Ectopic Crypt Formation and Other Histological Parameters in Relation to BRAF and KRAS Mutation Status of Dysplastic Serrated and Non-Serrated Colorectal Polyps
Michael J O'Brien, Robert D Odze, Sandra Cerda, Huihong Xu, Katherine Downey, Bohdana Burke, Christopher S Huang, Francis A Farraye, Shi Yang. Boston University Medical Center, Boston, MA; Brigham & Women's Hospital, Harvard Medical School, Boston, MA
Background: Activating BRAF/KRAS mutations are determinants in the serrated pathway of carcinogenesis and are present in a variety of neoplastic polyp precursors. To date, no studies have related specific morphologic features of dysplastic serrated polyps (DSP) to oncogene status. Thus, we evaluated histological features of putative DSP and correlated them to BRAF/KRASmut status.
Design: 118 polyps, including 78 that exhibited both a serrated component and dysplasia and/or eosinophilic cell atypia and 34 conventional adenomas, the latter by size category, were sequentially accessioned from our dept. files. DNA was assayed for KRAS (codon 12&13) and BRAF (exon 15) mutations. The following histological parameters were assessed independently of the WHO diagnostic category of polyp: 1) Background (contiguous) hyperplastic polyp or SSA/P (HP/SSAP), 2) Presence or absence of ectopic crypt formation (ECF), 3) %Serration, 4) %Eosinophilic Cell Atypia, 5) %Villi and 6) High grade dysplasia/intramucosal adenocarcinoma (HGD). The reviewer was blinded to the BRAF/KRASmut status of the polyps.
Results: Among all polyps studied (118), ECF was present in 17/38 (44.7%) KRASmut, 13/50 (26%) BRAFmut and 4/30 (13.3%) KRAS/BRAFwt. ECF correlated significantly with villous component >25% (27/49 (55.1%) vs 7/69 (10.4%) [p<.0001]) but not with the presence of serration (58/93 (62.4%) vs 10/25 (40%) [p=ns]). Among DSP only, background HP/SSAP was present in 23/ 50 (46%) that were BRAFmut compared to 3/38 (7.9%) KRASmut (p<.0001). Among DSP without background HP/SSAP, ECF did not correlate with oncogene mutation status (8/15 (53.3%) KRASmut vs 10/21(47.6%) BRAFmut). The frequency of HGD was 2/22(9.1%) and 2/19(10.5%) in KRASmut and BRAFmut DSP without HP/SSA, respectively, compared to 3/29(10.3%) BRAFmut DSP with contiguous HP/SSAP (p=ns).
Conclusions: The presence of ECF correlated with villous morphology but not serration and was not specific to either KRAS or BRAF mutations in DSP. Furthermore, ECF did not discriminate KRASmut DSP from BRAFmut DSP or KRASmut non-serrated adenoma. Contiguous HP/SSAP, on the other hand, was a robust indicator of BRAFmut status. The overall findings have implications for the classification of DSP and question the validity of ECF as the histological hallmark of TSA.
Tuesday, March 20, 2012 8:30 AM
Platform Session: Section D, Tuesday Morning