- Email: [email protected]
Prediction of Endometriosis Fertility Index in patients with endometriosis-associated infertility after laparoscopic treatment
Accepted Manuscript Title: Prediction of endometriosis fertility index in patients with endometriosisassociated infertility after laparoscopic treatment Author: Xianghui Zhang, Dong Liu, Wei Huang, Qiushi Wang, Xue Feng, Jing Tan PII: DOI: Reference:
S1472-6483(18)30108-1 https://doi.org/10.1016/j.rbmo.2018.03.012 RBMO 1918
To appear in:
Reproductive BioMedicine Online
Received date: Revised date: Accepted date:
14-5-2017 12-3-2018 14-3-2018
Please cite this article as: Xianghui Zhang, Dong Liu, Wei Huang, Qiushi Wang, Xue Feng, Jing Tan, Prediction of endometriosis fertility index in patients with endometriosis-associated infertility after laparoscopic treatment, Reproductive BioMedicine Online (2018), https://doi.org/10.1016/j.rbmo.2018.03.012. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Short title: Endometriosis Fertility Index endometriosis-associated infertility Prediction of Endometriosis Fertility Index in patients with endometriosisassociated infertility after laparoscopic treatment Xianghui Zhang,1a Dong Liu,1a Wei Huang,a* Qiushi Wang,a Xue Feng,a Jing Tana,1 a
Department of Reproductive Endocrinology, West China Second University Hospital, Sichuan University, Chengdu 610041, PR China *Corresponding author. E-mail address: [email protected] (W. Huang).
1
Co-first author.
Key message This retrospective study confirmed the predictive value of the Endometriosis Fertility Index (EFI) developed by Adamson ad Pasta (2010) among 1097 infertile Chinese women with endometriosis-associated infertility. Time to pregnancy was inversely related to the EFI score.
Page 1 of 22
Author biography Dr Wei Huang received her PhD in 1992 from West China Medical University, Chengdu Sichuan, China, and has worked as full Professor of Obstetrics and Gynaecology. Her research focus is human decidual stem cells, mroRNA, intrauterine devices, proliferation interaction at the maternal–fetal interface, endometriosis and female infertility.
Abstract The effect of endometriosis on fecundity is unclear. Although surgery plays a vital role in the treatment of related infertility, pregnancy outcomes after laparoscopy are poorly correlated with the currently used staging system. To address this, the Endometriosis Fertility Index (EFI) was developed. This retrospective study was designed to assess the predictive value of the EFI for patients with endometriosisassociated infertility. A total of 1325 patients with endometriosis-associated infertility were eligible for inclusion, 1097 of whom were successfully interviewed and 228 lost
Page 2 of 22
to follow-up (17.21%). Cumulative pregnancy incidence (CPI) according to EFI scores were calculated by life table and by Kaplan–Meier survival curve analyses. The log-rank test was used to evaluate difference between the EFI groups. Receiver operating characteristic curves were plotted to obtain the optimal cut-off point for pregnancy prediction. A total of 505 (46.03%) patients conceived naturally. The difference in CPI among EFI scores 10, 7–9, 4–6, and 2–3 was statistically significant (P < 0.001) and increased with increasing EFI score. The optimal cut-off point was 7.5: sensitivity (68.51%); specificity (52.20%). This retrospective study indicates the value of the EFI score for predicting spontaneous pregnancy in women treated by laparoscopy for endometriosis-associated infertility.
KEYWORDS: endometriosis, endometriosis Fertility Index, fecundity, infertility, laparoscopy, spontaneous pregnancy
Introduction Endometriosis is a chronic and diverse disease characterized by the presence of endometrial-like tissue outside the uterine cavity; it afflicts 35–50% women of reproductive age by causing related infertility, chronic pelvic pain and other physical problems (Giudice and Kao, 2004). Although many researchers have focused on the causal links between endometriosis and infertility, a definitive answer is not yet available (Bulletti et al., 2010). It is, however, clear that the fecundity rebounds to some degree after medical management. Among the various measures for treating endometriosis, laparoscopic surgery is widely used for treating endometriosisassociated infertility for both mild to moderate and severe patients (Marcoux et al., 1997; Jacobson et al., 2010; Shimizu et al., 2010). Moreover, it is the only approach to clarify the diagnosis by the visible presence of typical lesions and histological confirmation. Apart from the high cost, which is beyond the scope of medical insurance, the average outcomes of assisted reproductive technology (ART) for endometriosis-related infertile patients drive most couples in China to seek alternatives to surgery initially, and then try to conceive naturally.
Page 3 of 22
Pregnancy outcome after laparoscopy, however, is not consistent when using the most widely used staging system of endometriosis, namely the revised American Fertility Society (rAFS) score (Vercellini et al., 2006; Zeng et al., 2014); this drastically lowers the clinical value of rAFS in prognosis forecasts. To resolve this issue, the Endometriosis Fertility Index (EFI), which predicts spontaneous conception after surgery, was published in 2010 (Adamson and Pasta, 2010). The key elements of this scoring system include historical factors at the time of surgery, adnexal function after the intervention of surgery and assessment of endometriosis that was derived from rAFS. This index has been validated externally in endometriosis-associated infertile patients in Belgium (Tomassetti et al., 2013), China (Zeng et al., 2014), France (Boujenah et al., 2015), and Italy (Garavaglia et al., 2015). To demonstrate the value of the EFI score in predicting spontaneous pregnancy, and to determine an optimal cut-off point for positive outcome prediction, a larger sample size of Chinese women with endometriosis-associated infertility was included in this study.
Materials and methods Patients and inclusion criteria The medical Ethics Committee of West China Second University Hospital of Sichuan University indicated on 1 May 2017 that ethical approval was not needed owing to the retrospective nature of our study. Patients all provided informed consent for the use of their data, according to Ethics Committee requirements. Our study involved women with endometriosis-associated infertility who were undergoing laparoscopic treatment from January 2008 to December 2012 at the Department of Obstetrics and Gynecology in West China Second University Hospital of Sichuan University. The hospital represents a tertiary referral centre for endometriosis and reproductive surgery.
Women with endometriosis who were ovulating normally, had at least one patent tube according to hysterosalpingography, and had not conceived over a 12-month period
Page 4 of 22
were included in the study. Diagnosis was based on visualization of typical lesions, i.e., red, white and black implants in the peritoneal, ovarian and other organs, histopathological diagnosis for ovarian endometrioma and deep infiltrative lesions (DIE), or both. Deep infiltrative lesions are defined as lesions extending deeper than 5 mm under the peritoneal surface or those involving bowel, bladder, ureter or other sites. Diagnosis was based on criteria that included patients’ symptoms, careful physical examination, ultrasound, MRI, or both.
During physical examination, special attention was paid to the adnexal mass, nodules in retro-vaginal wall, sacrouterine ligaments, pelvic wall and the rectum during vaginal and rectal palpation. Additional criteria were normal sexual function and normal results for male partners’ semen analysis according to World Health Organization criteria (WHO, 2010). Exclusion criteria were as follows: ovulation disorders, such as polycystic ovary syndrome, luteinized unruptured follicle syndrome, hyperprolactinaemia and congenital defects, such as uterus septum, uterine leiomyoma, intrauterine adhesion and severe systemic disease.
