Analysis of the incidence and risk factors associated with ectopic pregnancy following in-vitro fertilization and embryo transfer

Transcription

1 Human Reproduction vol.0 no.l pp , 995 Analysis of the incidence and risk factors associated with ectopic pregnancy following invitro fertilization and embryo transfer Samuel F.Marcus and Peter R.Brinsden Bourn Hall Clinic, Bourn, Cambridge CB3 7TR, UK 'To whom correspondence should be addressed Ectopic pregnancy is a well known complication of invitro fertilization (IVF) and embryo transfer. From March 983 to December 993, 3000 clinical pregnancies were achieved at Bourn Hall Clinic, including 35 ectopic pregnancies (.5%). Of these ectopics 20 were heterotopic, eight ovarian, six bilateral tubal and the remainder were singleton tubal pregnancies. The main risk factor identified in the series was a history of pelvic inflammatory disease (P < 0.00). The data also showed that ectopic pregnancy is at present more prevalent among patients in whom tubal damage is the reason for treatment. There was slight statistical evidence (P = 0.05) that patients having ectopic pregnancies received a higher volume of culture medium than those having normal deliveries. There was also an apparent trend (P = 0.07, not significant) that high progesterone/oestradiol ratio on the day of embryo transfer was associated with ectopic pregnancy. There was no statistical evidence of association between ectopic pregnancy and a history of ectopic pregnancy, abortion, still birth, termination of pregnancy, neonatal death, tubal surgery, ovarian stimulation protocol, plasma concentration of oestradiol, luteinizing hormone and progesterone, number of oocytes retrieved, number or quality of embryos transferred, administration of general anaesthesia for embryo transfer, and the number of patent Fallopian tubes. Awareness of the risk factors associated with ectopic pregnancy plays an important part in the early diagnosis of this potentially fatal condition. Key words: ectopic/ivf/risk factors Introduction The first pregnancy reported after invitro fertilization (TVF) was ectopic (Steptoe and Edwards, 976) and ever since there have been several reports of ectopic pregnancies occurring after IVF and embryo transfer. However, most of these reports have been based on a small number of patients or in collaborative multicentre studies and the reported incidence of ectopic pregnancy after IVF and embryo transfer varies from 2 to % of pregnancy (Smith et al., 982; Lopata, 983; Cohen et al., 986; Martinez and Trounson, 986; Correy et al, 988; Karande et al., 99; Dubuisson et al., 99; Verhulst et al., 993). Data on the risk factors associated with ectopic pregnancies after FVF and embryo transfer are conflicting. Yovich et al. (985) reported a higher incidence of ectopic pregnancy when the embryos were replaced higher than mid cavity. Martinez and Trounson (986) could not identify any risk factor, while Cohen et al. (986) identified two factors, the therapeutic use of clomiphene citrate and the number of patent Fallopian tubes at the time of transfer. Karande et al. (99) identified prior ectopic pregnancy as a risk factor and Dubuisson et al. (99) identified preexisting tubal pathology as a risk factor. Verhulst et al. (993) identified tubal damage and the use of clomiphene citrate combined with human menopausal gonadotrophin (HMG) as risk factors. We believe this study to be the largest singlecentre series reported to date. The aim of the study was to analyse the incidence and risk factors associated with ectopic pregnancy. Materials and methods Data on 35 ectopic pregnancies were available for retrospective study taken from records at Bourn Hall Clinic covering the period In order to identify variables which may be associated with ectopic pregnancy, these data were compared with those from random samples of 35 singleton deliveries taken during the same period of The data may, therefore, be regarded as random samples of ectopic pregnancies and of normal singleton deliveries. Table I shows a comparison of patient characteristics in the two groups. Ovarian stimulation and monitoring All patients