Missed Cancers in the Upper Gastrointestinal Tract After Esophagogastroduodenoscopy

G&H You and a colleague recently published a commentary on a study that examined missed cancers in the upper gastrointestinal tract after esophagogastroduodenoscopy. What was the definition of a missed cancer in this setting?

RE Raftopoulos and colleagues at the University of Western Australia in Perth, Australia conducted a retrospective cohort study to better define missed cancer rates in esophagogastroduodenoscopy (EGD) and the natural history of upper gastrointestinal cancer in a Western population. This study, along with our commentary, was published in The American Journal of Gastroenterology earlier this year. Two categories of missed cancer were established in the study. A possible missed cancer was defined as a cancer that developed 1–3 years after a seemingly normal EGD, whereas a definite missed cancer was defined as a cancer that developed within 1 year following the index normal EGD. A cancer diagnosed more than 3 years after an EGD was classified as a new cancer.

G&H Could you discuss the study design and findings?

RE Data were derived from an Australian registry of 28,064 patients who had undergone an EGD at the study institution in Western Australia. Between 1990 and 2004, 822 cancers were diagnosed in this group of patients. The researchers reviewed the database for patients who had no record of cancer on EGD but who developed an upper gastrointestinal cancer within the subsequent 3 years (ie, patients with definite or possible missed cancers).

The researchers found that approximately 3.5% of patients developed an upper gastrointestinal tract cancer within 1 year of a normal EGD. Another 3.2% of patients developed an upper gastrointestinal cancer 1–3 years after an initial EGD that demonstrated no cancer findings. In total, approximately 6.7% of patients in the database developed an upper gastrointestinal cancer over a 3-year period following an EGD that showed no initial cancer findings.

These 2 groups of patients were then examined to determine a possible explanation for their cancer. The researchers found that approximately 70% of these patients had abnormalities noted at a site during their initial upper gastrointestinal evaluation. However, these abnormalities had not been definitively identified as cancer at the time of the procedure–in other words, suspicious findings were noted, but a diagnosis of cancer was not definitively established until a later date. In addition, over 60% of these patients experienced alarm symptoms such as anemia, weight loss, or dysphagia, which were suggestive of a serious underlying pathology.

The researchers concluded that particular attention should be given to any abnormalities noted on EGD, as most of the 6.7% of patients who had missed and possible missed lesions on upper gastrointestinal evaluation had abnormalities that were endoscopically noted but not definitively diagnosed as cancer based on their index procedure.

G&H How does this study compare to earlier studies examining the miss rates of upper gastrointestinal cancers?

RE Only a few other studies have looked at this issue in a similar patient population. EGDs have not been assessed in great detail in Western populations. However, there have been several studies from Japan, as well as one study from Scotland. In Japanese studies, the rates of missed upper gastrointestinal lesions have reportedly been as high as 19%. These figures are considerably higher than those reported in the study by Raftopoulos and associates as well as those reported in the study from Scotland.

G&H What lessons can be learned from these studies?

RE These studies have many parallels to colonoscopy and demonstrate that it is possible to miss lesions on EGD. Most of the missed lesions in the study by Raftopoulos and coworkers were associated with the presence of questionable or "mildly abnormal" findings. Thus, based upon due diligence, suspicious findings or alarm symptoms should prompt careful follow-up and even repeat endoscopy to address any ongoing concerns or questions.

G&H Could you expand on the parallels between missed upper gastrointestinal cancer after EGD and missed colonic polyps and cancers after colonoscopy?

RE EGD and colonoscopy have considerable parallels in terms of their cancer miss rates. A significant number of lesions, particularly right-sided lesions, appear to be missed during colonoscopy. A number of tandem studies have been conducted in which colonoscopies were immediately repeated, and missed lesions were found to be quite common. However, these missed lesions were usually not advanced; they tended to be small polyps. EGD has demonstrated the same finding, except that many of the patients undergoing this procedure appear to have an abnormality that is seen but not recognized as significant or, alternatively, they have symptoms suggestive of a serious underlying pathology.

Many of the missed lesions in colonoscopy have been either biopsied or sampled and then incorrectly labeled as a benign lesion or incompletely removed (ie, a polyp). Similarly, many of the missed upper gastrointestinal cancer lesions were noted to be abnormal but were incorrectly identified as benign without recognizing their potential for cancer. Thus, there are many parallels to colonoscopy, and it behooves endoscopists to perform EGD thoroughly and to follow up carefully with these patients.

G&H What role do adjunctive techniques such as chromoendoscopy have in the detection of subtle upper gastrointestinal lesions, including cancerous and precancerous lesions?

RE The benefit of ancillary techniques such as different types of imaging (eg, narrow-band imaging or chromoendoscopy) is controversial. Some studies evaluating ancillary techniques have reported very positive findings, whereas other studies have not been as encouraging. Nevertheless, these techniques all offer methods of assessing suspicious lesions in a slightly different manner. Thus, they facilitate the due diligence of endoscopists to assess all suspicious lesions completely and adequately in order to ascertain the absence of any underlying malignancies that are developing or will develop in the near future. Therefore, ancillary techniques play a significant role in lesion detection, particularly in the context of suspicious findings.

G&H What implications do these studies have for the training of future endoscopists?

RE These studies stress the importance of quality EGD. Often, the importance of EGD is dismissed, and the procedure is performed quickly, without striving for quality and thoroughness. Although the learning curve for EGD is not as steep as for colonoscopy, lesions can be missed just as easily in the upper gastrointestinal tract as in the colon. Thus, the quality of EGD is just as important as for lower colonoscopy. The presence of any endoscopic abnormalities or clinical alarm symptoms should encourage endoscopists to be particularly thorough. In order to emphasize the importance of quality to the younger generation of endoscopists, practitioners should be taught to perform EGD similar to how colonoscopy is performed–ideally, in a thorough manner and under the guidance of a report card system.

G&H Are there any upcoming or ongoing studies evaluating missed gastrointestinal cancers?

RE I am not aware of any such studies currently underway, but this area is certainly an important one in which other database-driven studies will be conducted in the future. These studies will likely evaluate gastrointestinal cancer miss rates by examining cancer databases in conjunction with endoscopy databases, just like in the studies discussed above.