R ROMAN1, M VERDU1,3, A VIDAL3,5, M CALVO4, X SANJUAN5, M JIMENO6, A SALAS7, J AUTONELL8, I TRIAS9, N RODON1, M GONZALEZ1, B GARCIA1 and X PUIG1,2,3.
1BIOPAT.Biopatologia Molecular, Barcelona, Spain; 2Hospital de Barcelona, Barcelona, Spain; 3Histopat Laboratoris, Barcelona, Spain; 4Statistics Department, Universitat de Barcelona, Barcelona, Spain; 5Hospital Universitari de Bellvitge, Barcelona, Spain; 6Hospital del Mar, Barcelona, Spain; 7Hospital Mútua de Terrassa, Terrasa, Spain; 8Hospital General de Vic, Vic, Spain and 9Hospital Plató, Barcelona, Spain.
Background: High microsatellite instability (MSI-H) allows the identification of a subset of colorectal adenocarcinomas (CRC) associated with good prognosis and Lynch syndrome. The objective of this work was to assess the interobserver variability of a MSI-H prediction model based on phenotypic features. This model was previously generated using a series of 204 cases.
Design: The validation series from five different hospitals included 265 primary CRC. The eight parameters that integrate our model, namely tumor location, growth, solid, mucinous and cribiform patterns, presence of Crohn-like reaction, Ki-67 and p53 immunophenotype, were evaluated in the corresponding centers and our prediction model applied to each case. The instability status was evaluated using a panel of 11 microsatellites. Simultaneously a prospective series of 148 cases was collected and identically assessed in our institution.
Result: Homogeneity assessment revealed significant differences between hospitals in the estimation of Ki-67, Crohn-like reaction, growth and cribiform patterns. Despite this observation, our model was globally able to predict MSI-H with a negative predictive value of 97.0%. The results obtained with the prospective series of 148 cases collected in our institution were equivalent, achieving a negative predictive value of 97.8%.
Conclusion: The complexity of the MSI study has triggered the development of models based on pathological features to predict instability status. A recently published model looks at pathology features associated with MSI in a series of CRC diagnosed before age 60. The model we present focus on a non-selected population of patients aiming at identifying not only possible Lynch syndrome candidates, but any CRC exhibiting an MSI-H phenotype and thus a better prognosis. The high negative predictive value achieved by our model allows the reduction of the cases to be tested for MSI to less than 10%. Furthermore, the easy evaluation of the parameters included in the model renders it a useful tool for the routine practice.
Texto completo
Abstract exhibited at the:United States and Canadian Academy of Pathology
98th Annual Meeting
Boston, USA, March 7-13, 2009
98th Annual Meeting
Boston, USA, March 7-13, 2009
Abstract in:
Modern Pathology 2009; 22 (suppl. 1): 147A