Introduction: Endometrial carcinoma (EC) is the most common gynecological cancer among women, and in more than 90% of cases, the initial manifestation of the disease is postmenopausal bleeding. Unfortunately, despite early diagnosis and treatment, EC often recurs. Among the many serum tumor markers studied, human epididymis protein 4 (HE4) and cancer antigen 125 (CA125) show the most promise as tools for EC diagnosis, prognosis, and monitoring.

Aim: The aim of the current study was to evaluate the clinical usefulness of HE4 and CA125, tested either as single markers or in combination by including them in the Risk of Ovarian Malignancy Algorithms (ROMA) and in Copenhagen index (CPH-I).

Material and methods: In this retrospective study, 1262 women (74 with confirmed EC) were included. The patients with EC had significantly higher values for HE4, CA125, ROMA, and CPH-I (p<0.001) than the healthy women and patients with benign diseases. The subgroup analysis based on the histological type of EC revealed that the highest markers and algorithms were recorded in type II EC group.

Results: The ROC curve analysis showed that the best diagnostic performance for detecting EC among patients with benign diseases was the ROMA index (AUC=0.869; 95% CI: 0.818-0.920), followed by CPH-I (AUC=0.822; 95% CI: 0.757-0.887), and HE4 (AUC=0.816; 95% CI: 0.750-0.881). Tested alone, CA125 presented unsatisfactory results for this purpose. Both algorithms proved to have a correlation with the disease stage and progression better than the markers alone (HR=1.046 vs. HR=1.018).

Conclusion: In summary, the ROMA index, CPH-I, and, to a lesser extent, standalone HE4 testing can supplement imaging methods as reliable tools for diagnosing and distinguishing patients with EC from those with benign conditions. They have potential as prognostic markers for advanced disease and could help gynecological oncologists to develop a therapeutic strategy.

gynecological malignancies, tumor markers

Endometrial carcinoma (EC) is the most common gynecological cancer in women. Over the past 30 years, the number of newly diagnosed cases of EC has increased by over 55%, indicating an increasing trend in its prevalence. Its estimated incidence globally is 8.4/100,000.^[1] Risk factors for endometrial carcinoma include obesity, type 2 diabetes, lack of physical activity, use of estrogen after menopause, increased life expectancy, and the pursuit of organ-preserving surgical treatment for benign uterine diseases.‌^{[2, 3]} Usually, the first clinical symptom is bleeding during the postmenopausal period. Clarke and colleagues^[4] published data from a meta-analysis encompassing 129 studies with 40,790 women (6,358 of them with EC), according to which in 91% (95% CI, 87%-93%) of EC cases, postmenopausal bleeding was the initial clinical manifestation of the disease. The generalized risk for EC among patients with postmenopausal bleeding was determined to be 9% (95% CI, 8%-11%). This makes it possible to detect it early and achieve an 85% 5-year survival rate. Unfortunately, EC often recurs (in 13%-17% of cases), despite early diagnosis and timely treatment.^[5]

Endometrial carcinoma is classified into two categories with different histology and prognosis:

• Type 1 includes the first and second stages of endometrioid endometrial carcinoma. Pathogenetically, it is associated with chronic estrogen stimulation and arises based on atypical hyperplasia. This type of EC is diagnosed relatively early and has a good prognosis.
• Type 2 consists of the third and fourth stages of endometrioid EC and non-endometrioid neoplasms, which develop based on atrophic mucosa and usually have a poorer prognosis – the recurrent EC has a 5-year survival rate of 15%-17%, regardless of treatment.

