- Academic Editor
Background: Determine potential triggers for
ovarian endometriosis recurrence to provide individualized
long-term management and follow-up for improved patient outcomes.
Methods: Relevant data were acquired through systematic retrieval from
PubMed, Embase, Web of Science, and Cochrane Library before
October 2022. We determined the odd ratio or the mean difference with their
corresponding 95% confidence interval (CI) to explore the relationship between
relevant risk factors and postoperative endometriosis relapse. Results:
This meta-analysis ultimately covered 6388 patients from 18
trials, and the findings demonstrated that postoperative endometriosis recurrence
was associated with the age at surgery [mean difference (MD)
(95% CI) = –0.69 (–1.33~–0.05), p = 0.04], family
history [odds ratio (OR) (95% CI) = 2.18 (1.10~4.29), p
= 0.02], preoperative carbohydrate antigen (CA-125) [MD (95% CI) = 24.08
(–7.55~40.61), p = 0.004], laterality of endometriosis
(EMs) [OR (95% CI) = 1.19 (1.00~1.40), p = 0.04],
presence of adenomyosis [OR (95% CI) = 1.53 (1.11~2.11),
p = 0.009], presence of myoma [OR (95% CI) = 1.44
(1.07~1.94), p = 0.02], previous endometriosis-related
surgery [OR (95% CI) = 1.90 (1.45~2.51), p
Endometriosis is a common disease affecting women of childbearing age, where the endometrium tissue (both interstitial and glandular) grows out of the uterine cavity. The periodic bleeding from the ectopic endometrium and fibrosis of the surrounding tissue often form ectopic nodules and other lesions. Pathologically, endometriosis presents benign morphological changes but invasive, metastatic, and recurrent biological behaviors of malignant tumors, earning it the moniker of benign cancer. The major symptoms include dysmenorrhea, chronic pelvic pain, irregular menstruation, and infertility [1]. Endometriosis can occur in all pelvic tissues and organs, particularly the ovaries [2].
Surgery is the standard treatment for ovarian endometriosis, but the postoperative recurrence rate is high. Recurrent endometriosis is the regeneration or reappearance of lesions after standard surgery and drug treatment, with clinical symptoms returning to or even exceeding the level before treatment after symptom relief. According to previous studies, the five-year recurrence rate after conservative surgery is over 40% [3]. The high postoperative recurrence rate severely affects patients’ physical and mental health and quality of life [4, 5] and increases their social and economic burden, making it a challenging issue clinically. Preventing or delaying postoperative recurrence of ovarian endometriosis is a difficult problem to tackle. However, the pathogenesis of endometriosis has not been fully elucidated, and no single theory can explain its occurrence [6]. Therefore, it is necessary to determine the factors associated with its clinical recurrence rate.
This study collected data on relapse-related factors from long-term follow-ups of childbearing-age patients with ovarian endometriosis. Assessing the association between these factors and ovarian endometriosis recurrence to identify the risk factors may help to provide individualized long-term management and follow-up based on clinical conditions and reduce postoperative recurrence.
Systematical searches were performed on PubMed, Embase, Web of Science, and Cochrane Library for studies published online before October 2022 on postoperative recurrence of ovarian endometriosis. The search terms included “endometriosis” (e.g., “endometrioses”, “endometrioma”, and “endometriomas”), “recurrence” (e.g., “recurrences”, “relapse”, “relapses”, and “recrudescence”) and “factors” (e.g., “reasons” and “element”), as shown in Supplementary Table 1.
(1) Studies of factors related to the postoperative recurrence in patients with ovarian endometriosis;
(2) Cohort studies or case-control studies;
(3) With subjects concerning patients with ovarian endometriosis confirmed by postoperative pathological examination;
(4) With a postoperative follow-up time of at least six months;
(5) The postoperative recurrence was confirmed by ultrasound and reoperation.
(1) Reviews, abstracts, case reports, meta-analyses, and other non-clinical studies;
(2) Postmenopausal endometriosis;
(3) Endometriosis in adolescents;
(4) Studies with incomplete original data;
(5) Repeated studies.
With the search terms and inclusion and exclusion criteria, relevant studies were retrieved from the above databases. The relevant literature was initially screened by reading the titles and abstracts, and those that failed to meet the requirements were excluded. Then, the full texts were read, and those that failed to meet the requirements were further excluded. Disagreements on literature inclusion were resolved through discussion or judgment by a third researcher. Studies with similar contents were carefully checked in terms of the title, author, and affiliation to determine whether they were duplications. Only one of the confirmed duplications was included.