received IVF treatment. Ovarian stimulation was performed with either clomiphene citrate (Serophene; Serono Laboratories, Welwyn Garden City, Herts, UK) and human menopausal gonadotrophin (HMG; Pergonal; Serono Laboratories) with or without follicular stimulating hormone (FSH; Metrodin; Serono Laboratories) or luteinizing hormonereleasing hormone (LHRH) agonist buserelin (Suprefact; Hoechst, Hounslow, UK) in either ultrashort protocol or long protocol (Marcus et al., 993). The method of ovarian stimulation, monitoring, sperm preparation and embryo culture has been previously described (Purdy, 982). Ovulation was induced with human chorionic gonadotrophin (HCG; Profasi; Serono Laboratories), 5000 U i.m. Oocyte recovery and embryo transfer Oocyte retrieval was performed by the vaginal ultrasoundguided method in 78% of the patients, and by laparoscopic or transabdominal oocyte recoveries in the remainder. In all, 8 fresh and 7 cryopreserved/thawed embryo transfer cycles were carried out. The cryopreserved/thawed embryos were transferred either Oxford University Press 99

2 S.F.Marcus and P.R.Brinsden Table I. Comparison of patient characteristics in study and control groups Variable Ectopic group Delivered group Results (control) No. patients Age (mean ± SEM) 33. ± ± 0.36 Duration of infertility (mean ± SEM) 3.03 ± ± 0.06 Proportion primary infertility (7/35) 0.5 (73/35) = not significant. Table n. The incidence and site of ectopic pregnancy in the study group Site of ectopic gestation No. of Incidence per Incidence per cases clinical pregnancy total ectopics Heterotopic pregnancy Bilateral tubal pregnancy Ovarian pregnancy Singleton tubal pregnancy Total % (20/3000) 0.2% (6/3000) 0.27% (8/3000) 3.% (0/3000) 35.5% (35/3000).8% (20/35).% (6/35) 5.9% (8/35) 7.8% (0/35) in natural monitored cycles or buserelin/hormone replacement cycles (Sathanandan et al., 99). Generally, embryos were transferred 5 8 h after oocyte recovery, 3 days after plasma luteinizing hormone (LH) surge or 3 days after commencement of progesterone in buserelin hormone replacement therapy (HRT) cycles. Replacements were performed in the dorsal lithotomy position with the head tilted downwards and the patient remaining in the position for 2 h after transfer. A maximum of four embryos were transferred transcervically into the uterine cavity about cm short of the fundus using Wallace catheters (Wallace, Colchester, Essex, UK) in 96% of cases, and Bourn or metal catheters in the remainder. General anaesthetic was used for difficult transfers. Embryo transfers were classified according to the degree of difficulty, i.e. smooth or difficult. The volume of culture medium injected varied from 0 to 60 /d. Luteal phase support Luteal phase support in the series varied from none in the natural cycles to natural progesterone (Gestone; Paines & Byrne, London, UK; or Utrogestan Laboratoires BesinsIscovesco, Paris, France) or HCG. Diagnosis of ectopic pregnancy Urinary or plasma concentration of HCG and plasma progesterone were measured 3 days after embryo transfer and if positive were repeated at 5 day intervals. At 7 weeks of gestation an ultrasound evaluation was performed to check the number, site and viability of pregnancy. Histological examination of the surgical specimens confirmed the diagnosis in all cases operated upon. Statistical analysis Continuous variables such as duration of infertility, hormone concentrations, etc. were analysed by a nonorthogonal analysis of variance with two factors, chronological time and type of pregnancy (ectopic/). Since the time span was over Table in. Proportion of ectopic pregnancies among patients in control and study group, grouped by previous medical history Previous medical Proportion of patients Proportion of patients Result event with negative history with positive history Previous live births Previous abortion Previous ectopic pregnancy Previous termination Previous stillbirth Previous neonatal death Previous tubal surgery 0.83 (02/2) 0.83 (/230) (/222) 0.8 (5/239) 0.87 (27/26) 0.90 (30/265) (69/25) (29/36) (20/36) 0.55 (20/) (6/27) (/5).000 (/) 0.35 (60/38) = no statistical significance. due