Pregnancy Outcome in Women With Endometriomas After IVF
Pregnancy Outcome in Women With Endometriomas After IVF
Eutopic endometrium of women with endometriosis was reported to be abnormal (Braun and Dmowski, 1998; Pritts and Taylor, 2003) and these abnormalities may impair decidualization and placentation in women affected (Fujimoto et al., 1996). Since these latest processes are crucial for pregnancy implantation and development, it has been hypothesized that several complications of pregnancy may be more frequent in women with endometriosis (Fernando et al., 2009). Available clinical data are not fully consistent with this hypothesis. The literature on the relationship between endometriosis and abortion rate is conflicting, with some studies reporting increased risk (Fitzsimmons et al., 1987; Matorras et al.,1998; Omland et al., 2005) and others failing to document any association (Pittaway et al., 1988; Candiani et al., 1991). Stephansson et al. recently evaluated the risk for obstetrics complications in a large cohort of women with a previous diagnosis of endometriosis. They showed an increased risk of preterm birth, pre-eclampsia, antepartal bleeding/placental complications and Cesarean section. There was conversely no association with small-for-gestational-age (SGA) birth or stillbirth (Stephansson et al., 2009). In contrast, Brosens et al. (2007) reported that the rate of pre-eclampsia is reduced in women with endometriosis. Moreover, Fernando et al. (2009) failed to document an association between endometriosis in general and SGA and preterm birth. This latter study is particularly illuminating in regard of the topic of the present study. Indeed, it is the first report presenting data separately for the group of women with ovarian endometriomas. Of relevance here is that the study suggested an increased risk of preterm birth and SGA in this specific group. The adjusted OR for preterm birth was 1.98 (95% CI: 1.09–3.62) when women with endometriomas were compared with fertile subjects. The authors however failed to confirm this risk when using women with other forms of endometriosis or causes of infertility as controls. Similarly, they documented and adjusted OR for SGA of 1.95 (95% CI: 1.06–3.60) when comparing women with endometriomas to those who underwent ART for non-endometriosis causes of infertility. Again, however, this enhanced risk was not confirmed when using women with other forms of endometriosis or subfertile subjects as controls (Fernando et al., 2009).
Results of our study do not support the notion that women with ovarian endometriomas achieving pregnancy through IVF are at increased risk of preterm birth or SGA. There is even a trend to a lower risk. The adjusted ORs were 0.47 (0.14–1.54) and 0.56 (0.12–2.56), respectively. We also failed to observe any association with other obstetrical complications. Moreover, the live birth rate was not reduced in women with ovarian endometriomas. The adjusted OR of live birth in affected cases was 0.79 (95% CI: 0.38–1.68). The differing data in these studies may be explained by the differences in study designs. Fernando et al. used national IVF and perinatal databases to identify 95 women with ovarian endometriomas who got pregnant through IVF. The study power was thus similar to the one of our study but the accuracy of collected information was inevitably lower since we retrieved data with an active follow-up and by consulting patients' charts whereas they relied on databases. For instance, Fernando et al. could not provide information on the risk of pregnancy loss since this outcome was not included in the perinatal database that they used for their analysis. Moreover, the choice of the group for comparison may influence the results. In our study, we used women without endometriosis achieving pregnancy through IVF as comparators. This choice is, in our view, more appropriate to investigate the impact of ovarian endometriomas since infertile women may be per se at higher risk of obstetrical complications (Wisborg et al., 2010; Kalra and Barnhart, 2011). Conversely, Fernando et al. used several control groups. In fact, when specifically focusing on the comparison with the ART non-endometriosis group, they detected a statistical significant difference only for the risk of preterm birth. On the other hand, it has to be recognized that our study design may have lead to underestimation of some associations because some women in the unexposed group may be affected by some undiagnosed forms of the disease (adhesions or peritoneal implants). It has however to be pointed out that the present study is mainly aimed at giving a clinical response. In particular, we wanted to disentangle whether endometrioma had to be removed because of obstetrics concerns. Moreover, the magnitude of this inaccuracy is presumably mild since only a minority of women belonging to the unexposed group is expected to hide endometriosis.
Some further limitations of this study have to be recognized.
Firstly, the sample size was relatively small impeding definite conclusions. Even if the estimate of ORs for preterm birth and SGA is <1, the 95% CI is large. This limit is of even more relevant for other major obstetrical complications such as pre-eclampsia, placenta previa, abruptio placenta and fetal growth abnormalities. Further evidence is therefore required. Nevertheless, given that our aim was to provide information for clinical practice we were mainly interested on the association of relevant magnitude. Considering the risk and costs associated with surgery for ovarian endometriomas, only relevant concerns regarding the obstetric outcome would tip the balance in favor of the intervention.
Secondly, the dimension of ovarian endometriomas was relatively small (the mean diameter was 22 mm). We therefore cannot rule out that larger cysts may be detrimental. Further evidence is also required to address this point. Nonetheless, we believe that our results are of clinical relevance since they reflect what physicians face daily in clinical practice.
In conclusion, women with endometriomas achieving pregnancy through IVF do not seem to be exposed to a significant increased risk of obstetrical complications. Fear about this possibility is not substantiated and thus removal of the endometriomas to reduce pregnancy complications is not justified. However, considering that previous publication on this topic reports opposing results, our data must be considered preliminary, and further larger evidence is required to definitively answer the research question and rule out any detrimental effects in women with endometriomas achieving pregnancy through IVF.
