1 INSERM U569-IFR69, Hôpital de Bicêtre, 82 avenue du Général Leclerc, 94276 Le Kremlin Bicêtre, 2 Hôpital Antoine Béclère, Service de Gynécologie Obstétrique, 92141 Clamart and 3 CHU Hôtel-Dieu, Service de Gynécologie Obstétrique, BP 69-63003 Clermont-Ferrand, France
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: ectopic pregnancy/fertility/localization/population-based/recurrence
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Several risk factors for EP have been identified (Chow et al., 1987; Coste et al., 1991
; Fernandez et al., 1991a
; Job-Spira et al., 1993
): pelvic inflammatory disease (PID)especially for infections involving Chlamydia trachomatissmoking, previous pelvic surgery, previous EP, past and current use of an intrauterine device (IUD).
The site of implantation of EP has been little studied, at least in large population-based samples. Only a few articles, published >15 years ago, report descriptive results concerning the site of EP in large samples either from populations defined geographically, (Hallatt, 1982; Hallatt and Grove, 1985
; Atrash et al., 1986
; Herbertsson et al., 1987
) or from hospital-based populations (Martin et al., 1988
; Al-Meshari et al., 1993
). The other studies on this subject are case reports or deal with very small series, with a review of previous studies. Moreover, all these papers focus on a specific site, none considering implantation of EP as a whole. To our knowledge, only one article has suggested a possible link between EP site and subsequent fertility (Pouly et al., 1991
) or EP recurrence, although improvements in the diagnosis and management of EP have shifted clinicians concerns away from the immediate health of the woman, towards preserving her subsequent fertility.
The site of EP implantation merits study as it affects the severity of the condition, and the immediate and delayed side-effects. In addition, the effect of the site of implantation on subsequent fertility should be investigated.
The aim of this study was to investigate the distribution of EP sites in a population-based sample, and its variation over time. We then aimed to study the immediate complications and factors determining the site of EP. Finally, we determined the rates of subsequent fertility and EP recurrence.
![]() |
Population and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The women were interviewed about reproductive events, by telephone, every 6 months after EP for 2 years, and then once per year until the age of 45 years. We asked questions about the desire for a new pregnancy, cumulative period of trying to become pregnant, pregnancies achieved, treatments for infertility and use of contraception.
This study was conducted on the 1933 EPs registered between January 1992 and December 2001. Women with medical treatment only (n = 133) were excluded from the study because the site of implantation could not be determined with certainty. Among the remaining 1800 subjects, the site of EP was determined by the operating surgeon and was known for 1679 subjects (93%), providing precise information concerning the distribution of EP implantation for this population-based sample. Six sites were recorded: interstitial part of the tube, isthmus, ampulla, fimbrium, ovary and abdomen.
For studies of subsequent fertility, we excluded the EP cases recorded in 2001 because we did not have enough time to record all the data for the first follow-up call for these cases. We also excluded 159 women who had undergone sterilization for contraception or therapeutic bilateral salpingectomy (with no desire for IVF). In addition, 133 women were lost to follow-up (8.7%). Thus, the analysis concerned 1228 women.
Factors studied and statistical methods
As very little is known about the factors determining the site of EP, we investigated the association between EP site and most of the known risk factors for EP. The treatment of EP was defined as radical if the tube in which the pregnancy occurred was removed, i.e. salpingectomy. Treatment was otherwise considered to be conservative. All French clinicians use similar guidelines when deciding between radical and conservative treatment (Pouly et al., 1991). Haemoperitoneum of >100 ml was classified as profuse (Fernandez et al., 1991c
). We tested possible associations between EP site and these factors by means of
2-tests and, when appropriate,
2-tests for trend (Armitage and Berry, 1987
) to take into account the ordered nature of the sites. Multivariate analyses were performed by multinomial logistic regression (Hosmer and Lemeshow, 2000
).
