJ, Mahieu LM, and van Bever HP: The role of inflammation in the development of chronic lung disease in neonates. Eur J Pediatr 160:457-463 (2001)
doi:10.1007/s004310100785
PMid:11548181
24. Leighton BL and Halpern SH: Epidural analgesia: effects on labor progress and maternal and neonatal outcome. Semin Perinatol 26:122-135 (2002)
doi:10.1053/sper.2002.32201
PMid:12005470
25. Anim-Somuah M, Smyth R, and Howell C: Epidural versus non-epidural or no analgesia in labor. Cochrane Database Syst Rev 19:CD000331 (2005)
26. Bailey PL, Egan TD, and Stanley TH: Intravenous opioid anesthetics. In: Miller RD (Ed.). Anesthesia, 5th ed, vol 1. Philadelphia: Churchill Livingstone; 273-376 (2000)
27. Kan RE, Hughes SC, Rosen MA, Kessin C, Preston PG, and Lobo EP: Intravenous remifentanil. Placental transfer, maternal and neonatal effects. Anesthesiology 88:1467-1474 (1998)
doi:10.1097/00000542-199806000-00008
PMid:9637638
28. Hill D: Remifentanil in obstetrics. Curr Opin Anaesthesiol 21:270-274 (2008)
doi:10.1097/ACO.0b013e3282f8e26c
PMid:18458540
29. Westgren M, Lindahl SG, and Nordén NE: Maternal and fetal endocrine stress response at vaginal delivery with and without an epidural block. J Perinat Med 14:235-241 (1986)
doi:10.1515/jpme.1986.14.4.235
PMid:3029362
30. Vogl SE, Worda C, Egarter C, Bieglmayer C, Szekeres T, Huber J, and Husslein P: Mode of delivery is associated with maternal and fetal endocrine stress response. BJOG 113:441-445 (2006)
doi:10.1111/j.1471-0528.2006.00865.x
31. Georgeson GD, Szony BJ, Streitman K, Varga IS, Kovacs A, Kovacs L, and Laszlo A: Antioxidant enzyme activities are decreased in preterm infants and in neonates born via caesarean section. Eur J Obstet Gynecol Reprod Biol 103:136-139 (2002)
doi:10.1016/S0301-2115(02)00050-7