Purpose of the investigation: To determine the diagnostic value of serum vascular endothelial growth factor (VEGF) in the preoperative assessment of the nature of ovarian masses.
Materials and methods: A prospective cohort study was conducted from August 2001 to September 2002 on 40 premenopausal and 23 postmenopausal patients with ovarian masses. During preoperative workup, patient age, serum Ca-125 levels, serum VEGF levels, and tumor volume based on ultrasonographic examination were determined. Laparoscopic (n=23) or laparotomic (n=39) approaches were undertaken to obtain the final pathologic result. According to the final ovarian pathology, follicular cysts, corpus luteum cysts and endometriomas were grouped as non-neoplastic ovarian masses (n=40, group I). Serous or mucinous cyctadenomas, dermoid tumors and fibromas were allocated into the neoplastic benign ovarian mass group (n=10, group II). Primary malignant ovarian neoplasms were categorized into the neoplastic-malign group (n=12, group III).
Results: Mean ages of cases among groups I, II and III were 39.0 +/- 2.0, 42.2 +/- 5.2 and 56.9 +/- 4.2, respectively. As age of the cases enrolled in this sudy increased, the more likely was the occurrence of neoplastic malign ovarian pathologies (p < 0.001). Among postmenopausal cases diagnosed with an ovarian mass, serum Ca-125 levels were 113.5 +/- 20 IU/ml compared to those in premenopausal cases (85.8 +/- 16.0, p = 0.05). The values for serum VEGF values among pre- and postmenopausal ovarian masses were 46.2 +/- 6.7 pg/ml and 68.2 +/- 7.9, respectively (p = 0.04). In group I, serum VEGF levels of endometriomas (56.5 +/- 1.5 pg/ml) were higher compared to those of follicular or corpus luteum cysts (30.6 +/- 2.8, p = 0.05). In contrast, tumor size appeared to be larger in non-endometriotic. non-neoplastic cysts (10.1 +/- 2.0 cm), compared to endometriomas (6.4 +/- 0.6 cm, p < 0.01). Serum VEGF levels of group III were higher than other groups (p < 0.001). With respect to the discriminating benign or malign nature of the mass, with a specific cut-off value of serum VEGF level of 68.7 pg/ml, the sensitivity, specificity, positive and negative likelihood ratios were 92.3%, 88.0%, 3.3 and 0.1, respectively. For serum Ca-125 levels, the sensitivity, specificity, positive and negative likelihood ratio with a statistically relevant cut-off value of 102 IU/ml were, 76.9%, 76.0%, 3.21 and 0.3, respectively. Area under curve (AUC) for serum VEGF and Ca-125 values were 0.938 (95% CI: 0.81-0.96) and 0.769 (95% CI: 0.64-0.86), respectively (p = 0.02). Among the postmenopausal group, AUC for serum VEGF and Ca-125 was detected as 0.902 (95% CI: 0.70-0.98) and 0.873 (95% CI: 0.66-0.91) (p = 0.14).
Conclusion: Serum VEGF has the potential to be considered as a tumor marker with a good diagnostic relevance in differentiating the nature of ovarian masses.