Endometrial carcinoma risk prediction and gynecological comorbidity in adenomyosis: a 5648-case study

A total of 5,648 women with adenomyosis, diagnosed by histopathological examination of specimens from both uterine-sparing and non-uterine-sparing surgeries, were enrolled in this study. A subset of these patients presented with concurrent gynecological conditions, the detailed classification of which is shown in Table  1 (patients with multiple concurrent conditions were counted once per respective category). Isolated adenomyosis (without other gynecological pathologies) was noted in 942 cases (16.7%), while 82.3% (4,647/5,648) of patients had at least one coexisting gynecological disease, and 31.9% (1,806/5,648) had two or more concurrent gynecological pathologies (Table 2 and 3 ). Table 1 Categorical description of concurrent gynecological conditions Benign Premalignant Malignant Leiomyomas Endometrial atypical hyperplasia Endometrial cancer Endometriosis Borderline ovarian tumer Cervical cancer Endometrial polyps Cervical intraepithelial neoplaisa Uterine sarcoma Endometrial simple hyperplasia Ovarian cancer Endometrial complex hyperplasia Benign ovarian cyst Table 2 Coexistence of benign or premalignant gynecological pathologies with adenomyosis Benign or premalignant gynecological pathologies No (% and 95% CI) Coexistence with leiomyomas 3227 (57.1%) Coexistence with endometriosis 605 (10.7%) Coexistence with endometrial polyps 432 (7.6%) Coexistence with endometrial simple and complex hyperplasia 191 (3.4%) Coexistence with atypical hyperplasia 234 (4.1%) Coexistence with benign ovarian cyst 792 (14.0%) Coexistence with borderline ovarian tumer 57 (1.0%) Coexistence with cervical intraepithelial neoplaisa (CIN II/III) 469 (8.30%, 7.58% – 9.02%) This table describes the frequency of benign and premalignant gynecological pathologies in 5648 women with adenomyosis. No: number of cases, %: percentage, 95% CI: 95% confidence interval Table 3 Coexistence of adenomyosis with two or more gynecological pathologies Gynecological pathologies No (%) Total number of cases with coexisted pathology 4647/5648 (82.3%) Coexistence with leiomyomas and endometriosis 356 (6.3%) Coexistence with leiomyomas and benign ovarian cyst 249 (4.4%) Coexistence with leiomyomas and borderline ovarian tumer 23 (0.4%) Coexistence with leiomyomas and ovarian cancer 99 (1.8%) Coexistence with leiomyomas and endometrial polyps 254 (4.5%) Coexistence with leiomyomas and endometrial hyperplasia 106 (1.9%) Coexistence with leiomyomas and atypical hyperplasia 114 (2.0%) Coexistence with leiomyomas and endometrial cancer 344 (6.1%) Coexistence with leiomyomas and uterine sarcoma 22 (0.4%) Coexistence with leiomyomas and cervical intraepithelial neoplaisa 210 (3.7%) Coexistence with leiomyomas and cervical cancer 207 (3.7%) This table describes the incidence of more than one gynecological pathologies in 5648 women with adenomyosis. No: number of cases, %: percentage Categorical description of concurrent gynecological conditions Coexistence of benign or premalignant gynecological pathologies with adenomyosis This table describes the frequency of benign and premalignant gynecological pathologies in 5648 women with adenomyosis. No: number of cases, %: percentage, 95% CI: 95% confidence interval Coexistence of adenomyosis with two or more gynecological pathologies This table describes the incidence of more than one gynecological pathologies in 5648 women with adenomyosis. No: number of cases, %: percentage The mean age of patients at the time of surgery was 49.0 ± 9.0 years. Patients who underwent hysterectomy had a significantly higher mean age (50.0 ± 9.0 years) compared with those in the uterine-sparing surgery group (40.0 ± 7.0 years) ( P  = 0.002). Among all concomitant gynecological pathologies, uterine leiomyoma was the most prevalent (57.1%, 3,227/5,648), followed by endometrial cancer (11.81%, 95% CI: 10.97% – 12.65%), endometriosis (10.7%), cervical cancer (9.67%, 95% CI: 8.90% – 10.44%), cervical intraepithelial neoplasia (CIN; 8.30%, 95% CI: 7.58% – 9.02%), benign ovarian cysts (7.7%), endometrial polyps (7.6%), endometrial atypical hyperplasia (4.1%), ovarian cancer (3.9%), and endometrial hyperplasia (3.4%) (Tables2  and 4 ). All patients with CIN II/III and cervical cancer underwent human papillomavirus (HPV) testing, with HPV-positive rates of 98% and 97.4% in these two subgroups, respectively (Table  5 ). Table 4 Coexistence of gynecological malignancy pathologies with adenomyosis Gynecological malignancy pathologies No (% and 95% CI) Coexistence with endometrial