Surgical procedure and EFI score All patients underwent standard surgery after conventional preoperative preparation. After induction of pneumoperitoneum of 14 mmHg, a thorough inspection was made to establish the presence of endometriotic lesions on the oviduct, ovary and pelvic peritoneum (especially uterosacral ligament), and the extent of pelvic adhesions. All visible peritoneal implants were excised or coagulated; normal anatomic structure of the pelvic organs was restored as much as possible by adhesiolysis. For endometrioma, the cyst was stripped carefully to avoid unnecessary damage to the healthy ovarian tissue. For patients with DIE, nodules were resected thoroughly. The presence of endometriotic lesions and adhesions was scored and staged according to the rAFS classification system of the American Society for Reproductive Medicine (ASRM, 1997). Before completion of surgery, the tubal patency test was carried out using methylene blue diluent.
Page 5 of 22
The following historical data were collected: patient age, duration of infertility and prior history of pregnancy. Total historical factor scores were then calculated (Figure 1). Surgery factors included assessment of endometriosis, which was derived from rAFS (endometriosis lesions score and total score), and adnexal function (least function). Least function score was based on the postoperative function of tube, fimbriate and ovary. The EFI scores were finally obtained after combining historical and surgical scores. For patients with advanced age（>40 years）or compromised tubal function, ART was proposed as the treatment. A recommendation was made to the other patients to attempt to conceive naturally for 12 months. Subsequently, patients who failed conception underwent ART.
Follow-up Between March 2013 and December 2015, each patient was contacted, either by letter or telephone interview to follow up postoperative outcomes. A spontaneous pregnancy was defined by a beta-HCG level over 25 UI/l. All patients were followed up for 36 months. For those who became pregnant during that interval, the follow-up period was extended to delivery, and the pregnancy outcomes were included in the data analysis. Women with a negative serum pregnancy test were not considered to be pregnant at the end of the follow-up period. Clinical data and information on the postoperative condition of all patients were collected and analysed.
Statistical analysis SPSS statistical software 21.0 (IBM Corp., USA) was used to assess relevant EFI score, rAFS classification, endometriosis type and pregnancy outcome after surgery. Data were presented as mean ± SE; chi-squared-test, life table and Kaplan–Meier survival analysis were used for comparing pregnancy rate and cumulative pregnancy incidence (CPI) at different time points. Patients were censored when they were lost
Page 6 of 22
to follow-up or underwent ART. Receiver operator characteristic curves were plotted to determine an optimal cut-off point of the EFI score for pregnancy prediction. For all analyses, P < 0.05 was considered as statistically significant.
Results Between March 2013 and December 2015, 1325 women were identified as eligible for study inclusion. A total of 228 (17.21%) women who were lost to follow-up were excluded; therefore, 1097 patients were eligible for EFI validation analysis. Among 1097 patients, 181 women had a complication of adenomyosis. The EFI score was calculated as previously reported, with EFI scores ranging from 0 for the poorest prognosis to 10 for the best prognosis. The median EFI score was 8. The distribution of patients according to their EFI score is presented in Figure 2. The most common EFI scores were 8 (238) and 9 (263). Baseline characteristics of infertile women related to age at operation, body mass index, duration of infertility, primary infertility, and r-AFS classification (Table 1). Laparoscopic surgery was carried out with no complications. Median operating time was 45 min (15–175). Patients with high rAFS stage and DIE underwent longer and more difficult operations. Most (27/31) DIE lesions were located within the uterosacral ligament or rectovaginal septum; others were located on the bowel (2) and bladder (2). The lesions were resected thoroughly without cystectomy or rectectomy. The median operating time of these patients was 78 min (55–175). The duration of follow-up was 14–36 months (median 26.34). After surgery, 176 patients used oral contraceptives for 1–3 months; 330 used gonadotrophin-releasing hormone agonists for 2–6 months; and 71 women took gestrinone for 2–3 months; all patients followed clinical advice according to their intraoperative condition. A total of 202 patients tried traditional Chinese medicine autonomously. Fifty women were advised to start ART, but none of them agreed to treatment in the first year after surgery. Eighty-three patients underwent ART after 12 months and 51 of them became pregnant (six by IUI and 45 by IVF and embryo transfer). A total of 505 (46.03%) women became pregnant naturally. Cumulative spontaneous clinical pregnancy rates calculated using the life table method for all
Page 7 of 22
patients was 42.57% ± 0.02%, 45.98% ± 0.02%, and 47.79% ± 0.02% at 12, 24, and 36 months, respectively. Among the 505 women who conceived naturally, 62 (12.28%) suffered from miscarriage and 24 (4.75%) patients had an ectopic pregnancy. The overall live birth rate was 38.20% (419/1097). Two neonatal deaths occurred among 34 premature babies (gestational age range 28+6 weeks to 36+4 weeks). Median time between surgery and conception was 7 months, with a range of 1 to 36 months.
The life table analysis stratified the EFI into five classes (Table 2). A significant relationship was found between EFI and time to achieving pregnancy (P < 0.001) (Table 2); however, no association was found between endometriosis type and pregnancy rate. The pregnancy rate of rAFS classifications I to Ⅳ was 50.57%, 45.58%, 47.33% and 37.00%, respectively. Paired comparison by chi-squared test showed no difference between stages Ⅱ and Ⅲ. The pregnancy rate of patients with stage I was significantly higher than that of patients with stages Ⅱto Ⅳ (P < 0.001). Most pregnancies were achieved within 2 years of surgery. After 36 months of followup, the pregnancy rate of patients was compared by chi-squared test. The pregnancy rate of patients with EFI 10 (70.31%) was significantly higher than that of patients with EFI 9 (52.47%) (P < 0.001). The aggregated pregnancy rate was 51.10% among patients with EFI 8–9; the pregnancy rates for EFI scores 8 and 9 (49.58% and 52.47% respectively) were not significantly different. Similar results were obtained for the pregnancy rate among patients with EFI 6 and 7 (33.33%, 43.87%). The aggregated pregnancy rate (39.94%) of patients with EFI 6–7 was lower than that of patients with EFI 8–9 (P = 0.001). No difference was found among patients with EFI score 2–5 (14.29%, 8.33%, 20.93%, 19.12%). The number of patients with EFI 2–3 was underrepresented (n = 19), and only two pregnancies were achieved among them. The Kaplan–Meier survival analysis result for CPI at different time points among EFI 10, 7–9, 4–6 and 2–3 was statistically significant (P < 0.001) (Figure 3). As expected, the CPI did not correspond to r-AFS staging (I-IV), and no significant difference was found in CPI between different types of endometriosis (Table 2). As shown in Figure 4, the area under the curve for EFI as pregnancy predictor was 0.648 (95% CI 0.616
Page 8 of 22
to 0.680; P < 0.001). The best cut-off point was 7.5 (sensitivity, 68.51%, specificity, 52.20%), whereas Youden index was 0.207 (Youden index = sensitivity + specificity – 1).