The diagnostic algorithm for detecting EC includes transvaginal ultrasound examination, endometrial biopsy and, in some cases, hysteroscopy. These examinations demonstrate good sensitivity but have poor specificity (ultrasound examination) or are invasive and painful for patients.^[6] Therefore, different proteins, nucleic acids, or genetic expressions are being actively studied to improve the medical management of this disease, even though there are currently no established diagnostic or prognostic blood biomarkers for EC in clinical practice, nor predictive markers for response to systemic treatment or monitoring of recurrence.^[7]

Human epididymis protein 4 (HE4) is an acidic whey protein, first isolated from the epithelium of the distal epididymis.^[8] Its biological function is not fully understood, but studies report data that it is overexpressed in >90% of EC cases, accelerating tumor proliferation, invasion, and growth.^{[9, 10]} HE4 shows better sensitivity and specificity compared to the traditionally used tumor marker cancer antigen (CA125) for detecting EC, and it has been established that its concentration correlates with histopathological markers for disease severity, survival, and recurrence.‌^[11] In a meta-analysis conducted by Li and colleagues^[12] on 23 studies, HE4 had a pooled sensitivity of 65%, specificity of 91%, and an area under the curve (AUC) of 0.84 for diagnosing EC compared to a control group of healthy women and patients with benign disease. Similar results are reported by Liu et al. in a meta-analysis of 17 studies with a pooled sensitivity of 65%, specificity of 91%, and an AUC of 0.75.^[13] Another meta-analysis by Li et al. that included data from 12 studies showed that HE4 was more capable of diagnosing EC patients than CA125.^[14] The ability to make firm recommendations is limited by the substantial data heterogeneity found in all three of the cited meta-analyses.

Despite the promising results to date, the reported sensitivity of HE4 varies significantly between studies. Reasons for this may include a small number of included patients, a wide variety of cut-off values adopted by research teams, and the choice of different control groups in different studies, making comparisons challenging.

There is an intensive search for different diagnostic approaches and the combination of imaging methods with biomarker testing to improve the diagnosis and preoperative staging of EC.^[15] There is no definitive answer to the question of whether the simultaneous determination of serum concentrations of CA125 and HE4 would have an advantage over their individual measurement. There are also few studies analyzing the applicability of diagnostic algorithms such as the Risk of Ovarian Malignancy Algorithm (ROMA) and the Copenhagen Index (CPH-I), established in epithelial ovarian carcinoma in patients with EC. Further research is needed to determine whether these or similar models could be validated for use and could have clinical significance in patients with EC.^[16]

The aim of the present study was to assess the diagnostic reliability of the serum tumor markers HE4 and CA125, both individually and in combination, with their inclusion in the algorithms ROMA and CPH-I for the diagnosis of EC.

Endometrial carcinoma is the most common gynecological cancer, and although it is diagnosed at an early stage in a large percentage of cases, it recurs in 15-17% of patients. The ability to make a timely diagnosis, as well as to build a prognosis regarding disease stage and risk of recurrence or progression, are the reasons for the intensive search for new or already established diagnostic modalities for other localizations.

The ROMA index is a calculator initially created and validated to differentiate ovarian carcinoma from benign diseases by combining CA125, HE4, and menopausal status.‌^{[18, 19]} CA125 is an established serum tumor marker in the diagnosis and follow-up of patients with epithelial ovarian carcinoma. Subsequently, it is often applied as a marker for EC detection. Studies have reported its elevation in 24.6% of patients with EC but only in 10% of patients with early-stage disease.^[20] In recent years, HE4 increasingly stands out as a promising biomarker for early detection, diagnosis, prognosis building, and recurrence monitoring in ovarian and endometrial carcinoma.^[21] Its advantage over CA125 is that benign pelvic tumor masses, inflammation, endometriosis, adenomyosis, and pregnancy have less of an impact on its serum concentration.

CPH-I was validated later than the ROMA index in patients with ovarian carcinoma, again using serum concentrations of HE4 and CA125 but considering the patient’s age instead of menopausal status.^[17] Publications from recent years report its superiority over the ROMA index and its implementation in diagnostic nomograms.^[22] In any case, however, to date, there are very few studies reporting data on the application of the ROMA index and/or CPH-I in diagnosing patients with EC.

