Peritoneal Washings and Ovary

Sampling Techniques

The sampling procedure used can influence quantity and quality of cells obtained and therefore the diagnostic accuracy of the procedure.

Laparotomy , as the traditional method for obtaining cytologic material and staging of malignant neoplasms of the abdomen and pelvis, is very effective, provided an adequate surgical incision is made that allows systematic visualization and palpation of peritoneal and visceral surfaces. However, lesions involving recesses such as the cul-de-sac region or posterior diaphragm may still be overlooked. Smears from liver, diaphragm, and parietal surfaces can be used to provide additional information from these recessed locations.

Fiberoptic technology has revolutionized the use of laparoscopic/robotic procedures in gynecologic and surgical oncology. Laparoscopy is extensively used in the diagnosis of clinically benign pelvic disease such as endometriosis. The procedure involves the periumbilical insertion of a laparoscopic trocar through which diagnostic material is procured by way of fine needle, forceps cytologic brushing, or peritoneal lavage. Detailed examination of the celomic cavity is facilitated by the stretching of the peritoneal surfaces induced by gas insufflation, but effectiveness is limited if adhesions are present as is common with second-look procedures.

Implantable intraperitoneal chemotherapy port systems can be used as sampling devices to collect washings for detection of persistent intraperitoneal disease in ovarian carcinoma.

Tumor implantation at the laparoscopy site is uncommon but is more likely in patients with malignant effusion than in those with low-grade tumors without ascites. Dissemination of malignant cells via fallopian tube into the peritoneal cavity has been reported following hysteroscopy for patients with endometrial carcinoma.

Specimens

The peritoneal cavity is normally completely lined by cuboidal mesothelial cells overlying a subserosal stroma well supplied by blood vessels and lymphatics. Detection of malignant tumor cells in washings and free fluid depends on exfoliation of cells and is influenced by such factors as extent of peritoneal involvement , localization by adhesion formation , and grade of tumor. Spread of malignant cells tends to occur by way of the vascular channels and lymphatics as well as across the peritoneal cavity, where deposits tend to lodge in dependent areas such as the cul-de-sac and paracolic gutters.

If there is no peritoneal involvement by tumor on gross inspection, any accumulated fluid should be aspirated and sent for cytologic evaluation. Otherwise, saline washings are submitted unfixed to the laboratory for processing as soon as possible; the specimen is well mixed, and an aliquot is resuspended in cytology fixative for preparation of cytocentrifuge or liquid-based preparations. Cytologic smears may also be prepared from surfaces such as the diaphragm, pelvic side wall, and solid viscera using a wooden spatula or cytobrush and spray fixed in 95% ethanol. The technique is convenient and cost-effective and enables a relatively large surface area of peritoneum to be sampled. Cell block preparations are useful in the workup of patients with suspicious lesions in cytocentrifuge or liquid-based preparations as they may provide optimal assessments of preserved architectural features and as a medium for multiple immunocytochemical staining procedures.

Cytology

Peritoneal washings and brushings differ from effusions with respect to cellularity and composition such as the presence of detached sheets of mesothelial cells, which may give rise to problems in interpretation ( Fig. 12-1 ). The size of the sheets indicates that they are benign fragments of mesothelium. Flatness of the sheets, the orderly mosaicism of their cells, and the absence of nuclear features of malignancy are further indications of their benignity. The diagnosis of metastatic carcinoma in peritoneal washings, as with effusions, depends on identification of single cells and papillary cell groups showing obvious criteria for malignancy ( Fig. 12-2 ).

Atypical Cells and Pitfalls for Diagnosis

Mesothelial cells are usually readily recognized as uniform-appearing cells with bland nuclear chromatin, a normal nuclear to cytoplasmic (N : C) ratio, and sheet-like arrangements with intercellular fenestrations. Difficulties may arise in the interpretation of loose groups of cells with nuclear enlargement and nucleoli ( Fig. 12-3 ). Frequently, there is evidence of a transition to normal-appearing mesothelial cells.

Cytologic atypia in mesothelial cells from peritoneal washings may be due to reactive changes associated with cyst or adnexal torsion , visceral adhesions , organizing hematoma , and inflammatory lesions such as pancreatitis and chronic salpingitis . Intraperitoneal chemotherapy may give rise to cytologic abnormalities that mimic malignancy.

Peritoneal washings from pelvic endometriosis may show loose clusters of atypical cells with decreased N : C ratios and prominent nucleoli. Characteristic hemosiderin-laden macrophages and spindle-shaped stromal cells may not always be present to assist diagnosis ( Fig. 12-4 ). Eosinophilic metaplastic change in endometriosis may also present as a diagnostic pitfall. Although well-differentiated endometrial adenocarcinoma has loose but usually 3-dimensional clusters of cells with coarse chromatin and nuclear membrane irregularities, the N : C ratio may not be increased if, as is sometimes the case, the cells have abundant cytoplasm. In endometriosis, however, the cells are low columnar with a bland nuclear chromatin and uniform nuclear membranes.

Rupture of ovarian or adnexal cysts often results in a distinctive cell population that can readily be differentiated from the surrounding mesothelial cells. Paratubal cysts have cohesive ciliated cuboidal cells that are recognizably benign. However, washings associated with ruptured endometriotic cysts may show considerable cytologic atypia and be difficult to differentiate from adenocarcinoma. Clinical correlation with reference to celioscopic and histologic findings may be necessary.

Papillary Structures

The other morphologic pattern that can give rise to significant diagnostic problems is the presence of papillary structures in peritoneal washings. Simple papillary structures are associated with Müllerian inclusions such as endosalpingiosis , frequently located on the peritoneal surface of the uterus, pelvic side wall, and cul-de-sac. These tightly cohesive tubular structures have a scanty layer of cuboidal or short columnar cells often arranged around psammoma bodies. The nuclear size is uniform with a vesicular chromatin pattern, and atypia is mild with small inconspicuous nucleoli.

