Abundance of mitochondrial superoxide dismutase is a negative predictive biomarker for endometriosis-associated ovarian cancers

The incidence of ovarian cancer cases has been increasing, and the current rate is fourfold higher than the rate 40 years ago in Japan [1]. Endometrioid ovarian carcinomas and clear cell ovarian carcinomas account for 16.7% and 23%, respectively, of all ovarian cancer cases in Japan, and they are proportionally much more common than with ovarian carcinoma subtypes in Europe and the USA [2, 3]. Endometriosis-associated ovarian cancer (EAOC) usually occurs in younger women, with high recurrence and mortality rates [4]. In the clinical and pathological fields of EAOC, alterations in mismatch repair proteins and polymerase ε likely identified a subset of EAOCs with excellent outcomes, and abnormal p53 levels were related to a worse prognosis of EAOC [5]. Hormone receptor expression was associated with longer survival with EAOC [6], and glypican-3 was related to a worse prognosis of ovarian clear cell carcinomas [7]. However, further biomarkers predicting the prognosis are essentially required.

Inflammation and oxidative stress induced by heme and iron seem to be important triggers in the malignant transformation of endometriosis into EAOC [8]. EAOC is regarded as a stress-tolerant cancer because it arises from chocolate cysts with a highly oxidative microenvironment, induced by excess heme and chronic inflammation. In clear cell ovarian carcinoma, abnormalities are often found in genes responding to oxidative stress and metabolism of reactive oxygen species (ROS) [9].

Mitochondrial superoxide dismutase (SOD2) is an enzyme that metabolizes superoxide in mitochondria and plays an important role in maintaining mitochondrial function through oxidative stress tolerance. SOD2 overexpression correlates with poor prognosis in several cancers [10‐12]. Suppression of SOD2 enhances ROS production in ovarian cancer cells and results in increased apoptosis, inhibition of proliferation, and enhanced sensitivity to chemotherapy [13]. However, in ovarian cancer, especially in EAOC, the correlation between SOD2 expression and prognosis remains unknown. It is unclear whether overexpression of SOD2 is strongly involved in ROS resistance and related to prognosis in EAOC. In the present study, we investigated the correlation between SOD2 expression and prognosis in a cohort of EAOC cases based on immunohistochemical staining of SOD2 and compared the results with clinicopathological data.

Formalin-fixed paraffin-embedded tissue was used for SOD2 immunohistochemical analysis. Samples were sliced into 5-μm sections. After dewaxing, the sections were autoclaved at 120 °C for 1 min in 10 mM sodium citrate buffer (pH 6.0) and immersed in 0.3% H₂O₂. They were then incubated overnight at 4 °C with primary antibodies against SOD2 (diluted 1:1000, #06-984, Millipore Corporation, Billerica, MA). Sections were rinsed with phosphate-buffered saline and incubated with secondary antibody conjugated with horseradish peroxidase (Simple Stain MAX-PO, Nichirei, Tokyo, Japan) at room temperature for 1 h. Sections were then stained with 3.3′-diaminobenzidine tetrahydrochloride and counter-stained with hematoxylin. All slides were scored by two pathologists blinded to the clinical data, using a standard light microscope. SOD2 expression in each EAOC was evaluated by relative comparison to the SOD2 expression of normal ovarian tissue stromal cells on the same slide. We categorized tumors into two groups on the basis of the intensity of SOD2 expression: (1) the high SOD2 group, wherein the intensity of the tumor cell staining was stronger than that of normal ovarian stromal cells, and (2) the low SOD2 group, wherein the intensity of the tumor cell staining was less or similar to normal ovarian stromal cells. As the S/N ratio of immunohistochemical SOD2 expression was quite high, the scoring of a case was inconsistent between the two pathologists, and the kappa statistic of SOD2 evaluation was 0.9567682. In one case with inconsistent scoring, obviously high SOD2 cell levels accounted for about 10% of tumor cells. We considered that such a population had a high probability of becoming seeds of recurrence or metastases; therefore, this case was eventually classified into the high SOD2 group.

In this study, we identified two clinically important issues. First, the high SOD2 expression of cancer cells is a poor prognostic factor for EAOC. Second, the current standard therapeutic protocol is insufficient to achieve improved survival of patients with EAOC showing high SOD2 expression.

