Search

Background: Confocal laser endomicroscopy (CLE) has been shown to reliably predict histology during ongoing endoscopy. To unmask lesions for CLE, chromoendoscopy has been mandated. Usually fluorescein then serves as a contrast agent for CLE, but it does not allow direct nuclear visualization, must be injected, leads to a transient skin discoloration, and may have allergic side effects., Objective: To establish a single topical dye, cresyl violet (CV), for simultaneous chromoendoscopy and in vivo CLE of the lower GI tract., Design: Animal preclinical study, prospective clinical trial., Setting: Mainz University Clinic (tertiary care center). PATIENTS, METHODS, AND INTERVENTIONS: To establish the staining characteristics and optimal concentration of CV, the ileum and colon of 7 BL6 mice were stained with CV (0.1%-2%), and in vivo confocal imaging was performed with FIVE1. In a subsequent clinical trial, 67 sites in 36 patients were topically stained with CV 0.13%, and subsurface serial images were generated at different depths with an endomicroscope., Main Outcome Measurements: Prediction of histology according to the Mainz confocal classification and nuclear visualization with topical CV., Results: Endomicroscopy with topical CV yielded (sub-)cellular details of normal mucosa, and regenerative and neoplastic changes at variable imaging depths in high resolution comparable to those with intravenous fluorescein. By cytoplasmic enrichment of CV, nuclear morphology could be negatively visualized. Reliable differentiation of nonneoplastic versus neoplastic changes during ongoing endoscopy and a high interobserver agreement based on the microscopic images generated in vivo could be achieved., Limitations: Single-center study, nonrandomized, limited number of patients., Conclusions: CV can be applied topically and allows simultaneous chromoendoscopy and endomicroscopy with accurate prediction of histology with visualization of nuclear morphology. It may therefore be a single-agent alternative to chromoendoscopy and fluorescein in endomicroscopy.

Background: EUS-guided FNA (EUS-FNA) is an established tissue-acquisition technique, with most studies concentrating on cytologic analyses of specimens, with only few data existing on histologic assessment., Objective: To assess the sensitivity of a combined analysis of histologic followed by cytologic tissue diagnosis., Design: A retrospective 3-center study., Methods: In consecutive patients undergoing FNA of solid pancreatic masses, core specimens were harvested for histology; residual tissue was examined cytologically. Only unequivocally positive results were regarded as malignant. Criterion standards were positive results from EUS-FNA or other histologic findings, or, if negative, clinical follow-up data (minimum 12 months)., Results: Among 192 patients (110 men; mean age 63 years) with mostly pancreatic-head masses (72.4%), overall, adequate tissue was obtained in 98.9% of all cases, with a mean of 1.88 needle passes and an overall sensitivity of 82.9% (95% CI, 76.0%-88.5%). Histology and subsequent cytology provided adequate tissue and sensitivities of 86.5% and 60%, and 92.7% and 68.1%, respectively. Excluding cases with inadequate specimens, sensitivities rose by 4% to 10%. Histology showed a trend for superiority over cytology only in characterizing nonadenocarcinoma tumor types. No differences in sensitivity were found between the centers involved., Limitations: Retrospective design, different processing of cytologic specimens., Conclusions: At EUS-FNA in pancreatic masses, combined histologic-cytologic analysis achieved a sensitivity of more than 80%, despite a low number of needle passes and may thus save time. Histology alone did not reach higher sensitivity than cytology. In particular situations, eg, rare tumors, histology may still be required.