These lesions need to be distinguished from well-differentiated papillary carcinoma or serous borderline tumors , which may show both simple and complex papillary groupings as well as occasional single cells. Complex papillary groups show cellularity with overlapping, increased N : C ratios, increased cell size, coarsening of the chromatin, irregular cell membranes, and prominent nucleoli ( Fig. 12-5 ).

Key Features of Papillary Carcinoma

• Complex papillary groups

• Increased cell size

• Overlapping nuclei

• Irregular cell membranes

• Prominent nucleoli.

Ancillary Techniques

The evaluation of ascitic fluid is concerned as much with identifying the site of origin as with confirming the presence of malignancy. The objective of peritoneal washings is to confirm malignancy; an abdominal/pelvic mass has already been identified. To that end, immunohistochemistry may assist in differentiating neoplastic cells from reactive mesothelial cells. Overinterpretation can be avoided if even a limited panel is used and correlation is made with the primary resected tumor. The panel could include conventional immunostains (BerEp4, LeuM-1, Calretinin, MOC-31, D2–40). In peritoneal washings of gastrointestinal and gynecologic cancers, molecular methods have potential to detect small quantities of residual disease not apparent by morphology alone. In situ hybridization techniques for detection of carcinoembryonic antigen (CEA) or human telomerase RNA components have potential for detection of small numbers of cancer cells from body fluids in patients with abdominal malignancies.

Diagnostic Accuracy

Peritoneal lavage cytology is a specific but a relatively insensitive procedure for detection of peritoneal surface involvement by malignant cells. To be useful, therefore, the cytologic findings should be evaluated in conjunction with clinical, serologic, and laparoscopic findings. The sensitivity of peritoneal cytology for detection of malignant cells varies from 48% to 88% for patients undergoing primary laparotomy for ovarian carcinoma, but is less for second-look procedures and histologically confirmed microscopic involvement in patients without gross peritoneal disease. This relates to reduced tumor burden or poor distribution of cells due to localization by adhesions. Separate collection of washing samples from multiple intra-abdominal sites does not appear to increase sensitivity either. Because brushings and smear specimens are usually obtained from a limited number of sites in the peritoneal cavity in patients without celioscopic evidence of tumor, the sensitivity of the technique is understandably low. Sensitivity of washing for gastric carcinoma is also low but use of adjunctive molecular techniques may help.

Sensitivity of peritoneal washing cytology for ovarian tumors of low malignant potential is high, and correlates with surface ovarian involvement and peritoneal implants. With that exception, ovarian carcinomas of high grade and stage were more likely to yield neoplastic cells in peritoneal washings than low-grade tumors.

A small but significant false-positive rate has been noted for cytologic washings in ovarian carcinoma, and several of the contributing factors have been discussed. Histologic correlation, preparation of cell block, and/or immunostains may help to reduce this discrepancy in peritoneal washings.

In some cases, it is possible that inadequate histologic sampling may account for what has been called erroneous false-positive results. Positive washings are infrequent in stage I endometrial carcinoma. Some of these results could be attributed to lymphatic spread or fallopian tube extravasation after endometrial biopsy procedures. Spread from lymphatics during hysterectomy procedures has been suggested as the reason for the presence of tumor cells only on postprocedure washings.

Sampling Techniques

Ovarian aspiration may be performed transvaginally, transrectally, transabdominally (via direct palpation or radiographically guided), laparoscopically, or directly at the time of laparotomy. Transrectal FNA is now used rarely, as a result of concerns regarding infection and the availability of other techniques. The transvaginal approach involves the insertion of a 14–22-gauge needle along a vaginal probe guided via real-time sonography. Strict sonographic criteria are used to identify cysts that can be drained safely in this manner. These criteria include size <10 cm; unilaterality; low echogenicity; absence of ascites; thin, regular walls in unilocular cysts, or thin, smooth septa in multinodular cysts; and no hypervascular foci on Doppler ultrasound. Cysts with low-level echogenicity, consistent with a hemorrhagic or endometriotic cyst, may also be included in patients with known or suspected endometriosis.

Intraoperative cytology (IOC) is a method of intraoperative consultation with the help of cytology smears or imprint specimens. Advantages of IOC are its simple, inexpensive, excellent preservation of cellular details, with no loss of tissue as occurs with cryostat sections and adequacy of surgical margins.

Basic Histology

The ovaries are paired, ovoid organs that lie near the pelvic sidewall and lateral to the uterus. They are draped by a fold of the peritoneal mesothelium, with which the ovarian surface epithelium shares its embryologic lineage. The single or focally stratified layer of columnar to cuboidal surface epithelium that covers the ovaries may invaginate focally into ovarian stroma, resulting in the formation of epithelial inclusion cysts. Being of Müllerian derivation, the surface epithelium is capable of phenotypic alteration to resemble serous (tubal), mucinous (endocervical or goblet cell), endometrial, or transitional (urothelial) cells. Each cell type may be recapitulated in epithelial ovarian tumors.

The millions of ova present in the neonatal ovary dwindle in number throughout reproductive life, ultimately to be depleted by menopause. The developing ova are surrounded by a shell of granulosa cells that secrete estradiol (E ₂ ) during folliculogenesis. The outer layers of the shell are composed of theca interna cells that produce a variety of hormones, including E ₂ and androstenedione. After ovulation, the follicle-lining cells become luteinized, acquire abundant eosinophilic cytoplasm, and produce copious amounts of progesterone.

The ovarian stroma is largely composed of spindled cells that resemble fibroblasts, collagen, and various stromal cells with an enormous capacity for hormone and enzyme production.