After the literature selection, two researchers extracted the data from the
included literature using the prepared table. Any discrepancies were settled by a
third researcher. The extracted data included the first author, year of
publication, age at surgery, age of menarche, dysmenorrhea, postoperative
pregnancy, infertility, endometriosis operation history, combined myoma, combined
adenomyosis, postoperative medication, family history, preoperative carbohydrate antigen (CA-125),
ovarian cyst, body mass index (BMI), laterality of endometriosis (EMs), r-American Fertility Society (r-AFS) stage, and extent
of surgery. The included studies were evaluated using the Newcastle-Ottawa Scale
(NOS). Studies with a NOS score
Research | Selection | Comparability | Outcome | Score |
Vercellini et al. 2006 [7] | * * * * | * * | 6 | |
Sengoku et al. 2013 [8] | * * * * | * * * | 7 | |
Selcuk et al. 2016 [9] | * * * * | * | * * * | 8 |
Küçükbaş et al. 2018 [10] | * * * * | * * | * * | 8 |
Hidari et al. 2019 [11] | * * * * | * * * | 7 | |
Li et al. 2019 [12] | * * * * | * | * * * | 8 |
Zhang et al. 2019 [13] | * * * * | * | * * * | 8 |
Won et al. 2020 [14] | * * * * | * | * * * | 8 |
Del Forno et al. 2021 [15] | * * * * | * * | * * | 8 |
Wacharachawana et al. 2021 [16] | * * * * | * | * * * | 8 |
Yu et al. 2022 [17] | * * * * | * * | * | 7 |
Tarumi et al. 2022 [18] | * * * * | * | * * * | 8 |
Huang et al. 2022 [19] | * * * * | * * | * * | 8 |
Refer to the Newcastle-Ottawa Scale, from lowest to highest * (one point) to **** (four points).
Research | Selection | Exposure | Outcome | Score |
Hayasaka et al. 2011 [20] | * * * * | * * * | 7 | |
Campo et al. 2014 [21] | * * * * | * | * * | 7 |
Chon et al. 2016 [22] | * * * * | * * | * * | 8 |
Han et al. 2017 [23] | * * * * | * | * * * | 8 |
He et al. 2018 [24] | * * * * | * * | * * | 8 |
Refer to the Newcastle-Ottawa Scale, from lowest to highest * (one point) to **** (four points).
Review Manager 5.4 (the International Cochrane Collaboration, Nordic Cochrane
Centre, Oxford, United Kingdom) was employed for the analysis. Efficacy analysis
statistics were expressed by odds ratio (OR) for dichotomous variables and by
mean difference (MD) and 95% confidence interval (CI) for continuous variables
to evaluate the impact of related factors on recurrence. The test level was
p
The literature retrieval process yielded 6768 potentially eligible studies. In addition, 3 studies were identified from the reference lists of the included studies. A total of 4615 studies were screened out by two reviewers independently after excluding duplicate studies. After reading the titles and abstracts, 4542 studies not associated with recurrent risk factors for ovarian endometriosis were further excluded. The remaining 73 studies were then fully reviewed. Among them, 5 were excluded due to insufficient data; 3 studies were excluded due to inconclusive postoperative pathologic findings; 25 studies were disqualified due to the lack of a control group or a non-recurrent group as the control group; 19 studies were excluded as cyst recurrence was not selected as an outcome; and 14 studies were disregarded due to insufficient postoperative follow-up. Finally, this meta-analysis included 18 studies satisfying the inclusion criteria. The flow chart of the literature selection process is shown in Fig. 1.
Flow chart of the literature selection process.
