The Emerging Role of HE4 in the Evaluation of Epithelial Ovarian and Endometrial Carcinomas

HE4 (human epididymis protein 4) is overexpressed in both ovarian and endometrial cancers. Levels of the shed HE4 protein are elevated in sera from ovarian and endometrial cancer patients. HE4 is less frequently elevated than cancer antigen 125 (CA 125) in benign gynecologic conditions and is found in a fraction of endometrial and ovarian cancers that lack CA 125 expression. Consequently, HE4 has emerged as an important biomarker that complements CA 125 in discriminating between benign and malignant pelvic masses, monitoring response to treatment, and detecting recurrences of both ovarian and endometrial cancer. The “risk of ovarian malignancy algorithm” (ROMA) incorporates CA 125, HE4, and menopausal status to distinguish benign from malignant adnexal masses, and has been approved by the US Food and Drug Administration to aid in referring patients who are likely to have ovarian cancer to specially trained gynecologic oncologists for surgery. HE4 also promises to augment the sensitivity of CA 125 for detecting early-stage ovarian cancer. In this review, we discuss the discovery and biologic significance of HE4 and evaluate available evidence regarding the utility of HE4 as a biomarker for ovarian and endometrial cancer.

Introduction

Some 22,280 women were diagnosed with ovarian cancer in the United States in 2012, and 15,500 died from the disease.[1] While the cure rate for all stages of ovarian cancer has remained approximately 30%, 5-year survival rates have improved significantly (P < .05)-from 37% in the 1970s to 45% at the turn of the century-as a result of improved surgery and combination chemotherapy with carboplatin and paclitaxel. In contrast to surgical management of cancers at some other sites, gynecologic oncologists attempt to remove as much of an ovarian cancer as possible, even when all macroscopic disease cannot be resected. Prognosis relates to the maximum size of tumor nodules that remain after surgery. Gynecologic oncologists are specially trained to perform “surgical cytoreduction” prior to chemotherapy. Outcomes are improved when primary operations are performed by gynecologic oncologists, but less than half of women are referred to these subspecialists. To facilitate appropriate referral of patients who are likely to have cancer, a “risk of malignancy index” (RMI) was developed in 1990 to identify patients with malignant pelvic masses; it used menopausal status, appearance of the pelvic mass during ultrasonography, and levels of the biomarker cancer antigen 125 (CA 125).[2] In different studies, the sensitivity of RMI for detecting malignant masses ranged from 71% to 88%, and specificity from 74% to 97%,[3] leaving room for improvement in sensitivity, while retaining specificity.

CA 125 is overexpressed in 80% of ovarian cancers. Serum CA 125 levels have been elevated in 50% to 60% of patients with stage I ovarian cancer and in 90% of patients with stage III/IV disease, related to release of CA 125 not only from cancer cells, but also from the inflamed peritoneum. In patients with elevated CA 125 levels, changes in biomarker levels have tracked tumor burden with greater than 90% accuracy. Persistent elevation of CA 125 following chemotherapy has correlated with residual ovarian cancer in 90% of cases, leading to approval of CA 125 by the US Food and Drug Administration (FDA) for detection of disease that has survived primary chemotherapy. Monitoring CA 125 levels in patients with a complete clinical response can detect recurrence of cancer in up to 70% of patients, with an average lead time of 3 to 4.8 months. The clinical value of early detection of disease recurrence has been questioned based on a single study with significant limitations, but it does provide time for patients to receive multiple conventional drugs and to participate in clinical trials with novel agents.[4]

Biomarkers could have the greatest impact on survival as part of an effective screening strategy. When ovarian cancer is detected at an early stage, where the disease is still contained within the ovaries (stage I), 5-year survival rates can approach 90% with optimal surgery and currently available combination chemotherapy. By contrast, ovarian cancer that has spread throughout the peritoneal cavity or outside the abdomen (stages III and IV) is associated with 5-year survival of less than 30%.[1] At present, less than 25% of ovarian cancers are detected in an early stage.[5] Although promising trials are underway in the United States and the United Kingdom, there is no established screening strategy for early detection of ovarian cancer in women at conventional risk for the disease. In women with a strong family history, CA 125 levels and transvaginal sonography (TVS) are recommended for early detection, although there is not yet evidence that this improves prognosis.[5] Prophylactic surgery is recommended for women from BRCA1, BRCA2, or Lynch syndrome kindreds when they have completed their families. Surveillance with CA 125 in premenopausal women is complicated by the fact that antigen levels can be elevated by several physiological and benign conditions, including endometriosis, adenomyosis, uterine fibroids, and normal menstruation, compromising specificity. To achieve a positive predictive value of at least 10% (ie, 10 operations for each case of ovarian cancer detected), a screening strategy must have high sensitivity (> 75%) for preclinical early-stage disease, and an extraordinarily high specificity (> 99.6%).[6] While neither CA 125 nor TVS alone can meet these standards, a two-stage strategy in which a rising CA 125 level triggers TVS in a small fraction of patients has achieved sufficient specificity that only three operations are required for each case of ovarian cancer detected.[7,8] With this strategy, early-stage disease has been detected, but there is still a need to improve sensitivity, since CA 125 testing will miss at least 20% of cases. Use of multiple biomarkers, including human epididymis protein 4 (HE4), could fill this need.

