Laparoscopic ultrasonography for staging gastroesophageal cancer.

BACKGROUND

TNM staging of gastroesophageal cancer is improved by the use of laparoscopy for the detection of occult metastases and endoscopic ultrasonography for T and possibly N staging. Laparoscopic ultrasonography may combine the strengths of both of these techniques. The purpose of this study was to compare TNM staging by means of laparoscopic ultrasonography (Lap US), laparoscopy, and conventional computed tomography (CT).

METHODS

TNM stage was determined by using Lap US, laparoscopy, and CT scanning in 26 surgical candidates with gastric or esophageal cancer in a blinded fashion. CT scans were also reviewed separately by an expert radiologist. Validation of findings was by means of final pathologic examination.

RESULTS

Resectability for potential cure was determined by means of Lap US with a sensitivity of 100% and a specificity of 91% versus 100% and 73% for laparoscopy and 75% and 60% for CT, respectively. Overall TNM staging was 82% accurate for Lap US versus 67% for laparoscopy and 47% for CT (chi-squared, 8.64, p < 0.005, and 10.4, p < 0.005, respectively). Accuracies for Lap US versus laparoscopy versus CT for N stage were 92% versus 84% versus 70%, respectively, for T stage were 92% versus 42% versus 60%, respectively, and for M stage were 89% versus 44% versus 62%, respectively.

CONCLUSIONS

T and N staging by Lap US were comparable to published results for endoscopic ultrasonography, and overall TNM staging was better. Lap US may provide the optimal preoperative staging for gastric cancer and has the potential to complement endoscopic ultrasonography in staging esophageal cancer.