Two reproductive outcomes were used to evaluate fertility: the recurrence of EP and intrauterine pregnancy (IUP). In both cases, survival analysis methods were used with person-time being the time to pregnancy. This is the cumulative period of time during which a woman was trying to become pregnant until she became pregnant or was censored. Cumulative rates and confidence intervals (CI) were calculated by the KaplanMeier estimator (Breslow and Day, 1987) for each of the sites. The curves obtained were compared by log-rank tests for univariate analysis, and by Cox regression (Cox and Oakes, 1990
) to take into account confounding variables. Ties in time to pregnancy data were handled using a published method (Efron, 1977
). The confounding factors had been identified as being associated with subsequent fertility either in our previous analyses (Job-Spira et al., 1996
; Bernoux et al., 2000
) or by other teams (Thorburn et al., 1988
; Yao and Tulandi, 1997
): age, educational level, prior tubal damage, history of infertility, tubal rupture and treatment (conservative or radical). For women with several successive pregnancies, the time to pregnancy was calculated until the first occurrence of the type of pregnancy studied (EP or IUP). Only natural fertility was studied, and follow-up was censored if a woman began an IVF programme or had declared that she was not trying to become pregnant again (Bernoux et al., 2000
). In terms of methodology, there is no entirely satisfactory way to take IVF into account because this event is not independent of a womans fertility (IVF is prescribed for less fertile women). The removal of women undergoing IVF from the sample analysed would therefore result in an overestimate of fertility. Moreover, such exclusion is always incomplete because we cannot identify women who will undergo IVF treatment in the future. In our study, we censored follow-up only when the woman actually underwent IVF. Thus, we did not exclude the entire follow-up period for women undergoing IVF, limiting the overestimation of infertility.
In cases of EP recurrence, we investigated whether the sites of the two successive EP were identical or close to one another, by calculating weighted -statistics (Fleiss, 1981
).
-statistics are parameters of agreement that take chance agreement into account (agreement or concordance between the sites of the two EP in this study). The weighted
-statistic is based on the idea that if two sites differ by more than one category, then the disagreement should be given more weight than if they differ by only one category. Quadratic (dis)agreement weights were used. Values of
>0.8 indicate almost perfect agreement, values of 0.60.8 indicate substantial agreement, with 0.40.6 indicating moderate agreement, 0.20.4 fair agreement, and <0.2 poor agreement (Landis and Koch, 1977
).
Statistical analyses were performed with STATA software (StataCorp, 2001).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
Results similar to those in Table II were obtained if the main characteristics of the women (such as age, smoking, previous sexually transmitted disease, current IUD use and educational level) were taken into account as potential confounders.
Determinant factors
The relationship between EP site and the characteristics of the woman is shown in Table III. The association with current use of an IUD was particularly strong: the proportion of IUD users increased from interstitial to ovarian sites (in the population studied, about one woman in three was a current IUD user). However, the distribution of EP sites was similar in women who had used an IUD only in the past and women who had never used an IUD.
|
In multivariate analysis, the only factor that remained significantly associated with EP site was current use of an IUD, although prior EP was also close to the significance threshold.
Subsequent fertility
For the 1228 women followed up, mean follow-up time was 3.2 years, and 693 women had tried to become pregnant again. It was not possible to study subsequent fertility among women with an abdominal pregnancy because only nine of these women had tried to become pregnant again. Of the remaining 684 women, 78 experienced another EP. The 2 year cumulative rate of recurrent EP was 0.22 (95% CI 0.170.28). No significant differences in EP recurrence rates were observed according to the site of the current EP (Table IV).
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Some abdominal pregnancies are not primary abdominal pregnancies, but result instead from ruptured tubal or ovarian pregnancies (Hallatt and Grove, 1985; Costa et al., 1991
; Dover and Powell, 1995
). Even if the primary site of implantation of such pregnancies was known (for instance fimbrium), we classified them as abdominal because they could not be considered to be similar to other fimbrial EP: even if only secondarily, their implantation was abdominal (Martin et al., 1988
). We therefore retained the abdominal category and studied the characteristics of EP for all six sites.
