PubMedDate not supplied
This paper evaluates the usefulness of determining the level of serum Tissue Polypeptide Specific Antigen (TPS) by TPS ELISA in the diagnosis of ovarian malignancies and compares it with the results of histological examination of the ovaries. The study covers 92 patients who had been clinically diagnosed with ovarian tumors by history, physical examination and pelvic examination with or without pelvic ultrasonography. All of them underwent surgical treatment by exploratory laparotomy in Siriraj Hospital between May 1, 1996 and 31 March, 1997. None of the patients had been treated with chemotherapy or hormonal therapy, and in no case were there any previously diagnosed malignancies. TPS was measured in the serum of 92 patients who were preoperatively diagnosed with ovarian tumor: 52 patients had benign pelvic masses and 40 patients had malignant ovarian tumors. Using the criterion for TPS positivity defined by the manufacturer (80 U/L), TPS levels were elevated in 28.8% of benign pelvic mass patients and in 90% of malignant ovarian tumor patients. Statistical analysis using a two-by-two table at every cut-off TPS level and a receiver operating characteristic (ROC) curve, the optimal accuracy, sensitivity, specificity, positive predictive value, negative predictive value, false positive rate and false negative rate were 79.4, 90.0, 71.2, 70.6, 82.9, 28.9 and 10.0%, respectively using the positivity criterion of 80 U/L. We conclude that determination of serum TPS level by TPS ELISA in the diagnosis of ovarian malignancy is good and clinically acceptable. A TPS level greater than 80 U/L is a useful positivity criterion for screening for malignant ovarian tumors, while a TPS level greater than 180 U/L is a positivity criterion for differentiating malignant ovarian tumor from benign pelvic mass. Because of its high false positive rate, any patient with TPS greater than 80 U/L should be further investigated for malignant ovarian tumor.