The included literature was retrospective studies published between 2006 and
2022, which covered 6388 patients with endometriosis, with 1138 and 5250 in the
relapsing and non-relapsing groups. The basic information of the included
literature is presented in Table 3 (Ref. [7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24)]. The comprehensive analysis of
relapsing-related factors is listed in Table 4. The analyzed potential risk
factors included age of menarche, age (years), infertility, postoperative
pregnancy, BMI (kg/m
Study cohort | Year | Affiliation | Study design | Inclusion criteria | Exclusion criteria | Recurrence/Non-recurrence | Indicators of observation |
Vercellini et al. [7] | 2006 | The First Department of Obstetrics and Gynaecology of the University of Milan | Retrospective Study | (1) Consecutive women with any type of endometriotic lesion. | (1) Previous clinical or endoscopic diagnosis of endometriosis; | 69/660 | ⑨ |
(2) Participated in controlled clinical trials; | |||||||
(3) Previous abdominal surgery except appendectomy; | |||||||
(4) Diagnosis of gastrointestinal, urological, or orthopedic diseases with potential pain irradiation to the pelvic area; | |||||||
(5) Known psychiatric disturbances. | |||||||
Hayasaka et al. [20] | 2011 | Tohoku University Hospital or Sendai Social Insurance Hospital | Retrospective Study | (1) At least one year of postoperative follow-up after undergoing laparoscopic excision of an ovarian endometrioma; | (1) Any prior surgery for ovarian endometriomas. | 78/95 | ①②④⑤⑩⑫⑮ |
(2) Patients were followed up every six months for the first year after the operation, followed by any follow-up as needed. | |||||||
Sengoku et al. [8] | 2013 | Asahikawa Medical University Hospital | Retrospective Study | (1) At least two years of postoperative follow-up; | (1) No follow-up within two years; | 73/175 | ②③④⑤⑨⑩⑪⑫⑬⑮⑯ |
(2) Underwent laparoscopic excision of ovarian endometriomas of |
(2) Irregular menstruation cycles; | ||||||
(3) Malignancy. | |||||||
Campo et al. [21] | 2014 | The Department of Obstetrics and Gynecology of the Catholic University of the Sacred Heart | Retrospective Study | (1) At least 1 ovarian endometrioma |
(1) Minimal or mild endometriosis; | 27/121 | ②③⑤⑦⑧⑬⑮ |
(2) Rectovaginal endometriosis, neoadjuvant treatment with estroprogestins; | |||||||
(3) Gonadotropin-releasing hormone (GnRH) analogs; | |||||||
(4) Follow-up period under 12 months. | |||||||
Selcuk et al. [9] | 2016 | Zeynep Kamil Training and Researching Hospital | Retrospective Study | (1) Underwent laparoscopic ovarian endometriotic cystectomy; | (1) Underwent bilateral oophorectomy at the time of endometrioma surgery; | 27/36 | ②③⑥⑧⑨⑫⑬ |
(2) Regular postoperative follow-up period. | (2) Ovarian endometrioma treated with drainage, ablation, fenestration techniques, or incompletely stripped; | ||||||
(3) Postmenopausal status; | |||||||
(4) No postoperative follow-up; | |||||||
(5) Postoperative medical therapy for endometriosis. | |||||||
Chon et al. [22] | 2016 | Department of Obstetrics and Gynecology, Gil Hospital, Gachon University College of Medicine | Retrospective Study | (1) Suspected of endometrioma using ultrasonography and pathologically diagnosed with endometrioma. | (1) Not diagnosed with pathologically; | 37/199 | ②⑤⑥⑧⑨⑫⑬⑮ |
(2) Without endometrioma; | |||||||
(3) Previously underwent surgery due to endometrioma in other hospitals. | |||||||
Han et al. [23] | 2017 | Kyungpook National University Hospital | Retrospective Study | (1) Underwent surgical operation including at least ovarian cystectomy for ovarian endometrioma; | (1) Incomplete medical records; | 27/77 | ①②③⑤⑥⑧⑨⑩⑪⑫⑬⑭ |
(2) Loss of follow-up before 24 months after primary surgery; | |||||||
(2) Reproductive ages (20–40 years old). | (3) Concurrently hysterectomized; | ||||||
(4) Menopaused; | |||||||
(5) History of previous endometriosis-related surgery. | |||||||
Küçükbaş et al. [10] | 2018 | The Zeynep Kamil Research and training hospital in Turkey | Retrospective Study | (1) Underwent laparoscopic unilateral/bilateral endometrioma cystectomy; | (1) Underwent oophorectomy or hysterectomy; | 56/113 | ②⑤⑥⑦⑨⑫⑬⑯ |
(2) Diagnosis of malignancy; | |||||||
