Utility of HE4 in the Evaluation of Adnexal Masses Among Premenopausal and Postmenopausal Women

Adnexal masses, encompassing a spectrum of benign and malignant conditions, present a diagnostic challenge in clinical practice. The timely and accurate assessment of these masses is paramount for guiding appropriate management strategies and ensuring optimal patient outcomes. In this context, Human Epididymis Protein 4 (HE4) has emerged as a promising biomarker, offering valuable insights into the differential diagnosis and risk stratification of adnexal masses. This study aims to elucidate the evolving role of HE4 in the evaluation of adnexal masses by delineating the diagnostic accuracy, prognostic significance, and clinical implications of HE4 alone or as a part of an algorithm in this context. This study was conducted in the Department of Obstetrics and Gynaecology of JIPMER from September 2017 to January 2019 among women who presented with adnexal mass to gynaecological OPD and were planned for surgery. CA-125 levels were estimated using the direct chemiluminescence method, and HE4 levels were estimated using the ELISA method. RMI and ROMA scores were calculated, and results were compared with the final histopathological report. In the present study, among 160 patients studied, the average age of the study population was 45 years. On histopathology, 59% were reported as benign, whereas 5% and 35% of patients were borderline and malignant, respectively. Among malignant tumours, the majority were epithelial tumours (92%), and among epithelial, the majority were serous (67.6%). In the present study, the sensitivity and specificity of the RMI and ROMA models in the prediction of malignancy was 87.7%, 71.6% and 93.8%, 81.1%, respectively. HE4 independently predicted malignancy in the adnexal masses when used with a cut-off of 70 pmol/L for premenopausal women and 140 pmol/L for post-menopausal women. The ROMA had good diagnostic power in predicting the risk of malignancy in an adnexal mass, with high sensitivity and AUC curve compared to RMI in this study. HE4 independently proved to be useful for predicting ovarian malignancy.

CA-125, Epithelial ovarian cancer, HE4, RMI, ROMA, Ultrasonography

Adnexal masses, encompassing a spectrum of benign and malignant conditions, present an important diagnostic challenge in clinical practice. The timely and accurate assessment of these masses is paramount for guiding appropriate management strategies and ensuring optimal patient outcomes. The landscape of adnexal mass evaluation is continuously evolving, propelled by advancements in molecular biology, imaging technology, and biomarker research. In this context, Human Epididymis Protein 4 (HE4) has emerged as a promising biomarker, offering valuable insights into the differential diagnosis and risk stratification of adnexal masses [1]. HE4, a glycoprotein overexpressed in epithelial ovarian cancer cells, has garnered increasing attention for its potential utility as a diagnostic and prognostic marker. Unlike traditional biomarkers such as CA-125, HE4 demonstrates superior specificity, particularly in discriminating between malignant and benign adnexal masses. Moreover, its expression patterns remain relatively unaffected by physiological conditions such as menstruation and endometriosis, enhancing its reliability in clinical settings [2]. Against this backdrop, a nuanced understanding of HE4's strengths, limitations, and clinical applicability is crucial for fostering evidence-based decision-making and personalized patient care. This study aims to elucidate the evolving role of HE4 in evaluating adnexal masses by delineating the diagnostic accuracy, prognostic significance, and clinical implications of HE4 in this context. Furthermore, we explore the potential synergies between HE4 and ultrasonography, CA 125, and risk algorithms (RMI and ROMA) to optimize patient management. Methodology The present study is a cross-sectional study conducted in the Department of Obstetrics and Gynaecology, JIPMER, from 2016 to 2019. Patients who presented to the JIPMER Gynaecology department with adnexal masses and planned for surgery were included in the study, and pregnant women, patients medically unfit for surgery, and those with a previous history of ovarian malignancy were excluded. A total of 160 subjects needed to be studied to compare the diagnostic accuracy of RMI and ROMA using the expected area under the receiver operating characteristic curve (ROC/AUC) with 0.80 and 0.70, respectively. A comparison of two correlated AUCs’ s was done under ROC curves to obtain a sample size with 80% power and a 5% level of significance. After obtaining written informed consent, a detailed history and examination were carried out, and USG, tumour markers, and other relevant investigations were done. RMI and ROMA scores were calculated. Risk of Malignancy Index The risk of malignancy index is calculated with a simplified regression equation obtained from the product of menopausal status score (M), ultrasonographic score (U), and the absolute value of serum CA-125. Menopausal status was scored as 1 point if pre-menopausal and 3 points if post-menopausal. USG features included multi-locular cysts, solid areas, metastases, ascites and bilateral lesions. No feature was scored as 0 points, 1 feature was scored as 1 point, and 2 or more features were scored as 2 points. CA-125 levels were estimated using the direct chemiluminescence method, and HE4 levels were estimated using the ELISA method. RMI value between 25–200 was indicative of moderate risk, and RMI > 200 is considered high risk for ovarian cancer [3]. ROMA Index Risk of ovarian malignancy algorithm (ROMA) incorporates cancer antigen 125 (CA125), human epididymal protein 4 (HE4), and menopausal status. HE4 and CA-125 values were input to the ovarian cancer risk assessment software, then automatically calculated the corresponding ROMA index [4]. The premenopausal calculation formula of the ROMA index is 12 + 2.38 × LN (HE4) + 0.062 6 × LN(CA-125) and the postmenopausal calculation formula of the ROMA index is 8.09 + 1.04 × LN (HE4) + 0.732 × LN (CA-125). ROMA value of ≥ 11.4 was considered a high risk for ovarian cancer in pre-menopausal women and ≥ 22.9 among post-menopausal women. Patients were followed up after surgery with the histopathological report, and the results of CA 125, HE4, RMI and ROMA were correlated with the final histopathological report [4].