Discussion To date, laparoscopic surgery has been the first-line treatment for most endometriosisrelated infertile patients (Marcoux et al., 1997; Jacobson et al., 2010). Medication has proved to be effective only for pain management and for reducing recurrence when used individually or when combined with laparoscopy (Olive and Pritts, 2001). The rAFS classification for description of visible lesions and adhesions in endometriosis has become a widely accepted staging system. It has, however, failed to predict postoperative pregnancy (Vercellini et al., 2006; Zeng et al., 2014). This may be due to the complicated influence of endometriosis on reproduction; moreover, visible lesions did not represent the severity of endometriosis. Laparoscopic surgery cannot resolve molecular, immune or other unknown factors that influence fertility; this makes it more difficult to develop a feasible and effective staging system. In this scenario, the EFI scoring system emerged as a useful method (Adamson and Pasta, 2010). As shown in Figure 2, we validated the EFI again by demonstrating different scores and corresponding fecundity rates. Apparently, the pregnancy rate is proportional to the increased score of the EFI. By considering historical factors, the EFI staging system provides a comprehensive evaluation of fecundity for endometriosis-related infertility after laparoscopic surgery for the first time. Adamson and Pasta (2010) used a different approach and considered historical factors that reflect the severity of the disease to some extent. They also included least function of bilateral adnexa after surgery, which plays a crucial role in natural conception. Endometriosis lesions occupied little weighting (2/10) in this system.
Although the EFI staging system ignored the assessment of uterine factors, adenomyosis, deep infiltrating disease and ovarian reserve, it still has vital clinical significance. Indeed, the probability of spontaneous conception seemed remote in
Page 9 of 22
patients who scored 2–3, albeit a small number. In fact, gynaecologists are quite capable of distinguishing such women with an intuitive definition of ‘severity’, such as tubal obstruction after surgery, advanced female age, or bot; these patients should be recommended ART immediately. In our opinion, the higher the EFI score, the higher the probability that a spontaneous pregnancy will be achieved; this has been confirmed in previous studies (Boujenah et al., 2015; Garavaglia et al., 2015). The CPI of 12 months and 24 months were in accordance with results reported by Tomassetti et al. (2013).
Natural conception is the first choice for women with EFI score 6–10, whereas, for patients with EFI score 4–5, natural conception or ART should be offered to infertile couples according to their pregnancy desire and financial situation. In particular, patients with an EFI score of 0–3 should be informed about the low possibility of spontaneous pregnancy, and ART should be strongly encouraged. In addition, we found that most pregnancies were achieved within 2 years of surgery; in this period, patients may benefit from the favourable effect of surgery on the anatomical factors altered by endometriosis. Therefore, ART should be recommended to patients who have failed to conceive naturally 2 years after laparoscopy, including those patients with a high EFI score. The area under the ROC curve demonstrated a low prediction accuracy. This may be because of the small number of patients with a score of 2–3. A pregnancy rate of 10.53% (2/19) in these patients may not represent the true picture. A similar pregnancy rate for patients with EFI score 4–5 was observed. In contrast to other studies (Tomassetti et al., 2013; Boujenah et al., 2015), all our patients chose natural conception first after surgery. This was probably due to economic restriction and societal perceptions in our country. For most patients, ART was expensive; therefore, this method has been treated with prejudice for some time. In view of this, we think our research is more representative, as patients with low EFI scores in developed countries chose ART directly (Garavaglia et al., 2015). Therefore, we can observe the real performance of the EFI score in predicting pregnancy in patients with low scores. In 19 patients with a EFI score of 0–3, only two conceived naturally after
Page 10 of 22
36 months. It is difficult to control for bias, however, owing to the small sample size in this group. Our study also indicates that the EFI score is a good staging system for pregnancy prediction for infertile women with endometriosis.
In China, many women prefer to try traditional Chinese medicine (TCM) orally after surgery for spontaneous pregnancy. They believe that it helps to increase their chances of conception. Some studies have confirmed the effectiveness of TCM in treating infertility (Flower et al., 2012). Considering the specificity of TCM, it is extremely difficult to judge the efficacy on fecundity (Guo et al., 2010). Therefore, it is unclear whether our results were influenced by this complicated treatment to some degree.
A few patients in our study accepted ovulation induction (81/1097), among which 31 became pregnant. This is distinct from research by Adamson and Pasta (2010); and Tomassetti et al. (2013), in which ovulation induction or ovarian stimulation combined with IUI was carried out generally after failure of spontaneous conception within 1 year of surgery. Nevertheless, the benefit of ovulation induction or ovarian stimulation combined with IUI after a certain time is also highly contentious (Gandhi et al., 2014). Therefore, we are unable to reach a definite conclusion at present.
The prevalence of DIE in our study is significantly lower than the 12–20% reported in other studies (Chapron et al., 1999; Lagana et al., 2016; Iversen et al., 2017). Several reasons may contribute to this difference. First, different ethnicity of the study populations may be a factor, as data from the Peking Union Medical College Hospital of China also revealed a low incidence of DIE (5.80%, 615/10597) in their 10-year experience (Li et al., 2011) in Chinese women, although this is higher than in the present study. Second, subject selection criteria differ from other studies, as patients in the present study were included primarily on the basis of their history of infertility caused by endometriosis, not chronic pelvic pain, dysmenorrhoea or pelvic mass, although some of these symptoms may also have been present in the selected patients. Third, some cases of DIE could have been missed owing to the lack of experience of
Page 11 of 22
doctors looking for the lesion during surgery, therefore under-representing the true incidence.
The present study has some limitations. It is retrospective in design, although the data necessary for calculation of the EFI were collected prospectively, as with the study by Tomassetti et al. (2013). Confounding factors included different surgeons and different post-surgical medications. In conclusion, despite these limitations, however, the EFI score seems to be relevant for the prediction of the natural conception rate in infertile women with endometriosis treated with laparoscopic surgery. This will be useful in counselling patients of reproductive age.
Acknowledgements We thank Dr Lu Zhou and Dr Licong Shen for their help in this study. We also thank Mr Ruibo Zhang for proofreading this article. Wei Huang designed the study; Dong Liu, Jing Tan, Xue Feng and QiuShi Wang collected reviews of medical history and surgical records; Xianghui Zhang and Dong Liu were responsible for follow up and data analysis; Xianghui Zhang and Dong Liu wrote the paper.
