This study retrospectively analyzed 380 patients with ovarian space-occupying who underwent surgical treatment at The Affiliated Huai’an Hospital of Xuzhou Medical University between August 2020 and August 2023. Among them, 344 patients with ovarian space-occupying conformed to the inclusion criteria, 59 were excluded according to the exclusion criteria, 25 had incomplete clinical data, and 260 patients with ovarian space-occupying were eventually enrolled in this study. According to the results of the pathological examination, 110 patients with EOC were assigned as the EOC group (7 cases of endometrial carcinoma, 22 cases of mucinous, 81 cases of serous), and 150 patients without EOC as the N-EOC group (9 patients with non-epithelial OC, including 5 cases of immature teratoma, 3 cases of anaplastic cell tumor, and 1 case of yolk sac tumor among ovarian germ cell tumors, and the others were patients with benign ovarian diseases). The diagnosis and classification of female genital tumors were based on the classification criteria of the World Health Organization (WHO) for female genital tumors in 2020 [13, 14, 15].

The inclusion criteria were as follows: (1) first diagnosis of ovarian space-occupying without any treatment; (2) had accepted surgical treatment and preoperative MRI examination, with definite ovarian space-occupying type determined by postoperative pathology; (3) complete clinical data; (4) $>$18 years old.

The exclusion criteria were as below: (1) complications of other benign or malignant tumors in women; (2) complications of cardiovascular, cerebrovascular, blood system and liver/kidney function diseases, as well as infectious and infectious diseases; (3) complications of endocrine diseases such as diabetes and hyperthyroidism; (4) had received blood transfusions, consumed alcohol, developed acute infections, or used antibiotics or anticoagulants within two weeks before retaining serum tumor markers; (5) a history of abdominal surgery in the past 3 months; (6) menstrual, pregnancy, lactation women.

EOC pathological diagnostic criteria [13, 14, 15, 16, 17]: The diagnosis was made according to the 2020 WHO classification criteria for female genital tumors. Pathologic examination revealed hyperplastic epithelium protruding to form papillae, showing multiple papillary branches, dense papillary hyperplasia, papillary ectodermal or extraperitoneal implants, regular nipple structure and nucleus, atypical hyperplasia of epithelial cells, different nucleus size and shape, and increased chromatin and nuclear division.

The body mass index (BMI), age, tumor family history, gravidity, parity, menopausal status, interleukin (IL)-6, C-reactive protein (CRP), follicle stimulating hormone (FSH), luteinizing hormone (LH), anti-mullerian hormone (AMH), apparent diffusion coefficient (ADC) value, carbohydrate antigen 125 (CA125), and human epididymis protein 4 (HE4) of all subjects were recorded. The enrolled patients had 5 mL of elbow vein blood drawn on an empty stomach in the early morning of the next day of admission. The blood was then centrifuged at 3000 $\times$g for 8 min within 3 h of collection, and the supernatants were kept in a refrigerator at –80 °C for storage and centralized testing. Out of the samples, 3 mL were utilized for enzyme-linked immunosorbent assay (ELISA), while the remaining samples were employed for chemiluminescence immunoassay (CLIA) detection.

ELISA was used to determine the expression levels of IL-6, CRP, FSH, LH and human AMH in the serum of all subjects. CLIA was utilized to measure the expression patterns of CA125 and HE4 in serum. Specific procedures were strictly carried out following the operating instructions of the IL-6 kit (orb1532209, Biorbyt, Waterbeach, Cambridge, UK), CRP kit (LM-EL-1865, LMAI Bio, Shanghai, China), FSH kit (LM-EL-6467, LMAI Bio, Shanghai, China), LH kit (LM-EL-6364, LMAI Bio, Shanghai, China), AMH kit (LM-EL-1041, LMAI Bio, Shanghai, China), CA125 kit (MZ095925, Beetle Biological Products Co., Ltd., Suzhou, Jiangsu, China) and HE4 kit (AT03CHE4, Anti Biotechnology Co., Ltd., Shenzhen, Guangdong, China).

