Vitreous evaluation: a diagnostic challenge.

PURPOSE: To categorize vitrectomy cytologic diagnoses and ancillary tests to address appropriate processing of low-volume vitreous samples. DESIGN: Retrospective case series. PARTICIPANTS: Five thousand seven hundred thirty-six vitreous samples. METHODS: Cytologic diagnoses of therapeutic and diagnostic vitrectomy samples and their processing protocols from 3 teaching institutions were reviewed. MAIN OUTCOME MEASURES: Diagnostic results were categorized as negative for malignancy, suspicious for malignancy, and positive for malignancy. All ancillary studies performed were documented, including special stains, immunohistochemistry analysis, cytokine levels, and polymerase chain reaction (PCR) analysis. RESULTS: Of the 5736 vitreous samples analyzed, 4683 (81.64%) were from Tufts Medical Center (TMC), 955 (16.65%) were from Boston Medical Center (BMC), and 98 (1.70%) were from Massachusetts Eye Research and Surgery Institution (MERSI). Cases from TMC and BMC were therapeutic and diagnostic vitrectomies, and MERSI cases were diagnostic vitrectomies. Most vitrectomies showed negative results for malignancy: 99.47% of TMC cases, 99.89% of BMC cases, and 79.6% of MERSI cases. These included vitreous hemorrhage and inflammatory or infectious findings. Ancillary studies performed in this category included Periodic Acid-Schiff staining for fungi, PCR analysis for toxoplasmosis, cytomegalovirus, Epstein-Barr virus (EBV), herpes simplex virus I and II, and vitreous cultures for infections (coagulase-negative Staphylococcus, Candida, Fusarium, and Propionibacterium species). Interleukin (IL) 10-to-IL-6 ratios were performed on 38.7% of cases from MERSI. Fourteen cases from TMC were suspicious for malignancy based on cytologic evaluation. Eleven cases from TMC, 1 case from BMC, and 20 cases from MERSI showed positive results for malignancy and included B-cell lymphoma, retinoblastoma, melanoma, and metastatic adenocarcinoma. The ancillary testing included PCR for heavy chain immunoglobulin gene rearrangements, immunohistochemistry for EBV, in situ hybridization for κ and λ light chains, and cytogenetics. CONCLUSIONS: This is the largest data pool of reported cytologic diagnoses of diagnostic and therapeutic vitrectomy samples. Cytologic evaluation of therapeutic vitrectomy samples provides a valuable baseline of nonpathologic findings that assist in differentiation between malignancy, infections, and inflammatory conditions. Allocation of small-volume vitreous samples to select ancillary testing from the plethora of available diagnostic tests requires preoperative communication between surgeons and pathologists to ensure appropriate and timely treatment methods.