Tumor Budding Detection by Immunohistochemical Staining is Not Superior to Hematoxylin and Eosin Staining for Predicting Lymph Node Metastasis in pT1 Colorectal Cancer

BACKGROUND: Tumor budding is recognized as an important risk factor for lymph node metastasis in pT1 colorectal cancer. Immunohistochemical staining for cytokeratin has the potential to improve the objective diagnosis of tumor budding over detection based on hematoxylin and eosin staining. However, it remains unclear whether tumor budding detected by immunohistochemical staining is a significant predictor of lymph node metastasis in pT1 colorectal cancer. OBJECTIVE: The purpose of this study was to clarify the clinical significance of tumor budding detected by immunohistochemical staining in comparison with that detected by hematoxylin and eosin staining. DESIGN: This was a retrospective study. SETTINGS: The study was conducted at Niigata University Medical & Dental Hospital. PATIENTS: We enrolled 265 patients with pT1 colorectal cancer who underwent surgery with lymph node dissection. MAIN OUTCOME MEASURES: Tumor budding was evaluated by both hematoxylin and eosin and immunohistochemical staining with the use of CAM5.2 antibody. Receiver operating characteristic curve analyses were conducted to determine the optimal cutoff values for tumor budding detected by hematoxylin and eosin and CAM5.2 staining. Univariate and multivariate analyses were performed to identify the significant factors for predicting lymph node metastasis. RESULTS: Receiver operating characteristic curve analyses revealed that the cutoff values for tumor budding detected by hematoxylin and eosin and CAM5.2 staining for predicting lymph node metastases were 5 and 8. On multivariate analysis, histopathological differentiation (OR, 6.21; 95% CI, 1.16-33.33; p = 0.03) and tumor budding detected by hematoxylin and eosin staining (OR, 4.91; 95% CI, 1.64-14.66; p = 0.004) were significant predictors for lymph node metastasis; however, tumor budding detected by CAM5.2 staining was not a significant predictor. LIMITATIONS: This study was limited by potential selection bias because surgically resected specimens were collected instead of endoscopically resected specimens. CONCLUSIONS: Tumor budding detected by CAM5.2 staining was not superior to hematoxylin and eosin staining for predicting lymph node metastasis in pT1 colorectal cancer.