References Adamson, G.D., Pasta D.J. 2010. Endometriosis fertility index: the new, validated endometriosis staging system. Fertil Steril 94, 1609-15. ASRM. 1997. Revised American Society for Reproductive Medicine classification of endometriosis: 1996. Fertil and Steril 67, 817-821. Boujenah, J., Bonneau, C., Hugues, J. N., Sifer, C. and Poncelet, C. 2015. External validation of the Endometriosis Fertility Index in a French population. Fertil Steril 104, 119-23. Bulletti, C., Coccia, M. E., Battistoni, S. and Borini, A. 2010. Endometriosis and infertility. J Assist Reprod Genet 27, 441-7. Chapron, C., Fritel, X. and Dubuisson, J. B. 1999. Fertility after laparoscopic management of deep endometriosis infiltrating the uterosacral ligaments. Hum
Page 12 of 22
Reprod 14, 329-32. Flower, A., Liu, J. P., Lewith, G., Little, P. and Li, Q. 2012. Chinese herbal medicine for endometriosis. Cochrane Database Syst Rev, CD006568. Gandhi, A. R., Carvalho, L. F., Nutter, B. and Falcone, T. 2014. Determining the fertility benefit of controlled ovarian hyperstimulation with intrauterine insemination after operative laparoscopy in patients with endometriosis. J Minim Invasive Gynecol 21, 101-8. Garavaglia, E., Pagliardini, L., Tandoi, I., Sigismondi, C., Vigano, P., Ferrari, S. and Candiani, M. 2015. External validation of the endometriosis fertility index (EFI) for predicting spontaneous pregnancy after surgery: further considerations on its validity. Gynecol Obstet Invest 79, 113-8. Giudice, L. C. and Kao, L. C. 2004. Endometriosis. Lancet 364, 1789-99. Guo, S., He, W., Zhao, T., Liu, X. and Zhang, T. 2010. Clinical trials and trial-like studies on the use of traditional Chinese medicine to treat endometriosis. Obstet Gynecol 5, 533-555. Iversen, M. L., Seyer-Hansen, M. and Forman, A. 2017. Does surgery for deep infiltrating bowel endometriosis improve fertility? A systematic review. Acta Obstet Gynecol Scand 96, 688-693. Jacobson, T. Z., Duffy, J. M., Barlow, D., Farquhar, C., Koninckx, P. R. and Olive, D. 2010. Laparoscopic surgery for subfertility associated with endometriosis. Cochrane Database Syst Rev, CD001398. Lagana, A. S., Vitale, S. G., Trovato, M. A., Palmara, V. I., Rapisarda, A. M., Granese, R., Sturlese, E., De Dominici, R., Alecci, S., Padula, F., Chiofalo, B., Grasso, R., Cignini, P., D'Amico, P. and Triolo, O. 2016. Full-Thickness Excision versus Shaving by Laparoscopy for Intestinal Deep Infiltrating Endometriosis: Rationale and Potential Treatment Options. Biomed Res Int 2016, 3617179. Li, L., Leng, J. H., Dai, Y., Zhang, Y., Li, X. Y. and Lang, J. H. 2011. Characteristics and trends of surgical management of pelvic endometriosis in different ages. Chin J Pract Gynecol Obstet, 2011. 27(3): p. 188-191. Marcoux, S., Maheux, R. and Berube, S. 1997. Laparoscopic surgery in infertile
Page 13 of 22
women with minimal or mild endometriosis. Canadian Collaborative Group on Endometriosis. N Engl J Med 337, 217-22. Olive, D. L. and Pritts, E. A. 2001. Treatment of endometriosis. N Engl J Med 345, 266-75. Shimizu, Y., Takashima, A., Takahashi, K., Kita, N., Fujiwara, M. and Murakami, T. 2010. Long-term outcome, including pregnancy rate, recurrence rate and ovarian reserve, after laparoscopic laser ablation surgery in infertile women with endometrioma. J Obstet Gynaecol Res 36, 115-8. Tomassetti, C., Geysenbergh, B., Meuleman, C., Timmerman, D., Fieuws, S. and D'Hooghe, T. 2013. External validation of the endometriosis fertility index (EFI) staging system for predicting non-ART pregnancy after endometriosis surgery. Hum Reprod 28, 1280-8. Vercellini, P., Fedele, L., Aimi, G., De Giorgi, O., Consonni, D. and Crosignani, P. G. 2006. Reproductive performance, pain recurrence and disease relapse after conservative surgical treatment for endometriosis: the predictive value of the current classification system. Hum Reprod 21, 2679-85. World Health Organization. WHO laboratory Manual for the Examination Processing
of
Human
Semen
Fifth
edition
.Geneva
:
World
and
Health
Organization ,2010.100 Zeng, C., Xu, J. N., Zhou, Y., Zhou, Y. F., Zhu, S. N. and Xue, Q. 2014. Reproductive performance after surgery for endometriosis: predictive value of the revised American Fertility Society classification and the endometriosis fertility index. Gynecol Obstet Invest 77, 180-5.
Declaration The authors report no financial or commercial conflicts of interest.
Page 14 of 22
Figure legends Figure 1 – Endometriosis Fertility index (EFI) surgery form (Tomassetti et al., 2013, reproduced with permission).
Figure 2 – Distribution of patients according to their Endometriosis Fertility index (EFI) score. The median EFI score was 8.
Figure 3 –
Kaplan–Meier survival analysis for cumulative pregnancy incidence
according to different Endometriosis Fertility index (EFI) score groups. The Kaplan– Meier survival analysis result for cumulative pregnancy incidence at different time points among EFI 10, 7–9, 4–6 and 2–3 was statistically significant (P < 0.001).
Figure 4 – receiver operator characteristic curve. The area under the curve for EFI as pregnancy predictor was 0.648 (95% CI 0.616 to 0.680, P < 0.001). The best cut-off point was 7.5 (sensitivity, 8.51%, specificity, 52.20%), Youden index was 0.207 (Youden index = sensitivity + specificity –1).
Page 15 of 22
Table 1– Baseline characteristics of infertile women. Mean ± SD, mean ± SD (range) or n (%) Age at operation (years) ≤35 years 36–39 years ≥40 years Body mass index Duration of infertility (years) ≤3 years >3 years Primary infertility, n (%) Secondary infertility, n (%) Least function score High score (7–8) Moderate score (4–6) Low score (1–3) AFS endometriosis lesion score <16 ≥16 AFS total score <71 ≥71
29.84 4.29 (20–46) 989 (90.15) 87 (7.93) 21(1.91) 20.34 2.57 (13.67–33.73) 3.02 2.64 (1–18) 794 (72.38) 303 (27.62) 537 (48.95) 560 (51.05) 385 (35.10) 574 (52.32) 138 (12.58) 894(81.49) 203 (18.51) 1008 (91.89) 89 (8.11)
AFS, American Fertility Society.
Page 16 of 22
Table 2 – Life table for cumulative pregnancy incidence according to different parameters over different time periods. Patients (n) Parameter
Cumulative pregnancy rate（%） 1 year 2 years 3 years
P-value
rAFS I
352
48.18 ± 2.68
51.01 ± 2.70
51.64 ± 2.74
II
283
41.23 ± 2.96
45.63 ± 3.02
48.03 ± 3.19
III
262
43.16 ±3.08
46.50 ± 3.12
49.76 ± 3.33
IV
200
33.76 ± 3.36
36.98 ± 3.44
38.07 ± 3.55
Peritoneal
693
43.32±1.90
47.44 ± 1.92
48.08 ± 1.95
Ovarian
373
40.44±2.56
42.82 ± 2.60
46.93 ± 2.84
DIE
31
51.61±8.98
51.61 ± 8.98
51.61 ± 8.98
2–3
19
10.53 ± 7.04
10.53 ± 7.04
10.53 ± 7.04
4–5
111
15.52 ± 3.46
19.41 ± 3.81
21.37 ± 4.19
6–7
338
36.32 ± 2.64
38.96 ± 2.69
42.85 ± 2.95
8–9
501
48.04 ± 2.25
51.59 ± 2.27
52.49 ±2.31
10
128
65.90 ± 4.21
70.89 ± 4.06
70.89 ± 4.06
<0.001
Pathology type NS
EFI <0.001
DIE, deep infiltrating endometriosis; EFI, Endometriosis Fertility Index; NS, non significant; rAFS, revised American Fertility Society.