Few articles describing the distribution of all EP implantation sites have been published. Moreover, the distribution is often cited with respect to all pregnancies, rather than with respect to EP, and may therefore vary according to the proportion of EP among pregnancies. Most of the remaining papers are devoted to specific sites (mainly interstitial, ovarian or abdominal EP), often studied from hospital-based data. However, the published figures are consistent with those reported here: 311% for interstitial EP (Fernandez et al., 1991b; Kabukoba and de Courcy-Wheeler, 1992
; Khalifa et al., 1994
; Tulandi et al., 1995
); 16% for ovarian EP (Gérin-Lajoie, 1951
; Boronow et al., 1965
; Hallatt, 1982
; Grimes et al., 1983
; Steiner et al., 1983
; Herbertsson et al., 1987
; Sandvei et al., 1987
; Cabero et al., 1989
; Raziel et al., 1990
; Al-Meshari et al., 1993
; Chahtane et al., 1993
); 0.91.4% for abdominal EP (Radman, 1978
; Delke et al., 1982
; Atrash et al., 1987
).
An increase in the incidence of ovarian EP over time has been reported (Steiner et al., 1983; Sandvei et al., 1987
; Cabero et al., 1989
; Raziel et al., 1990
). Some authors account for this increase by improvements in diagnosis (Cabero et al., 1989
; Raziel et al., 1990
). This is consistent with our observations because the incidence of ovarian EP increased until 1997, with the proportion of ovarian EP remaining fairly steady thereafter. Another possible interpretation is based on the role of IUD in interfering with the transport of ova from the ovary into the tube and variations in IUD use over time (Herbertsson et al., 1987
). This interpretation is not supported by our observations because IUD use has decreased in the Auvergne region since 1992 (Coste et al., 2000
).
We observed that the increase over time in the proportion of EP treated conservatively was greater for interstitial EP than for EP at other sites. This is consistent with the fact that before the 1990s, interstitial pregnancy was treated by cornual resection or hysterectomy probably due to late diagnosis in many cases (Parente et al., 1983; Tulandi et al., 1995
). Since then, advances in transvaginal sonography and in medical treatment with methotrexate have opened up new avenues for conservative treatment by laparoscopy (Fernandez et al., 1991b
).
In multivariate analysis, the only factor significantly associated with the site of EP was current contraception with an IUD, which was more frequent in distal EP cases. This is consistent with previous findings (McMorries et al., 1979; Hallatt, 1982
; Sandvei et al., 1987
; Raziel et al., 1990
). The IUD may cause mild inflammation, resulting in deciliation of the endosalpinx and therefore delayed ovum transport, leading to ectopic implantation (Herbertsson et al., 1987
). This finding both confirms the role of IUD in the aetiology of EP (Bouyer et al., 2000
), and suggests that there may be tubal factors involved in the aetiology of ovarian pregnancy (Sandvei et al., 1987
). IUD seem to play a role only in the short term, because no association with the site of EP was found if the woman had used an IUD only in the past.
The higher proportion of women with markers of tubal damage, such as previous EP or previous tubal surgery, that we found in proximal EP cases is consistent with previous reports (Boronow et al., 1965; Herbertsson et al., 1987
; Cabero et al., 1989
; Raziel et al., 1990
; Yarali et al., 1994
; Lau and Tulandi, 1999
), but this association was found to be weaker after adjustment for other determinant factors. Moreover, similar results were obtained regardless of the site of the previous EP (homo- or contralateral), suggesting that tubal surgery per se did not determine the location of a subsequent EP.
The crude difference in subsequent fertility observed between sites was close to the threshold of statistical significance and became non-significant if conservative treatment was considered. It is therefore possible that subsequent fertility does not depend on the EP site. However, as treatment was more frequently conservative in cases of distal EP, we cannot rule out the possibility that treatment plays the role of an intermediate factor between EP site and subsequent fertility. This would imply that the site itself plays a role, as previously suggested for ovarian pregnancies (Boronow et al., 1965; Hallatt, 1982
; Raziel et al., 1990
).
The concordance between the sites of recurrent homolateral EP, but not between the sites of contralateral EP, is interesting. As the infectious histories of women with homo- and contralateral EP were similar, this result suggests that surgical treatment of the tube may induce recurrence at the same location. In cases of conservative management of EP, this, in turn, may be an argument for favouring medical treatment whenever possible.
