[1189] Recognition of Nonkeratinizing Morphology in Oropharyngeal Squamous Cell Carcinoma – A Prospective Cohort and Interobserver Variability Study.

James S Lewis, Raja MA Khan, Ramya Masand, Rebecca D Chernock, Qin Zhang, Nasser Said Al-Naief, Susan Muller, Jonathan B McHugh, Manju Prasad, Margaret Brandwein-Gensler, Bayardo Perez-Ordonez, Samir K El-Mofty. Washington University, St. Louis, MO; University of the Pacific, San Francisco, CA; Emory University, Atlanta, GA; University of Michigan, Ann Arbor; Yale University, New Haven, CT; University of Alabama at Birmingham; University of Toronto, ON, Canada; Baylor University, Houston, TX

Background: Nonkeratinizing morphology in oropharyngeal squamous cell carcinoma (NK SCC) is increasingly recognized as strongly correlating with p16 expression. However, NK SCC is not widely recognized by pathologists. We have developed a histologic classification system typing tumors as type 1 (keratinizing SCC), type 2 (hybrid SCC), or type 3 (NK SCC). Our aims were to 1) Correlate the histologic types with p16 expression and 2) Examine the reproducibility of diagnoses for a group of head and neck pathologists applying these criteria.
Design: For aim 1, we captured 6 months of prospective data on routine, clinical SCC cases which were classified using the histologic criteria by the 3 individual central institutional pathologists. For aim 2, from a large research database, 40 test cases were randomly selected and an instruction sheet and 3 example slides circulated for review to 6 head and neck pathologists not previously familiar with the system. p16 immunohistochemistry was performed on all cases with 50% of tumor cells (nuclear + cytoplasmic) staining as the lower cutoff for positivity.
Results: For aim 1, there were 51 cases. All cases called type 3 were p16 positive.

Prospective Cohort: Histology-p16 Correlation

Type	Total Cases	p16 positivity
Keratinizing (1)	13	5 (38.5%)
Hybrid (2)	12	12 (100%)
Nonkeratinizing (3)	26	26 (100%)

For aim 2, agreement was highest between the 6 pathologists for types 1 and 3 (kappa values 0.62 and 0.56, respectively, considered moderate to good agreement; p<0.0001 for both) and lowest for type 2 (kappa 0.35, considered poor agreement; p<0.0001). 33 of the 40 cases were p16 positive. Every case classified as type 3 by any reviewer (21 cases) was p16 positive.
Conclusions: Head and neck pathologists can recognize NK SCC as a distinct subtype of oropharyngeal SCC. In particular, our data again show that when a pathologist classifies a tumor as type 3, this reliably predicts p16 positivity.
Category: Head & Neck

Tuesday, March 1, 2011 1:00 PM

Poster Session IV # 136, Tuesday Afternoon