cancer 667 (11.81%, 10.97% – 12.65%) Coexistence with cervical cancer 546 (9.67%, 8.90% – 10.44%) Coexistence with uterine sarcoma 42 (0.7%) Coexistence with ovarian cancer 219 (3.9%) This table describes the frequency of gynecological malignancy pathologies in women with adenomyosis. No: number of cases, %: percentage, 95% CI: 95% confidence interval Table 5 Age and HPV infection status in patients with CINII/III and cervical cancer Total number CINII/III Cervical cancer 469 546 Age 51.23 ± 8.79 50.49 ± 8.64 HPV positive rate 460(98.0%) 532 (97.4%) This table describes the age and HPV positive rate of patients with CIN and cervical cancer Coexistence of gynecological malignancy pathologies with adenomyosis This table describes the frequency of gynecological malignancy pathologies in women with adenomyosis. No: number of cases, %: percentage, 95% CI: 95% confidence interval Age and HPV infection status in patients with CINII/III and cervical cancer This table describes the age and HPV positive rate of patients with CIN and cervical cancer Subgroup analyses further revealed that patients with adenomyosis complicated by endometrial carcinoma had significantly higher age and postmenopausal status than those without endometrial carcinoma (all P  < 0.001, Table  6 ). Stratified analyses by age and menopausal status demonstrated a significant increase in the proportion of adenomyosis patients with concurrent endometrial carcinoma with advancing age: the proportion was 5.56% in patients aged < 45 years and reached 20.91% in those aged ≥ 60 years. The proportion of endometrial carcinoma was also significantly higher in postmenopausal patients than in premenopausal counterparts (17.68% vs. 6.72%) (Table 7 ). Table6 Baseline characteristics comparison between endometrial cancer and Non-Endometrial cancer patients Endometrial cancer Non-endometrial cancer P value Age 53.28 ± 9.36 48.97 ± 8.63 < 0.001 Menopausal status 464/667(69.57%) 2161/4981(43.39%) < 0.001 Table7 Endometrial cancer proportion in subgroups stratified by age and menopausal status Age(year) No (%) 95% CI P value Menopausal No (%) 95% CI P value < 45 84/1486 (5.56%) 4.48%–6.83% < 0.001 Yes 464/2625 (17.68%) 16.22%–19.14% <0.001 45 ~ 59 441/3483 (12.66%) 11.56%–13.77% No 203/3023 (6.72%) 5.82%–7.61% ≥ 60 142/679 (20.91%) 17.85–23.97% This table describes the proportion of endometrial cancer in each subgroup following stratification by age and menopausal status. No: number of cases, %: percentage, 95% CI: 95% confidence interval Baseline characteristics comparison between endometrial cancer and Non-Endometrial cancer patients Endometrial cancer proportion in subgroups stratified by age and menopausal status 84/1486 (5.56%) 4.48%–6.83% This table describes the proportion of endometrial cancer in each subgroup following stratification by age and menopausal status. No: number of cases, %: percentage, 95% CI: 95% confidence interval Given that age and menopausal status were identified as independent risk factors for endometrial carcinoma in patients with adenomyosis, binary logistic regression analysis was conducted to develop predictive models for this malignancy in the study cohort. The constructed models included single-factor models (age-only, menopausal status-only) and dual-factor combined models (age-menopausal status). Furthermore, the composite index of age and menopausal status was validated to have robust predictive efficacy for endometrial carcinoma risk in patients with adenomyosis (AUC = 0.772, Table 8 ). Table8 Sensitivity, specificity, correct classification rate and AUC values for different models of linear discriminant analyses for human data. AUC: area under the curve Model No. Model Sensitivity Specificity Sensitivity + Specificity Correct rate AUC 1 Age 0.762 0.685 1.447 0.698 0.756 2 Menopause status 0.696 0.566 1.262 0.581 0.598 3 Age + Menopause 0.792 0.703 1.495 0.711 0.772 Sensitivity, specificity, correct classification rate and AUC values for different models of linear discriminant analyses for human data. AUC: area under the curve In addition, abnormal ectopic endometrial lesions were identified in 50 patients with adenomyosis, including 30 cases of ectopic endometrial atypical hyperplasia, 3 cases of ectopic adenocarcinoma, and 17 cases of atypical polypoid adenomyoma (APA) (Table 9 ). Stratified analyses by menopausal status also showed that endometrial cancer and ovarian cancer were more prevalent in postmenopausal women, whereas benign gynecological lesions were predominantly observed in younger premenopausal women (all P  < 0.05). Table 9 Ectopic endometrial pathologies in adenomyosis Ectopic endometrium pathologies No (%) Normal ectopic endometrium 5598 (99.0%) Atypical hyperplasia arising in adenomyosis 30 (0.53%) Ectopic endometrium carcinoma 3 (0.05%) APA (atypical polypoid adenomyoma) 17 (0.3%) This table describes the frequency of abnormal ectopic endometrial pathologies in 5648 women with adenomyosis. No: number of cases, %: percentage Ectopic endometrial pathologies in adenomyosis This table describes the frequency of abnormal ectopic endometrial pathologies in 5648 women with adenomyosis. No: number of cases, %: percentage