This study is the first to describe the sites of implantation of EP in a large population-based sample. It confirms that the complications of EP vary considerably according to site. It shows that current IUD use protects against interstitial pregnancies, which are the most difficult type of EP to manage. Finally, it shows that subsequent fertility tends to be higher for distal and extratubal EP and that there is fair concordance between the sites of two successive EP if they are homolateral.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Notes |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Anonymous (1995) Ectopic pregnacyUnited States, 19701992. Morbid. Mortal. Wk. Rep., 44, 4648.
Armitage, P. and Berry, G. (1987) Statistical Methods in Medical Research. Blackwell, Oxford.
Atrash, H.K., Hughes, J.M. and Hogue, C.J.R. (1986) Ectopic pregnancy in the United States, 19701983, CDC surveillance summaries. Morbid. Mortal. Wk. Rep., 35, 2937.
Atrash, H.K., Friede, A. and Hogue, C.J.R. (1987) Abdominal pregnancy in the United States. Frequency and maternal mortality. Obstet. Gynecol., 69, 333.[Abstract]
Bernoux, A., Job-Spira, N., Germain, E., Coste, J. and Bouyer, J. (2000) Fertility outcome after ectopic pregnancy and use of an intrauterine device at the time of the index ectopic pregnancy. Hum. Reprod., 15, 11731177.
Boronow, R., McElin, T.W., West, R.H. and Buckingham, J.C. (1965) Ovarian pregnancy. Report of four cases and a thirteen-year survey of the English literature. Am. J. Obstet. Gynecol., 91, 10951106.[ISI][Medline]
Bouyer, J., Rachou, E., Germain, E., Fernandez, H., Coste, J., Pouly, J.L. and Job-Spira, N. (2000) Risk factors for extrauterine pregnancy in women using an intrauterine device. Fertil. Steril., 74, 899908.[ISI][Medline]
Breslow, N.E. and Day, N.E. (1987) Statistical Methods in Cancer Research. Volume II. The Design and Analysis of Cohort Studies. Oxford University Press, London.
Cabero, A., Laso, E., Lain, J.M., Manas, C., Escribano, I. and Calaf, J. (1989) Increasing incidence of ovarian pregnancy. Eur. J. Obstet. Gynecol. Reprod. Biol., 31, 227232.[ISI][Medline]
Chahtane, A., Dehayni, M., Rhrab, B., Kharbach, A., El Amrani, S. and Chaoui, A. (1993) La grossesse ovarienne. A propos de 4 observations avec revue de la littérature [Ovarian pregnancy: based upon four cases, with a review of the literature]. Rev. Fr. Gynécol. Obstet., 88, 3538.[Medline]
Chow, W.H., Daling, J.R., Cates, W. and Greenberg, R.S. (1987) Epidemiology of ectopic pregnancy. Epidemiol. Rev., 9, 7094.[ISI][Medline]
Costa, S.D., Presley, J. and Bastert, G. (1991) Advanced abdominal pregnancy. Obstet. Gynecol. Surv., 46, 515525.[Medline]
Coste, J., Job-Spira, N., Fernandez, H., Papiernik, E. and Spira, A. (1991) Risk factors for ectopic pregnancy: a casecontrol study in France, with special focus on infectious factors. Am. J. Epidemiol., 133, 839849.[Abstract]
Coste, J., Job-Spira, N., Aublet-Cuvelier, B., Germain, E., Glowaczover, E., Fernandez, H. and Pouly, J.L. (1994) Incidence of ectopic pregnancy. First results of a population-based register in France. Hum. Reprod., 9, 742745.[Abstract]
Coste, J., Aublet-Cuvelier, B., Bouyer, J., Germain, E. and Job-Spira, N. (1995) Evaluation de lexhaustivité du registre des grossesses extra-utérines dAuvergne par la méthode capture-recapture [Evaluation of the completeness of the Auvergne register of ectopic pregnancy with the capturerecapture method]. Rev. Epidemiol. Santé Publique, 43 (Suppl. 1), 10.