to missing data for this field. Table IV. Relationship between a history of pelvic inflammatory disease (PID) and ectopic pregnancy in the study group Period Proportion of patients with a history of PID (7/3) (2/8) (2/6) Total (3/7) a Proportion of patients with no history of PID (0/2) (/52) 0.32 (5/33) 0.03 (9/87) a Significantly different (P < 0.00). Table V. Proportion of patients in study and control groups having tubal damage, arranged according to time period Patient group Ectopic Delivered Significance 0.53 (9/35) (9/2) (38/58) 0.5 (37/68) (29/38) (7/3) P = = not significant. Table VI. The association between pelvic inflammatory disease (PID) and tubal damage presented as the reason for IVF treatment of ectopic patients and patients Patient group History Proportion of patients Statistical of PID with tubal damage significance Ectopic group (n = 0) Delivered group (n = 99) yes no yes no (29/3) 0. (/9).000 (6/6) 0.2 (35/83) P < 0.00 P = 0.00 years, this period was divided, for purposes of analysis, into early (983986), middle (987990) and late (99 993). Also, in view of the extreme ranges, the hormone concentrations were

3 Ectopic pregnancy following FVF and embryo transfer Table VII. Ectopic and normal deliveries divided according to ovarian stimulation protocol Natural HRT C + P ± M SBG LBG Total (0.97) 22 (0.97) 3 (0.565) 0 (0.35) 2 29 (0.20) 0 (0.580) (0.52) 0 (0.76) 0 (0.370) 7 (0.630) 20 (0.35) 26 (0.565) (0.60) 68 (0.50) 38 (0.69) 3 (0.53) HRT = hormone replacement therapy; LBG = long buserelin protocol; P = human menopausal gonadotrophin; M = follicular stimulating hormone; SBG = ultrashort buserelin protocol; C = clomiphene citrate. transformed onto the logarithmic scale prior to analysis. The analysis was intended to derive unbiased estimates of the factor mean values, removing any distortion caused by other factors (such as 'time' in the present case). Categorical variables such as number of embryos transferred, quality, etc., were formed into contingency tables and analysed using loglinear methods, again correcting for time period. Results Of the 3000 clinical pregnancies achieved by IVF and embryo transfer during the period from March 983 to December 993, 35 resulted in ectopic pregnancies (.5%). Table II summarizes the incidence and site of ectopic pregnancy. There was no statistical evidence of an association between ectopic pregnancy and past history of abortion, termination of pregnancy, still birth, neonatal death, and tubal surgery (Table IH). History of pelvic inflammatory disease was positively associated with ectopic pregnancy, the incidence of ectopic pregnancy amongst patients with pelvic inflammatory disease was (3/7), whereas the incidence amongst the remaining patients was 0.03 (9/87) (P < 0.00). The regularity of this effect over the three time periods may be noted in Table IV, which displays the proportion of ectopic pregnancies in the samples. Table V shows the relationship between tubal damage and ectopic pregnancy. There was clear evidence in the late period (99 993) that tubal damage is associated with ectopic pregnancies. There was strong evidence of an association between pelvic inflammatory disease and tubal damage as the reason for IVF treatment. Table VI gives the proportion of patients with tubal damage as the cause of infertility. For both the ectopic patients and the normal deliveries the proportion with tubal damage was significantly higher in the patients with a history of pelvic inflammatory disease. There was no statistical evidence of an association between the type of ovarian stimulation protocol and ectopic pregnancy (Table VII). There was no significant difference in the ectopic pregnancy rate according to oestradiol, LH and progesterone concentrations at the time of ovulation induction. The progesterone/oestradiol ratio at the time of embryo transfer was not significant (P = 0.007, Table VIII). As the technique of embryo transfer did not change during the period of the study, none of the parameters related to the Table VOL Hormone concentrations on day of ovulation induction according to the outcome of IVF, i.e. ectopic pregnancy" Variable Ectopic Delivered Result Oestradiol 6.6 (±0.06) [70] 6.73 (±0.06) [833] LH.6 (±0.08) [3.9].08 (±0.08) [2.95] Progesterone 0.56 (±0.06) [0.57] 0.6 (±0.06) [0.53] Progesterone/ 3.0 (±0.06) [0.09] 3.6 (±0.09) [0.03] oestradiol ratio = not significant. a Analysis carried out on log transformed data. The regenerated mean values on the original scale are given in square brackets. Values are mean ± SEM. Table IX. Variables having no evidence of association with ectopic pregnancy Variable Ectopic group Delivered group Result No. oocytes recovered 8.26 (±0.) 