The combined evaluation of ultrasound, serum CA 125 assay and serum CA 19-9 assay in discriminating a benign from a malignant ovarian mass

Pelvic ultrasound and serum CA 125 assay were performed preoperatively in 225 patients with ovarian masses undergoing laparotomy at the Department of Gynecology and Obstetrics of the University of Pisa. All ultrasound examinations were scored using a personal morphological echostructural evaluation system. Each mass was graded between 0 and 16; the value of 9 was considered as the cut-off limit of the score in discriminating a benign from a malignant ovarian mass. Ultrasound score > 9, serum CA 125 levels > 35 U/ml and serum CA 125 levels > 65 U/ml were respectively found in 48 (75.0%), 54 (84.4%) and 52 (81.3%) of the 64 patients with ovarian carcinoma, and in 8 (5.0%), 48 (29.8%) and 15 (9.3%) of the 161 patients with benign ovarian pathology. When the cut-off limit of CA 125 was increased from 35 U/ml to 65 U/ml, the sensitivity (SE) of this antigen remained essentially unaltered (84.4% vs 81.3%, p = NS), while the specific (SP) increased from 70.2% to 90.7% (p <0.05); therefore 65 U/ml should be preferred as the cutoff limit for CA 125 in distinguishing a benign from a malignant ovarian mass. Among the 108 patients over 45 years, the combined evaluation of ultrasound and CA 125 assay, with 65 U/ml as cut-off limit, had a greater SE (94.8% vs 77.6%, p <0.05) and a similar SP (90.0% vs 92.0%, p = NS) with respect to ultrasound alone. Among the 48 patients over 45 years having both an ultrasound score <9 and serum CA 125 levels <65 U/ml, CA 19-9 values were > 40 U/ml in 3 of the 3 patients with ovarian carcinoma (two of them had a mucinous tumor and one a serous tumor) and in 2 (4.4%) of the 45 patients with benign ovarian pathology. In conclusion, even if serum CA 125 assay alone cannot be used to define the nature of an ovarian mass, this test improves the diagnostic reliability of ultrasound in discriminating a benign from a malignant ovarian pathology in patients over 45 years. When both these tests are negative, we think that serum CA 19-9 should be measured. In fact CA 19-9 assay seems to be able to reveal other cases of ovarian carcinoma, particularly of mucinous histotype, with a very low rate of false positive values.