Page 17 of 22
Figure 1.
Page 18 of 22
Figure 2.
Page 19 of 22
Figure 3.
Page 20 of 22
Figure 4.
Page 21 of 22
photo.png
Page 22 of 22
S1472-6483(18)30108-1 https://doi.org/10.1016/j.rbmo.2018.03.012 RBMO 1918
To appear in:
Reproductive BioMedicine Online
Received date: Revised date: Accepted date:
14-5-2017 12-3-2018 14-3-2018
Please cite this article as: Xianghui Zhang, Dong Liu, Wei Huang, Qiushi Wang, Xue Feng, Jing Tan, Prediction of endometriosis fertility index in patients with endometriosis-associated infertility after laparoscopic treatment, Reproductive BioMedicine Online (2018), https://doi.org/10.1016/j.rbmo.2018.03.012. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Short title: Endometriosis Fertility Index endometriosis-associated infertility Prediction of Endometriosis Fertility Index in patients with endometriosisassociated infertility after laparoscopic treatment Xianghui Zhang,1a Dong Liu,1a Wei Huang,a* Qiushi Wang,a Xue Feng,a Jing Tana,1 a
Department of Reproductive Endocrinology, West China Second University Hospital, Sichuan University, Chengdu 610041, PR China *Corresponding author. E-mail address: [email protected] (W. Huang).
1
Co-first author.
Key message This retrospective study confirmed the predictive value of the Endometriosis Fertility Index (EFI) developed by Adamson ad Pasta (2010) among 1097 infertile Chinese women with endometriosis-associated infertility. Time to pregnancy was inversely related to the EFI score.
Page 1 of 22
Author biography Dr Wei Huang received her PhD in 1992 from West China Medical University, Chengdu Sichuan, China, and has worked as full Professor of Obstetrics and Gynaecology. Her research focus is human decidual stem cells, mroRNA, intrauterine devices, proliferation interaction at the maternal–fetal interface, endometriosis and female infertility.
Abstract The effect of endometriosis on fecundity is unclear. Although surgery plays a vital role in the treatment of related infertility, pregnancy outcomes after laparoscopy are poorly correlated with the currently used staging system. To address this, the Endometriosis Fertility Index (EFI) was developed. This retrospective study was designed to assess the predictive value of the EFI for patients with endometriosisassociated infertility. A total of 1325 patients with endometriosis-associated infertility were eligible for inclusion, 1097 of whom were successfully interviewed and 228 lost
Page 2 of 22
to follow-up (17.21%). Cumulative pregnancy incidence (CPI) according to EFI scores were calculated by life table and by Kaplan–Meier survival curve analyses. The log-rank test was used to evaluate difference between the EFI groups. Receiver operating characteristic curves were plotted to obtain the optimal cut-off point for pregnancy prediction. A total of 505 (46.03%) patients conceived naturally. The difference in CPI among EFI scores 10, 7–9, 4–6, and 2–3 was statistically significant (P < 0.001) and increased with increasing EFI score. The optimal cut-off point was 7.5: sensitivity (68.51%); specificity (52.20%). This retrospective study indicates the value of the EFI score for predicting spontaneous pregnancy in women treated by laparoscopy for endometriosis-associated infertility.
KEYWORDS: endometriosis, endometriosis Fertility Index, fecundity, infertility, laparoscopy, spontaneous pregnancy
Introduction Endometriosis is a chronic and diverse disease characterized by the presence of endometrial-like tissue outside the uterine cavity; it afflicts 35–50% women of reproductive age by causing related infertility, chronic pelvic pain and other physical problems (Giudice and Kao, 2004). Although many researchers have focused on the causal links between endometriosis and infertility, a definitive answer is not yet available (Bulletti et al., 2010). It is, however, clear that the fecundity rebounds to some degree after medical management. Among the various measures for treating endometriosis, laparoscopic surgery is widely used for treating endometriosisassociated infertility for both mild to moderate and severe patients (Marcoux et al., 1997; Jacobson et al., 2010; Shimizu et al., 2010). Moreover, it is the only approach to clarify the diagnosis by the visible presence of typical lesions and histological confirmation. Apart from the high cost, which is beyond the scope of medical insurance, the average outcomes of assisted reproductive technology (ART) for endometriosis-related infertile patients drive most couples in China to seek alternatives to surgery initially, and then try to conceive naturally.
Page 3 of 22
Pregnancy outcome after laparoscopy, however, is not consistent when using the most widely used staging system of endometriosis, namely the revised American Fertility Society (rAFS) score (Vercellini et al., 2006; Zeng et al., 2014); this drastically lowers the clinical value of rAFS in prognosis forecasts. To resolve this issue, the Endometriosis Fertility Index (EFI), which predicts spontaneous conception after surgery, was published in 2010 (Adamson and Pasta, 2010). The key elements of this scoring system include historical factors at the time of surgery, adnexal function after the intervention of surgery and assessment of endometriosis that was derived from rAFS. This index has been validated externally in endometriosis-associated infertile patients in Belgium (Tomassetti et al., 2013), China (Zeng et al., 2014), France (Boujenah et al., 2015), and Italy (Garavaglia et al., 2015). To demonstrate the value of the EFI score in predicting spontaneous pregnancy, and to determine an optimal cut-off point for positive outcome prediction, a larger sample size of Chinese women with endometriosis-associated infertility was included in this study.
Materials and methods Patients and inclusion criteria The medical Ethics Committee of West China Second University Hospital of Sichuan University indicated on 1 May 2017 that ethical approval was not needed owing to the retrospective nature of our study. Patients all provided informed consent for the use of their data, according to Ethics Committee requirements. Our study involved women with endometriosis-associated infertility who were undergoing laparoscopic treatment from January 2008 to December 2012 at the Department of Obstetrics and Gynecology in West China Second University Hospital of Sichuan University. The hospital represents a tertiary referral centre for endometriosis and reproductive surgery.
Women with endometriosis who were ovulating normally, had at least one patent tube according to hysterosalpingography, and had not conceived over a 12-month period
Page 4 of 22
were included in the study. Diagnosis was based on visualization of typical lesions, i.e., red, white and black implants in the peritoneal, ovarian and other organs, histopathological diagnosis for ovarian endometrioma and deep infiltrative lesions (DIE), or both. Deep infiltrative lesions are defined as lesions extending deeper than 5 mm under the peritoneal surface or those involving bowel, bladder, ureter or other sites. Diagnosis was based on criteria that included patients’ symptoms, careful physical examination, ultrasound, MRI, or both.
During physical examination, special attention was paid to the adnexal mass, nodules in retro-vaginal wall, sacrouterine ligaments, pelvic wall and the rectum during vaginal and rectal palpation. Additional criteria were normal sexual function and normal results for male partners’ semen analysis according to World Health Organization criteria (WHO, 2010). Exclusion criteria were as follows: ovulation disorders, such as polycystic ovary syndrome, luteinized unruptured follicle syndrome, hyperprolactinaemia and congenital defects, such as uterus septum, uterine leiomyoma, intrauterine adhesion and severe systemic disease.