We performed a retrospective analysis of medical records from patients who underwent gynecological surgery at the Obstetrics and Gynecology Hospital of Fudan University between January 2021 and May 2023. The study was conducted in line with the Declaration of Helsinki, and data were retrieved from the hospital’s electronic medical record system and pathological database. Inclusion criteria: All patients underwent gynecological surgery (laparoscopy or laparotomy), with thorough intraoperative exploration of the abdominopelvic cavity to exclude residual lesions. All patients had histologically confirmed adenomyosis, either as a primary diagnosis or an incidental finding during other gynecological procedures. All patients were aged 20 years or older. Exclusion criterion: patients without histological confirmation of adenomyosis, patients under 20 years of age (i.e., adolescents), those with incomplete clinical or pathological data, and patients with non-gynecological malignant tumors. After strict screening, the final cohort included 5,648 cases with histologically verified adenomyosis. Data were extracted and verified by two independent researchers to reduce bias. The primary surgical indications included abnormal uterine bleeding, chronic pelvic pain, uterine enlargement with menstrual irregularities, anemia, and other associated symptoms, uterine leiomyomas, as well as premalignant or malignant gynecological pathologies of the reproductive system. Of these, 558 patients were suspected of having endometrial cancer and 532 of cervical cancer preoperatively. The corresponding primary surgical approaches comprised hysterectomy, uterine lesion resection combined with diagnostic curettage, adnexectomy or cyst enucleation, and radical hysterectomy. Uterine specimens, uterine lesion specimens, curettage specimens, and cyst or adnexal specimens were collected for pathological assessment. Collected data included patient age, surgical approaches, concurrent gynecological pathologies (benign, premalignant, malignant), and pathological characteristics. The final diagnosis of adenomyosis was confirmed by postoperative histological examination (preoperative ultrasound/MRI was performed). All pathological evaluations were independently conducted by two blinded senior pathologists in strict accordance with the criteria established by the International Federation of Gynecology and Obstetrics (FIGO). Pathological diagnosis of endometrial, cervical and ovarian carcinoma was based on the FIGO Classification of 2023 Revised FIGO Staging System for Endometrial Cancer, FIGO Staging System for Cervical Cancer (2018) and FIGO Staging System for Ovarian, Fallopian Tube and Peritoneal Cancer (2014), respectively. This study was approved by the Ethics Committee of the Obstetrics and Gynecology Hospital Affiliated with Fudan University (Approval No.: 2023 − 169). Statistical analyses were performed using SPSS Statistics software (Version 26.0). Categorical variables were presented as absolute numbers and percentages, while continuous variables were expressed as mean ± standard deviation (SD) if normally distributed (e.g., age) or median (interquartile range) if non-normally distributed. The Kolmogorov-Smirnov test was used to verify the normality of continuous variables. For intergroup comparisons, the independent-samples t-test was applied to normally distributed continuous data, the chi-square test to categorical data, and the Mann-Whitney U test to non-normally distributed continuous data. Binary logistic regression analysis was conducted to construct predictive models for endometrial cancer in patients with adenomyosis, including single-factor models (age-only, menopausal status-only) and two-factor combined models (age + menopausal status. Discrimination was evaluated by receiver operating characteristic (ROC) curve analysis, and the area under the ROC curve (AUC) was calculated to quantify predictive performance. A two-tailed P value < 0.05 was considered statistically significant.

