Endoscopic esophageal ultrasound understages many early-stage esophageal cancers, possibly leading to undertreatment for many patients, a study has shown.
A review of more than 100 patients with esophageal cancer found that the procedure correctly staged tumor depth in just 39% of pT1a tumors and 51% of pT1b tumors, Dr. Edward Bergeron and his colleagues reported online Dec. 5 in the Journal of Thoracic and Cardiovascular Surgery (2013 [doi:10.1016/j.jtcvs.2013.10.003]).
"Esophageal ultrasound has [been] shown to be inadequate for the definitive evaluation of lymph nodes in early-stage esophageal cancers," wrote Dr. Bergeron and his coauthors. "As a result, [it] should have a limited role beyond the initial staging examination to rule out more advanced lesions that would require neoadjuvant therapy before surgery."
Dr. Bergeron of the University of Michigan, Ann Arbor, and his team conducted a retrospective study of 107 patients who underwent esophagectomy for early esophageal cancers. The patients’ mean age was 66 years. The majority of the tumors (91%) were in the distal esophagus. Most (83%) were adenocarcinomas, and 63% showed signs of Barrett’s metaplasia.
For those with visible lesions, the average tumor size was 9 mm. More than half of the tumors (55%) were confined to the mucosa; 33% were pT1b with submucosal involvement.
Esophageal ultrasound understaged tumor depth in 32% of the pT1a tumors and 49% of the pT1b tumors. Ultrasound overstaged 29% of the pT1a and 51% of the pT1b tumors.
"In particular, three tumors on esophageal ultrasound invaded into but not through the muscularis mucosa," the authors wrote. "Two of these tumors were staged as pT2 [adenocarcinomas], and one of these tumors was staged as a pT3 squamous cell carcinoma on histology."
The sensitivity of esophageal ultrasound for detecting cT1a tumor invasion was 42%; the specificity was 81%.
Of the 1,083 lymph nodes harvested, 14 from nine patients showed metastatic disease. All of these patients had adenocarcinomas that were understaged by esophageal ultrasound, with no evidence of lymph node involvement.
Among these, disease had spread to the muscularis mucosa in one, to the submucosa in six, and to the muscularis propria in two.
"Thus, eight of the nine patients with positive pathologic lymph nodes had tumor invasion into the submucosal level or further," the authors wrote.
In addition, they noted, two of the nine patients also had multiple metastatic nodes, which led to reclassification as pN2 disease. One of these had tumor invasion into the submucosa and one into the muscularis propria.
"It is notable that 48 patients with pathologic T1a-1p disease had no evidence of lymph node spread," the authors wrote.
Esophageal ultrasound understaged 10% (9/90) of patients with lymph node metastases who had been preoperatively staged as cN0. In addition, 17 patients proved to be node negative after ultrasound staged them as having nodes suspicious of metastasis.
All patients underwent esophagectomy. The median blood loss was 250 mL. One patient needed a splenectomy after an intraoperative injury. The median length of stay was 8 days. One patient died after an unexplained respiratory arrest on postoperative day 2.
None of the authors declared any relevant financial conflicts.