Surgical procedure and EFI score All patients underwent standard surgery after conventional preoperative preparation. After induction of pneumoperitoneum of 14 mmHg, a thorough inspection was made to establish the presence of endometriotic lesions on the oviduct, ovary and pelvic peritoneum (especially uterosacral ligament), and the extent of pelvic adhesions. All visible peritoneal implants were excised or coagulated; normal anatomic structure of the pelvic organs was restored as much as possible by adhesiolysis. For endometrioma, the cyst was stripped carefully to avoid unnecessary damage to the healthy ovarian tissue. For patients with DIE, nodules were resected thoroughly. The presence of endometriotic lesions and adhesions was scored and staged according to the rAFS classification system of the American Society for Reproductive Medicine (ASRM, 1997). Before completion of surgery, the tubal patency test was carried out using methylene blue diluent.
Page 5 of 22
The following historical data were collected: patient age, duration of infertility and prior history of pregnancy. Total historical factor scores were then calculated (Figure 1). Surgery factors included assessment of endometriosis, which was derived from rAFS (endometriosis lesions score and total score), and adnexal function (least function). Least function score was based on the postoperative function of tube, fimbriate and ovary. The EFI scores were finally obtained after combining historical and surgical scores. For patients with advanced age（>40 years）or compromised tubal function, ART was proposed as the treatment. A recommendation was made to the other patients to attempt to conceive naturally for 12 months. Subsequently, patients who failed conception underwent ART.
Follow-up Between March 2013 and December 2015, each patient was contacted, either by letter or telephone interview to follow up postoperative outcomes. A spontaneous pregnancy was defined by a beta-HCG level over 25 UI/l. All patients were followed up for 36 months. For those who became pregnant during that interval, the follow-up period was extended to delivery, and the pregnancy outcomes were included in the data analysis. Women with a negative serum pregnancy test were not considered to be pregnant at the end of the follow-up period. Clinical data and information on the postoperative condition of all patients were collected and analysed.
Statistical analysis SPSS statistical software 21.0 (IBM Corp., USA) was used to assess relevant EFI score, rAFS classification, endometriosis type and pregnancy outcome after surgery. Data were presented as mean ± SE; chi-squared-test, life table and Kaplan–Meier survival analysis were used for comparing pregnancy rate and cumulative pregnancy incidence (CPI) at different time points. Patients were censored when they were lost
Page 6 of 22
to follow-up or underwent ART. Receiver operator characteristic curves were plotted to determine an optimal cut-off point of the EFI score for pregnancy prediction. For all analyses, P < 0.05 was considered as statistically significant.
Results Between March 2013 and December 2015, 1325 women were identified as eligible for study inclusion. A total of 228 (17.21%) women who were lost to follow-up were excluded; therefore, 1097 patients were eligible for EFI validation analysis. Among 1097 patients, 181 women had a complication of adenomyosis. The EFI score was calculated as previously reported, with EFI scores ranging from 0 for the poorest prognosis to 10 for the best prognosis. The median EFI score was 8. The distribution of patients according to their EFI score is presented in Figure 2. The most common EFI scores were 8 (238) and 9 (263). Baseline characteristics of infertile women related to age at operation, body mass index, duration of infertility, primary infertility, and r-AFS classification (Table 1). Laparoscopic surgery was carried out with no complications. Median operating time was 45 min (15–175). Patients with high rAFS stage and DIE underwent longer and more difficult operations. Most (27/31) DIE lesions were located within the uterosacral ligament or rectovaginal septum; others were located on the bowel (2) and bladder (2). The lesions were resected thoroughly without cystectomy or rectectomy. The median operating time of these patients was 78 min (55–175). The duration of follow-up was 14–36 months (median 26.34). After surgery, 176 patients used oral contraceptives for 1–3 months; 330 used gonadotrophin-releasing hormone agonists for 2–6 months; and 71 women took gestrinone for 2–3 months; all patients followed clinical advice according to their intraoperative condition. A total of 202 patients tried traditional Chinese medicine autonomously. Fifty women were advised to start ART, but none of them agreed to treatment in the first year after surgery. Eighty-three patients underwent ART after 12 months and 51 of them became pregnant (six by IUI and 45 by IVF and embryo transfer). A total of 505 (46.03%) women became pregnant naturally. Cumulative spontaneous clinical pregnancy rates calculated using the life table method for all
Page 7 of 22
patients was 42.57% ± 0.02%, 45.98% ± 0.02%, and 47.79% ± 0.02% at 12, 24, and 36 months, respectively. Among the 505 women who conceived naturally, 62 (12.28%) suffered from miscarriage and 24 (4.75%) patients had an ectopic pregnancy. The overall live birth rate was 38.20% (419/1097). Two neonatal deaths occurred among 34 premature babies (gestational age range 28+6 weeks to 36+4 weeks). Median time between surgery and conception was 7 months, with a range of 1 to 36 months.
The life table analysis stratified the EFI into five classes (Table 2). A significant relationship was found between EFI and time to achieving pregnancy (P < 0.001) (Table 2); however, no association was found between endometriosis type and pregnancy rate. The pregnancy rate of rAFS classifications I to Ⅳ was 50.57%, 45.58%, 47.33% and 37.00%, respectively. Paired comparison by chi-squared test showed no difference between stages Ⅱ and Ⅲ. The pregnancy rate of patients with stage I was significantly higher than that of patients with stages Ⅱto Ⅳ (P < 0.001). Most pregnancies were achieved within 2 years of surgery. After 36 months of followup, the pregnancy rate of patients was compared by chi-squared test. The pregnancy rate of patients with EFI 10 (70.31%) was significantly higher than that of patients with EFI 9 (52.47%) (P < 0.001). The aggregated pregnancy rate was 51.10% among patients with EFI 8–9; the pregnancy rates for EFI scores 8 and 9 (49.58% and 52.47% respectively) were not significantly different. Similar results were obtained for the pregnancy rate among patients with EFI 6 and 7 (33.33%, 43.87%). The aggregated pregnancy rate (39.94%) of patients with EFI 6–7 was lower than that of patients with EFI 8–9 (P = 0.001). No difference was found among patients with EFI score 2–5 (14.29%, 8.33%, 20.93%, 19.12%). The number of patients with EFI 2–3 was underrepresented (n = 19), and only two pregnancies were achieved among them. The Kaplan–Meier survival analysis result for CPI at different time points among EFI 10, 7–9, 4–6 and 2–3 was statistically significant (P < 0.001) (Figure 3). As expected, the CPI did not correspond to r-AFS staging (I-IV), and no significant difference was found in CPI between different types of endometriosis (Table 2). As shown in Figure 4, the area under the curve for EFI as pregnancy predictor was 0.648 (95% CI 0.616
Page 8 of 22
to 0.680; P < 0.001). The best cut-off point was 7.5 (sensitivity, 68.51%, specificity, 52.20%), whereas Youden index was 0.207 (Youden index = sensitivity + specificity – 1).
