Autofluorescence Imaging Versus Zoom-Narrow-Band Imaging for Endoscopic Imaging in BE
Video-autofluorescence imaging (AFI) and magnification narrow-band imaging (zoom-NBI) provide multimodal-ity imaging with a single endoscope, allowing for greater detection of high-grade intraepithelial neoplasia (HGIN) in patients with BE. Whereas zoom-NBI is interpreted based upon the presence of regular versus irregular patterns, AFI is interpreted based upon color changes and, thus, may be easier to interpret. To further evaluate differences between the 2 modalities, Kim and colleagues undertook a study comparing interobserver agreement in images obtained via AFI and zoom-NBI in patients with BE. Images with corresponding biopsies were obtained from a prospective trial of tandem AFI and zoom-NBI that used a prototype multimodality endoscope capable of switching between the 2 modalities. The current study compared findings determined by 6 endoscopists, including 3 experts and 3 trainees. The participants first underwent an hour-long structured teaching session using 8 AFI/NBI images and then evaluated a set of 36 AFI images (17 with high-grade dysplasia or cancer) and 44 zoom-NBI images (21 with high-grade dysplasia or cancer) obtained from 25 patients. The endoscopists all reported a median image quality score of 3 (good) on a scale of 1—5. Overall, interobserver agreement was good for both AFI (mean kappa value, 0.48) and zoom-NBI (mean kappa value, 0.50). Mean kappa values for prediction of histology were 0.48 and 0.50 for AFI and zoom-NBI, respectively. No differences in interpretation were noted between experts and nonexperts for images obtained with AFI (mean kappa values, 0.48 and 0.44, respectively). However, for images obtained with zoom-NBI, kappa values were lower in experts versus nonexperts (0.39 and 0.63, respectively). Based upon these findings, the researchers suggested that AFI is easier to interpret and has a shorter learning curve. The sensitivity, specificity, and accuracy for detecting HGIN was 79%, 80%, and 80%, respectively, for AFI, and 89%, 68%, and 77%, respectively, for zoom-NBI. The investigators suggested that further improvements are needed to increase the accuracy of detection with both modalities.
Diagnostic Yield Similar With 22-Gauge Versus 25-Gauge Needles During Endoscopic Ultrasound-Guided Fine-Needle Aspiration
Conway and colleagues presented results of a prospective, randomized trial comparing the diagnostic yield of 22-gauge needles, which are preferred by many endosonographers, and 25-gauge needles during endo-scopic ultrasound—guided fine-needle aspiration (EUS-FNA). This ongoing study enrolled patients undergoing EUS-FNA of solid lesions between January 2009 and November 2009. All lesions underwent 4 passes—2 with each needle size—using the EchoTip Ultra needle (Cook Medical). Block randomization determined the order of needle use. Cytotechnicians applied 10 cc of suction during each pass. Neither the on-site technician nor the cytopathologist had knowledge of the sampling technique. Of the 75 patients included in this interim analysis, 62% were male, the mean age was 66 years, and 83% were white. Patients assigned to receive the 22-gauge needle first were older than patients in the other group (mean age, 69 vs 62 years; P=.02). Lesions were primarily biopsied from the pancreas (49%) and lymph nodes (23%). The average lesion diameter was 31 mm. Diagnoses included adenocarcinoma (41%), reactive adenopathy (17%), gastrointestinal stromal tumors (8%), and suspicious for cancer (7%); 12% were nondiagnostic. The investigators reported no significant differences in diagnostic yield between the 2 needles in pancreatic masses, lymph nodes, or other lesions. Overall, the diagnostic yield, defined as the acquisition of adequate cells for the pathologist to render a diagnosis, was similar with the 22-gauge and 25-gauge needles (83% and 84%, respectively). Blood was present in 80% and 76% of samples, respectively; clots were present in 61% and 63%, respectively; and high cellularity was noted in 60% and 64% of samples, respectively. No immediate complications or needle failures occurred. The researchers suggested that either needle could be used, as diagnostic yield appears to be independent of needle size. order levitra