Discussion
Eutopic endometrium of women with endometriosis was reported to be abnormal (Braun and Dmowski, 1998; Pritts and Taylor, 2003) and these abnormalities may impair decidualization and placentation in women affected (Fujimoto et al., 1996). Since these latest processes are crucial for pregnancy implantation and development, it has been hypothesized that several complications of pregnancy may be more frequent in women with endometriosis (Fernando et al., 2009). Available clinical data are not fully consistent with this hypothesis. The literature on the relationship between endometriosis and abortion rate is conflicting, with some studies reporting increased risk (Fitzsimmons et al., 1987; Matorras et al.,1998; Omland et al., 2005) and others failing to document any association (Pittaway et al., 1988; Candiani et al., 1991). Stephansson et al. recently evaluated the risk for obstetrics complications in a large cohort of women with a previous diagnosis of endometriosis. They showed an increased risk of preterm birth, pre-eclampsia, antepartal bleeding/placental complications and Cesarean section. There was conversely no association with small-for-gestational-age (SGA) birth or stillbirth (Stephansson et al., 2009). In contrast, Brosens et al. (2007) reported that the rate of pre-eclampsia is reduced in women with endometriosis. Moreover, Fernando et al. (2009) failed to document an association between endometriosis in general and SGA and preterm birth. This latter study is particularly illuminating in regard of the topic of the present study. Indeed, it is the first report presenting data separately for the group of women with ovarian endometriomas. Of relevance here is that the study suggested an increased risk of preterm birth and SGA in this specific group. The adjusted OR for preterm birth was 1.98 (95% CI: 1.09–3.62) when women with endometriomas were compared with fertile subjects. The authors however failed to confirm this risk when using women with other forms of endometriosis or causes of infertility as controls. Similarly, they documented and adjusted OR for SGA of 1.95 (95% CI: 1.06–3.60) when comparing women with endometriomas to those who underwent ART for non-endometriosis causes of infertility. Again, however, this enhanced risk was not confirmed when using women with other forms of endometriosis or subfertile subjects as controls (Fernando et al., 2009).
Results of our study do not support the notion that women with ovarian endometriomas achieving pregnancy through IVF are at increased risk of preterm birth or SGA. There is even a trend to a lower risk. The adjusted ORs were 0.47 (0.14–1.54) and 0.56 (0.12–2.56), respectively. We also failed to observe any association with other obstetrical complications. Moreover, the live birth rate was not reduced in women with ovarian endometriomas. The adjusted OR of live birth in affected cases was 0.79 (95% CI: 0.38–1.68). The differing data in these studies may be explained by the differences in study designs. Fernando et al. used national IVF and perinatal databases to identify 95 women with ovarian endometriomas who got pregnant through IVF. The study power was thus similar to the one of our study but the accuracy of collected information was inevitably lower since we retrieved data with an active follow-up and by consulting patients' charts whereas they relied on databases. For instance, Fernando et al. could not provide information on the risk of pregnancy loss since this outcome was not included in the perinatal database that they used for their analysis. Moreover, the choice of the group for comparison may influence the results. In our study, we used women without endometriosis achieving pregnancy through IVF as comparators. This choice is, in our view, more appropriate to investigate the impact of ovarian endometriomas since infertile women may be per se at higher risk of obstetrical complications (Wisborg et al., 2010; Kalra and Barnhart, 2011). Conversely, Fernando et al. used several control groups. In fact, when specifically focusing on the comparison with the ART non-endometriosis group, they detected a statistical significant difference only for the risk of preterm birth. On the other hand, it has to be recognized that our study design may have lead to underestimation of some associations because some women in the unexposed group may be affected by some undiagnosed forms of the disease (adhesions or peritoneal implants). It has however to be pointed out that the present study is mainly aimed at giving a clinical response. In particular, we wanted to disentangle whether endometrioma had to be removed because of obstetrics concerns. Moreover, the magnitude of this inaccuracy is presumably mild since only a minority of women belonging to the unexposed group is expected to hide endometriosis.
Some further limitations of this study have to be recognized.
Firstly, the sample size was relatively small impeding definite conclusions. Even if the estimate of ORs for preterm birth and SGA is <1, the 95% CI is large. This limit is of even more relevant for other major obstetrical complications such as pre-eclampsia, placenta previa, abruptio placenta and fetal growth abnormalities. Further evidence is therefore required. Nevertheless, given that our aim was to provide information for clinical practice we were mainly interested on the association of relevant magnitude. Considering the risk and costs associated with surgery for ovarian endometriomas, only relevant concerns regarding the obstetric outcome would tip the balance in favor of the intervention.
Secondly, the dimension of ovarian endometriomas was relatively small (the mean diameter was 22 mm). We therefore cannot rule out that larger cysts may be detrimental. Further evidence is also required to address this point. Nonetheless, we believe that our results are of clinical relevance since they reflect what physicians face daily in clinical practice.
In conclusion, women with endometriomas achieving pregnancy through IVF do not seem to be exposed to a significant increased risk of obstetrical complications. Fear about this possibility is not substantiated and thus removal of the endometriomas to reduce pregnancy complications is not justified. However, considering that previous publication on this topic reports opposing results, our data must be considered preliminary, and further larger evidence is required to definitively answer the research question and rule out any detrimental effects in women with endometriomas achieving pregnancy through IVF.