Correlation of rapid on-site evaluation with final diagnosis in the evaluation of renal lesions

Abstract

Rapid on-site evaluation (ROSE) increases adequacy and diagnostic yield of cytology procedures and provides information for rapid clinical decisions. Cytology procedures with ROSE (fine-needle aspiration [FNA] and touch preparation with core biopsy [TP + CB]) are used to evaluate renal lesions, especially prior to concomitant ablation. Consecutive image-guided procedures of FNA and TP + CB of renal lesions with ROSE were reviewed for a ten year period. ROSE diagnoses were correlated with final diagnoses and clinical course. Diagnoses were considered in five categories: positive, atypical/suspicious, nonspecific (including adequate, lesional, cellular, and oncocytic descriptors), negative, and nondiagnostic. Statistical analysis was performed using Fisher's exact test. A total of 209 procedures with 226 ROSE (73 FNA, 119 TP + CB, 17 FNA + TP + CB) were performed. FNAs had more nondiagnostic specimens than CBs by both ROSE and final diagnosis (19 of 90 versus 7 of 136 for ROSE [P = 0.0004], 15 of 90 versus 5 of 136 for final [P = 0.0013]). More FNAs than CBs were positive by ROSE (33 of 90 versus 23 of 136, P = 0.0009), with no difference in positive final diagnoses (66 of 90 versus 106 of 136, P = 0.43). Treatment following diagnosis included ablation (67,with 42 concomitant after ROSE), surgical resection (50), chemotherapy/radiation (42), re-biopsy (5), serial imaging (15), no treatment/other (15), and lost to follow up (15). All lesions with positive ROSE had positive final diagnoses. For cases with positive final diagnoses, ROSE diagnoses were relatively evenly distributed among positive, atypical/suspicious, and nonspecific. TP + CB had higher adequacy than FNA by both ROSE and final diagnosis, although FNAs more often had positive ROSE diagnoses.