EUS is an essential part of a comprehensive staging workup for esophageal cancer and, when used appropriately in conjunction with CT and/or PET scan, is generally considered the most accurate tool for staging [31
]. Further, one of the most important parts of EUS in staging esophageal cancer is to ascertain the presence of celiac node involvement [32
]. The echoendoscope needs to be able to pass the malignant stricture in order to provide celiac node assessment. The need for dilating a malignant stricture in order to pass the echoendoscope is reported to be 10–38% [1
]. This emphasizes the importance of dilation in completing accurate and comprehensive EUS staging that includes celiac node assessment. Without dilation, these are patients that would have suboptimal staging, which in turn can negatively affect treatment decisions.
Early studies concluded that dilating malignant strictures for the purposes of EUS staging of esophageal cancer was dangerous, leading to unacceptably high rates of perforation [1
]. More recent studies have refuted this and confirmed that dilating with smaller diameter Savary or through-the-scope dilators, one can safely dilate malignant esophageal strictures in order to complete EUS staging [2
These previous safety studies focused on perforation rates in dilating malignant esophageal strictures. No studies to date report any relationship between dilating malignant strictures and the subsequent development of metastases. It is well established that dilating esophageal strictures carries a relatively high rate of transient bacteremia [27
], presumably through the breakdown of tissue planes allowing direct seeding of bacteria into the blood stream. One could postulate that a similar mechanism could lead to cancer cells seeding the bloodstream during dilation of a malignant stricture, but there are no data to support this theory. This theory was considered at our institution because of the observation of the rapid development of metastases in unusual locations for a number of patients who had dilation of their malignant strictures for EUS staging as part of their otherwise negative pretreatment staging. Besides the analogy to bacteremia during dilation as stated above, the possibility of malignant spread may also be supported by reports of iatrogenic periprocedural spread of tumor cells being well described in other procedures, including seeding needle tracts in breast biopsies [12
], diagnostic and therapeutic procedures for hepatocellular carcinoma [15
], cutaneous seeding in laparoscopic cholecystectomy [18
], and seeding tracts with FNA in pancreatic, esophageal, and thyroid lesions [22
Malignant cells have been noted in the blood stream after various invasive procedures. Prostate cells, both benign and malignant, have been isolated in the circulation following transrectal prostate biopsy and transurethral resection of the prostate [33
]. Furthermore, iatrogenic periprocedural spread of tumor cells directly into the bloodstream has been documented following percutaneous ethanol injection and transarterial chemoembolization in primary liver cancer [34
We did find that among all patients (not matched for similar stage at the time of EUS), those who required dilation had a trend towards an overall higher rate of metastases at any time during followup (71% versus 46%) (). However, this is not completely unexpected as those patients with higher-grade strictures who required dilation generally have more advanced disease [1
], and therefore the need for dilation may be a marker for more advanced or aggressive disease. Thus, these patients may be more likely to develop metastases at an earlier time regardless of dilation.
Because of this observation, we evaluated patients in the dilated and nondilated groups to match those with similar staging—locally advanced disease. When these groups were analyzed, we saw no difference in the rates of metastases at any time interval studied (). Furthermore, there is no distinct pattern of metastatic spread unique to either group, and no distinct pattern of location of metastases based on the time frame they were detected. There is also no difference in the median time to detection of metastases in either group ().
Survival data show a nonsignificant trend towards lower survival in the dilated group when we look at all patients regardless of initial staging. A similar nonsignificant trend is also seen among those without distant metastases at initial staging. However, when the dilated and nondilated groups were matched for similar initial staging, locally advanced disease, five-year survival rates were similar.
The possibility remains that dilation results in seeding the bloodstream and metastatic spread of disease in a small number of cases. However, this theory has not been proven in any previous model or study in this particular clinical setting. While limited by the total number of patients, our study appears to refute that dilation of malignant strictures leads to increased rates of metastases. A more likely explanation is simply that the cases of early metastases in each group represent metastases that were present but not detectable at the time of initial staging.
Our study was limited by a small sample size, retrospective nature of the study, and being a single-center experience. The small sample size makes it difficult to draw any definitive conclusions; however, our study provides valuable information about the natural progression of metastases in patients who undergo EUS for staging of esophageal cancer.
EUS provides the most accurate locoregional staging for esophageal cancer, and dilation may often be necessary to complete EUS staging. Dilating malignant strictures in order to complete EUS staging does not clearly lead to a higher rate of metastases.