(2) Followed for more than 6 months. | (3) Intraoperatively diagnosed adnexal pathology requiring subsequent unilateral or bilateral oophorectomy. | ||||||
He et al. [24] | 2018 | Department of Obstetrics and Gynecology at Peking Union Medical College Hospital | Retrospective Study | (1) Postoperative histopathological diagnosis of ovarian endometriosis; | (1) The combination of malignant or borderline tumors. | 45/180 | ①②⑥⑫⑮ |
(2) Aged 45 years and over at the time of surgery; | |||||||
(3) The clinical and pathological data were complete. | |||||||
Hidari et al. [11] | 2019 | The University of Tokyo Hospital | Retrospective Study | (1) Age between 35 and 45 years; | (1) The presence of bilateral endometriomas; | 8/42 | ②③④⑧⑩⑪⑫⑬ |
(2) Preoperative magnetic resonance imaging (MRI)-confirmed unilateral ovarian endometrioma. | (2) Previous unilateral or bilateral oophorectomy; | ||||||
(3) Pathological diagnosis of malignancy. | |||||||
Li et al. [12] | 2019 | Peking Union Medical College Hospital | Retrospective Study | (1) Diagnosis confirmed by pathologists; | (1) Age |
68/290 | ②③④⑤⑥⑧⑨⑩⑪⑬⑭⑮ |
(2) Conducted ultrasonography to determine endometrioma recurrence at least 6 months after surgery; | (2) Having undergone bilateral oophorectomy or hysterectomy; | ||||||
(3) Patients were observed without postoperative medications or were treated with postoperative GnRHa injections for 3–6 cycles; | (3) Intra-operative conversion to laparotomy. | ||||||
(4) The duration of follow-up was at least 5 years. | |||||||
Zhang et al. [13] | 2019 | The Gynaecology and Obstetrics Department of Tianjin Centre Hospital | Retrospective Study | (1) Patients who previously underwent surgery; | (1) Patients with primary ovarian carcinoma or ovarian metastases; | 39/74 | ②⑥⑨⑫ |
(2) Ovarian endometriosis malignancy or non-malignant ovarian endometriosis confirmed by postoperative pathology; | (2) Patients who refused surgical therapy or accepted conservative treatment; | ||||||
(3) Patients with complete clinical data. | (3) Patients who had incomplete clinical data. | ||||||
Won et al. [14] | 2020 | CHA Gangnam Medical Center | Retrospective Study | (1) Patients were surgically treated and displayed pathologically confirmed ovarian endometriosis. | (1) Gynecologic malignancy; | 362/2319 | ②③④⑤⑥⑨⑫⑬⑭⑮⑯ |
(2) Underwent bilateral oophorectomy; | |||||||
(3) Menopause; | |||||||
(4) Revised American Society of Reproductive Medicine (rASRM) stage of I or II (n = 18) at initial surgery; | |||||||
(5) Follow-up duration of | |||||||
Del Forno et al. [15] | 2021 | Gynecology and Human Reproduction Physiopathology, IRCCS Azienda Ospedaliero-Universitaria di Bologna | Retrospective Study | (1) The presence of a histological diagnosis of ovarian endometrioma; | (1) Bilateral oophorectomy; | 45/161 | ⑨ |
(2) A minimum follow-up period of 12 months. | (2) Hormonal therapies other than EPs and Ps pregnancy; | ||||||
(3) Menopausal state diagnosed postoperatively or during follow-up. | |||||||
Wacharachawana et al. [16] | 2021 | Bhumibol Adulyadej Hospital | Retrospective Study | (1) Underwent laparoscopic surgery with a provisional diagnosis of ovarian endometriosis; | (1) Non-ovarian endometriosis from pathologic reports; | 24/82 | ②③⑤⑧⑨⑩⑪⑫⑯ |
(2) Follow-up visits of at least one full year postoperation. | (2) Incomplete follow-up (follow-up of less than one full year); | ||||||
(3) Incomplete medical data. | |||||||
Yu et al. [17] | 2022 | Department of Gynaecology, Zhoushan Women and Children Hospital | Retrospective Study | (1) Not pregnant for more than 1 year without contraception; | (1) Cognitive dysfunction or mental illness; | 30/128 | ②④⑤⑧⑨⑬⑮ |
(2) Ultrasonography showed that there was growth and infiltration outside the endometrium, and repeated bleeding formed nodules and masses, which caused pain; | (2) Accompanied by dysfunction of the heart, liver, kidney, and other organs; | ||||||
(3) Patients who did not take hormone drugs within 6 months before the study; | (3) With contraindications for operation; | ||||||
(4) The clinical data are complete. | (4) There are drug contraindications; | ||||||
(5) Failed to meet the follow-up conditions. | |||||||