of Statistical Analysis The data about categorical variables such as clinical characteristics, co-morbidity, and histopathology were expressed as frequency and percentage. The discrete/continuous variables, such as the age of the patient, serum CA-125 levels, HE4, RMI, ROMA, etc., were expressed as mean with standard deviation or median with range. The diagnostic accuracy of RMI and ROMA was estimated with sensitivity and specificity and compared with ROC curves. Statistical analysis was carried out at a 5% level of significance, and p-value < 0.05 was considered significant. All the analyses were done using IBM SPSS statistics version 19.0. Ethical Approval Institute Ethics Committee clearance was obtained—Reg no: JIP/IEC/2016/1106 dated 30th of January 2017. Informed written consent was obtained from all participants.

The present study comprised 160 patients with an average age of 45 years. According to final histopathology, 95 patients had benign tumours (59.3%), malignant tumours were present in 57 patients (35.6%) and borderline tumours were in 8 patients (5%). The majority of the patients, about 58%, were in the age group of 31–50 years. About 20 (12.5%) tumours were found among women less than 30 years old, and 41% of tumours were in the post-menopausal age group. The highest incidence of malignancy was found among women more than 40 years old, and women between 21 and 40 had the lowest incidence (Table 1). Table 1. | Age group (n) | No. of malignancies, n (%) | |---|---| | 60 years (20) | 10 (50%) | | Total | 57 (100%) | Most of the patients presented with symptoms of mass per abdomen, pain abdomen, non-specific gastrointestinal symptoms, or were asymptomatic. Most women were parous (92%) and premenopausal (59%) in the study group. It was observed that among 94 premenopausal women, 23 patients had malignancy (24.4%), and among 66 post-menopausal women, 34 patients had malignancy (51.5%). On ultrasound, among 160 patients, 55% of the patients (n = 88) had solid areas on USG, and 66% (n = 105) were multiloculated cysts. Among 88 patients with solid areas, about 48 patients had malignancy (54.5%), and about 52 out of 105 patients with multiloculated cysts had malignancy (50%). The mean CA125 levels were 210, and the mean HE4 levels were 197 among the 160 population. Histopathology of the surgical specimen revealed that among benign tumours, the majority were serous, mucinous and germ cell tumours. In malignant tumours, the majority were epithelial tumours, of which serous types were common. Among patients with malignant tumours, 53% were in the early stage (stages 1 & 2) and 47% were in the late stage (stages 3 & 4). The average of CA125, HE4, RMI, and ROMA were analysed among benign, borderline, and malignant tumours, and the findings are noted in Table 2. A statistically significant association was established with values of CA125, HE4, RMI, and ROMA in pre and postmenopausal age groups with risk of malignancy (Table 3). Table 2. | Parameter | Benign | Borderline | Malignant | |---|---|---|---| | Numbers (pre-menopausal, post-menopausal) | 95 (67, 28) | 8 (4,4) | 57 (23,34) | | Age (median, range) years | 40 (16–77) | 46.9 (32–65) | 50 (13–75) | | CA125 (median, range) (U/ml) | 56 (9.8–890) | 95.35 (13–1500) | 153 (24–2214) | | HE4 (median, (range)) (pmol/L) | 18 (4.4–287) | 151.00 (42–1120) | 214 (15–1860) | | ROMA (median, range) | 3.4 (0.0–392) | 57.050 (5.5–99.0) | 77 (5–99.8) | | RMI (median, range) | 116 (13–12600) | 544.50 (96–13500) | 909 (77–19926) | Table 3. | Tumor marker | Histopathology | Total number of cases (n = 94) | Statistical significance | || |---|---|---|---|---|---| | Benign | Borderline | Malignant | ||| | CA125 values (pre-menopausal) | ||||| | 200 U/ml | 4 (30.8%) | 1 (7.7%) | 8 (61.5%) | 13 | | | CA125 values (post-menopausal) | ||||| | 35 U/ml | 18 (34%) | 2 (3.8%) | 33 (62.3%) | 53 | | | HE4 values (pre-menopausal) | ||||| | 70 pmol/L | 4 (14.8%) | 3 (11.1%) | 20 (74.1%) | 27 (100.0%) | | | HE4 values (post-menopausal) | ||||| | 140 pmol/L | 1 (3.3%) | 2 (6.7%) | 27 (90.0%) | 29 (100.0%) | | | RMI values | ||||| | 200 | 29 (33.7%) | 6 (7.0%) | 51 (59.3%) | 86 | | | ROMA values (pre-menopausal) | ||||| | < 11.4% | 56 (94.9%) | 1 (1.7%) | 2 (3.7%) | 59 (100.0%) | p = 0.001 | | ≥ 11.4% | 11 (31.4%) | 3 (8.6%) | 21 (60%) | 35 (100.0%) | | | ROMA values (post-menopausal) | ||||| | 22.9 | 6 (14.6%) | 3 (7.3%) | 32 (78%) | 41 (100.0%) | There is also a significant association between HE4 postmenopausal and stages of malignancy (p < 0.05), which was not seen with CA 125 (pre or postmenopausal), HE4 premenopausal, RMI, ROMA (pre and postmenopausal). The analysis of agreement between the RMI, ROMA and histopathological risk of malignancy showed a significant agreement between both. The sensitivity and specificity of the RMI and ROMA model in the prediction of risk of malignancy was 87.7%, 71.6% and 93.8%, 81.1%, respectively (Table 4). Table 4. | Variables | Sensitivity (%) | Specificity (%) | AUC | |---|---|---|---| | RMI | 87.7 | 71.6 | 0.847 | | ROMA | 93.8 | 81.1 | 0.921 |