The literature has established associations between adenomyosis and various gynecological diseases, and our study further confirms a significantly increased incidence of endometrial cancer in affected patients. Adenomyosis is characterized by the chronic infiltration of endometrial glands and stroma into the myometrium, resulting in uterine enlargement, tissue hyperplasia, and hypertrophy [ 13 ]. The exact pathogenesis remains incompletely understood, though recent studies suggest the involvement of multiple inflammatory factors, sex steroid hormone receptors, extracellular matrix enzymes, growth factors, and neuroangiogenic factors in disease development [ 3 ]. Additionally, genetic alterations in signaling pathways have been identified as potential pathogenic contributors. Previous research indicates that gynecological disease occurrence may be influenced by factors including inflammatory responses, dietary habits, genetic predisposition, and environmental exposures [ 14 ]. In recent years, investigating the shared pathogenesis and causal relationships between adenomyosis and other benign/malignant gynecological diseases has emerged as a key research focus.

Regarding ovarian cancer, a large-scale study from the Netherlands found that the incidence of ovarian cancer was increased in patients with histologically confirmed endometriosis and adenomyosis [ 39 ], whereas contradictory findings from another study suggested no such elevated risk [ 21 ]. According to data from the National Cancer Institute, the lifetime probability of being diagnosed with endometrial cancer is approximately 3.1% in women, compared to less than 2% for ovarian cancer and 0.6% for cervical cancer [ 31 ]. Matalliotakis et al. reported that ovarian cancer accounted for 1.4% of adenomyosis patients, followed by cervical cancer (0.8%) [ 9 ]. Recent research indicates that both adenomyosis and endometriosis may increase endometrial cancer risk, while endometriosis alone may elevate ovarian cancer risk, with neither condition appearing to affect cervical cancer incidence [ 20 ]. To date, limited studies have investigated the association between adenomyosis and cervical cancer risk. In our study, cervical cancer was present in 9.7% of adenomyosis patients, followed by cervical intraepithelial neoplasia (CIN) (8.3%) and ovarian cancer (3.9%). Based on these findings, we cannot conclusively establish that adenomyosis increases cervical cancer risk. This discrepancy may stem from the study’s inclusion of patients who underwent surgery for cervical cancer or CIN, in whom adenomyosis was an incidental finding; another contributing factor is the low uptake of HPV vaccination. Given that endometriosis is a known risk factor for ovarian cancer, and considering that our study did not exclude adenomyosis patients with concurrent endometriosis, the slightly higher observed risk of ovarian cancer may be attributable to this co-occurrence. Adenomyosis-related growth, invasion and angiogenesis may induce malignant transformation [ 40 ], a condition more common in postmenopausal/elderly women and those with benign gynecological comorbidities [ 10 ], accounting for 6.8% of endometrial cancer cases [ 41 ]. Endometrial epithelial carcinogenesis may also drive adenomyosis malignant progression and tumorigenesis [ 10 ]. Recent studies show adenomyosis modulates endometrial cancer aggressiveness: though not an independent prognostic factor [ 42 ], its coexistence with endometrial cancer correlates with better prognoses, with a 2022 study verifying higher survival rates in histologically confirmed cases [ 43 ], possibly due to earlier diagnosis, favorable pathological features and more optimal treatment strategies. However, a large preliminary study found no significant differences in clinical and prognostic characteristics between endometrial cancer patients with and without adenomyosis [ 28 ]. Systematic reviews and meta-analyses confirm adenomyosis and endometrial cancer as distinct pathological entities [ 29 ], while de novo and synchronous malignant transformation of adenomyotic lesions may significantly impact adenocarcinoma prognostic evaluation [ 23 ]. Atypical polypoid adenomyoma (APA), a rare uterine lesion first reported by Mazur in 1981 [ 44 ], may indicate or precede endometrial adenocarcinoma, or coexist with it [ 45 , 46 ]. As