All patients underwent pelvic MRI examination before operation. The intestinal tract was emptied and cleaned before examination. The pelvis was scanned using a 3.0 T MRI scanner (MAGNETOM Prisma, SIEMENS AG FWB, Munich, Germany). The instrument parameters were set as follows: (1) T1-weighted imaging (T1WI): fast spin echo sequence was utilized, the repetition time (TR) = 545 ms, echo time (TE) = 25 ms, the number of signal averages (NSA) = 2, flip angle = 90°, layer thickness = 5 mm, layer spacing = 1 mm, matrix = 256 $\times$ 256, field of view = 300 mm $\times$ 300 mm; (2) T2-weighted imaging (T2WI): fast spin echo sequence was adopted, TR = 4000 ms, TE = 100 ms, NSA = 4, flip angle = 90°, slice thickness = 5 mm, slice spacing = 1 mm, matrix = 256 $\times$ 256, field of view = 300 mm $\times$ 300 mm; (3) Dusion-weighted imaging (DWI): spin echo-planar imaging sequence was applied, TR = 5800 ms, TE = 47 ms, NSA = 4, flip angle = 90°, layer thickness = 5 mm, layer spacing = 1 mm, matrix = 128 $\times$ 128, field of view = 300 mm $\times$ 300 mm, diffusion sensitivity coefficient b value of 0 and 800 s/mm², imaging time of 1 min 22 s. The region of interest (ROI) was obtained, including each solid lesion region with a diameter greater than 3 mm and each cystic region with a diameter greater than 20 mm, and ROIs were placed at the largest level of each region. As per this principle, ROI was put on the DWI image with b = 800 s/mm² and copied to the apparent diffusion coefficient (ADC) map at the same level. The ADC values were determined and averaged after three measurements at the same level and position.

MRI positive was determined when the tumor lump was large and lobulated, with papillary protrusions or solid masses in the cystic cavity, adhesion to adjacent organs, an irregular boundary, and an unclear boundary; the microvascular morphology was uneven, which was substantive or cystic solid, or there were pelvic organ infiltration, lymph node enlargement, ascites, and planting; the T1WI and T2WI signals were complex, with diverse density and apparent enhancement (Supplementary Figs. 1,2).

SPSS 21.0 (IBM Corp., Armonk, NY, USA), MedCalc 19.0 (MedCalc Software Ltd., Ostend, Belgium) and GraphPad Prism 8.01 software (GraphPad Software, San Diego, CA, USA) were utilized for data analysis and graph plotting. The Kolmogorov-Smirnov test was used to assess the conformity of the data to normal distribution. The normally distributed measurement data were exhibited as mean $\pm$ standard deviation (SD), with the t-test used for comparisons between two groups, and one-way analysis of variance (ANOVA) conducted for comparisons among multiple groups. Measurement data of non-normal distribution were expressed as median values (minimum, maximum). The Mann-Whitney U test was used for comparisons between two groups, and the Kruskal-Wallis rank sum test was used for comparisons among groups. Counting data were denoted as the number of cases, and comparisons of binary variables of counting data between two groups were performed using the Chi-square test (Chi-square test/Chi-Square Goodness-of-Fit Test). The receiver operating characteristic (ROC) curve was plotted to assess the diagnostic effectiveness of the parameters and determine the threshold value. The clinical diagnostic value of MRI, HE4, CA125, and their combination in EOC patients was investigated. Test level a = 0.05, and p values were obtained from a bilateral test, with p 0.05 indicating a statistically significant difference.

The clinical data of patients with ovarian tumors (n = 260) who underwent surgical treatment in The Affiliated Huai’an Hospital of Xuzhou Medical University from August 2020 to August 2023 were collected for analysis (Table 1). Among them, 110 patients with EOC were included in the EOC group, and 150 patients without EOC were included in the N-EOC group. There were no statistically significant differences in age, BMI, or menopausal status when comparing the EOC group with the N-EOC group (all p $>$ 0.05). The differences between the EOC and N-EOC groups were statistically significant in terms of tumor family history, gravidity, parity, and levels of serum IL-6, CRP, FSH, LH and AMH, as well as the ADC values of the MRI examinations (all p 0.05).

Clinicopathological characteristics of OC patients were analyzed by pathological examination. The results (Table 2) showed that the EOC group and the N-EOC group exhibited distinct disparities in terms of tumor stage, pathological type, degree of differentiation, tumor location, tumor diameter, and lymph node metastasis (all p 0.05).