Coste, J., Bouyer J., Germain E., Ughetto S., Pouly J.L. and Job-Spira, N. (2000) Recent declining trend in ectopic pregnancy in France: evidence of two clinicoepidemiologic entities. Fertil. Steril., 74, 881886.[ISI][Medline]
Cox, D.R. and Oakes, D. (1990) Analysis of Survival Data. Chapman & Hall, London.
Delke, I., Veridiano, N.P. and Tancer, M.L. (1982) Abdominal pregnancy: review of current management and addition of 10 cases. Obstet. Gynecol., 60, 200.[Abstract]
Dover, R.W. and Powell, M.C. (1995) Management of a primary abdominal pregnancy. Am. J. Obstet. Gynecol., 172, 16031604.[ISI][Medline]
Efron, B. (1977) The efficiency of Coxs likelihood function for censored data. J. Am. Stat. Assoc., 76, 312319.
Egger, M., Low, N., Smith, G.D., Linblom, B. and Herrmann, R. (1998) Screening for chlamydial infections and the risk of ectopic pregnancy in a county in Sweden: ecological analysis. Br. Med. J., 316, 17761780.
Fernandez, H., Coste, J. and Job-Spira, N. (1991a) Controlled ovarian hyperstimulation as a risk factor for ectopic pregnancy. Obstet. Gynecol., 78, 656659.[Abstract]
Fernandez, H., De Ziegler, D., Bourget, P., Feltain, P. and Frydman, R. (1991b) The place of methotrexate in the management of interstitial pregnancy. Hum. Reprod., 6, 302306.[ISI][Medline]
Fernandez, H., Lelaidier, C., Thouvenez, V. and Frydman, R. (1991c) The use of a pretherapeutic, predictive score to determine inclusion criteria for the non-surgical treatment of ectopic pregnancy. Hum. Reprod., 6, 995998.[Abstract]
Fleiss, J.L. (1981) Statistical Methods for Rates and Proportions. Wiley, New York.
Gérin-Lajoie, L. (1951) Ovarian pregnancy. Am. J. Obstet. Gynecol., 62, 920929.[ISI][Medline]
Grimes, H.G., Nosal, R.A. and Gallagher, J.C. (1983) Ovarian pregnancy: a series of 24 cases. Obstet. Gynecol., 61, 174180.[Abstract]
Hallatt, J.G. (1982) Primary ovarian pregnancy: a report of twenty-five cases. Am. J. Obstet. Gynecol., 143, 5560.[ISI][Medline]
Hallatt, J.G. and Grove, J.A. (1985) Abdominal pregnancy: a study of twenty-one consecutive cases. Am. J. Obstet. Gynecol., 152, 444449.[ISI][Medline]
Herbertsson, G., Magnusson, S. and Benediktsdottir, K. (1987) Ovarian pregnancy and IUCD use in a defined complete population. Acta Obstet. Gynecol. Scand., 66, 607610.[ISI][Medline]
Hosmer, D.W. and Lemeshow, S. (2000) Applied Logistic Regression. Wiley, New York.
Job-Spira, N., Collet, P., Coste, J., Brémond, A. and Laumon, B. (1993) Facteurs de risque de la grossesse extra-utérine. Résultats dune enquête cas-témoins dans la région Rhône-Alpes [Risk factors for ectopic pregnancy. Results of a case control study in the Rhone-Alpes region]. Contracept. Fertil. Sex. 21, 307312.[ISI][Medline]
Job-Spira, N., Bouyer, J., Pouly, J.L., Germain, E., Coste, J., Aublet-Cuvelier, B. and Fernandez, H. (1996) Fertility after ectopic pregnancy: first results of a population-based cohort study in France. Hum. Reprod., 11, 99104.[Abstract]
Kabukoba, J.J. and de Courcy-Wheeler, R.H.B. (1992) Hysteroscopy in the diagnosis of suspected interstitial pregnancy. Int. J. Gynecol. Obstet., 37, 121126.[ISI][Medline]
Khalifa, Y., Redgment, C.J., Yazdani, N., Taranisi, M. and Craft, I.L. (1994) Intramural pregancy following difficult embryo transfer. Hum. Reprod., 9, 24272428.[Abstract]
Landis, J.R. and Koch, G.G. (1977) The measurement of observer agreement for categorical data. Biometrics, 33, 159174.[ISI][Medline]
Lau, S. and Tulandi, T. (1999) Conservative medical and surgical managment of interstitial ectopic pregnancy. Fertil. Steril., 72, 207215.[ISI][Medline]
Mäkinen, J.I. (1987) Ectopic pregnancy in Finland 196783: a massive increase. Br. Med. J., 294, 740741.[ISI][Medline]
Mäkinen, J.I. (1996) Ectopic pregnancy falls in Finland. Lancet, 348, 129130.
