OBJECTIVE: To determine the appropriate evaluation of women with cervi
cal cytologic smears suggestive of adenocarcinoma. STUDY DESIGN: Of 1,
192,972 smears obtained between July 1989 and March 1992, 57 (0.005%)
were interpreted as adenocarcinoma, adenocarcinoma in situ or ''rule o
ut'' adenocarcinoma. Review of medical records allowed the determinati
on of pathologic diagnoses in 46 of 57 women. Thirty of the 46 cytolog
ic smears were reviewed. RESULTS: Fifteen of the 46 women were premeno
pausal and < 50 years of age. Twenty-three were asymptomatic, 22 had a
bnormal vaginal bleeding, and I had increasing abdominal girth. Twenty
-nine (63%) of the 46 women had cancer, 12 (26%) had cervical intraepi
thelial neoplasia 2 (GIN 2) or 3, and 5 (11%) had CIN 1, condyloma or
no pathology. Of the 29 women with cancer, 11 had cervical cancer, I h
ad vaginal cancer, 13 had uterine cancer, and 4 had extrauterine cance
r. Nineteen or 22 women (86%) with abnormal vaginal bleeding had cance
r; 9 of 23 asymptomatic women had cancer (39%) (chi(2) = 9.84, P < .01
). DISCUSSION Women with smears suggestive of adenocarcinoma require b
iopsy of cervical or vaginal masses, colposcopy with directed biopsy,
endocervical curettage and endometrial biopsy. If cancer is not diagno
sed, cervical conization with dilatation and curettage (D&C) is indica
ted. If conization with D&C is negative, laparoscopy is indicated to e
xclude extrauterine cancer.