COMPARISON BETWEEN DIFFERENT DIAGNOSTIC MODALTTIES FOR ASSESSMET OF PELVIS MASSES

Abstract

The study population consisted of 22,000 postme•1opausal women with a median age of 56 (range 45-85) years at the time of recruitment. The postal questionnaire was not returned by 636 (2.9%) volunteers. In all, 49 index cancers had been identified in the

study population by 31 '1 July 1998, of which 16 were stage I, 4 stage II, I 22 stage III, and 7 stage IV at the time of diagnosis.

A total of 22,000 volunteers were recruited over 5 years: 8049 attended the Royal London Hospital, 4110 participated at work, and 9841 attended their general practitioner's surgery.

The length of amenorrhoea ranged from I to 43 years with a median of

6.2 years. The mean number of pregnancies beyond 28 weeks' gestation was 2.2 (range 0-14). Of the women, 4450 had had hysterectomy with conservation of one or both ovaries and 892 women had a history of malignancy (403 of breast cancer) but no evidence of recurrent disease at recruitment.

Of the 22,000 volunteers, 767 (3.5%) had at least one serum CA125 measurement 2:30 U/ml during the study, and of the 47,755 serum samples, 1180 (2.5%) had concentrations 2:30 U/ml. Seventy-two screens were performed among volunteers, with a subsequent diagnosis of an index cancer in 49 volunteers (36 had a single screen before diagnosis, 6 had two screens, 4 had three screens, and 3 had four scrccns). The

sensitivity of serum CA 125 for detection of index cancers at one, three, and five years of follow up was 75%, 54%, and 48% respectively.