Search

We retrospectively analyzed 155 urine cytology samples (78 from patients treated with indinavir; 77, no indinavir) from 90 HIV+ patients to evaluate possible association between human polyomavirus and hematuria and to describe indinavir-associated urinary cytologic findings. The CD4 count also was recorded. Variables studied included the presence of cellular viral changes consistent with polyomavirus infection (PVCs), microscopic hematuria, multinucleated cells, indinavir crystals, neutrophils, and eosinophils. Twenty-two samples (15.8%) from patients with CD4 counts of more than 200/µL (>200 × 10 6 /L) showed PVCs. Multinucleated cells, of presumed histiocytic origin based on morphologic features and selective immunocytochemical findings, were present in a higher percentage of samples from indinavir-treated patients. Neutrophils were present in a higher percentage of indinavir-treated patients. Indinavir crystals were identified in 9 samples (12%) from patients receiving indinavir. The lower percentage of PVCs in HIV+ patients with high CD4 counts likely represents an indirect antipolyomavirus indinavir effect by boosting immunity. Multinucleated cells (presumably histiocytic) and acute inflammation are associated with indinavir therapy. Indinavir crystals have a characteristic fan or circular lamellate appearance. Because indinavir crystals may be associated with genitourinary disease, recognizing and reporting them is clinically relevant in HIV+ patients. Genitourinary disease will develop in one third of HIV+ patients or those with AIDS, 1 with hematuria reported in more than 25%. 2 The cause of hematuria in immunocompromised patients may be related to infection with or reactivation of the human polyomavirus (PV), based on the previously reported association of the BK virus (BKV) with hemorrhagic cystitis in an HIV+ patient and in bone marrow transplant recipients 3-5 and microscopic hematuria following bone marrow transplantation. 6 Protease inhibitors are used commonly for the treatment of HIV+ patients. During treatment with one such drug, indinavir, cystitis and pyuria have developed. We have observed the presence of indinavir crystals and multinucleated cells morphologically suggestive of a histiocytic origin in urine samples from patients receiving indinavir therapy. During the last several years, routine urinary cytologic studies in HIV+ patients for the evaluation of microscopic hematuria, crystals, pyuria, drug reaction, and cystitis have been done in our laboratory. We performed a retrospective study on these samples to evaluate the presence of PV cells and its possible association with hematuria and to analyze the cytologic features associated with indinavir therapy.

The standard of practice in cytopathology does not include an individual specimen triage (IST) for sample optimization, but rather prescribes a uniform procedure, e.g., for smears, cell blocks, and cytospins. IST requires additional resources. We sought to evaluate whether IST would result in enhanced diagnostic accuracy and specimen turnaround time in effusions. In order to evaluate the efficacy of IST, 50 effusion samples (31 pleural, 16 peritoneal, and 3 pericardial), each with a minimum volume of 50 ml, were utilized, Each sample was prepared via IST to include at least two initial prepared Diff-Quik-stained cytospins on which the IST was based, as well as a standard cytopreparation protocol for nontriaged samples (NTS) which was limited to 3 smears (2 Papanicolaou-stained, and 1 Diff-Quik-stained) and a hematoxylin-eosin (H&E)-stained cell block section. All triaged and NTS were reviewed retrospectively to determine if IST offered any advantages over the standard cytopreparation protocol for effusion samples. Each was evaluated for diagnostic concordance, turnaround time for final diagnosis, and optimal preparation. In 46 cases, diagnoses in IST and NTS were 100% concordant. Four cases showed minor discrepancies between the original and the NTS diagnoses. In general, the discordant cases were due to sparse cellularity in a specimen composed largely of blood. There was no difference in turnaround time for final diagnosis. Based on a review of all samples, the combination of cell block preparation and cytospins (stained with Diff-Quik and Papanicolaou stains) were considered optimal for microscopic evaluation. IST offers no practical advantage over the NTS standard specimen preparation in relation to the accuracy of final diagnosis or turnaround time. The lysing of grossly bloody fluids with subsequent preparation of cytospins yielded superior preparations for microscopic evaluation over NTS. The standard preparation of effusion samples should include the preparation of a cell block, and cytospins stained with Diff-Quik and Papanicolaou stains, for optimal microscopic evaluation.