Discussion To date, laparoscopic surgery has been the first-line treatment for most endometriosisrelated infertile patients (Marcoux et al., 1997; Jacobson et al., 2010). Medication has proved to be effective only for pain management and for reducing recurrence when used individually or when combined with laparoscopy (Olive and Pritts, 2001). The rAFS classification for description of visible lesions and adhesions in endometriosis has become a widely accepted staging system. It has, however, failed to predict postoperative pregnancy (Vercellini et al., 2006; Zeng et al., 2014). This may be due to the complicated influence of endometriosis on reproduction; moreover, visible lesions did not represent the severity of endometriosis. Laparoscopic surgery cannot resolve molecular, immune or other unknown factors that influence fertility; this makes it more difficult to develop a feasible and effective staging system. In this scenario, the EFI scoring system emerged as a useful method (Adamson and Pasta, 2010). As shown in Figure 2, we validated the EFI again by demonstrating different scores and corresponding fecundity rates. Apparently, the pregnancy rate is proportional to the increased score of the EFI. By considering historical factors, the EFI staging system provides a comprehensive evaluation of fecundity for endometriosis-related infertility after laparoscopic surgery for the first time. Adamson and Pasta (2010) used a different approach and considered historical factors that reflect the severity of the disease to some extent. They also included least function of bilateral adnexa after surgery, which plays a crucial role in natural conception. Endometriosis lesions occupied little weighting (2/10) in this system.
Although the EFI staging system ignored the assessment of uterine factors, adenomyosis, deep infiltrating disease and ovarian reserve, it still has vital clinical significance. Indeed, the probability of spontaneous conception seemed remote in
Page 9 of 22
patients who scored 2–3, albeit a small number. In fact, gynaecologists are quite capable of distinguishing such women with an intuitive definition of ‘severity’, such as tubal obstruction after surgery, advanced female age, or bot; these patients should be recommended ART immediately. In our opinion, the higher the EFI score, the higher the probability that a spontaneous pregnancy will be achieved; this has been confirmed in previous studies (Boujenah et al., 2015; Garavaglia et al., 2015). The CPI of 12 months and 24 months were in accordance with results reported by Tomassetti et al. (2013).
Natural conception is the first choice for women with EFI score 6–10, whereas, for patients with EFI score 4–5, natural conception or ART should be offered to infertile couples according to their pregnancy desire and financial situation. In particular, patients with an EFI score of 0–3 should be informed about the low possibility of spontaneous pregnancy, and ART should be strongly encouraged. In addition, we found that most pregnancies were achieved within 2 years of surgery; in this period, patients may benefit from the favourable effect of surgery on the anatomical factors altered by endometriosis. Therefore, ART should be recommended to patients who have failed to conceive naturally 2 years after laparoscopy, including those patients with a high EFI score. The area under the ROC curve demonstrated a low prediction accuracy. This may be because of the small number of patients with a score of 2–3. A pregnancy rate of 10.53% (2/19) in these patients may not represent the true picture. A similar pregnancy rate for patients with EFI score 4–5 was observed. In contrast to other studies (Tomassetti et al., 2013; Boujenah et al., 2015), all our patients chose natural conception first after surgery. This was probably due to economic restriction and societal perceptions in our country. For most patients, ART was expensive; therefore, this method has been treated with prejudice for some time. In view of this, we think our research is more representative, as patients with low EFI scores in developed countries chose ART directly (Garavaglia et al., 2015). Therefore, we can observe the real performance of the EFI score in predicting pregnancy in patients with low scores. In 19 patients with a EFI score of 0–3, only two conceived naturally after
Page 10 of 22
36 months. It is difficult to control for bias, however, owing to the small sample size in this group. Our study also indicates that the EFI score is a good staging system for pregnancy prediction for infertile women with endometriosis.
In China, many women prefer to try traditional Chinese medicine (TCM) orally after surgery for spontaneous pregnancy. They believe that it helps to increase their chances of conception. Some studies have confirmed the effectiveness of TCM in treating infertility (Flower et al., 2012). Considering the specificity of TCM, it is extremely difficult to judge the efficacy on fecundity (Guo et al., 2010). Therefore, it is unclear whether our results were influenced by this complicated treatment to some degree.
A few patients in our study accepted ovulation induction (81/1097), among which 31 became pregnant. This is distinct from research by Adamson and Pasta (2010); and Tomassetti et al. (2013), in which ovulation induction or ovarian stimulation combined with IUI was carried out generally after failure of spontaneous conception within 1 year of surgery. Nevertheless, the benefit of ovulation induction or ovarian stimulation combined with IUI after a certain time is also highly contentious (Gandhi et al., 2014). Therefore, we are unable to reach a definite conclusion at present.
The prevalence of DIE in our study is significantly lower than the 12–20% reported in other studies (Chapron et al., 1999; Lagana et al., 2016; Iversen et al., 2017). Several reasons may contribute to this difference. First, different ethnicity of the study populations may be a factor, as data from the Peking Union Medical College Hospital of China also revealed a low incidence of DIE (5.80%, 615/10597) in their 10-year experience (Li et al., 2011) in Chinese women, although this is higher than in the present study. Second, subject selection criteria differ from other studies, as patients in the present study were included primarily on the basis of their history of infertility caused by endometriosis, not chronic pelvic pain, dysmenorrhoea or pelvic mass, although some of these symptoms may also have been present in the selected patients. Third, some cases of DIE could have been missed owing to the lack of experience of
Page 11 of 22
doctors looking for the lesion during surgery, therefore under-representing the true incidence.
The present study has some limitations. It is retrospective in design, although the data necessary for calculation of the EFI were collected prospectively, as with the study by Tomassetti et al. (2013). Confounding factors included different surgeons and different post-surgical medications. In conclusion, despite these limitations, however, the EFI score seems to be relevant for the prediction of the natural conception rate in infertile women with endometriosis treated with laparoscopic surgery. This will be useful in counselling patients of reproductive age.
Acknowledgements We thank Dr Lu Zhou and Dr Licong Shen for their help in this study. We also thank Mr Ruibo Zhang for proofreading this article. Wei Huang designed the study; Dong Liu, Jing Tan, Xue Feng and QiuShi Wang collected reviews of medical history and surgical records; Xianghui Zhang and Dong Liu were responsible for follow up and data analysis; Xianghui Zhang and Dong Liu wrote the paper.
