Background: Tumor budding along the advancing front of colorectal adenocarcinoma is an early event in the metastatic process. A reproducible, prognostic budding scoring system based on outcomes in early stage colorectal cancer has not been established.
Design: One hundred twenty-eight T3N0M0 colorectal carcinoma patients with known outcome were identified. Tumor budding was defined as isolated tumor cells or clusters of <5 cells at the invasive tumor front. Tumor bud counts were generated in 5 regions at 200x by 2 pathologists (conventional bud count method). The median bud count per case was used to divide cases into low (median=0) and high budding (median > or =1) groups. Forty cases were reevaluated to assess reproducibility using the conventional and a novel rapid bud count method.
Results: Fifty-seven (45%) carcinomas had high and 71 (55%) had low budding scores. High budding was associated with an infiltrative growth pattern (P<0.0001) and lymphovascular invasion (P=0.005). Five-year cancer-specific survival was significantly poorer in high compared with low budding groups: 63% versus 91%, respectively, P<0.0001. Multivariate analysis demonstrated tumor budding to be independently prognostic (hazard ratio=4.76, P<0.001). Interobserver agreement was at least equivalent comparing the conventional to the rapid bud count methods: 87.5% agreement (kappa=0.75) versus 92.5% agreement (kappa=0.85), respectively.
Conclusions: Tumor budding is a strong, reproducible, and independent prognostic marker of outcome that is easily assessed on hematoxylin and eosin slides. This may be useful for identifying the subset of T3N0M0 patients at high risk of recurrence who may benefit from adjuvant therapy.