8.8 (±0.2) Proportion with good quality (79/3) 0.67 (9/35) embryos Proportion of patients needing 0.092(2/9) 0.052(7/35) general anaesthesia for embryo transfer Proportion of smooth embryos transferred 0.877(/30) 0.895(9/33) Proportion using Wallace (26/3).000 (35/35) catheter = not significant. When the numbers do not add up to the sample size, the discrepancy is due to missing data in this field. Table X. The relationship between the outcome of ectopic pregnancy Ectopic group Delivered group Totals Status of tubes Both normal condition of Fallopian tube and the One blocked Both blocked When the numbers do not add up to the sample size, the discrepancy is due to missing data in the field

4 S.F.Marcus and P.R.Brinsden technique of embryo transfer were analysed, apart from the smoothness compared to the difficulty encountered at embryo transfer which was not statistically significant. There was some statistical evidence that the ectopic patients had received a higher volume of culture medium than the injected patients: 22.7 ± 0.58 /* versus 2.0 ± 0.3 fd (P = 0.05). There was no significant difference observed between the two groups according to the treatment given for luteal phase support. Table IX shows the variables with no statistical evidence of an association with ectopic pregnancy. There was no statistical evidence of an association between the number of patent Fallopian tubes and ectopic pregnancies (Table X), and the condition of the right and left tubes was examined in conjunction with the site of ectopic gestation (right and left) but showed no statistical evidence of systematic patterns. Discussion Ectopic pregnancy is a known complication of IVF and embryo transfer. The overall incidence of ectopic pregnancy after IVF in this series was 35/3000 (.5%). Although the incidence is comparable with other reports (Seppala, 985; Cohen etal., 986; Martinez and Trounson, 986; Correy et al, 988; World Collaborative Report, 989; Karande et al., 99; Dubuisson et al., 99; Verhulst et al., 993), it is much higher than that following natural conception (Mishell, 987). We have shown that one in four ectopic pregnancies following IVF and embryo transfer are heterotopic, ovarian or bilateral tubal pregnancies. In this series 75% of the ectopic pregnancies were single tubal, 5% were heterotopic, 6% ovarian and % were bilateral tubal pregnancies. This should be considered whenever the diagnosis of ectopic pregnancy is made. The aetiological factors associated with ectopic pregnancies are complex. In our study we observed that history of pelvic inflammatory disease was the main risk factor. Not surprisingly perhaps, there was strong evidence of an association between pelvic inflammatory disease and tubal damage presented as the reason for the IVF treatment. Despite the changing pattern of the relationship between ectopic pregnancy and tubal damage per se over the period of time in the study, we may say that ectopic pregnancy is at present more prevalent among patients when tubal damage is the reason for treatment. The progressive decline in the incidence of tubal damage as the cause for treatment among normal deliveries may simply reflect the fact that other causes of infertility, such as male infertility, have, over the years, become more responsive to IVF treatment. The apparent increase of tubal damage in the ectopic group during the study period probably reflects changing operational practices which have, in fact, increased the likelihood of ectopic pregnancy. Tubal diathermy and division near the cornu was advocated by Steptoe and Edwards (976) to minimize the risk of ectopic pregnancy. This may have helped to reduce the incidence of ectopic pregnancy after IVF and embryo transfer. Tubal occlusion prior to IVF is no longer performed at Bourn Hall. Dubuisson et al. (99) found the incidence of ectopic pregnancy in patients with bilateral salpingectomy was % as compared with.2% in patients who had hydrosalpinx and 9.9% in patients with pathological but patent tubes. These authors also reported 00% pathological lesions in their series of ectopic pregnancies, even when IVF had been done for endometriosis or unexplained infertility. Verhulst et al. (993) reported tubal damage was a major risk factor towards development of an ectopic pregnancy after IVF and embryo transfer. There was statistical evidence that the ectopic group had received a higher volume of culture medium than the control group (P = 0.05). This result confirms the finding of Knutzen et al. (989) which showed that after injection of 50 /tl of radioopaque fluid through a standard embryo transfer catheter, the material was passed either totally or partially into the Fallopian tubes in % of patients, suggesting that the chance of the embryo being carried into the tube immediately after transfer is high. The fact that most ectopic pregnancies occur in patients with damaged tubes supports the hypothesis that the embryo(s) is not transported back to the uterine cavity because of tubal dysfunction. We could not identify any risk factor related to prior ectopic pregnancy per se; this is in contrast to Karande et al. (99) who identified prior ectopic as the main risk factor for developing ectopic pregnancy after IVF and embryo transfer. There was no statistical evidence of an association between ectopic pregnancy and different ovarian stimulation protocols. This result confirmed the results of the 990 US Registry (Medical Research International Society, 992) and Dubuisson et al. (99); However, Cohen et al. (986) identified the use of clomiphene citrate as a significant risk factor which increased the rate of ectopic pregnancy from 3 to 6%. Similarly, Verhulst etal. (993) reported increased risk of ectopics with the use of clomiphene citrate and HMG compared with gonadotrophinreleasing hormone and HMG (2.8 and 0.8% respectively). They suggested that the reason for this is due to the antioestrogen effect of clomiphene citrate on the tubal oestrogen receptors with alteration of the local progesterone/oestradiol ratio and subsequently a disturbed oviductal peristalsis. In this study, there was an apparent trend for patients who suffered ectopic pregnancy to have a higher progesterone/oestradiol ratio on the day of embryo transfer (P = 0.07, not significant). SalatBaroux et al. (985) suggested that the early rise of progesterone in stimulated cycles would predispose to ectopic pregnancy by favouring the opening of the isthmus and myorelaxation of the tube, and Pulkkinen and Talo (98) suggested that there is a critical value of progesterone/ oestradiol ratio which increased electrical activity. Animal studies have shown that high oestrogen concentrations are associated with tubal blocking and arrest of the embryo in the tube (Pulkkinen and Talo, 98). However, we did not observe any significant difference in the oestrogen concentration on the day of ovulation induction between the ectopic and the delivery group. There was no statistical evidence of an association between the number of patent Fallopian tubes and ectopic pregnancy. This is in contrast to Cohen et al. (986) who found a relationship between the number of patent Fallopian tubes and the ectopic pregnancy rate from 3 or % (with no or two patent tubes) to 3% (with one patent tube). We conclude that patients with a history of pelvic inflammatory disease and/or tubal damage are at a higher risk of ectopic 202