References Adamson, G.D., Pasta D.J. 2010. Endometriosis fertility index: the new, validated endometriosis staging system. Fertil Steril 94, 1609-15. ASRM. 1997. Revised American Society for Reproductive Medicine classification of endometriosis: 1996. Fertil and Steril 67, 817-821. Boujenah, J., Bonneau, C., Hugues, J. N., Sifer, C. and Poncelet, C. 2015. External validation of the Endometriosis Fertility Index in a French population. Fertil Steril 104, 119-23. Bulletti, C., Coccia, M. E., Battistoni, S. and Borini, A. 2010. Endometriosis and infertility. J Assist Reprod Genet 27, 441-7. Chapron, C., Fritel, X. and Dubuisson, J. B. 1999. Fertility after laparoscopic management of deep endometriosis infiltrating the uterosacral ligaments. Hum
Page 12 of 22
Reprod 14, 329-32. Flower, A., Liu, J. P., Lewith, G., Little, P. and Li, Q. 2012. Chinese herbal medicine for endometriosis. Cochrane Database Syst Rev, CD006568. Gandhi, A. R., Carvalho, L. F., Nutter, B. and Falcone, T. 2014. Determining the fertility benefit of controlled ovarian hyperstimulation with intrauterine insemination after operative laparoscopy in patients with endometriosis. J Minim Invasive Gynecol 21, 101-8. Garavaglia, E., Pagliardini, L., Tandoi, I., Sigismondi, C., Vigano, P., Ferrari, S. and Candiani, M. 2015. External validation of the endometriosis fertility index (EFI) for predicting spontaneous pregnancy after surgery: further considerations on its validity. Gynecol Obstet Invest 79, 113-8. Giudice, L. C. and Kao, L. C. 2004. Endometriosis. Lancet 364, 1789-99. Guo, S., He, W., Zhao, T., Liu, X. and Zhang, T. 2010. Clinical trials and trial-like studies on the use of traditional Chinese medicine to treat endometriosis. Obstet Gynecol 5, 533-555. Iversen, M. L., Seyer-Hansen, M. and Forman, A. 2017. Does surgery for deep infiltrating bowel endometriosis improve fertility? A systematic review. Acta Obstet Gynecol Scand 96, 688-693. Jacobson, T. Z., Duffy, J. M., Barlow, D., Farquhar, C., Koninckx, P. R. and Olive, D. 2010. Laparoscopic surgery for subfertility associated with endometriosis. Cochrane Database Syst Rev, CD001398. Lagana, A. S., Vitale, S. G., Trovato, M. A., Palmara, V. I., Rapisarda, A. M., Granese, R., Sturlese, E., De Dominici, R., Alecci, S., Padula, F., Chiofalo, B., Grasso, R., Cignini, P., D'Amico, P. and Triolo, O. 2016. Full-Thickness Excision versus Shaving by Laparoscopy for Intestinal Deep Infiltrating Endometriosis: Rationale and Potential Treatment Options. Biomed Res Int 2016, 3617179. Li, L., Leng, J. H., Dai, Y., Zhang, Y., Li, X. Y. and Lang, J. H. 2011. Characteristics and trends of surgical management of pelvic endometriosis in different ages. Chin J Pract Gynecol Obstet, 2011. 27(3): p. 188-191. Marcoux, S., Maheux, R. and Berube, S. 1997. Laparoscopic surgery in infertile
Page 13 of 22
women with minimal or mild endometriosis. Canadian Collaborative Group on Endometriosis. N Engl J Med 337, 217-22. Olive, D. L. and Pritts, E. A. 2001. Treatment of endometriosis. N Engl J Med 345, 266-75. Shimizu, Y., Takashima, A., Takahashi, K., Kita, N., Fujiwara, M. and Murakami, T. 2010. Long-term outcome, including pregnancy rate, recurrence rate and ovarian reserve, after laparoscopic laser ablation surgery in infertile women with endometrioma. J Obstet Gynaecol Res 36, 115-8. Tomassetti, C., Geysenbergh, B., Meuleman, C., Timmerman, D., Fieuws, S. and D'Hooghe, T. 2013. External validation of the endometriosis fertility index (EFI) staging system for predicting non-ART pregnancy after endometriosis surgery. Hum Reprod 28, 1280-8. Vercellini, P., Fedele, L., Aimi, G., De Giorgi, O., Consonni, D. and Crosignani, P. G. 2006. Reproductive performance, pain recurrence and disease relapse after conservative surgical treatment for endometriosis: the predictive value of the current classification system. Hum Reprod 21, 2679-85. World Health Organization. WHO laboratory Manual for the Examination Processing
of
Human
Semen
Fifth
edition
.Geneva
:
World
and
Health
Organization ,2010.100 Zeng, C., Xu, J. N., Zhou, Y., Zhou, Y. F., Zhu, S. N. and Xue, Q. 2014. Reproductive performance after surgery for endometriosis: predictive value of the revised American Fertility Society classification and the endometriosis fertility index. Gynecol Obstet Invest 77, 180-5.
Declaration The authors report no financial or commercial conflicts of interest.
Page 14 of 22
Figure legends Figure 1 – Endometriosis Fertility index (EFI) surgery form (Tomassetti et al., 2013, reproduced with permission).
Figure 2 – Distribution of patients according to their Endometriosis Fertility index (EFI) score. The median EFI score was 8.
Figure 3 –
Kaplan–Meier survival analysis for cumulative pregnancy incidence
according to different Endometriosis Fertility index (EFI) score groups. The Kaplan– Meier survival analysis result for cumulative pregnancy incidence at different time points among EFI 10, 7–9, 4–6 and 2–3 was statistically significant (P < 0.001).
Figure 4 – receiver operator characteristic curve. The area under the curve for EFI as pregnancy predictor was 0.648 (95% CI 0.616 to 0.680, P < 0.001). The best cut-off point was 7.5 (sensitivity, 8.51%, specificity, 52.20%), Youden index was 0.207 (Youden index = sensitivity + specificity –1).
Page 15 of 22
Table 1– Baseline characteristics of infertile women. Mean ± SD, mean ± SD (range) or n (%) Age at operation (years) ≤35 years 36–39 years ≥40 years Body mass index Duration of infertility (years) ≤3 years >3 years Primary infertility, n (%) Secondary infertility, n (%) Least function score High score (7–8) Moderate score (4–6) Low score (1–3) AFS endometriosis lesion score <16 ≥16 AFS total score <71 ≥71
29.84 4.29 (20–46) 989 (90.15) 87 (7.93) 21(1.91) 20.34 2.57 (13.67–33.73) 3.02 2.64 (1–18) 794 (72.38) 303 (27.62) 537 (48.95) 560 (51.05) 385 (35.10) 574 (52.32) 138 (12.58) 894(81.49) 203 (18.51) 1008 (91.89) 89 (8.11)
AFS, American Fertility Society.
Page 16 of 22
Table 2 – Life table for cumulative pregnancy incidence according to different parameters over different time periods. Patients (n) Parameter
Cumulative pregnancy rate（%） 1 year 2 years 3 years
P-value
rAFS I
352
48.18 ± 2.68
51.01 ± 2.70
51.64 ± 2.74
II
283
41.23 ± 2.96
45.63 ± 3.02
48.03 ± 3.19
III
262
43.16 ±3.08
46.50 ± 3.12
49.76 ± 3.33
IV
200
33.76 ± 3.36
36.98 ± 3.44
38.07 ± 3.55
Peritoneal
693
43.32±1.90
47.44 ± 1.92
48.08 ± 1.95
Ovarian
373
40.44±2.56
42.82 ± 2.60
46.93 ± 2.84
DIE
31
51.61±8.98
51.61 ± 8.98
51.61 ± 8.98
2–3
19
10.53 ± 7.04
10.53 ± 7.04
10.53 ± 7.04
4–5
111
15.52 ± 3.46
19.41 ± 3.81
21.37 ± 4.19
6–7
338
36.32 ± 2.64
38.96 ± 2.69
42.85 ± 2.95
8–9
501
48.04 ± 2.25
51.59 ± 2.27
52.49 ±2.31
10
128
65.90 ± 4.21
70.89 ± 4.06
70.89 ± 4.06
<0.001
Pathology type NS
EFI <0.001
DIE, deep infiltrating endometriosis; EFI, Endometriosis Fertility Index; NS, non significant; rAFS, revised American Fertility Society.
Page 17 of 22
Figure 1.
Page 18 of 22
Figure 2.
Page 19 of 22
Figure 3.
Page 20 of 22
Figure 4.
Page 21 of 22
photo.png
Page 22 of 22