Routine esophagogastroduodenoscopy before upper endoscopic ultrasound for non-luminal indications: clinical value and findings from a large single-center experience

Introduction

Endoscopic ultrasound (EUS) has evolved over the past decades, transitioning from a purely diagnostic tool to a set of interventional procedures on intestinal and extraintestinal organs. In the United States alone, more than 500,000 EUS procedures are conducted annually, highlighting the critical role of EUS in the medical landscape [1].

Despite its widespread use, EUS procedures are limited by the oblique field of view of echoendoscopes, which can restrict complete luminal visualization, potentially missing significant lesions [2,3]. To mitigate this limitation, esophagogastroduodenoscopy (EGD) is sometimes performed before EUS, providing a comprehensive visual assessment of the upper gastrointestinal tract and improving evaluations in gastrointestinal diagnostics [4-6]. However, there is no consensus on the need for, or the value of, routine performance of EGD before EUS, largely because of the variability in practices across institutions and the lack of prospective data supporting this approach [2,7]. This study aimed to address this gap, by evaluating the proportion of clinically meaningful lesions detected when EGD is routinely performed before EUS. This study hypothesized that performing EGD prior to EUS would significantly increase the detection of clinically relevant lesions that would otherwise go undetected with EUS alone, thereby potentially altering patient management and improving diagnostic accuracy.

Results

Data were collected from the medical records of 2 endoscopists who conducted EUS procedures between January 2017 and December 2022. Initially, 863 patients were identified; however, after the exclusion criteria had been applied and those with missing data excluded, the final analysis included 626 patients who underwent EUS for non-luminal indications. Among those excluded were 10 patients who had EUS for evaluation of submucosal lesions. Additionally, of these 626 patients, a total of 568 had an EGD performed immediately prior to EUS.

The primary indications for EUS included pancreatic cysts (31.2%, n=177), pancreatic neoplasms (23.5%, n=133) and common bile duct dilation (17.7%, n=100). Acute pancreatitis accounted for 7.6% (n=43) of cases, while chronic pancreatitis and elevated liver enzymes were less frequent, representing 4.0% (n=23) and 5.1% (n=29), respectively. Mediastinal lesions and liver masses were the least common indications, each found in less than 1% of the patients (Table 1).

Table 1 Patient demographics and procedure indications

The most prevalent EGD findings were reflux esophagitis (32.7%, n=36), gastritis (17.3%, n=19), and Barrett’s esophagus (12.7%, n=14). Notably, duodenal ulcers and gastric ulcers were also detected, albeit less frequently (7.3% and 2.7%, respectively). On histopathological evaluation, benign squamous mucosa with reflux changes (24.7%, n=22) and intestinal metaplasia of the granular type (19.1%, n=17) were the most common findings. Changes in management or additional diagnostics were frequently necessitated by these findings, with tissue biopsy (54.3%, n=70) and medication initiation (36.6%, n=47) being the most common interventions (Table 2).

Table 2 Clinically significant findings and management

Among the 568 patients who underwent preliminary EGD, 95 (16.8%) had clinically significant findings. In addition, 94 patients (16.6%) had findings on EGD that could potentially influence the subsequent EUS examination. Only a small fraction of the patients (4.6%, n=26) had undergone an EGD within the last 6 months prior to the current examination. Of these, 6 patients were found to have clinically significant findings, including erosive esophagitis, Barrett’s esophagus, duodenal stenosis, large hiatal hernia >5 cm and esophageal candidiasis.

Discussion

Our study found that 16.5% of patients undergoing EGD prior to EUS had clinically significant findings, demonstrating that luminal examination should be considered for upper EUS exams in otherwise asymptomatic individuals. This rate is lower than the 29.7% and 22.1% reported by Kim et al and Sahakian et al, respectively, probably because of differences in patient populations, study designs and definitions of clinically significant findings [2,7]. For instance, our study excluded patients with dysphagia or upper gastrointestinal symptoms, while Kim et al excluded pregnant patients and those with significant comorbidities [7]. Additionally, our study included a broader range of findings, such as strictures and large hiatal hernias, which could impact the safety and feasibility of subsequent EUS procedures.