Tarumi et al. [18] | 2022 | Department of Obstetrics and Gynecology, Kyoto Prefectural University of Medicine | Retrospective Study | (1) Reproductive age; | (1) Less than 2-year postoperative follow-up period; | 36/96 | ②⑤⑥⑨⑩⑪⑫⑬ |
(2) Underwent total laparoscopic cystectomy for ovarian endometriosis with or without excision of deep endometriosis (DE). | (2) Pregnancy within 12 months after surgery. | ||||||
Huang et al. [19] | 2022 | The Department of Gynecology and Tumors, Changzhou Maternal and Child Health Care Hospital, Affiliated with Nanjing Medical University | Retrospective Study | (1) Aged 18 to 45 years at the time they underwent surgery; | (1) Patients treated with radical surgery or having gynecologic malignancies or other malignancies; | 49/240 | ②③④⑨⑩⑪⑫⑬ |
(2) Surgery was laparoscopic ovarian cyst debulking, which is a fertility-preserving surgery that is done without removing fallopian tubes and ovaries; | (2) Patients with incomplete medical records or lost visits or who were unwilling to cooperate with a follow-up visit; | ||||||
(3) Complete clinical history data and follow-up information were available; | (3) Patients who had laparoscopic intermediate open surgery; | ||||||
(4) Postoperative follow-up time was more than 12 months. | (4) Patients with pelvic tuberculosis, ovulation disorders, genital malformations, and other causes of infertility. |
Indicators of observation: ① Age of menarche; ② Age (years); ③ Infertility; ④ Postoperative pregnancy; ⑤ Body mass index (kg/m
Correlation factors | Number of studies | Recurrence /Non-recurrence | Aggregate effect | Heterogeneity | Effect model | |||
OR/MD | 95% CI | p value | p value | I | ||||
Age at surgery | 12 | 844/3767 | –0.69 | –1.33 |
0.04 | 0.02 | 52% | Random-effects model |
Age of menarche | 3 | 150/352 | –0.18 | –0.44 |
0.18 | 0.72 | 0% | Fixed-effects model |
Dysmenorrhea | 6 | 153/608 | 1.95 | 0.95 |
0.07 | 0.04 | 57% | Random-effects model |
Postoperative pregnancy | 7 | 687/3368 | 0.40 | 0.19 |
0.01 | 83% | Random-effects model | |
Infertility | 8 | 638/3305 | 0.95 | 0.71 |
0.71 | 0.10 | 41% | Fixed-effects model |
Endometriosis operation history | 4 | 515/2689 | 1.90 | 1.45 |
0.14 | 45% | Fixed-effects model | |
Combined myoma | 8 | 363/1097 | 1.44 | 1.07 |
0.02 | 0.12 | 39% | Fixed-effects model |
Combined adenomyosis | 7 | 285/1002 | 1.53 | 1.11 |
0.009 | 0.40 | 4% | Fixed-effects model |
Postoperative medication | 8 | 720/3507 | 1.64 | 1.02 |
0.04 | 0.001 | 71% | Random-effects model |
Family history | 2 | 83/234 | 2.18 | 1.10 |
0.02 | 0.17 | 47% | Fixed-effects model |
Preoperative CA-125 | 5 | 550/2998 | 24.08 | –7.55 |
0.004 | 0.001 | 78% | Random-effects model |
Size of ovarian cyst | 10 | 737/3403 | 0.48 | –0.21 |
0.17 | 85% | Random-effects model | |
BMI | 9 | 752/3471 | –0.21 | –0.46 |
0.08 | 0.58 | 0% | Fixed-effects model |
Laterality of EMs | 11 | 792/3794 | 1.19 | 1.00 |
0.04 | 0.04 | 48% | Fixed-effects model |
r-AFS stage | 13 | 1005/4652 | 0.57 | 0.48 |
0.46 | 0% | Fixed-effects model | |
Cystectomy of the affected site | 3 | 457/2686 | 3.94 | 0.07 |
0.50 | 98% | Random-effects model |
OR, odds ratio; MD, mean difference; BMI, body mass index; CI, confidence interval; EMs, endometriosis; r-AFS, r-American Fertility Society.
Age at surgery: In total, 12 studies [8, 10, 11, 12, 13, 14, 16, 20, 21, 22, 23, 24] with
4611 patients reported the connection between age at surgery and postoperative
endometriosis recurrence. Since the heterogeneity was significant (I
Pooled results for potential risk factors.
Age of menarche: A total of 3 studies [20, 23, 24] reported the
connection between the age of menarche and postoperative recurrence (I
Dysmenorrhea: The effect of dysmenorrhea on recurrence was reported in
6 studies [9, 11, 16, 17, 21, 22]. With significant heterogeneity between the studies
(I
Postoperative pregnancy: There were 7 studies [8, 12, 14, 17, 19, 20, 21]
investigating the effect of postoperative pregnancy on recurrence. As the
heterogeneity among the studies was significant (I
Infertility: Exactly 8 studies [8, 9, 11, 12, 14, 16, 19, 21] covered the
relationship between infertility and postoperative endometriosis recurrence.
