We found 15 incidental GISTs in the UGI tract of 11 patients during esophagectomy or gastrectomy, leading to an incidence of 5.3% (11 of 207), which was much higher than previously reported by Liu and colleagues,9
who found an incidence of 0.74% (49 of 6585) among patients with UGI epithelial malignancies. Although the reason for this discrepancy is not known, differences in the patient demographic characteristics — a highly heterogeneous population in Montréal, Canada, versus a Chinese population in Sichuan Province, China — may have played a major role. We also found a significantly higher proportion of adenocarcinomas (78% [161 of 207] v. 64% [4203 of 6585], p
< 0.001), especially in the esophagus (25% [51 of 207] v. 0.5% [35 of 6585], p
< 0.001). Based on studies reported by Kawanowa and colleagues7
and Abraham and colleagues,8
incidental GISTs are preferentially located in the upper part of the stomach and GEJ of patients with UGI epithelial malignancies. Surgical specimens resected with distal gastrectomy for distal gastric adenocarcinomas or esophagectomy with minimal gastric resection for more proximal diseases may not include this entire UGI segment, hence decreasing the likelihood of identifying incidental GISTs. Likewise, surgical approach may influence the intraoperative detection of a synchronous GIST. Esophagectomy using a transthoracic approach without entering the abdomen may leave synchronous GISTs undetected intraoperatively at a more distal site. Since information regarding surgical procedures was missing in the report by Liu and colleagues,9
no comparison can be made between their results and ours. When extensive pathological examination with microscopic evaluation of the entire specimen was undertaken, Kawanowa and colleagues7
and Abraham and colleagues8
reported small GISTs in the surgical specimen in 35% and 10%, respectively, of patients undergoing resection of UGI epithelial cancer. Since most of the incidental GISTs were less than 1 cm in diameter, serial sections skipping a depth greater than 1 cm could leave these tumours undetected. Thus, the detection rate greatly depends on the number of histological sections per specimen examined. This may be another plausible contributing factor leading to the apparent difference. Based on our results and those of other published studies, incidental GISTs can be found frequently, and their prevalence is likely dependent on the detection methods and perhaps on the patient population.
Previous studies reported low rates of preoperative and intraoperative detection of incidental GISTs (0%–6%),7–9
whereas the intraoperative detection rate in the present study was 33% (5 of 15 incidental GISTs). The GIST size in the present series was not significantly larger than that reported by Liu and colleagues,9
but was much larger than those reported by Kawanowa and colleagues7
and Abraham and colleagues.8
Nevertheless, those identified pre- or intraoperatively in all these studies, including ours, were present either on the serosal or intraluminal surface, which could have been easily detected by preoperative diagnostic laparoscopy or gastroscopy, respectively. Thus, location rather than size of the incidental GISTs determined their intraoperative detection. The higher intraoperative detection rate in our study might only have been due to coincidence.
With respect to pre- or intraoperative identification of incidental GISTs in the context of a primary UGI epithelial malignancy, alteration of surgical therapy must be considered. If the decision is to remove the incidental GIST, surgical outcome may differ in 2 ways. First, obtaining a negative margin for the GIST can compromise the integrity of the gastric conduit depending on its size and location. Second, although the rate of staple line leak varies in the literature depending on the techniques and materials used, any additional staple line theoretically increases the chance of postoperative leak. By contrast, leaving an incidental GIST behind is associated with a risk of tumour progression, although these incidental GISTs all had a low risk of malignant potential in our study and in other published series.7–9
In addition, residual incidental GIST may be mistaken for recurrence or metastasis of the primary epithelial malignancy at follow-up. Consequently, patients may be subjected to unnecessary additional therapy when no recurrence truly exists. In light of these possibilities, it has been recommended that incidental GISTs be removed en bloc with other tumours when possible. Alternatively, local resection should be performed.10–12
Owing to the presence of incidental GIST, 5 (45%) patients with primary epithelial malignancies in our series required additional procedures intraoperatively. Simple excision or wedge resection was performed in all cases involving the eventual gastric conduit used to restore intestinal continuity. No cases of anastamotic leak were observed in this group of patients. However, because of the likelihood of much slower progression of incidental GIST relative to the primary epithelial malignancies in patients with gastroesophageal cancers in our study and in the literature, and given the poor prognosis associated with the primary epithelial malignancy in these patients (45% of patients died of recurrence of their primary cancer during the follow-up period of 6 to 36 mo), we suggest that excision of incidental GISTs be carried out only if this additional resection does not compromise the integrity of the gastric conduit and the intestinal continuity.
Surgical resection remains the treatment of choice for localized GISTs. Currently, no evidence exists supporting the use of adjuvant or neoadjuvant therapy in the treatment of non–high risk GISTs in which surgical resection is possible.13
All incidental GISTs found in our series and in the published reports7–9
were determined to have low or very low risk of malignant potential. None of the patients in our series or that by Liu and colleagues9
had GIST recurrence despite the lack of adjuvant therapy for GISTs. These patients typically died of a recurrence of their primary epithelial cancer because most of them had stage II to IV disease. Because prognosis is greatly dictated by the primary epithelial cancers rather than the GISTs, adjuvant therapy should be focused primarily on the more aggressive disease.