Tumor Budding Score Based on 10 High-Power Fields (HPFs) Is a Reliable and Reproducible Scoring System in Colorectal Cancer
Alessandro Lugli, Agathi Kondi-Pafiti, Viktor Koelzer, Inti Zlobec, Eva Karamitopoulou Diamantis. Clinical Pathology Division, Bern, Switzerland; University of Athens, Athens, Greece
Background: One of the main challenges in colorectal cancer (CRC) research is to improve the prognostic stratification power of the TNM staging system, especially in Stage II CRC. Although recognized by the UICC/AJCC as an additional prognostic factor, tumor budding remains unreported in daily diagnostic work due to the absence of a standardized scoring method. The aim was therefore to assess the reproducibility of tumor buds measurement by evaluating 10-high-power fields (HPFs) at the invasive front and to confirm the prognostic value of tumor buds in our CRC setting.
Design: The study design was based on the REMARK guidelines published by the JNCI in 2005. Whole tissue sections of 215 CRCs with full clinico-pathological data as well as treatment and follow-up information were stained with cytokeratin AE1/AE3. Two pathologists scored the presence of buds across 10-HPFs at the invasive front and the measurements were correlated to patient and tumor characteristics.
Results: The 10-HPF scoring system was reproducible as underlined by an excellent inter-observer agreement (p<0.0001). Low-grade, moderate- and high-grade budding were defined as an average of <5 buds (n=78), 5-19 buds (n=108) and ≥20 buds (n=29), per 10-HPFs respectively. In univariate analysis, high-grade tumor budding was associated with higher TNM-stage (p=0.0003), vascular invasion (p<0.0001), higher tumor grade (p<0.0001), infiltrating tumor border configuration (p<0.0001) and reduced survival (p<0.0001). Multivariable analysis confirmed its independent prognostic (p=0.03) effect not only when adjusting for TNM-stage, L and V stage and tumor grade, but also when including adjuvant therapy (p=0.007). Additionally, tumor budding (HR=1.48) was a stronger prognostic parameter than tumor grade (HR=1.02).
Conclusions: Using the 10-HPF scoring system, tumor budding leads to an excellent inter-observer agreement and is a strong independent prognostic factor in CRC in accordance with the literature. Additionally, tumor budding seems to have a stronger prognostic power than tumor grade which makes it a valuable candidate parameter to better stratify CRC patients into prognostic subgroups.
Tuesday, March 20, 2012 1:00 PM
Poster Session IV # 72, Tuesday Afternoon