5 pregnancy following FVF and embryo transfer and recommend injecting the minimum volume of culture medium during embryo transfer. Acknowledgements We would like to thank Dr D.E.Walters for carrying out the statistical analysis of the data and Sue Wordingham for typing the manuscript. References Cohen,J., Mayaux,M., GuihardMoscato,M. and Schwartz,D. (986) Invitro fertilization and embryo transfer, a collaborative study of 63 pregnancies. Hum. Reprod.,, Correy,J.F., Watkins.R.A., Bradfield.G.F., Garmen,S., Watson.S. and Gray,C. (988) Spontaneous pregnancies and pregnancies as a result of treatment on an invitro fertilisation program terminating in ectopic pregnancies or spontaneous abortion. Fertil. Sterii, 50, Dubuisson,J.B., Aubriot,F.X., Mathieu,L, Foulot,H., Mandelbrot,L. and Bouquet de Joliniere,J.B. (99) Risk factors for ectopic pregnancy in 556 pregnancies after invitro fertilisation: implications for preventative management. Fertil. Sterii., 56, Karande.V.C, Flood,J.T., Heard.N., Veeck,L. and Muasher.S.J. (99) Analysis of ectopic pregnancies resulting from invitro fertilization and embryo transfer. Hum. Reprod., 3, 6 9. Knutzen.V., ScotoAlbors,C.E., Fuller,D., Sher.G., Shynock.K. and Behr,B. (989) Mock embryo transfer (MET) in early luteal phase, the cycle prior to invitro fertilisation and embryo transfer. Presented at the 5th Annual Meeting of the American Fertility Society, San Francisco, California, 36 November 989. Published by the American Fertility Society in the program supplement, p. S52 (Abstract P. 229). Lopata,A. (983) Concepts in human invitro fertilisation and embryo transfer. Fertil. Sterii, 0, 289. Marcus,S.F., Brinsden,P., Macnamee.M., Rainsbury,P., Elder,K. and Edwards,R.G. (993) Comparative trial between an ultra short and long protocol of luteinizing hormonereleasing hormone agonist for ovarian stimulation in invitro fertilization. Hum. Reprod., 8, Martinez,F. and Trounson.A. (986) An analysis of factors associated with ectopic pregnancy in a human invitro fertilisation program. Fertil. Sterii, 5, Medical Research International Society for Assisted Reproductive Technology, The American Fertility Society (992) In vitro fertilization embryo transfer in the United States: 990 results from IVFET Registry. Fertil. Sterii, 57, 52. Mishell,D.R., Jr (987) Ecotopic pregnancy. In Drogegemuller,W., Herbst,A.L., Mishell.D.R., Jr and Stenchever.M.A. (eds), Comprehensive Gynecology. C.V.Mosby Co., St Louis, MO, p. 06. Pulkkinen.M.O. and Talo.A. (98) Myoelectrical activity in the human oviduct with tubal pregnancy. Am. J. Obstet. Gynecol, 8, 55. Purdy,J.M. (982) Methods for fertilization and embryo culture invitro. In Edwards,R.G. and Purdy.J.M. (eds), Human Conception Invitro. Academic Press, London, pp SalatBaroux.J., Giacomini,P., Cornet,D., Pereira Coelho.A., Mandelbaum,J. and Plachot.M. (985) Grossesses extraute'rines apres fecondation invitro dont deux associees a une grossesse intraute'rine evolutive. J. Gynecol. Obstet. Biol. Reprod. (Paris),, Sathanandan,M., Macnamee.M., Rainsbury,P., Wick,K., Brinsden.P. and Edwards,R.G. (99) Replacement of frozen thawed embryos in artificial cycle and natural cycle; a prospective semirandomized study. Hum. Reprod., 6, Seppala,M. (985) The world collaborative report on invitro fertilisation Ectopic pregnancy following IVF and embryo transfer and embryo transfer, current state of the art in January, 98. Am. NYAcad. Sci., 2, Smith,D.H., Pike.I., Tucker.M., Sinosich.M.J., KempJ.F., Picker,R.H. and Sanders,D.M. (982) Tubal pregnancy occurring after successful invitro fertilisation and embryo transfer. Fertil. Sterii., 38, Steptoe.P.C. and Edwards,R.G. (976) Reimplantation of a human embryo and subsequent tubal pregnancy. Lancet, i, Testart,J., Lassalle,B., BelaischAllert,J., Hazout.A., Forman,R., Daniel RalinholmJ. and Frydman.R. (986) High pregnancy rate after early embryo freezing. Fertil. Sterii, 6, Testart,J., Plachot,M., Mandelbaum,J., SalatBaroux,J., Frydman.R. and Cohen,J. (992) World Collaborative Report (989) on IVFET and GIFT. Hum. Reprod., 7, Verhulst,G., Camus,M., Bollen,M., Van Steirteghem,A. and Devroey,P. (993) Analysis of the risk factors with regard to the occurrence of ectopic pregnancy after medically assisted procreation. Hum. Reprod., 8, Yovich,J.L., Turner,S.R. and Murphy,A.J. (985) Embryo transfer technique as a cause of ectopic pregnancies in invitro fertilisation. Fertil. Sterii,, Received on November 30, 993; accepted on August 8,