* Objective.-Peritoneal washings are routinely performed during gynecologic surgery. The presence or absence of malignant cells in washings helps determine the stage of the malignancy. However, the efficacy of this procedure has not been studied recently.

Design.-All intraoperative washings for gynecologic disease at our hospital from 1992 through 1994 (901 cases) were reviewed. Of these, 380 were gynecologic malignancies that were reviewed for changes in staging based on the presence of malignant cells.

Results.-Histologically, 380 cases were gynecologic

malignancies, 521 benign, 79 nongynecologic, and 25 had no accompanying surgical pathology. Of the malignancies, 125 had a diagnosis of cancer on washings. In 12 cases (3.1 %), a change in stage resulted.

Conclusions.-In a small but significant number of cases, malignant cells in the washings changed postoperative staging, impacting therapeutic measures and prognosis for these patients greatly. Peritoneal washings remain a simple yet effective tool in the evaluation and management of gynecologic malignancies.

(Arch Pathol Lab Med. 1997;121:604-606)

Cytologic sampling of peritoneal fluid from the pouch of Douglas at the time of surgery, as a concept, was introduced in 1958. Peritoneal washings are now commonly performed during any exploratory laparotomy for gynecologic disease because peritoneal involvement can be undetectable by visual inspection alone. The application of peritoneal lavage in laparotomies serves three purposes: detection of occult tumor, determination of recurrent or persistent tumor, and staging. The presence of peritoneal tumor indicates a worse prognosis, and thus the results of peritoneal washing cytology are incorporated in staging and treatment decisions. In this era of managed care and cost containment, the significance of this procedure is being tested anew.

Cytologic preparations of all intraoperative washings from 1992 through 1994 at The Johns Hopkins Hospital, Baltimore, Md, were studied retrospectively. A total of 901 peritoneal washings from women undergoing laparotomy for gynecologic disease were reviewed, and 380 of these cases were reviewed for changes in staging.

All cytologic specimens were obtained intraoperatively; some had washings separately collected from multiple intra-abdominal sites, and these fluids were interpreted as all positive or all negative in all patients. All cytologic preparations (cytospin, Millipore filter preparations; Millipore Corp, Bedford, Mass) were subsequently stained with modified Papanicolaou and/or DiffQuik stains. All cell blocks were stained with hematoxylin-eosin.

In cases with documented malignancy, the results of peritoneal cytology and corresponding malignant histology, as well as pertinent clinical data, were reviewed for each patient to determine whether the cytopathologic diagnosis changed the FIGO (International Federation of Gynecology and Obstetrics,1988) stage of disease.

RESULTS

Histologically, 380 of the 901 cases were gynecologic malignancies, and 521 were benign. Of the 521 washings from patients with histologically benign genital disease, no false positives were found. This excellent concordance may be a result of good cytohistologic correlation, which enhanced the accuracy of interpretation. Of the gynecologic malignancies reviewed, a diagnosis of cancer was made in 125 cases, and in 12 cases (3.1%), a change in stage resulted (Table). Four of the 12 cases were endometrioid carcinomas of the uterus (one clear cell carcinoma [Fig 1], three endometrial cancers), two were clear cell carcinomas of the ovary, two were serous carcinomas, and one was a serous tumor of low malignant potential. The remaining cases involved three rare tumors, namely, a small cell carcinoma of the uterus (Fig 2), a mixed mullerian tumor, and a fallopian tube carcinoma. All but one of these 12 cases (the fallopian tube carcinoma) were primary neoplasms and not recurrences. Over the short follow-up period (at time of submission, 1992-1996), none of these cases had, nor subsequently developed, a second primary tumor, confirming that the malignant cells seen were not from an occult tumor. Four of the patients are dead of disease, six are disease-free, and one has evidence of disease. One patient was lost to follow-up. It is interesting to note that two of the patients with apparently limited disease (cases 2 and 4 in the Table) are dead of disease. The other two deceased patients (cases 10 and 11 in the Table) had minimal disease, but aggressive tumors with poor prognoses. On review of all the gynecologic malignancies, 25 were recurrences or second-look operations.

COMMENT

Pelvic malignancies with accompanying exudates have been evaluated since as early as 1867. Peritoneal washing cytology as a diagnostic tool was proposed in 1958. The prognostic value for pelvic tumors was shown by Morton et al in 19611 and 10 years later by Creasman and Rutledge in 1971.23

Today, peritoneal washings have become an accepted adjunct to the pathologic evaluation of gynecologic malignancies.3 This was considered necessary because the rationale for peritoneal washing cytology since its inception has included the concept that some cases will be cytologically diagnostic before a clinically suspicious focus is found.2,5 The information obtained by this procedure is used to plan adjuvant therapy and management for patients with cytopathologic evidence of malignancy established by positive peritoneal washings.