Subsequently, the expression levels of CA125 and HE4 in the serum of patients with ovarian space-occupying were determined by CLIA. As shown in Fig. 1 the expression of CA125 in the EOC group [127.33 (21.84, 293.22) U/mL] was dramatically higher than that in the N-EOC group [36.55 (17.91, 150.51) U/mL], and notably higher HE4 expression was also observed in the EOC group [112.16 (42.45, 196.07) pmol/mL] than in the N-EOC group [52.01 (19.10, 160.50) pmol/mL] (p 0.01). These findings suggest high expression patterns of CA125 and HE4 in EOC patients.

Fig. 1.

Carbohydrate antigen 125 (CA125) and human epididymis protein 4 (HE4) were highly expressed in epithelial ovarian cancer (EOC) patients. The expression of CA125 (A) and HE4 (B) was detected by chemiluminescence immunoassay (CLIA), and the Kolmogorov-Smirnov test was used to test for normal distribution, and the non-normally distributed measures were expressed as the median (minimum, maximum), and comparisons between groups were made using the Mann-Whitney U test, *** p 0.01. N-EOC, patients without EOC.

We analyzed the clinicopathological features of EOC patients by pathological examination and analyzed the expression of CA125 in EOC patients with varying clinicopathological features. The results (Table 3) depicted notable differences in CA125 expression in EOC patients with distinct tumor stages, tumor diameters, tumor sites, and pathological types (all p 0.05). Thereafter, HE4 expression in EOC patients with various clinicopathological features was also analyzed. The results (Table 3) demonstrated that there were significant differences in the expression of HE4 in EOC patients with varying tumor diameters, tumor stages, pathological types, and tumor sites (all p 0.05). The ADC value of patients with EOC was measured using a 3.0 T nuclear magnetic resonance scanner, and the ADC value of EOC patients with various clinicopathological characteristics was analyzed. As displayed in Table 3, there were significant differences in ADC values between EOC patients with different tumor diameters (all p 0.05). On top of these results, CA125 and HE4 expression levels in EOC patients were strongly connected with the clinicopathological features of EOC.

The examination using a 3.0 T magnetic resonance scanner revealed the following results (Table 4): there were 95 positive patients, out of which 75 were true positive and 20 were false positive; there were 156 negative patients, out of which 35 were false negative and 121 were true negative. As illustrated in Fig. 2, ROC curve analysis using SPSS found that MRI [area under the curve (AUC) = 0.77] had certain diagnostic value for EOC. Furthermore, ROC curve analysis (Fig. 2, Table 5) found that both CA125 (AUC = 0.85, cut off at 53.23) and HE4 (AUC = 0.90, cut off at 79.51) had certain diagnostic value for EOC. The clinical diagnostic value of MRI combined with CA125 and HE4 (combination) for EOC patients was significantly higher than that of MRI, CA125, or HE4 alone, and the AUC was 0.97 (p 0.05). On the basis of the cut-off value of CA125 and HE4, the patients were categorized into the CA125/HE4 high expression group ($>$53.23)/($>$79.51) and the CA125/HE4 low expression group ($\le$53.23)/($\le$79.51). The incidence of EOC in ovarian space-occupying patients with different CA125 and HE4 expression levels was subsequently analyzed. The results (Table 6) revealed that the CA125 high expression group had a higher EOC prevalence than the CA125 low expression group, and the HE4 high expression group had a higher EOC prevalence than the HE4 low expression group (p 0.05). As a consequence, the prevalence of EOC in ovarian space-occupying patients with highly-expressed CA125/HE4 was higher, and the results also suggested that MRI coupled with CA125 and HE4 had higher clinical diagnostic value for EOC patients.

Fig. 2.

The clinical value of MRI, CA125, HE4, and the combined detection of three indexes in patients with EOC. ROC curves were employed to analyze the clinical diagnostic value of HE4, MRI, CA125, and combined triple-index testing in patients with EOC. ROC, receiver operating characteristic; CA125, carbohydrate antigen 125; HE4, human epididymis protein 4; EOC, epithelial ovarian cancer; MRI, magnetic resonance imaging.