[606] Mismatch Repair Status in a Cohort of Rectal Adenocarcinomas before and after Chemoradiation

AM Bellizzi, CD Crowder, WL Marsh, H Hampel, WL Frankel. The Ohio State University Medical Center, Columbus, OH; Brigham and Women's Hospital, Boston, MA

Background: At our institution we have performed mismatch repair immunohistochemistry (MMR IHC) as a screen for Lynch syndrome on all colorectal cancer resections since 2006. Anecdotally we have observed decreased staining in neoadjuvant-treated rectal adenocarcinomas (NTx RA). This diminution is particularly prominent with MSH6, often resulting in an "equivocal" interpretation. We performed this study to systematically evaluate MMR IHC in NTx RA.
Design: We assembled a cohort of 65 matched pretreatment biopsies (bx) and resections (res) from patients receiving NTx for RA. MMR IHC for MLH1, PMS2, MSH2, and MSH6 was performed. Bx and res tissue were stained on the same slide. Tumor IHC was evaluated for intensity (0, 1: less staining than is typical/less staining than internal control, 2: robust staining) and quantity (0-100%) and a score representing the product of these calculated. Intact staining in internal control was required for interpretation. In 14 (22%), res demonstrated a complete response. Mean data was analyzed using Wilcoxon matched-pairs tests, with P < 0.05 considered significant.
Results: Quantitative IHC data is summarized in the table.

Quantitative Mismatch Repair Immunohistochemistry (Score ± SD)

	Biopsy (n=65)	Resection (n=51)	P
MLH1	191 ± 34	171 ± 52	0.002
PMS2	189 ± 36	169 ± 54	0.0005
MSH2	191 ± 38	157 ± 67	0.0002
MSH6	165 ± 54	81 ± 64	<0.0001

score = intensity * percent; sd, standard deviation

Three (4.6%) cases displayed abnormal (absent) MMR IHC (2 MSH2/MSH6, 1 MLH1/PMS2/MSH6). Eight (16%) res demonstrated between 1-5% MSH6 staining (considered equivocal); the corresponding bx revealed 70-100% MSH6 staining. Equivocal results were often seen in tumors with atrophic glands/prominent nucleoli, and although nuclear staining was sparse, diffuse nucleolar staining was frequently observed.
Conclusions: NTx RA demonstrate significantly decreased MMR IHC expression, compared to matched pretreatment controls. MSH6 is especially prone to this phenomenon, not infrequently resulting in equivocal results (decrease did not alter interpretation of other stains). This finding may be due to altered morphology/proliferative activity in treated tumor. Also, limited residual tumor in about a quarter of cases precludes MMR IHC assessment. Before embarking on expensive molecular testing in these cases, staining the pretreatment biopsy represents an attractive alternative.
Category: Gastrointestinal

Tuesday, March 23, 2010 11:45 AM

Platform Session: Section E, Tuesday Morning