The fixed-effects model was used due to the absence of
significant heterogeneity among the 8 studies (I
Endometriosis operation history: There were 4 studies [8, 10, 14, 16]
investigating the effect of previous relevant surgery history on postoperative
endometriosis recurrence. Due to the little heterogeneity among the 4 studies
(I
Combined myoma: The effect of combined myoma on recurrence was reported
in 8 studies [8, 11, 12, 16, 18, 19, 20 23], with no significant heterogeneity
(I
Combined adenomyosis: In total, 7 studies [8, 11, 12, 16, 18, 19, 23]
with 1287 patients reported the connection between combined adenomyosis and the
postoperative recurrence rate of ovarian endometriosis, with minimal
heterogeneity (I
Postoperative medication: There were 8 studies [8, 12, 14, 17, 20, 21, 22, 24]
reporting the effect of postoperative medication on postoperative recurrence,
with significant heterogeneity (I
Family history: The effect of family history on recurrence was reported
in 2 studies [10, 21], which had no significant heterogeneity between them
(I
Preoperative CA-125: In total, 5 studies [10, 12, 14, 22, 23] with
3548 patients reported the connection between preoperative CA-125 and
postoperative recurrence rate of ovarian endometriosis. Since the heterogeneity
was significant (I
Size of ovarian cyst: There were 10 studies [8, 10, 11, 14, 16, 20, 21, 22, 23, 24] concerning
the relationship between cyst size and postoperative recurrence.
With significant heterogeneity among the 10 studies (I
BMI: The effect of BMI on recurrence was reported in 9 studies [8, 10, 12, 14, 16, 20, 21, 22, 23], which had no significant heterogeneity (I
Laterality of EMs: In total, 11 studies [8, 9, 10, 12, 14, 17, 18, 19, 21, 22, 23] with
4586 patients reported the effect of laterality of EMs on the recurrence rate.
Since the heterogeneity was significant (I
r-AFS stage: There were 6 studies [7, 10, 12, 13, 19, 22] that
reported on the difference between stages I and II versus stages III and IV on
endometriosis relapse, which had little heterogeneity (I
There were also 13 studies [7, 8, 9, 10, 12, 13, 14, 15, 16, 17, 18, 19, 22] reporting the difference between
stage III and stage IV on endometriosis relapse, with little heterogeneity
(I
Cystectomy of the affected site: The effect of cystectomy of the
affected site on postoperative recurrence was reported in 3 studies [12, 14, 23], which had significant heterogeneity (I
Sensitivity analysis was performed on postoperative pregnancy, infertility,
combined myoma, combined adenomyosis, endometriosis operation history, ovarian
cyst size, BMI, and r-AFS stage. After removing the data one by one, no
significant change was found in the analysis results, indicating the reliability
of the summary results. Sensitivity analysis was also performed on dysmenorrhea.
Following the exclusion of the study by Wacharachawana et al. [16], the
heterogeneity was statistically changed (I
The results revealed that dysmenorrhea had an effect on endometriosis recurrence, and the recurrence rate of patients with dysmenorrhea was high (OR = 2.44, 95% CI = 1.26–4.73, Z = 2.65, p = 0.008).
r-AFS stage was used for the funnel plot analysis to assess the publication bias. As shown in Fig. 3, all the included studies are in inverted funnel plots with good symmetry. Therefore, the publication bias of the included studies has relatively little impact on the results.
The funnel plot of studies included in the r-AFS stage analysis.
As a benign disease, the exact pathogenesis of endometriosis has not been fully elucidated. The numerous theories include endometrial implantation theory [25], lymphatic and venous dissemination theory [26], coelomic metaplasia theory [27], and immunity and inheritance theory [28]. However, it is difficult to explain the exact mechanisms of endometriosis with one theory, and many hypotheses are linked to the pathogenesis and progression of endometriosis. Currently, the most widely accepted theory of endometriosis is endometriotic implantation caused by retrograde menstruation, in which endometrial glandular epithelial and mesenchymal cells are ectopically implanted in the ovary by retrograde menstruation, where they continue to grow and spread to form ectopic lesions [6]. Endometriosis does not, however, occur in all women with retrograde menstruation. The hypothesis of abnormal eutopic endometrium partly explains why most women of childbearing age have retrograde menstruation and only a few develop endometriosis, which constitutes an important supplement and extension to the theory of ectopic implantation [29, 30]. This hypothesis may provide a satisfactory explanation for the same mutations carried by the true and ectopic endometrial tissues in patients. However, in the study by Koppolu et al. [31], no variants were found common to both tissues from the same patient, thus not supporting this hypothesis.
Recurrence is one of the many difficulties in the clinical management of endometriosis. Identifying the factors related to postoperative endometriosis recurrence can provide insights into the individualized treatment of endometriosis to help clinicians determine the risk of recurrence and formulate treatment plans from the perspective of personalized medicine.