High SOD2 expression was a worse prognostic factor for EAOC. SOD2 plays an important role in ROS removal, whose production is induced by chemotherapeutic treatments, and maintenance of mitochondrial function. In previous studies, high SOD2 expression is associated with poor prognosis in some carcinomas [10‐12]. Especially in renal clear cell carcinomas, which show pathological similarities with clear cell ovarian carcinomas, high SOD2 expression reflects better mitochondrial function and ROS resistance, and increased SOD2 expression correlates with poor prognosis [14]. The cancer genome atlas (TCGA) database cannot suggest that SOD2 is a prognostic indicator in ovarian cancer. However, as the database strongly depends on the overall incidence of cancer types, the provisional and final analyses of ovarian cancers have primarily focused on serous ovarian adenocarcinomas, rather than on EAOC including clear cell and endometrioid ovarian carcinomas. As EAOC often arises from endometriosis, a tissue abundantly exposed to inflammatory ROS, it is thought to acquire resistance to oxidative stress. Hemachandra et al. [15] revealed that SOD2 is more strongly expressed in ovarian clear cell carcinoma than in other epithelial ovarian cancer subtypes and that SOD2 is a pro-tumorigenic or metastatic factor. Hemachandra’s study conformed to the findings of the present study.

In the present cohort, no patients died in the low SOD2 expression group, even among patients with high stage and recurrence. All three patients including those with tumors with low SOD2 expression, who did not complete chemotherapy, have not had a relapse until now. In addition, two patients with high FIGO class and tumors with low SOD2 expression have survived without any cancer relapses. Regarding tumors with low SOD2 expression that are highly sensitive to chemotherapeutic ROS increments, the current standard therapeutic protocol, i.e., platinum-based chemotherapy following surgical resection, is considered sufficient. Increased ROS levels improve the antineoplastic effect of platinum-based chemotherapy [16]. This fact suggests that platinum-based chemotherapeutic agents are effective agents against EAOC with low SOD2 expression. Conversely, among cases with high SOD2 expression, 15 of 46 cases relapsed and 14 deaths were observed. These results indicate that the current therapeutic protocol should be considered insufficient in tumors with high SOD2 expression. With some anti-cancer agents, such as cisplatin, ROS are involved in the antitumor effect [17, 18]. As SOD2 is an antioxidant enzyme that can prevent oxidative redox-mediated damage of mitochondrial proteins and preserve mitochondrial function, EAOC with high SOD2 expression likely have strong resistance to oxidative stress caused by cancer treatment such as chemotherapy. In our cohort, SOD2 evaluation was performed from the sample before each chemotherapeutic treatment. Therefore, SOD2 may reflect the intrinsic aggressive character of the tumor in addition to the predisposed resistance to chemotherapy. EAOC, especially ovarian clear cell carcinomas, is well recognized to have greater resistance to platinum-based chemotherapy, a more aggressive clinical course, and more malignant behavior than other types of ovarian cancers [19‐21]. Ovarian clear cell carcinomas have poor prognosis because of resistance to current chemotherapy protocols, based on platinum and taxane [19].

This study confirmed the following findings: a novel therapeutic strategy for clear cell ovarian carcinomas and/or other EAOCs, in which SOD2 is strongly expressed, should be established. Metformin and other biguanides, which have been clinically used for the treatment of diabetes mellitus, can inhibit complex I of the mitochondrial electron transport chain and mitochondrial respiration [22]. In addition, they exert anticancer effects on patients with diabetes diagnosed with pancreatic cancer, colon cancer, and prostate cancer [23‐27]. By targeting tumor mitochondrial function, drug repositioning with the use of biguanides such as metformin has been proposed for some types of cancers [28]. Isono et al. [14] demonstrated that biguanide could improve the therapeutic effect in clear cell carcinomas in the kidney, which show similar pathology to ovarian clear cell carcinomas. Replacement therapy or drug repositioning to target tumor cell mitochondria, as can be achieved with the use of biguanides, should be investigated for the treatment of EAOC with abundant SOD2 expression, which is likely to retain stronger mitochondrial function.

As this study could not include a huge number of cases, it was impossible to analyze whether SOD2 was significantly predictive of the prognosis, after dividing into each category of FIGO clinical stage or histological type. Further analyses with more cases and a larger cohort of EAOC may be desired. Such studies will provide more supportive evidences for SOD2 as a predictive biomarker for EAOC.

This study demonstrated that SOD2 abundance is a predictive biomarker for the poor prognosis in EAOC and highlighted the insufficiency of the current therapeutic standard against these tumors. Mitochondrial SOD2 should be regarded as an important therapeutic target for EAOC.