Abstract
Background and aim
Epithelial ovarian cancer (EOC) is characterized by few early symptoms, presentation at an advanced stage and poor survival. As a result, it is the most frequent cause of death from gynecological cancer. Recently, not only tumor antigens but also antibodies produced in response to the disease have been considered as biomarkers for early detection of EOC. Kallikrein-related peptidases (KLK) are secreted serine proteases implicated in tumor progression. A recent study has demonstrated that the serum KLK5 (sKLK5) concentration is elevated in patients with EOC, but the existence of a humoral immune response to KLK5 has not been studied. Moreover, the presence of sKLK5 in other ovarian diseases has not been clearly determined.
The aim of this study was to examine the serological existence of anti-KLK5 antibodies, bound to KLK5 into circulating immune complexes (KLK5-lgM and KLK5-lgG ICs) or free antibodies (IgM and IgG), in healthy women and in patients with benign ovarian masses, borderline tumors and EOC. In the same patients we also assessed the levels of sKLK5.
Materials and methods
Serum samples were obtained from 50 healthy women, 50 patients with benign ovarian masses, 1 7 patients with borderline ovarian tumors, and 50 patients with EOC, before any surgical or chemotherapeutic treatment. Patients with a past or concomitant history of malignancy were excluded from the study. KLK5-lgM and KLK5-lgG ICs were detected with a sandwich ELISA using a polyclonal anti-KLK5 antibody (R&D Systems, Minneapolis, MN, USA) in capture and an anti-hlgM-HRP antibody (Sigma Aldrich Inc, St Louis, MO, USA) or anti-hlgG-HRP antibody (Sigma) in detection. Free anti-KLK5 IgM and IgG were detected with an indirect ELISA using recombinant human kallikrein 5 (R&D) in coating and anti-hlgM-HRP or anti-hlgG-HRP antibodies in detection. Finally, sKLK5 was detected with a sandwich ELISA, using a polyclonal anti-KLK5 antibody in capture and a biotinylated poyclonal anti-hK5 antibody (R&D) in detection.
Results
Elevated levels of KLK5-lgM and KLK5-lgG ICs were detected in 11% of patients with borderline tumors and 8% of patients with EOC, whereas elevated levels of free anti-KLK5 IgM and IgG were found in 1 7% of patients with borderline tumors and 8% of patients with EOC, resulting in a 100% specificity both in healthy women and patients with benign ovarian disease. Patients with EOC showed higher levels of sKLK5, whereas the protein was almost undetectable in women with other ovarian tumors (Kruskal-Wallis test: p<0.001). In particular, at 95% specificity in healthy controls, 52% of patients with EOC showed high sKLK5 levels. Interestingly, both the immune complexes and the free antibodies were elevated in patients with undetectable sKLK5 levels and borderline tumors with intraepithelial carcinoma.
Conclusion
Our results showed that sKLK5 is a potential new biomarker to be used in combination with other biomarkers for EOC detection. Moreover, the combination of sKLK5 and antibodies reactive to KLK5 might improve the sensitivity in EOC detection since these antibodies may be found in patients without detectable sKLK5 levels. In conclusion, the identification of an immune response that generally precedes the presence of high levels of circulating antigens may represent a novel useful tool for early EOC detection.