Assessment of pathologic interpretation of colorectal polyps by general pathologists in community practice.



B Denis, C Peters, C Chapelain, I Kleinclaus, A Fricker, R Wild, B Auge, D Chatelain, JF Fléjou.

The assessment of type and dysplasia of colorectal polyps is poorly reproducible in clinical practice.


Aim        

to assess the pathologic interpretation of colorectal polyps detected by a population-based FOBT CRC screening program.


Methods        

A FOBT screening program is underway in the Haut-Rhin area. The pathologic interpretation of colorectal polyps was performed by all 14 general pathologists (P1) practicing in the area. The slides of 300 polyps were reviewed by an expert GI academic pathologist (P2). All serrated adenomas and pTis and pT1 carcinomas were reviewed and the other polyps reviewed were randomly selected.


Results        

The final diagnosis by P2 of the 300 polyps was: hyperplastic polyp (HP) (n = 71), serrated adenoma (SA) (n = 9), tubular (TA) (n = 45), tubulovillous (TVA) (n = 136) and villous adenoma (VA) (n = 12). There was 38 pT1 and 38 pTis carcinomas. 61 adenomas had high-grade dysplasia (HGD) and 72 had low-grade dysplasia (LGD). The initial diagnosis of HP was confirmed by P2 in 88.9% of cases, that of SA only in 15.1% of cases (87.1% of SA diagnosed by P1 were HP according to P2). The initial diagnosis of TA, TVA and VA was confirmed by P2 in respectively 29.1%, 75.6% and 26.7% of cases. The villous component had been overread in 12.2% of TVA and 60% of VA. It had been underread in 69.8% of TA and 6.7% of TVA. LGD was confirmed by P2 in 72% of cases. The initial diagnosis of HGD and pTis carcinoma was confirmed in 66.9% of cases. It had been overestimated by P1 in 27.1% of cases and underestimated in 5.1%. The initial diagnosis of pT1 carcinoma was confirmed in 84.2% of cases. It had been overestimated in 15.8% of cases. The diagnosis of pT1 carcinoma by P2 had been underestimated by P1 in 15.8% of cases. Overall, there was agreement between P1 and P2 for the assessment of type and dysplasia in 40.3% of cases, disagreement without any clinical impact in 11% and with an impact on management in 48.7% of cases, leading to inappropriate endoscopic surveillance intervals (too short n = 73; too long n = 68) or to unnecessary surgical resection (n = 2). The degree of differentiation of pT1 carcinomas was specified in 68.4% of cases, the resection line margin in 73.7% and the presence or absence of lymphatic invasion in 44.7% of cases.


Conclusion        

Our results show a poor inter-observer agreement in the pathologic interpretation of colorectal polyps by general pathologists in community practice with an impact on clinical management in the half of cases. They show the frequent lack of adequate characterization of malignant polyps. They illustrate the need for a clarification of the nomenclature of colorectal polyps, especially for serrated polyps.



Vous pouvez nous contacter à tout moment en nous écrivant un email.

Site réalisé avec le concours de la Ligue contre le cancer