Age at surgery: Studies have revealed that the endometriosis recurrence rate decreases with age, while the age at the time of surgery is a risk factor for recurrence. Higher estrogen levels in young women induce persistent growth of residual lesions. The study by Yang et al. [32] also demonstrated that younger age at surgery was a significant risk for postoperative recurrence of ovarian endometriosis. Therefore, more attention should be attached to young female patients, and the number of follow-up visits should be increased.
Family history: This study suggests that family history has some impacts on postoperative endometriosis recurrence and confirms the existence of genetic trait changes in the pathogenesis of endometriosis. The genetic differences lead to molecular biological differences between the eutopic endometrium of patients and normal subjects.
Preoperative CA-125: Carbohydrate antigen CA-125 is a glycoprotein derived from coelomic epithelial cells and expressed in normal tissues, which could offer some insights into the diagnosis of endometriosis [33]. The above studies concluded that postoperative CA-125 was a risk factor for postoperative endometriosis recurrence. However, CA-125 is a sensitive but poorly specific indicator, and many other diseases also cause marked increases in CA-125. Therefore, CA-125 can be used as a follow-up index but not a diagnostic index of postoperative recurrence. In addition, CA-125 levels can vary with the menstrual cycle. An increased sample size is needed to improve the accuracy of this study.
Laterality of EMs: This study revealed that the recurrence rate of patients with bilaterality of cysts was significantly higher than that of patients with unilaterality. It is speculated that bilateral cysts may increase the possibility of residual lesions. This conclusion is consistent with the current clinical consensus but contrary to the conclusion of Han et al. [34] and Koga et al. [35], suggesting no significant association between bilateral cysts and postoperative recurrence. Therefore, it is necessary to expand the sample for further study.
Presence of adenomyosis: The results revealed that adenomyosis was also a risk factor for postoperative endometriosis recurrence as the pathogenesis of endometriosis and adenomyosis overlap [36, 37]. Therefore, the correct identification of endometriosis combined with adenomyosis and the formulation of personalized long-term postoperative management plans play a crucial part in recurrence prevention.
Presence of myoma: Fibroids in the uterus were regarded as estrogen-dependent tumors. An animal model of fibroid xenotransplantation revealed that steroids, including estradiol and progesterone, were necessary for tumor growth [37]. Therefore, the high estrogen in patients with uterine fibroids affects the endometriosis recurrence rate. Clinically, long-term management plans and regular follow-ups should be made for postoperative patients with uterine fibroids to prevent recurrence.
Previous endometriosis-related surgery: The results suggested that patients with previous endometriosis surgeries had a higher postoperative recurrence rate, possibly because of the more aggressive endometriosis type in patients with previous related surgeries. The mainstream view is that whether the lesion is completely removed during the operation determines the postoperative recurrence. Therefore, surgeons should remove the lesion as much as possible during the operation rather than relying on postoperative adjuvant drugs. Radical surgery is recommended for patients at high risk of recurrence and without fertility requirements.
r-AFS stage: r-AFS stage is determined based on endometriosis lesion size, location, degree of adhesion, etc. [1, 38]. This study concluded that the postoperative recurrence rate was higher with higher surgical scores and more severe stages. In short, extensive lesion adhesions during surgery tend to cause residual lesions and increase the risk of recurrence. Therefore, through clinical analysis, the r-AFS stage is of certain reference significance to evaluate the prognosis and disease severity. When selecting treatment plans, the patient’s r-AFS stage should be considered.
Postoperative pregnancy: The combined data analysis confirmed the benefit of postoperative pregnancy in preventing endometriosis recurrence, possibly because pregnancy and postpartum lactation can lead to ovarian non-ovulation and endometrium decidual-like changes, causing the necrosis of endometriosis lesions and reducing the postoperative recurrence rate [39]. Therefore, Vercellini et al. [40] also found that after endometriosis surgery, patients with fertility requirements were preferably impregnated sooner.
Postoperative medication: The integrated data revealed that patients treated with medication had a lower risk of endometriosis recurrence. Gonadotropin-releasing hormone agonist (GnRHa) can effectively inhibit the secretion of endogenous gonadotrophins and lead to a hypoestrogenic state in vivo, causing deciduosis of endometrium and necrosis of endometriosis lesions [41]. The study by Adachi et al. [42] showed that the endometriosis recurrence rate was significantly reduced after 24 months of dienogest (DNG) administration. In recent years, as an important means of long-term maintenance management after surgery, drug therapy has become a common clinical treatment. In the meantime, Chiu et al. [43] also showed that postoperative adjuvant plus maintenance therapy might be the most effective intervention to prevent endometriosis recurrence. However, the variety of drugs and the complex action mechanisms require more accurate and extensive studies to clarify the efficacy and safety of drugs [44].