a low-malignant-potential, low-incidence endometrial tumor, APA mainly affects reproductive-age patients [ 47 ]: in 237 patients from 11 retrospective studies, 85.5% were premenopausal, 62.9% nulliparous; 5.5% had APA with atypical hyperplasia, 5.9% concurrent endometrial cancer, with overall recurrence and malignant transformation risks of 28.9% and 16.6% respectively [ 48 ]. APA pathogenesis is linked to estrogen-related factors, obesity, diabetes, prolonged estrogen stimulation and HRT [ 47 ], yet its exact mechanism remains unclear due to lack of large-scale retrospective studies. In our 5,648 adenomyosis patients, 30 had adenomyosis-related atypical hyperplasia, 3 ectopic endometrial carcinoma and 17 APA; larger case numbers will enable further exploration of their molecular pathogenic mechanisms. The limitations of this study partly stem from its retrospective design. Given that we did not exclude cases with concurrent ovarian cancer, adenomyosis, and endometriosis, potential bias may exist in the reported incidence of ovarian cancer among patients with adenomyosis. Additionally, the study included patients who underwent surgery for cervical cancer or endometrial carcinoma, in whom adenomyosis was an incidental finding. Specifically, since our study is based on postoperative pathological findings, there may be errors caused by selection and referral bias. Nevertheless, this study has two notable strengths: its large sample size and histopathology-based analytical approach. This rigorous methodology enhances the reliability of our findings and strengthens the validity of the observed associations.

This retrospective study found a significantly elevated co-occurrence of endometrial carcinoma and adenomyosis in surgically treated patients, suggesting that adenomyosis may act as a risk factor for endometrial carcinoma. In contrast, the prevalence of other gynecological comorbidities in adenomyosis patients within this cohort was not significantly different from that in the general population. Due to selection and indication biases inherent to this surgery-based cohort, the high rate of cancer co-occurrence observed herein cannot be generalized to all adenomyosis patients, particularly those receiving conservative management. Additionally, age and menopausal status may serve as valuable indicators for risk stratification of endometrial carcinoma in adenomyosis patients. These findings highlight the need for further investigations in subsequent studies with more diverse adenomyosis cohorts.

Adenomyosis, an estrogen-dependent pathological condition, ranks among the most prevalent benign gynecological disorders in reproductive-aged women [ 1 ]. This condition is pathologically defined by the ectopic infiltration of endometrial glands and stroma into the myometrium, resulting in diffuse or focal uterine enlargement [ 2 – 4 ]. First documented in 1860 by Carl von Rokitansky as “cystosarcoma adenoides uterinum” [ 5 ], its reported incidence ranges from 10% to 70%, with peak prevalence observed in women aged 40–55 years [ 6 ]. Vercellini et al. estimated its potential presence in 30%−60% of hysterectomy specimens [ 7 ]. Emerging evidence has confirmed a rising recognition and diagnosis of adenomyosis in adolescents and young women in recent years. A retrospective analysis of ultrasound findings in 270 adolescent females aged 12–20 years identified 16 cases of ultrasound-diagnosed adenomyosis, accounting for 5.9% of the cohort [ 8 ]. Mounting evidence also indicates that adenomyosis frequently coexists with endometriosis and uterine leiomyomas. Although pathologically distinct, 35%−55% of women with leiomyomas manifest clinical features overlapping with adenomyosis, including irregular vaginal bleeding, dysmenorrhea, and chronic pelvic pain. Notably, adenomyosis has been consistently associated with endometrial hyperplasia, polyps, and endometrial carcinoma [ 9 – 12 ]. Risk factors for adenomyosis include: age in the mid-40s, multiparity, smoking, elevated hormone levels (such as follicle-stimulating hormone (FSH) and prolactin), previous cesarean delivery and other uterine surgeries, which may disrupt the junctional zone [ 5 ]. This study aims to evaluate the co-occurrence rates of benign, premalignant, and malignant gynecological pathologies with adenomyosis among surgically treated patients.