In 2012, an estimated 47,130 women were diagnosed with endometrial cancer and 8,010 women died from this disease.[1] In contrast to the lack of specific symptoms in ovarian cancer, endometrial cancer is typically associated with postmenopausal vaginal bleeding, leading to early-stage (stage I) detection in almost 70% of cases-which makes for an excellent prognosis.[9] While opportunities to improve early detection are not as great as in ovarian cancer, biomarkers for endometrial cancer might aid in prognostication, in the detection of recurrence, and in the monitoring of therapy for metastatic disease. Women with high-grade tumors, deep myometrial invasion, and lymphovascular space invasion have recurrence rates of 20% to 30%.[10] Current methodologies for detecting recurrence rely on imaging and symptom presentation for the diagnosis of recurrent disease.[9] CA 125, the FDA-approved biomarker for ovarian cancer, is elevated in only a small fraction (25%) of recurrent disease.[11]There is a clear need for other biomarkers that might complement CA 125 in the identification of women who are at a higher risk for recurrent endometrial cancer.

In the past decade, HE4 has emerged as a promising biomarker to address the unmet clinical needs in ovarian and endometrial cancer. In this review, we discuss the discovery and biologic significance of the biomarker, and we explore the promise of HE4 for addressing the various clinical needs in these cancers.

HE4: Discovery and Biologic Significance

HE4 was discovered by Kirchhoff et al in 1991 and belongs to the “four-disulfide core” family of proteins, which typically function as proteinase inhibitors.[12]In initial studies, HE4 messenger RNA (mRNA) was localized to the distal regions of the epithelial cells of the epididymal duct, indicating a possible role for HE4 in sperm maturation.[12] Its role as a potential biomarker for ovarian cancer emerged after cDNA comparative hybridization experiments found increased primary expression of HE4 in some ovarian cancers, relative to normal tissues.[13] Further studies using serial analysis of gene expression (SAGE) to analyze ovarian cell lines, tissues, and primary cancers corroborated HE4 as a potential marker that is upregulated in ovarian cancers.[14]

HE4 promotes migration and adhesion of ovarian cancer cells. In in vitro studies, HE4 knockdown resulted in tumor growth inhibition.[15] HE4 overexpression in endometrial cancer cell lines induced cancer cell proliferation in vivo and in vitro, supporting a function for HE4 in tumor progression.[16] In recent studies by LeBleu et al, HE4 was found to serve as a protease inhibitor, decreasing the activity of serine proteases Prss35 and Prss23, which degrade the type I collagen that accumulates in kidney fibrosis.[17] Fibrosis was inhibited in three mice models when HE4-neutralizing antibodies were administered, implicating HE4 as a therapeutic target in renal fibrosis.[17] HE4 may have an additional role in maintaining the innate immunity of the respiratory tract and the oral cavity.[18]

Tissue Expression

As with most biomarkers, HE4 is expressed in a number of normal and malignant tissues. Galgano et al completed a comprehensive study of HE4 mRNA and protein expression in normal and malignant tissues.[19] Normal glandular epithelium of the breast, female genital tract, epididymis, vas deferens, distal renal tubules, respiratory epithelium, colonic mucosa, and salivary glands all show HE4 immunoreactivity.[19] Among normal tissues, the highest levels of HE4 were found in the trachea and salivary gland. Lower expression was found in lung, prostate, pituitary gland, thyroid, and kidney. Positive HE4 immunoreactivity was prominent and consistent in ovarian cancer, while some positivity was observed in other cancers, including mesothelioma, lung, endometrial, breast, gastrointestinal, renal, and transitional cell carcinomas.[19] Among the different tumor sites, the highest expression levels were found in ovarian cancer; moderate levels in lung adenocarcinoma; and lower levels in breast, transitional cell, gastric, and pancreatic carcinomas.[19-21] Among malignant conditions, the highest levels of HE4 have been noted in ovarian cancer for women and in lung cancer for men.[22] Recently, LeBleu et al identified HE4 as a significantly-and the most frequently-upregulated gene in fibrosis-associated myofibroblasts in patients with kidney fibrosis.[17]

FIGURE

Normalized Expression Values of the HE4 Gene Across Different Cancers

Given the high expression levels noted for HE4 in ovarian cancer, Drapkin et al utilized immunohistochemistry to compare HE4 expression between malignant ovarian, benign, and nonovarian malignant tissues.[23] They found that a majority of nonovarian carcinomas did not express HE4 in their studies, with expression in normal tissues confined to the reproductive tracts and respiratory epithelium.[23] Interestingly, HE4 expression varied between histologic subtypes of ovarian cancer: HE4 was positive in 93% of serous tumors, in 100% of endometrioid tumors, and in 50% of clear-cell tumors, but in no mucinous tumors.[23] Expression has been found in primary tubal carcinomas and the normal fallopian tube epithelium, which may contribute to decreased specificity.[24] Similarly, in endometrial cancer, tissues from endometrial cancer patients had significantly upregulated HE4 in comparison with normal endometrium, implicating HE4 as a biomarker for endometrial cancer.[25]

Internal server error