Age of menarche: Ponomarenko et al. [45] concluded that some single nucleotide polymorphisms (SNPs) for the age of menarche were associated with endometriosis. Meanwhile, Vercellini et al. [40] found in their study on endometriosis that the early age of menarche in modern women was closely related to the high incidence of endometriosis. However, this meta-analysis suggested that the age of menarche had no significant effect on postoperative endometriosis recurrence. The increased incidence of contemporary endometriosis may be attributed to the improved detection rate. The understanding of the disease is gradually deepened, and the disease becomes treatable, thus increasing the detection rate of endometriosis.
Infertility: The mechanisms by which endometriosis induces infertility are multifactorial, involving mechanical, molecular, genetic, and environmental causes. New research has found that changes in gene expression and gene defects are also responsible [46]. However, infertility could be caused by a variety of factors, the most common one being female pelvic inflammatory diseases, which cause the fallopian tubes to adhere to prevent the ovum and sperm from bonding and have no significant association with endometriosis incidence. Therefore, the fallopian factor accounting for a large proportion of infertility may affect the accuracy of the study results.
Cystectomy of the affected site: It is generally accepted in clinical practice that lesion resection alone leads to an increased likelihood of endometriosis recurrence. However, the combined evidence here suggested that postoperative recurrence rates were not significantly increased in patients with lesion resection alone.
Cyst size: Jiang et al. [47] revealed that greater cyst diameters led to higher possibilities of endometriosis recurrence. On the one hand, the intraoperative residual lesions of large cysts are more likely to conceal the potential factors for recurrence. On the other hand, large cysts often represent greater ectopic cell vitality, thus increasing the recurrence rate. Nevertheless, the findings revealed no association between cyst size and recurrence. We concluded from the overall review that the possible cause was the lack of an objective standard for the intraoperative cyst size evaluation, which was mainly based on the subjective judgment of the surgeon, thus bringing errors to cyst size calculation. Therefore, improving the objective evaluation criteria for cyst size is suggested to increase the accuracy of results.
BMI: As the endometrial tissue grows outside the uterine cavity, the progesterone and estrogen signals are interfered with, leading to progesterone resistance and estrogen dominance in patients with endometriosis [27, 48]. Hormone disorders may lead to obesity in the patients and thus increased BMI. It can be inferred that BMI may be related to the recurrence of endometriosis. The study of the relationship between ovarian endometriosis and obesity here revealed no significant association between BMI and endometriosis recurrence.
Dysmenorrhea: Previous studies have shown that the direct and indirect effects of local bleeding in endometriosis tissue, the actions of celiac inflammatory cytokines, and stimulation or direct infiltration of pelvic floor nerve may be the causes of dysmenorrhea [49]. Meanwhile, patients with dysmenorrhea generally have more extensive lesions, which are difficult to completely remove by surgery, thus explaining the increased recurrence rates in patients with dysmenorrhea. However, this analysis suggested that dysmenorrhea was not a risk factor for recurrent endometriosis, contrary to previous findings by Chon et al. [22]. After removing the study by Wacharachawana et al. [16] with heterogeneous source, we found that patients with dysmenorrhea had a higher recurrence rate after surgery (OR = 2.44, 95% CI = 1.26~4.73, Z = 2.65, p = 0.008), which was consistent with the research findings of Chon et al. [22]. After carefully studying the research by Wacharachawana et al. [16], no cause for heterogeneity was found. Therefore, an increased sample is crucial for further exploration to enhance the reliability of the results.
In summary, the pooled evidence suggested that the following risk factors were significantly associated with endometriosis recurrence, namely, age of surgery, family history, preoperative CA-125, laterality of EMs, presence of adenomyosis, presence of myoma, previous endometriosis-related surgery, r-AFS stage, postoperative pregnancy, postoperative medication. The age of menarche, infertility, the extent of surgery, cyst size, BMI, and dysmenorrhea were not associated with endometriosis recurrence. The risk factors for relapse could be identified, and individualized long-term management and follow-up could be tailored to patient conditions. The sample size can be increased for large-scale studies to confirm the inevitable limitations of this meta-analysis.
All data generated or analyzed during this study are included in this published article.
JD and HD: conception and design of the research. CS: wrote and revised the manuscript. CS and JY: analysis and interpretation of the data and statistical analysis. All authors contributed to editorial changes in the manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work. All authors read and approved the final manuscript.
Not applicable.
The authors would like to thank all those who have contributed to the preparation of this manuscript.
This work was supported by the grants from National Natural Science Foundation of China (No. 82201820).
The authors declare no conflict of interest.
Publisher’s Note: IMR Press stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.