The most common EGD findings in our study were reflux esophagitis (32.7%), gastritis (17.3%) and Barrett’s esophagus (12.7%), consistent with prior studies. Baltz et al reported that 62% of patients had luminal abnormalities, with 9% having significant pathologies affecting management, including erosive esophagitis and Barrett’s esophagus [8]. Similarly, Kim et al found erosive esophagitis and Barrett’s esophagus in 10.9% and 6.3% of patients, respectively [7]. A recent study by Alkurdi et al demonstrated that routine EGD prior to EUS detects clinically significant luminal lesions in up to 38.78% of asymptomatic patients: the authors identified actionable pathologies (e.g., Barrett’s esophagus, peptic ulcers, stenosis) in a substantial proportion of patients, which might have remained undetected by oblique-viewing EUS alone [9]. These findings highlight the importance of thorough luminal evaluation, particularly in the esophagus and stomach, where lesions can be missed by non-forward-viewing endoscopes unless dedicated efforts to visualize the mucosa are undertaken [1,3].

Importantly, 16.6% of patients had EGD findings that could influence subsequent EUS procedures, such as strictures, large hiatal hernias or surgically altered anatomy. Sahakian et al reported similar findings, with 9.8% of patients having lesions impacting EUS [2]. A study by El-Dika et al demonstrated that upper endoscopy found luminal abnormalities (such as stenosis or diverticula) in 12% of patients referred for EUS examination of the pancreas [5]. Alkurdi et al reported that 6.93% of patients had findings detected by EGD (e.g., stenosis, upside-down stomach) that directly impacted the feasibility of endoscopic retrograde cholangiopancreatography or EUS and necessitated preprocedural dilatation in 2.4% of cases [9]. This underscores the clinical significance of these findings and the need to consider comprehensive luminal evaluation to ensure safe and effective EUS procedures. Sahakian et al also emphasized the importance of thorough diagnostic evaluations prior to complex procedures, noting that such evaluations can enhance patient safety by identifying conditions that might otherwise complicate subsequent interventions [2].

Management changes, such as medication initiation (36.6%), were common in our study, often prompted by diagnostic steps, such as tissue biopsy (54.3%), to further evaluate findings detected during EGD. These changes reflect the impact of early lesion detection on patient outcomes. Abu Ghanimeh et al reported management changes in 67% of asymptomatic patients undergoing EGD prior to EUS, while Sahakian et al noted changes in 15.7% of cases [2,10]. These findings suggest that EGD can influence clinical decisions, particularly in asymptomatic patients.

The statistical significance of our findings (P<0.005) supports the diagnostic utility of EGD in identifying conditions that may be missed by EUS alone. Ashby et al also found that performing forward-viewing endoscopy during EUS detected additional lesions in 20.5% of cases (P=0.00025) [1]. In contrast, Kim et al found no significant difference in miss rates between EGD and linear-array EUS for clinically meaningful lesions (P=0.39), suggesting that both modalities are equally effective [7].

Differences in detection rates across studies may stem from variations in patient demographics, endoscopic techniques, and operator expertise. For example, the higher rate of clinically significant findings in the study by Kim et al could be attributed to its multicenter design, which included a broader patient population [7]. Additionally, the prevalence of underlying risk factors for gastrointestinal conditions, such as alcohol or tobacco use, may influence detection rates, as suggested by Abu Ghanimeh et al [10].

Early detection of lesions, such as gastric ulcers or Barrett’s esophagus, can prevent complications like bleeding or cancer progression. Moreover, the low percentage of patients who had undergone EGD in the 6 months prior to our study suggests that many patients may go for extended periods without thorough luminal examination, increasing the risk of missed diagnoses. Consideration of routine EGD could enhance patient safety by identifying lesions that might otherwise complicate EUS procedures. Notably, our study demonstrated an additional diagnostic yield even among patients who had undergone an EGD within the prior 6 months. This suggests that repeating EGD at the time of EUS may still uncover clinically significant findings. However, future studies should evaluate whether a defined timeframe (e.g., 6-12 months) after a prior negative EGD might be sufficient to obviate the need for repeat examination

Our study had some limitations. High-quality esophageal/gastric/duodenal evaluation can be performed with echoendoscopes, and it is unknown how this compares to dedicated EGD. The study’s retrospective, single-center design may have introduced selection bias, limiting its generalizability. Additionally, the reliance on inpatient records may have underestimated the full impact of EGD findings, as outpatient follow-up data, including biopsy results and subsequent interventions, were not captured. Future prospective, multicenter studies are needed to validate our findings and explore the clinical benefits of routine EGD before EUS.

In conclusion, our study supports the value of a routine EGD before EUS in detecting clinically significant luminal lesions and influencing patient management. While EGD is not mandatory prior to EUS, it should be considered, as it can uncover relevant findings that may alter clinical decisions. Further research is needed to standardize practices and improve diagnostic accuracy in gastroenterology.