AB0 blood groups and rhesus factor expression as prognostic parameters in patients with epithelial ovarian cancer – a retrospective multi-centre study

Epithelial cancer of the ovary, the fallopian tube and the peritoneum (EOC) represent the fifth most common cancer type in women [1]. Due to the absence of early symptoms and an effective screening option, the majority of EOC is diagnosed at an advanced stage and hence, patients’ prognosis is poor with only 46% surviving five years [2]. According to the World Health Organization’s cancer statistics, EOC caused the death of 42.749 European women in 2012 [3]. The strongest predictors of survival are tumor stage, volume of residual tumor after surgery, age, histological subtype, amount of ascites, preoperative albumin level, performance status, and family history suggestive of breast/ovarian cancer [4]. Despite all these features, available biomarkers predicting clinical outcomes remain limited [4]. Therefore, there is an unmet need to identify features that reflect the biological and clinical aggressiveness of the tumor.

AB0 blood group antigens have been reported to be expressed in benign and malignant ovarian tumors [5, 6]. The presence of AB0 antigens on epithelial cells of early stage cancer has been suggested to play a role in carcinogenesis by increasing the cell’s resistance to apoptosis [7]. In addition, AB0 blood groups have been associated with tumor progression in several malignancies [8‐11]. A recent meta-analysis of 89 studies in 30 cancer sites suggested that blood group A is associated with a higher risk for developing various malignancies such as ovarian cancer [12, 13].

AB0 blood groups and Rhesus factor expression are both routinely assessed for patients with EOC prior to therapy. To date, only few data are available on the association of AB0 blood groups and prognosis in patients with EOC. A small retrospective study on 92 EOC patients suggested that the presence of blood group A is associated with worse prognosis [14]. A second study published recently reported similar results in a retrospective cohort of 256 EOC patients [15]. This cohort comprises only Chinese patients with advanced EOC who received optimal upfront cytoreductive surgery.

To the best of our knowledge, no data is available on the prognostic effect of Rhesus factor expression on EOC patients’ survival. We hypothesized that both the AB0 blood groups and the Rhesus factor expression might be associated with clinical outcomes of patients treated for EOC. To test this hypothesis, we performed a retrospective, multi-center study assessing the impact of AB0 blood groups and Rhesus factor expression on clinico-pathological features and prognosis of patients with EOC.

We evaluated the prognostic value of AB0 blood groups and Rhesus factor expression in a large multicenter study of patients with EOC. We rejected our hypothesis and found no association of either AB0 blood groups or Rhesus factor expression with clinical outcomes. We found that blood group antigen B was associated with a prolonged CSS in patients with serous or undifferentiated adenocarcinoma. However, this effect was not independent of that of standard prognostic parameters in multivariable analysis.

Unlike the findings of two previous studies we could not validate that patients with blood group A have shorter survival than those with other blood groups [14, 15]. As Zhou et al. included patients with advanced EOC who underwent optimal primary debulking surgery only, and as the authors pointed out that the effect was stronger in elder patients, we performed analyses in the subgroups of patients with FIGO III and IV and in patients older than 62 years of age. Even in these subgroups, we could not find an association between blood group A and the patients’ survival outcome. We assume the inconsistency might be related to the differences in the patients’ collectives, as Zhou et al. included Chinese patients only in their study [18]. Furthermore, in the study of Zhou et al., the effect of blood group A was not adjusted for the effect of residual tumor after surgery, one of the strongest known prognostic parameters [19]. Unfortunately, as the study by Marinaccio et al. was published in Italian language only, we could not assess further details [14].

Interestingly, we found associations of AB0 blood groups with survival in the subgroup of patients with high-grade serous carcinoma. We found that patients with blood groups B and AB had a significantly longer CSS compared to those with blood groups A or 0. Yet, this effect was no longer seen after adjustment for the effects of established prognostic parameters. We did not perform subgroup analyses of the effect of AB0 blood groups on survival in patients with low-grade serous, mucinous, endometrioid or clear cell carcinoma as the number of patients within these subgroups was too small.

The genomic structure of AB0 genes consists of seven exons that span approximately 19 kb of genomic DNA on the long arm of chromosome 9, band q34 [20]. Several studies could show that 48 to 59% of epithelial ovarian tumors display genetic imbalance on the long arm of chromosome 9 [21, 22]. These changes were even more common in serous, undifferentiated and late stage tumors. Alterations have been observed most frequently in sub-chromosomal region 9q32-q34 that has, therefore, been suggested to be a potential tumor suppressor locus. Furthermore, differences in blood group antigen expression on EOC tumor cells could be observed between histologic subgroups [6]. Compared to benign and low grade mucinous carcinoma, blood group antigens were less abundant in serous carcinoma tumor cells. The presence of AB-H antigens on uro-epithelial cells in women has been associated with a reduced risk of recurrent urinary tract infections. A protective effect of blood group antigens by preventing binding of microbial pathogens has been postulated [23]. AB-H antigens are furthermore found having a role in cell membrane integrity, cell adhesion, membrane transportation of molecules, and acting as receptors for extracellular ligands and enzymes [24]. It was postulated that the loss of blood group antigens results in improved tumor cell ability to move and circulate through the body because of a loss of cell adhesion proteins [25]. A previous study investigated the expression of AB-H blood group antigens on various sites of EOC, but could not detect a difference between the primary tumor and metastatic site [25]. Moreover, one study investigating 130 EOC tumors could not observe an association between expression of AB-H antigens on the tumor cells and the patients’ survival [26]. The exact role of AB0 blood group antigens and their expression or absence on tumor cells is yet unclear.

We also investigated the role of Rhesus factor expression as prognostic parameter in patients with EOC. The risk of developing various malignancies such as skin, esophageal, breast, lung, and endometrial cancer has been linked to a higher probability of being Rhesus factor-negative [27]. In the present study, we could not show any association between Rhesus factor expression and clinico-pathological features or survival outcomes. Similar results were shown in patients with bladder cancer [11].

To our knowledge, this is the largest study investigating the association between AB0 blood groups and survival in patients with EOC. Furthermore, this is the first study to investigate differences in histological subtypes and the role of Rhesus factor expression. The distribution of blood groups within the present study correlates with blood group distribution in the general Austrian population [28]. Patients’ characteristics were distributed equally in both study centers and pathological and survival outcomes are representative for EOC patients. Although multicenter data acquisition offers the opportunity to narrow study bias due to patient selection, the present study has several limitations inherent to its retrospective nature, such as incomplete data acquisition and heterogeneity in care. Differences in pathologic and surgical procedures may have masked small effects. Finally, information about blood group antigen expression on tumor cells is missing.

In conclusion, our findings do not suggest any prognostic value for AB0 blood groups or Rhesus factor expression in patients with EOC. We could not validate the findings of previous studies, but assume this is due to differences in patient cohorts and treatment modalities. Our study shows that in patients with high grade serous EOC, the presence of blood group antigen B is associated with a prolonged CSS. Yet, this requires further validation as this was only in univariate analysis and there was no association with other outcomes.