Using biomarkers and prediction models in preoperatively evaluating adnexal masses is essential for ensuring timely referral, appropriate surgical intervention, and improved patient outcomes. Among these tools, CA-125, HE4, RMI, and ROMA are widely studied individually or in combination, with varying results. In the present study, we evaluated CA-125, HE4, RMI, and ROMA to predict the risk of malignancy. ROMA demonstrated a sensitivity and specificity of 93.8% and 81.1%, respectively, compared to RMI, which showed a sensitivity of 87.7% and specificity of 71.6%. These findings underscore the superior diagnostic performance of ROMA in predicting malignancy. The results of our study were compared to available literature, as shown in Table 5. Table 5. | Studies | Nature of tumour | CA 125 Median (range) | HE4 Median (range) | RMI Median (range) | ROMA Median (range) | |---|---|---|---|---|---| | Karlsen et al. [4] | Benign | 28.7 (3–3586) | 28.7 (19–1426)) | 48 (2–17514) | 10.8 (0.9–99.6) | | Borderline | 87.6 (10–6410) | 81.8 (29–1109) | 360 (13–15264) | 32 (2.2–96.9) | | | Epithelial cancers | 647 (10–10000) | 436 (16–15000) | 3330 (6–153432) | 95.7 (0.6–1000) | | | Tobing et al. [10] | Benign | 17.1 (3–293) | 64 (12.7–325) | 30.3 (0–496.8) | 13.3 (0.8–41.5) | | Borderline | va26.4 (10.5–844.5) | 96 (24–193.6) | 168.3 (0–2534) | 24.9 (3–67.2) | | | Malignant Tumour | 138.7 (7.5–2856) | 166.6 (2.5–2427) | 648 (0–25704) | 61 (0–99.9) | | | Present study | Benign | 56 (9.8–890) | 18 (4.4–287) | 116 (13–12600) | 3.4 (0–392) | | Borderline | 95.35 (13–1500) | 151 (42–1120) | 544.50 (96–13500) | 57 (5.5–99) | | | Malignant | 153 (24–2214) | 214 (15–1860) | 909 (77–19926) | 77 (0.5–99.8) | A study conducted in Italy on 104 women with pelvic masses evaluated the performance of ROMA for assessing the risk of epithelial ovarian malignancy. In the premenopausal group, ROMA exhibited a sensitivity of 53.3% and a specificity of 80.6%. In the postmenopausal group, the sensitivity and specificity were higher, at 82.5% and 84.6%, respectively. Notably, this study reported that ROMA achieved greater sensitivity and specificity than CA-125 alone [2]. Similarly, a study involving 457 women with pelvic masses compared RMI and ROMA. At a set specificity of 75%, RMI demonstrated a sensitivity of 84.6%, while ROMA showed a significantly higher sensitivity of 94.3% for distinguishing benign from epithelial ovarian cancers (p = 0.0029). In cases of early ovarian disease, ROMA displayed a sensitivity of 85.3%, significantly outperforming RMI, which had a sensitivity of 64.7% (p < 0.0001) [5]. In another study by Karlsen et al., the diagnostic performance of CA-125, HE4, RMI, and ROMA was assessed. At a set sensitivity of 94.4%, the specificity values for CA-125, HE4, RMI, and ROMA were 62.2%, 63.2%, 81.5%, and 76.5%, respectively. The areas under the curve (AUC) for