Morphological US scoring aids diagnosis of small adnexal masses

Nov 6, 2005

Descriptive morphologic sonographic scoring may overcome the subjectivity of interpreting morphologic characteristics in small adnexal masses, wrote the authors of a prospective multicenter trial published in the November Journal of Ultrasound in Medicine.

"And, at the same time, it can incorporate criteria to avoid simplistic description of a complex mass," the Italian team added.

To assess the intercenter reproducibility and diagnostic accuracy of transvaginal grayscale sonographic examination of small adnexal masses by descriptive sonographic scoring, four teaching hospitals and two regional hospitals in Italy performed a prospective study (JUM, November 2005, Vol. 24: 11, pp. 1467-1473).

Employing homogenous standard ultrasound equipment and operator experience, the sites recruited 677 consecutive patients with small adnexal masses smaller than 5 cm; morphological scoring was obtained for each mass.

Management of the mass was based on local protocols, with the minimal requirement that surgery had to be performed for complex masses scoring 8 or higher. In addition, follow-up of at least 12 months had to be performed and recorded for those patients not admitted to surgery.

Sonographic results were then compared with pathologic results and follow-up findings. The researchers observed 52 malignant tumors (19 borderline, 15 stage I-II, 15 stage III-IV, two tubal carcinomas, and one ovarian lymphoma), 243 benign tumors at pathologic examination, and 382 masses defined as benign according to follow-up findings.

The group found that malignant tumors had a significantly higher mean ± SD morphologic score (11.2 ± 2.7) than benign masses (6.2 ± 2.5) (p = 0.001). A likelihood ratio of 0.06 (95% confidence interval [CI], 0.01-0.41) was found with a score of 0 through 4 (1 malignant versus 203benign).

With a score of 5 through 7, the likelihood ratio was 0.14 (95% CI, 0.05-0.41; 3 malignant versus 262 benign), and with a score of 8 or higher, the likelihood ratio was 3.61 (95% confidence interval, 3.09-4.21); sensitivity, 92%; specificity, 76.9%; and positive predictive value, 25.6%.

No difference was seen in the scoring assignment of malignant masses in different centers, although a difference was observed when evaluating the mean score of benign cysts. The authors stated that the disparity was probably due to different incidences of complex benign cysts such asendometriosis in different centers and differences in the ageof the population studied.

From the receiver operating characteristic curves calculated on the whole series, the study team determined that the best cutoff between benign and malignant lesions was a score value of 8. Overall sensitivity proved to be 92%. A logistic regression model proved a slight but significant correlation between the diameter of the cyst and malignancy (odds ratio,1.04; 95% CI, 1.01-1.07), according to the authors.

"These findings confirm on small masses the results so far published in series, in which the volume of malignancy was not limited by dimension," the authors concluded. "According to these findings, sonographic morphologic characteristics could represent a cornerstone for the differential diagnosis of small adnexal masses from their first observation. A simple step-by-step description of the characteristics of a lesion may be reproducible by other physicians and achieves high sensitivity, which is very close to the diagnostic accuracy achieved by complex triage methods for larger, often bilateral, masses."

By Erik L. Ridley
AuntMinnie.com staff writer
November 7, 2005

Ultrasound helps identify adnexal mass as ectopic pregnancy, May 19, 2005

Surgery delay often safe for adnexal masses in pregnancy, May 6, 2005

Adnexal mass in breast cancer patients often benign, February 9, 2005

'Kissing ovaries' a sign of severe endometriosis, January 6, 2005