Over 4.5 years, 158 patients with malignant ovarian neoplasms were evaluated al the Cancer Research and Treatment Center, University of New Mexico. Of these patients, 32 had preoperative sonograms; 25 at this institution and seven at outside institutions. The preoperative sonograms, surgical notes, and pathologic reports of all 32 patients were reviewed.Pelvic masses were identified sonographically in 31 (97 per cent) of 32 cases and were 3 to 21 cm in diameter. In the single case in which no discrete mass was identified, the patient had diffuse peritoneal tumor implants and the ovaries were normal size at surgery.Of the 31 cases, masses were characterized as solid in six (19 per cent), complex in 21 (68 per cent), and cystic in four (13 per cent). Sonography accurately characterized the gross appearance in 26 (84 per cent) of 31 cases.In comparing the gross pathologic appearance to the histologic type of tumor, 10 of 11 serous cystadenocarcimona were complex and one was purely cystic. Two of the mucinous cystadenocarcimomas were cystic and one endometrioid tumor was solid. Five of nine undifferen-tiated adenocarcinomas were complex and four were solid.Of the 32 patients, five had prior hysterectomies. The uterus could be separated from the pelvic mass by sonography and appeared normal in 14 patients. Of these 14 patients, 13 (93 per cent) had no tumor involvement of the uterus at surgery. In 11 patients, the uterus could not be separated from the pelvic mass by sonography, and in eight (74 per cent), the uterus was involved by tumor. Overall, nine (33 per cent) of 27 patients had uterine involvement by tumor.All patients were evaluated for ascites on sonography which was confirmed at surgery. Minimal ascites was found at surgery in five patients (16 per cent) who had sonograms negative for ascites, however, three of these five had pelvic sonograms only. There were no false-positive diagnoses by sonography. The total number of patients with ascites was 19 (58 per cent) of 32.In seven of 32 cases, sonography was limited to the pelvis. Four of these seven had omental and peritoneal metastases at surgery. In the 25 cases in which the abdomen was included, 22 were negative and three were positive for tumor; the positive cases were pathologically confirmed. Therefore, there were no false-positive diagnoses. In 12 (55 per cent) of the 22 sonographically negative cases, the omentum and peritoneum were studded with small (1 cm) tumor implants at surgery. The total of positive cases pathologically was 19 (59 per cent) of 32. Sonography detected only three (20 per cent) of 15 cases of omental and peritoneal metastases.Of the 25 patients in whom the retroperitoneal and pelvic nodal regions could be evaluated, only one scan was interpreted as positive and this was confirmed pathologically. Three patients with negative sonograms had proven lymph node metastases. In seven cases, the retroperitoneum was not examined at surgery. Therefore, four (16 per cent) of 25 patients had lymph node metastases with only one (25 per cent) detected by sonography.The liver was adequately examined in only seven cases, all of which were negative on sonography and at surgery. Sonography detected hydronephretic kidneys in four cases and right pleural effusions in two cases. Additional findings not detected sonographically were ovarian tumor involving the rectosigmoid colon in four cases, the bladder in two, and the vagina in one. Significant metastatic disease requiring resection was not detected by sonography in two cases involving the ileum and one involving the cecum.There was no correlation between the histological grade and the gross pathologic or sonographic appearance. There were 11 stage 1 tumors (tumor limited to ovaries) and these were equally distributed in grades O to III. There were no stage II tumors (ovarian involvement with pelvic extension). Seventeen tumors were stage III (introperitoneal or retroperitoneal metastases outside pelvis) and four were stage IV (distant metastases, including liver parenchyma). Of these last 21 tumors, six were grade II and 15 were grade III. The incidence of ascites in stage III to IV tumors was 62 per cent compared to a 45 per cent incidence in stage I tumors. Of 32 patients, 21 (65 per cent) had stage III or IV disease when diagnosed.In 24 patients who had abdominal and pelvic examination, the sonographic findings alone correctly suggested that four were stage III or IV. However, of 21 cases judged to be stage I or II by sonography, 13 (62 per cent) were actually stage III or IV. Therefore, sonography correctly stages only 12 (48 per cent) of 25 patients.
展开▼