CA-125, HE4, RMI, and ROMA were 0.854, 0.864, 0.905, and 0.897, respectively, highlighting the robust performance of RMI and ROMA in clinical evaluation [6]. ROMA has proven to be a valuable diagnostic tool for the preoperative work-up of patients with pelvic masses. In a study conducted on 248 patients treated at the Department of Gynaecologic Surgery and Oncology, Stettin (2011–2012), ROMA outperformed CA-125 and HE4 in detecting type II and advanced-stage malignancies. Moreover, ROMA was particularly effective in postmenopausal patients, whereas CA-125 performed better in premenopausal individuals [7]. A multicenter clinical study in the Korean population also demonstrated the diagnostic utility of ROMA. In this study, the area under the curve (AUC) for HE4 and ROMA was higher than that for CA-125 in the receiver operating characteristic (ROC) curve analysis. These findings supported the conclusion that HE4 and ROMA could be independent diagnostic markers for epithelial ovarian cancer [8]. Consistent with other studies, including the present study, the levels of CA-125, HE4, RMI, and ROMA progressively increased from benign to malignant tumours. RMI remains the most extensively studied risk model for predicting malignancy in adnexal masses, with reported sensitivities ranging from 30% to 99.9% and specificities between 40% and 89.6% across various studies [6, 9–11]. In our study, RMI exhibited a sensitivity of 87.7% and a specificity of 71.6%. One significant advantage of ROMA is its reliance solely on tumour markers (CA-125 and HE4) without incorporating ultrasound parameters. This approach circumvents the challenges of subjective ultrasound assessments and the necessity for expert opinions during evaluation. Across various studies, ROMA has demonstrated sensitivities ranging from 74 to 94% and specificities between 75 and 82% [6, 9, 12, 13]. In our study, ROMA achieved a sensitivity of 93.8% and a specificity of 81.1%.

A larger sample size would have strengthened the association between the evaluated parameters and the stages of malignancy, providing a more comprehensive understanding of their diagnostic performance.

The ROMA had good diagnostic power in predicting the risk of malignancy in an adnexal mass, with high sensitivity and AUC curve compared to RMI in our study. Individually, CA125 and HE4 performed well in predicting the risk of malignancy in adnexal masses. HE4 among post-menopausal women also showed significant association with stages of malignancy. Funding An intra-mural grant of 1,50,000 INR from JIPMER, Pondicherry, was used to buy the reagents to assess CA125 and HE4. Declarations Conflicts of Interest Authors do not report any conflicts of interest. Footnotes Key Message ROMA and HE4 independently proved to be better diagnostic models for predicting ovarian malignancy. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

