A. Usui1, K. Okita1, T. Nishidate1, K. Okuya1, E. Akizuki1, M. Ishii1, T. Satoyoshi1, I. Takemasa1 1Sapporo Medical University Hospital,Sapporo, Hokkaido, Japan
Introduction:
Lateral pelvic lymph node (LPLN) dissection in advanced lower advanced rectal cancer remains a subject of debate, and in order to justify this procedure, a reliable method of nodal characterization to assess preoperative nodal status is essential. Computed tomography (CT) of the pelvis is utilized to determine preoperative lymph node metastasis in rectal cancer, and morphological and internal structure of lymph nodes report to be promising factors. In this study, quantitative analysis of morphological and internal structure of LPLN in lower rectal cancer patients was performed to clarify the diagnostic value of these parameters in terms of metastasis.
Methods:
Data were retrospectively collected for 41 patients who had undergone total mesenteric excision with LPLN dissection for lower advanced rectal cancer from 2007 to 2014 at our institution. None of the patients had any treatment prior to surgery. All patients underwent preoperative enhanced CT. A single LPLN largest in short axis diameter was selected on the enhanced phase CT image for each patient, and the region of interest was manually drawn along the margin of the node. The number of voxels was calculated to determine the size of the lymph node and the mean signal intensity was recorded for the enhancement quality. Internal heterogeneity was quantified using kurtosis, skewness, and standard deviation of the pixel distribution histogram. All factors were compared between LPLN with and without metastasis pathologically proven with surgery.
Results:
An average of 16.9 LPLN was identified histopathologically for each patient (range: 2 to 39). Of the 41 patients, the LPLN detected on CT in 9 patients were too small for quantitative evaluation. Nine patients were histologically positive for LPLN metastasis. Compared to those without metastasis, the LPLN in these patients had a significantly increased number of voxels and significantly low mean signal intensity. Regarding heterogeneity, skewness was significantly higher and standard deviation was significantly lower in positive LPLN status. No significant difference was observed in kurtosis.
Conclusion:
LPLN with pathologically proven metastasis are presented on enhanced CT as larger in size, less enhanced, and CT enhancement was less heterogenous, compared to non-metastatic lymph nodes. Quantitative analysis with enhance CT using number of voxels, mean signal intensity, skewness, and standard deviation of the pixel distribution histogram are promising for discriminating metastatic lateral pelvic lymph nodes in lower rectal cancer.