BACKGROUND Malignant effusions are complications of metastatic malignant melanoma (MM). Differential diagnosis often involves distinguishing MM from adenocarcinoma and reactive mesothelial cells. Descriptions in the literature of the morphologic and immunocytochemical (IM) staining characteristics of MM in effusions are sparse. A combination of morphology and immunocytochemistry should yield the most accurate diagnostic results. The MART-1 antigen, a transmembrane protein, is specifically expressed in melanocytes and MM. A recently developed monoclonal antibody to the MART-1 antigen may represent a useful marker for the identification of MM in effusions. METHODS The authors conducted a retrospective review of 32 effusion samples diagnosed as MM. The review consisted of morphologic and IM analyses of the effusion samples with antibodies to MART 1, HMB45, S-100, and cytokeratins (AE1/AE3).IM stains were performed on cell block or cytospin material, depending on availability. In the morphologic review, emphasis was placed on Diff Quik-stained material, due to its enhanced cytoplasmic volume and detail. RESULTS Predominant cytologic features noted were lack of cellular cohesion (in 100% of cases), large eccentric nuclei with prominent nucleoli (in 100%), multinucleation (in 84%), variable cytoplasmic vacuolization (in 75%), pigment (in 72%), and cell-in-cell engulfment (in 47%). All immunoreactive cases with sufficient material stained with at least one of the markers used. Tumor cells were positive with IM stains to MART-1 in 78% of cases, HMB45 in 81%, and S-100 in 81%. Coexpression of MART-1, HMB45, and S-100 was noted in 63% of cases. Of cases that showed expression for only 1 of the 3 antigens, the MART-1 was positive in 1 case, and HMB45 and S-100 were positive in 2 cases each. Three cases showed immunoreactivity for cytokeratins in the melanoma cells. CONCLUSIONS The diagnosis of MM in effusions can be made reliably through a combination of morphologic and IM features. Differential diagnosis often involves distinguishing MM from adenocarcinoma or reactive mesothelial cells. Cytoplasmic vacuolization, multinucleation, prominent nucleoli, and cell-in-cell engulfment are cytologic features common to all three. The lack of IM staining for cytokeratins alone cannot reliably distinguish MM; 11% of cases showed positive staining with this antibody in the melanoma cells. The use of a panel of antibodies increases the accuracy of diagnosing MM. In this study, MART-1 proved a useful adjunct to the HMB45/S-100/cytokeratin panel for the diagnosis of MM in effusions, staining 78% of the immunoreactive cases, with positivity in 1 case that was negative for HMB45 and S-100. Cancer 1997; 81:57-63. © 1997 American Cancer Society.

The effects on morphology and diagnostic interpretation of delayed processing of refrigerated effusion samples have not been well documented. The potential for cellular degeneration has led many laboratories to reflexively fix samples rather than submit fresh/refrigerated samples for cytologic examination. We sought to determine if effusion specimens are suitable for morphologic, immunocytochemical, and DNA-based molecular studies after prolonged periods of refrigerated storage time. Ten fresh effusion specimens were refrigerated at 4 degrees C; aliquots were processed at specific points in time (days 0, 3, 5, 7, 10, 14). Specimens evaluated included four pleural (3 benign, 1 breast adenocarcinoma) and six peritoneal (2 ovarian adenocarcinomas, 1 malignant melanoma, 2 mesotheliomas, 1 atypical mesothelial) effusions. The morphology of the cytologic preparations from the 10 effusions was preserved and interpretable after 14 days of storage at 4 degrees C. The immunocytochemical profile of the samples (AE1/AE3, EMA, calretinin, and LCA) was consistent from day 0 to day 14. Amplifiable DNA was present in all samples tested on day 14. We conclude that cytopathologic interpretation of effusion samples remains reliable with refrigeration at 4 degrees C even if processing is delayed.

BACKGROUND Squamous atypia in postmenopausal (PM) cervical vaginal smears (CVS) only rarely is associated with a biopsy-proven squamous intraepithelial lesion (SIL), and thus most commonly represents an atrophy-associated benign reactive change. METHODS To distinguish atypical squamous cells of undetermined significance (ASCUS) and SIL from atrophy-associated benign reactive changes, a review of atypical atrophic PM CVS was performed. Ninety CVS exhibiting an atrophic smear pattern were considered appropriate for study. Repeat smears and/or biopsy after local estrogen therapy were requested to distinguish atrophic/reactive from dysplastic changes. RESULTS Generally, atrophic CVS exhibit uniform nuclear enlargement in the squamous cell population, which, using the criterion of nuclear enlargement alone, would qualify the majority of these cases to be classified as ASCUS. The nuclear enlargement associated with atrophy resolves with the local application of estrogen. Follow-up after local estrogen treatment was available for 84 of 90 patients and revealed 10 cases of SIL (12%) and 9 cases of ASCUS (11%), 6 of which were favored to be of a reactive etiology. Nuclear features most commonly noted in the cases considered to be ASCUS (nonreactive) and SIL were nuclear hyperchromasia and nuclear contour irregularities. CONCLUSIONS Nuclear enlargement alone is not sufficient for diagnosing ASCUS or SIL in PM CVS. Nuclear enlargement in squamous cells is an expected normal reactive change present in PM CVS that resolves with the application of local estrogen. Nuclear hyperchromasia and irregular nuclear contours remain the most reliable cellular characteristics for diagnosing SIL in atrophic CVS. [See editorial on pages 200-1, this issue.] Cancer (Cancer Cytopathol) 1998;84:218-225. © 1998 American Cancer Society.