Martin, J.N., Sessums, J.K., Martin, R.W., Pryor, J.A. and Morrison, J.C. (1988) Abdominal pregnancy: current concepts of management. Obstet. Gynecol., 71, 549557.[Abstract]
McMorries, K., Lofton, R.H., Stinson, J.C. and Cummings, R.V. (1979) Is the IUCD increasing the number of ovarian pregnancies? Contemp. Obstet. Gynecol., 13, 165169.
Parente, J.T., Ou, C.S., Levy, J. and Legatt, E. (1983) Cervical pregnancy analysis: a review and report of five cases. Obstet. Gynecol., 62, 7982.[Abstract]
Pouly, J.L., Chapron, C., Manhès, H., Canis, M., Wattiez, A. and Bruhat, M.A. (1991) Multifactorial analysis of fertility after conservative laparoscopic treatment of ectopic pregnancy in a series of 223 patients. Fertil. Steril., 56, 453460.[ISI][Medline]
Radman, H.M. (1978) Abdominal pregnancy: pathogenesis, diagnosis, and treatment. South. Med. J. 71, 670.[ISI][Medline]
Raziel, A., Golan, A., Pansky, M., Ron-El, R., Bukovsky, I. and Caspi, E. (1990) Ovarian pregnancy: a report of twenty cases in one institution. Am. J. Obstet. Gynecol., 163, 11821185.[ISI][Medline]
Sandvei, R., Sandstad, E., Steier, J.A. and Ulstein, M. (1987) Ovarian pregnancy associated with the intrauterine contraceptive device. Acta Obstet. Gynecol. Scand., 66, 137141.[ISI][Medline]
StataCorp (2001) Stata Statistical Software. Release 7.0. Stata Corporation, College Station, TX.
Steiner, E., Tabaste, J.L., Servaud, M., Catanzano, G., Boudinet, F., Dabir, P. and Ardilouze, J.L. (1983) Grossesses ovariennes: le point en 1982. A propos de 6 cas opérés, revue de la littérature [Ovarian pregnancy: six personal cases]. J. Gynecol. Obstet. Biol. Reprod., 12, 401406.[Medline]
Thorburn, J., Philipson, M. and Linblom, B. (1988) Fertility after ectopic pregnancy in relation to background factors and surgical treatment. Fertil. Steril., 49, 595601.[ISI][Medline]
Tulandi, T., Vilos, G. and Gomel, V. (1995) Laparoscopic treatment of interstitial pregnancy. Obstet. Gynecol., 85, 465467.
Weström, L., Bengtsson, L.P.H. and Mardh, P.-A. (1981) Incidence, trends and risks of ectopic pregnancy in a population of women. Br. Med. J., 282, 1518.[ISI][Medline]
Yao, M. and Tulandi, T. (1997) Current status of surgical and nonsurgical management of ectopic pregnancy. Fertil. Steril., 67, 421433.[ISI][Medline]
Yarali, H., McComb, P.F. and Lee, N.H. (1994) Evaluation of the sites of ectopic pregnancies subsequent to reconstructive tubal surgery. Hum. Reprod., 9, 9091.[Abstract]
Submitted on April 23, 2002; resubmitted on July 5, 2002; accepted on August 8, 2002.