A re-review of caspule endoscopies of patients referred for deep enteroscopy changes their management

Abstract

Introduction

Video capsule endoscopy (VCE) is widely used in the community for the diagnosis of small bowel disorders. However, deep enteroscopy is generally only performed at tertiary care centers. Thus, when lesions are found that require deep enteroscopy, patients are referred to a tertiary center, often in an open-access manner. Procedures are commonly scheduled, based on the referring physician’s interpretation of the capsule findings, but precise anatomic locations of lesions, route of insertion (i.e., antegrade vs. retrograde), and interpretation of capsule findings may be unspecified or incorrect.

At our institution, we reviewed capsule endoscopies prior to scheduling procedures to determine whether deep enteroscopy versus another endoscopic or nonendoscopic test is required, and determine the best route of insertion. Patients are referred in an open-access manner, although only those who will likely benefit from deep enteroscopy (as determined by the enteroscopist) are scheduled for the procedure. Our primary aim was to determine how often changes in clinical management are made for patients who are referred for deep enteroscopy when capsules are reviewed by the enteroscopist prior to scheduling the procedure.

Methods

Results

Outside VCE discs were available for 45 patients who were referred for deep enteroscopy. All patients were included in this analysis. The mean age of the patients was 61 years (18–79 years) and 23 (51%) patients were female (Table 1). Indications for VCE included obscure GI bleeding (37 patients), abnormal imaging (3 patients), abdominal pain (2 patients), Peutz-Jegher syndrome (2 patients), and weight loss (1 patient). The referring physician findings included polyps or masses (13 patients), angioectasia (13 patients), ulcers (9 patients), active bleeding (9 patients), nonspecific findings (8 patients), and normal (2 patients). Many patients had more than one finding. Thirty-eight (84.4%) patients were referred for antegrade single balloon enteroscopy (ASBE); 3 (6.7%) patients, for retrograde single balloon enteroscopy (RSBE); 2 (4.4%) patients, for push enteroscopy; and 2 (4.4%) patients, for a second opinion.

Capsule review led to disagreement in the findings of 13 (28.8%) patients (Table 2). Of these 13 patients with discordant findings, there were 11 overcalled lesions, five missed lesions, and three lesions with misidentified locations. Most patients had multiple discordant findings. Overcalled lesions included masses or polyps (5 patients; Figs. 1 and 2); ulcers (4 patients); and angioectasias (2 patients). Missed lesions included angioectasias (2 patients; Fig. 3); gastric antral vascular ectasia (GAVE)/portal hypertensive gastropathy and enteropathy (1 patient); and nonspecific enteropathy (1 patient).

Nine (20%) patients required a change in management (Table 3). Conservative management was recommended for two patients secondary to overcalled lesions. There were two additional patients who were also treated with conservative management because of overcalled lesions; however, these patients were referred for a second opinion and could not be counted as a change in management. Overcalled lesions included polyps, masses, or angioectasias that were actually folds, insignificant erythema, light reflex, or food debris. Two patients with multiple small bowel angioectasias were recommended to have conservative management because of multiple comorbidities and/or lack of the need for blood transfusion. One patient was deferred for enteroscopy in favor of cross-sectional imaging; magnetic resonance enterography revealed a carcinoid tumor and the patient subsequently underwent its surgical resection. Two patients underwent ASBE instead of push enteroscopy and one patient underwent RSBE instead of ASBE because of a misidentified lesion location. One patient underwent repeat colonoscopy instead of ASBE because of a lesion that was read as a small bowel polyp but was actually a colonic fold. In total, 37 patients underwent enteroscopy with a diagnostic yield of 48.8% and therapeutic intervention of 24.4%. There were no immediate or delayed adverse events.

Discussion

Capsule endoscopy is becoming increasingly common for the initial investigation of suspected small bowel pathology.2 The relative ease, low risk, and minimal training requirements have made it widely available in academic settings and in private practice settings. Abnormal findings often require deep enteroscopy, which is typically only available at tertiary care centers. Patients are often referred in an open-access manner, which is acceptable as long as there is a good indication. At our institution, we re-read capsules whenever possible, particularly if there is doubt about the indication.

In this study, we found that changes in management after rereading capsule endoscopies were because of lesion misdiagnosis, incorrect lesion location, or inappropriate route of insertion requested. Diagnostic errors may result from a lack of experience and training in capsule endoscopy, but even among experts there is interobserver variability in capsule interpretation.3–5 It has been previously suggested that a second reading by an experienced viewer may improve the diagnostic accuracy of VCE.6 An inaccurate lesion location may result from a lack of familiarity with the length of each small bowel segment or result from the use of gross time estimates rather than small bowel transit time, as discussed below. Improper recommended route of insertion is likely to be secondary to a lack of understanding of the depth limitations of antegrade and/or retrograde enteroscopy. Enteroscopists may have an advantage in determining lesion location/route of insertion and in recognizing small bowel pathology on VCE because they regularly visualize the correlating lesions during enteroscopy. Previous studies have not compared VCE interpretation between endoscopists who perform enteroscopy and those who do not.

Capsule endoscopy differs from other imaging modalities in that it can only give an estimate of where a lesion may be in the small bowel. To overcome this difficulty, the small bowel transit time was developed. By using the first duodenal image and first cecal image, each small bowel image is given an estimated percentage of small bowel covered out of the total small bowel transit. This technique has been validated for determining the route of insertion for enteroscopy. For example, lesions up to 36% through the small bowel can be reliably reached by ASBE.7 Most readers, however, still use gross time through the bowel, rather than the percent progress, to estimate a lesion’s anatomic location. This technique does not take into account gastric transit time or pauses in capsule movement because of stricture or lack of peristalsis. These errors may lead to an inaccurate estimation of lesion location and subsequent incorrect mode of enteroscopy. Disagreement in lesion location led to a change in management in two (4.4%) patients in this study.

There is a significant false positive rate with capsule endoscopy.3,8,9 This is particularly true with polyps and ulcers, although angioectasias and masses are also frequently overcalled.3,10–12 Criteria have been developed to help differentiate submucosal tumors from folds; however, the criteria remain a very challenging aspect of VCE.13,14 Our study was consistent with these observations in that the most common overcalled lesions were masses/polyps5 and ulcers.4 There were also three instances in which angioectasias were present that were missed or mistaken for erosions or ulcers. These patients underwent ASBE for a different indication than originally intended, but the overall management was not changed.

The limitations of this study include the small number of patients and readings were performed by a single enteroscopist at a single center. Our results nonetheless demonstrate that reviewing capsule endoscopies prior to enteroscopy results in a substantial number of changes in management. These changes are not insignificant in that they may lead to a different route of insertion, a different modality (e.g., cross-sectional imaging), or obviate the need for further work-up. If abstracted to a larger scale, re-reading could substantially decrease cost, resource utilization, and the performance of potentially unnecessary procedures. In conclusion, reviewing capsule endoscopies in patients referred for open-access enteroscopy changes management in a large proportion of patients. If not employed for all patients, this process should at least be performed when there is any doubt as to the indication for enteroscopy or route of insertion.

Article information

References