The current study analyzed the diagnostic potential of two serum tumor markers – HE4 and CA125 – and their combination in the algorithms ROMA and CPH-I. The results demonstrated the best diagnostic performance of the ROMA index (AUC=0.869), followed by CPH-I (AUC=0.822) and HE4 (AUC=0.816) for distinguishing benign processes from EC. Unlike its significance in epithelial ovarian carcinoma, CA125 has no diagnostic effectiveness in EC and cannot independently distinguish benign from oncological disease. Its addition to HE4, however, enhances its clinical significance.

Our data confirm the results from meta-analyses by Li et al., Liu et al., and Li J et al.^[12–14] regarding the diagnostic performance of HE4 in detecting EC. HE4 has good specificity, allowing it to discriminate EC from benign diseases. Unfortunately, the diagnostic sensitivity is not sufficient (65%-71%), limiting its application in early stages. Our data confirm the results from meta-analyses regarding the lack of diagnostic effectiveness of standalone application of CA125 in EC (AUC=0.64 in our study, AUC=0.58 in Li et al.’s meta-analysis).

In conducting the subgroup analysis, our study did not prove sufficient potential of the markers, neither individually nor in combination, to distinguish the histological type of EC, although higher values were found in the group of non-endometrioid carcinoma (mainly in patients with serous type carcinoma). Similar data are reported by Sun et al.^[23] In their study, HE4 is significantly higher in papillary serous carcinoma of the endometrium compared to patients with endometrioid carcinoma of the uterus.

When the relationship between their elevation and disease stage was examined, HE4 and the combination of the two markers showed a good correlation with EC advancement.^[24] This opens a horizon for building prognosis and prediction before treatment initiation. Currently, staging is performed according to FIGO criteria based on postoperative pathological assessment, and about 15%-20% of patients are clinically staged as earlier stage before treatment. ^[25] Therefore, proper preoperative evaluation is extremely important, especially in cases of women where fertility preservation is aimed or in inoperable patients. Additionally, proper preoperative staging can help in decision-making regarding neoadjuvant chemotherapy or the inability to map sentinel lymph nodes in advanced disease.

In our patient cohort, both the standalone use of markers HE4 and CA125, and even more explicitly the calculated indices with the addition of multivariate parameters such as age or menopausal status (ROMA and CPH-I), demonstrate potential as prognostic markers in terms of discrimination between early and late disease stage, as well as risk of progression.

To date, there are few scientific reports regarding the applicability of the ROMA index and CPH-I algorithms in patients with EC. The study published by Wang et al. in 2022 regarding the predictive significance of the ROMA index for advanced EC before surgical treatment compares biomarker data with several other variables such as age, body mass index, presence of type II diabetes, hormonal status, hypertension, tumor size, and degree of invasion.^[26] The authors conclude that the ROMA index, as a standard, reliable, easily measurable laboratory test, could be used by gynecologic oncologists as a tool to select an appropriate therapeutic plan for patients with advanced EC before surgical treatment.

Our study also established the superiority of the ROMA index, determined preoperatively, over the standalone determination of HE4 and CA125. The results regarding CPH-I show comparable diagnostic significance to that of the ROMA index. The advantage of using CPH-I should be sought in its independence from the hormonal status of patients (especially in the perimenopausal age group) and its closer correlation with HE4 values. The relationship of HE4 with age and the increase in its values, especially after 60 years of age, when the peak incidence of EC occurs, is known. Another advantage of CPH-I would be its ability to be easily implemented in the laboratory information system and automated result generation.

At the time of this publication, we did not find any reported data from another study evaluating the application of CPH-I in a cohort of patients with EC. For this reason, although highly optimistic, our results require validation of the algorithm in a larger group of patients in a prospective study design.

In summary, the ROMA index, CPH-I, and to a lesser extent, the standalone testing of HE4 can complement imaging methods as a reliable means of diagnosing patients with EC and distinguishing them from benign diseases. They have potential as prognostic markers regarding advanced disease and could assist gynecological oncologists in developing a therapeutic strategy.