Biopsies of colorectal clinical polyps--emergence of diagnostic information on deeper levels.

Although the occasional appearance of a normal histology of biopsies from endoscopic colorectal (CR) polyps is generally held knowledge, its prevalence has rarely been focused on, and the yield of additional sections in such cases has been previously addressed in merely four communications. Hitherto, this issue has not been discussed in the context of the clinical setting. The prime aim of this study was to evaluate the yield of step sectioning CR biopsies, considered non-diagnostic (non-diagnostic biopsies (NDB)) on routine sections. The results are correlated with the indications for endoscopy. Additionally, an appropriate, cost-effective approach for handling NDB was sought. Biopsies from 480 clinical polyps were prospectively evaluated by one of three gastrointestinal pathologists and classified as (a) diagnostic biopsies (DB), comprising neoplastic polyps, hyperplastic polyps (HP), sessile-serrated polyp, other diverse causes of polyp formation and (b) NDB comprising normal histology (group 1), suspicious of either adenoma (group 2) or HP (group 3). Material grouped 1-3 was subsequently step-sectioned (three sections prepared from each of nine additional levels). The biopsy specimens were obtained from 245 endoscopies and stratified in the following categories according to the clinical indications: relevant symptoms (symptomatic, n=127), previously documented sporadic large bowel neoplasia (follow-up, n=99), and documented or presumed hereditary condition that confer an increased risk of CRC (hereditary, n=19, including 15 hereditary non-polyposis colorectal cancer (HNPCC) cases). Sixty-five (13.5%) of the 480 samples were classified as NDB (normal morphology n=49, suspicious of adenoma n=5, suspicious of HP n=11), constituting roughly 10% of all biopsies from the symptomatic and the follow-up categories, 32.1% of samples from the hereditary cases, the difference between the hereditary and the non-hereditary cases being statistically significant (p<0.0001). Upon leveling the 65 NDB, a DB emerged in 24 (36.9%) cases, with no significant difference in the yield in relation to the delineated indication categories. Thereby, diagnostic information was obtained with three additional levels in 15 cases, the remaining 9 cases requiring additional sections, ranging from 4 to 8 levels. The present step sectioning approach implied an extra expense of about 112 US$ for each NDB converted to a DB. The higher prevalence of NDB in relation to genetic disorders probably reflects sampling of particularly diminutive lesions. Given the high yield of step sectioning NDB coupled with an acceptable price, our strategy delineated here is recommended in routine practice with the modification of an initial preparation of sections from merely three levels, and if still non-diagnostic, supplementation with additional five levels.