Sentinel lymph node biopsy in gastric cancer: possible applications and limits.

AIMS: To explore the feasibility and accuracy of sentinel lymph node (SLN) biopsy in gastric cancer. PATIENTS AND METHODS: Twenty-nine patients with clinical T1 and T2 N0 M0 gastric cancer less than 5 cm in diameter underwent SLN biopsy with the intraoperative Patent blue method. The procedure continued with radical gastrectomy and D2 lymphadenectomy. We investigated all technical aspects of the blue dye technique and determined the accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of the SLN technique. RESULTS: SLNs were detected in 28 of 29 patients; the total number of excised SLNs was 45, with a median of two (range 1-3). Seventeen patients had metastatic SLN, with 21 lymph nodes retrieved. Twenty-two patients had SLNs located at the first level. Four patients had SLNs at the second level, one at the first and second levels, and one at the first and third levels. Five patients had false negative SLNs. The ability of SLN biopsy to predict the status of the other lymph nodes was summarised by an accuracy of 75%, a sensitivity of 75%, a specificity of 75%, a positive predictive value of 88%, and a negative predictive value of 55%. CONCLUSIONS: Our study demonstrates that pick-up SLN biopsy in gastric cancer is technically feasible but has very low sensitivity. Regarding the utility of SLN navigation when attempting to detect the nodal basin, the high rate of false negative SLNs and lymph node level jumping warrant further studies with a large accrual before the method can be introduced into daily practice.

Reliability of pancreatic cancer staging classifications.

The prognostic value of UICC and JPS classifications was evaluated in a group of 74 patients undergoing resection for pancreatic carcinoma in the 1982-1992 period. The predictivity of peritoneal cytology and intraoperative liver biopsies was evaluated in a subgroup of 15 patients. Patient division by stage according, respectively, to UICC and JPS classifications was: stage I: 28 and 2 patients; stage II: 12 and 11 patients; stage III: 29 and 30 patients; stage IV: 5 and 31 patients. Median survival by UICC stages was: 17 months (m) in stage I, 10 m in stage II, 12 m in stage III, and 6 m in stage IV; by JPS stages it was 29 m in stage II, 14 m in stage III, and 7 m in stage IV. A radical resection was possible in 48 patients (65%); the survival was significantly worse in the nonradical resection group. The survival difference among UICC stages was not significant in a multivariate analysis with radicality, whereas among JPS stages a significant survival difference was found in both univariate and multivariate analysis. The tumor invasion of the retroperitoneal tissues and the presence of distant metastasis were indicators of a significantly worse prognosis. Lymph node involvement, tumor size, and serosal invasion had no significant effect on survival. Peritoneal cytology and liver biopsies did not provide further prognostic information.