|
Back to 2017 Program and Abstracts
ACCURACY OF DIFFERENT CRITERIA FOR LYMPH NODE STAGING IN RECTAL CANCER BY MAGNETIC RESONANCE IMAGING
Florian Loch*, Matthias Taupitz, Christoph Schmidt, Joern Groene, Martin E. Kreis
Charite Universitatsmedizin Berlin, Berlin, Germany
Introduction: Accuracy of preoperative staging of regional lymph nodes by high-resonance MR imaging in patients with rectal cancer is still limited. The aim of this retrospective analysis was to determine the sensitivity and specifity of expanded criteria in regional nodal staging of rectal cancer.
Methods: 60 patients with histologically confirmed rectal cancer underwent surgery without neoadjuvant chemoradiation. MRI of all patients were evaluated by a radiologist with more than 20 years of experience blinded for postoperative histopathological results. All visible lymph nodes were measured (short-axis diameter) and characterized in terms of border contour (lobulated, spiculated/indistinct and unaltered) and homogeneous or inhomogeneous signal intensitiy. An examination was considered node positive (cN+) when at least one lymph node was at least 6 mm of size, altered in border contour or of inhomogeneous signal intensity. Sensitivites and specifities were calculated for each criteria and compared by histopathological examination as reference standard.
Results: 25 of the 60 (41,7%) patients had histologically confirmed lymph node metastasis (pN+). By size criteria (6 mm) 66,7% of patholgoically positive lymph nodes were correctly identified (k=0,298, p=0,028) and 76,7% (k=0,497, p<0,001) by the criteria "spiculated/indistinct border contour.“ 13% versus 5% were overstaged and 20% versus 18,3% understaged by these criteria. Sensitivities and specifities for the size criterium were 52 % versus 77.1 %, 56% versus 85.7% for the size-independent criteria "spiculated/indistinct border contour“, and 56 % versus 91.4 % for the combined criteria "inhomogeneous signal intensitiy and spiculated/indistinct border contour“. T-stage was correctly identified in 40% of patients (24/60), 48,3% were overstaged and 11,7% understaged).
Summary: T-staging by MRI was inaccurate in the majority of patients. Accuracy of lymph nodes staging in rectal cancer is only minimally improved by the use of expanded criteria "border contour“ or "inhomogeneous signal intensitiy“ compared to standard size criteria. Presence of a "spiculated/indistinct border contour“ is correlated with positive nodal status but sensitivity is low. These limitations suggest to use caution when giving neoadjuvant chemoradiation merely based on preoperative positive lymph node staging.
Back to 2017 Program and Abstracts
|
