2010 Annual Meeting Abstracts
Back to Program
Nine out of ten patients with T1 rectal cancer do not need radical surgery. But who is the one that does?
*Harry M Salinas1, *C L Klos2, *Abdulmetin Dursun1, Paul Shellito1, Patricia Sylla1, David Berger1, Liliana Bordeianou1
1Massachusetts General Hospital, Boston, MA;2Universiteit van Amsterdam, Amsterdam, Netherlands
Objective: To determine predictors of regional lymph node metastasis for T1-T2 rectal cancers with radiologically negative nodes
Design: Retrospective review, 3/1998 - 9/2007
Setting: Tertiary care center
Patients: 82 consecutive patients treated with radical resection (LAR or APR) for histologically confirmed T1-T2 rectal cancer, age 39-91 years (mean 66). Records reviewed for: bleeding, obstruction, perforation, personal/family history of colorectal neoplasms, radiologic and histologic TNM stage, lymphovascular/perineural/or small vessel invasion, histologic subtype, and depth of invasion
Main Outcome Measures: Impact of above factors on rate of positive lymph nodes on univariate and logistic regression analysis.
Results: 58 patients underwent LAR and 19 APR. Despite preoperative staging with MRI, CT or PET, 11.4% (4/35) of T1 patients and 27.7% (13/47) of T2 patients had positive nodes. On univariate analysis, the only significant predictor in patients with radiologically negative nodes was the depth of invasion: 52.2% of patients with negative nodes vs 86.7% w/ positive nodes had tumors abuting the muscularis propria (p=0.018). On logistic regression analysis incorporating depth, size, distance from anal verge, differentiation, adjacent structure invasion, and lymphovascular/perineural/or vascular invasion; only depth of invasion remained significant. (OR: 0.13, P=0.033; OR: 0.6, P=0.16; OR: 0.89, P=0.084; OR: 0.11, P=0.302; OR: 0.62, p=0.586; OR: 1.35, p=0.826; respectively)
Conclusions: 88.6% of patients with T1 and 72.3% of patients with T2 disease undergo unnecessary radical surgery. Preoperative staging with MRI, CT and PET scans cannot identify these patients reliably. Markers of aggressive disease (differentiation, lymphovascular invasion) are not helpful. Thus, local resection for T2 rectal cancer is not justified, while resections for T1 disease should only be offered to motivated patients who will comply with aggressive postoperative surveillance
Back to Program