Objective: Morphologic diagnosis and grading of anal squamous intraepithelial lesions (ASILs) are challenging. In this study, we investigated interobserver variability and p16 utility in accurately grading anal SIL.
Materials and methods: Six pathologists evaluated the degree of SIL on hematoxylin and eosin slides from 146 anal biopsies, followed by the review of both p16 and hematoxylin and eosin slides in cases where p16 was previously performed. κ was calculated in the following 4 ways: (A) 4-tiered diagnosis (negative for SIL [NSIL], anal intraepithelial neoplasia [AIN 1, AIN 2, AIN 3]); (B) 3-tiered diagnosis (NSIL and AIN 1 [pooled], AIN 2, AIN 3); (A) 3-tiered diagnosis (NSIL, low-grade SIL, high-grade SIL [HSIL]); and (D) 2-tiered diagnosis (no HSIL, HSIL).
Results: There is only moderate agreement with a 4-tiered diagnosis with or without p16 (κ = 0.48-0.57). There is substantial agreement when AIN 2 and AIN 3 are pooled as HSIL in cases with or without p16 review (κ = 0.71-0.78). There is almost perfect agreement with a 2-tiered diagnosis of negative for HSIL and HSIL both in cases where p16 was used and where p16 was not required, with the best agreement for a 2-tiered diagnosis with concurrent p16 review.
Conclusions: This study highlights the importance of a judicious use of p16 for diagnosis. When there is no need for p16 by the Lower Anogenital Squamous Terminology guidelines, interobserver agreement was substantial to almost perfect with a 2-tiered diagnosis. However, when its use is indicated but it is not performed or reviewed, the agreement is much lower even with a 2-tiered diagnosis. Rational use of p16 will ensure diagnostic accuracy and the best possible patient care.