- 1.Braicu EI, Krause CL, Torsten U, Mecke H, Richter R, Hellmeyer L, Lanowska M, Müller B, Koch E, Boenneß-Zaloum J, Ames K, Chekerov R, Hasenbein K, Zimmermann M, Mangler M, Chen F, Tauber R, Sehouli J (2022) HE4 as a serum biomarker for the diagnosis of pelvic masses: a prospective, multicenter study in 965 patients. BMC Cancer 22(1):831. 10.1186/s12885-022-09887-5 [DOI] [PMC free article] [PubMed] [Google Scholar] - 2.Montagnana M, Danese E, Ruzzenente O, Bresciani V, Nuzzo T, Gelati M et al (2011) The ROMA (Risk of Ovarian Malignancy Algorithm) for estimating the risk of epithelial ovarian cancer in women presenting with pelvic mass: Is it really useful? Clin Chem Lab Med 49(3):5215 [DOI] [PubMed] [Google Scholar] - 3.Jacobs I, Oram D, Fairbanks J, Turner J, Frost C, Grudzinskas JG (1990) A risk of malignancy index incorporating CA 125, ultrasound and menopausal status for the accurate preoperative diagnosis of ovarian cancer. Br J Obstet Gynaecol 97(10):922–929. 10.1111/j.1471-0528.1990.tb02448.x [DOI] [PubMed] [Google Scholar] - 4.Moore RG, Brown AK, Miller MC, Skates S, Allard WJ, Verch T, Steinhoff M, Messerlian G, DiSilvestro P, Granai CO, Bast RC Jr (2008) The use of multiple novel tumor biomarkers for the detection of ovarian carcinoma in patients with a pelvic mass. Gynecol Oncol 108(2):402–408. 10.1016/j.ygyno.2007.10.017 [DOI] [PubMed] [Google Scholar] - 5.Moore RG, Jabre-Raughley M, Brown AK, Robison KM, Miller MC, Allard WJ et al (2010) Comparison of a novel multiple marker assay versus the risk of malignancy index for the prediction of epithelial ovarian cancer in patients with a pelvic mass. Am J Obstet Gynecol 203(3):228.e1-228.e 6 [DOI] [PMC free article] [PubMed] [Google Scholar] - 6.Karlsen MA, Sandhu N, Hogdall C, Christensen IJ, Nedergaard L, Lundvall L, Engelholm SA et al (2012) Evaluation of HE4, CA125, risk of ovarian malignancy algorithm (ROMA) and risk of malignancy index (RMI) as diagnostic tools of epithelial ovarian cancer in patients with a pelvic mass. Int J Gynecol Oncol 127:379–383 [DOI] [PubMed] [Google Scholar] - 7.Chudecka–Glaz, Cymbaluk A, et al (2013) Preoperative diagonstic preformance of ROMA in relation to etiopathogenesis of epithelial ovarian tumors. J Mol Biomark Diagn S4–003 - 8.Yon Cho H et al (2015) Comparison of HE4, CA125, and Risk of Ovarian Malignancy Algorithm in the Prediction of Ovarian Cancer in Korean Women. J Korean Med Sci 30:1777–1783 [DOI] [PMC free article] [PubMed] [Google Scholar] - 9.Al Musalhi K, Al Kindi M, Al Aisary F, Ramadhan F, Al Rawahi T, Al Hatali K et al (2016) Evaluation of HE4, CA-125, Risk of Ovarian Malignancy Algorithm (ROMA) and Risk of Malignancy Index (RMI) in the Preoperative Assessment of Patients with Adnexal Mass. Oman Med J 31(5):336–344 [DOI] [PMC free article] [PubMed] [Google Scholar] - 10.Oranratanaphan S, Wanishpongpan S, Termrungruanglert W, Triratanachat S (2018) Assessment of Diagnostic Values among CA-125, RMI, HE4, and ROMA for Cancer Prediction in Women with Nonfunctional Ovarian Cysts. Obstet Gynecol Int 1:1–6 [DOI] [PMC free article] [PubMed] [Google Scholar] - 11.Bouzari Z, Yazdani S, Shirkhani KZ, Abbaszadeh N (2011) Risk of malignancy index as an evaluation of preoperative pelvic mass. Caspian J Intern Med 2(4):331–335 [PMC free article] [PubMed] [Google Scholar] - 12.Tobing MDL, Nugraha RA, Susanto H (2016) Comparison of diagnostic accuracy between the risk of malignancy index (RMI) and the risk of ovarian malignancy algorithm (ROMA) to assume any malignancy in ovarian tumor. Int J Adv Res 4(10):2174–2179 [Google Scholar] - 13.Iyer VR, Lee SI (2010) MRI, CT, and PET/CT for Ovarian Cancer Detection and Adnexal Lesion Characterization. AJR 194:311–321 [DOI] [PubMed] [Google Scholar]