Introduction
Obesity represents a major health concern [1 ] leading to an increased risk of cardio-vascular, pulmonary, metabolic, and gallbladder disease and certain cancers, with a significant health care cost burden [2 ]
[3 ]. Bariatric surgery is the most effective treatment to date for reducing weight and maintaining the loss and resolving several comorbidities [4 ].
Preoperative diagnostic work-up of obese patients selected for bariatric surgery includes assessments by several health professionals, but the role of routine esophagogastroduodenoscopy (EGD) is still debated [5 ]
[6 ]
[7 ]
[8 ]. The rationale for performing EGD is to detect and treat lesions that could potentially affect the intervention being performed and which would result in complications or symptoms postoperatively [6 ]
[7 ]. Moreover, endoscopic findings may guide therapeutic decisions, such as the choice of intervention [8 ].
In spite of this, EGD may carry higher risks in obese patients, especially individuals with morbid obesity, because of the high rates of respiratory problems, such as sleep apnea, respiratory decompensation, and oxygen desaturation during sedation [4 ]
[5 ]
[6 ]. For these reasons, it is very important to correctly assess the real benefits of EGD and to determine whether they balance or outweigh the potential risks of the procedure in this group of patients.
Moreover, during the ongoing COVID-19 pandemic, the role of endoscopy is even more questionable, given the high risk of viral transmission related to the air droplets spread during the procedure [9 ]. Indeed, most international scientific societies have recommended avoiding non-urgent endoscopies; during the COVID-19 outbreak, a significant drop in the number of all kinds of endoscopies has been reported [10 ].
Recently, a new model of video-endoscopic capsule (VEC) has been introduced for the study of the upper gastrointestinal tract [11 ], presented as a comfortable, noninvasive, well-tolerated, and well-performing alternative to an upper endoscopy, which allows EGD to be performed on a select minority of patients who really need it for biopsies or treatment.
The aim of this study was to determine if a kind of new upper gastrointestinal VEC is an adequate diagnostic alternative to EGD in preoperative work-up of obese patients selected for bariatric surgery, particularly in the setting of COVID-19.
Patients and methods
Study design and population
This was a pilot, prospective, comparative, double-blind, noninferiority study carried out from January to November 2020. During that period, all patients selected for bariatric surgery at our university hospital were asked to undergo an upper gastrointestinal VEC evaluation and an EGD the next day. This sequence was chosen to avoid misinterpretation during VEC of mucosal traumas from an endoscope touching or sucking possible vascular lesions.
Images from each upper gastrointestinal VEC procedure were interpreted by two medical experts, separately and blinded to each other, to reach a higher level of agreement. For the same reason, the data from each EGD ware analyzed by two expert endoscopists, separately and blinded to each other: the first during the endoscopic of the procedure; the second one subsequently, by viewing the recorded file of the procedure.
Finally, the four reports (two from the upper gastrointestinal VEC and two from the EGD) were compared to evaluate the eventual noninferiority of the VEC compared with the EGD, in describing the endoscopic findings diagnosed.
Primary endpoint of the study was to define a noninferiority of upper gastrointestinal VEC compared to EGD in detecting upper gastrointestinal diseases or focal lesions in obese patients prior to bariatric surgery. The secondary endpoint was a comparison of the two procedures in terms of length of stay in the endoscopy room, total amount of drugs used, number of staff members involved, technical problems during the procedures; visual analog scale (VAS) discomfort gradation, and global cost of the procedures.
Inclusion criteria were age between 18 and 65 years, body mass index (BMI) ≥ 35 kg/m2 with obesity-related diseases or BMI ≥ 40, scheduled bariatric surgery intervention, and signing an informed consent.
Exclusion criteria were pregnancy, breastfeeding, use of alcohol/drugs, psychiatric disorders, and patients unwilling to undergo the procedures.
A full and informed consent was obtained from all participants before enrollment. At any time, patients could leave the protocol. All investigations complied with the principles of the Declaration of Helsinki (64th WMA General Assembly, Fortaleza, Brazil, October 2013).
Sample size calculation and statistical analysis
Assuming a diagnostic yield of 80 % and 50 % for EGD and VEC, respectively, as previously assessed [12 ]
[13 ], a sample size of 25 patients would be needed to demonstrate a statistically significant difference in diagnostic yield between the two modalities with 80 % power and 5 % two-sided significance [14 ]
[15 ].
Parametric continuous data (Shapiro-Wilk test, P = .05) are presented as means ± standard deviation and nonparametric continuous data (Shapiro-Wilk test, P < .05) as median and interquartile range (IQR). A rank-based nonparametric comparison was performed using the Kruskal-Wallis H test. Categorical data are presented as an absolute number percentage (%), and the McNemar test was used for comparison. Statistical significance was defined as P < .05 in this study.
The SPSS 20.0 software package for Windows (IBM, Chicago, Illinois, United States) and GraphPad Prism 3.0 for Windows (GraphPad Software, San Diego California, United States) were used for statistical analysis.
Endoscopic procedures
To optimize visualization of the digestive mucosa, 20 minutes before the upper gastrointestinal VEC and the EGD, all patients drank a liquid preparation (water 100 mL, acetylcysteine 6 mL, simethicone 5 mL), already validated to eliminate foam, bubbles, and mucus [16 ]
[17 ].
For both procedures, the stomach study was evaluated separately in the three segments: fundus, body, and antrum.
Endoscopic images were documented using, for both procedures, a predefined diagnostic score for reporting endoscopic findings listed in [Table 1 ].
Table 1
Endoscopic images, with respective scores, used to describe the findings of upper gastrointestinal VEC and EGD.
Score
Endoscopic findings
Details
0
Normal
1
Schatzki ring
Narrowing fold of the lower esophagus partially blocking the lumen
2
Esophagitis
According to the Los Angeles classification [18 ]
3
Suspected Barrett’s esophagus
Salmon-pink colored extension of mucosa growing into the esophagus above the esophageal gastric junction
4
Hiatal hernia
Anterograde picture of trans-cardiac prolapse of gastric folds; retroversion picture of the sub-cardia area in which the endoscope is not tightly gripped by the tissue or the lumen of the esophagus is open, often allowing the squamous epithelium to be viewed from below (respectively grade iii and grade iv according to the hill classification) [44 ]
5
Gastric biliary reflux
Presence of bile in the stomach
6
Esophageal varices
According to the Baveno criteria [19 ]
7
Gastric varices
According to the Baveno criteria [19 ]
8
Congestive gastropathy
Gastric mucosal mosaic like appearance, resembling snake-skin
9
Chronic gastropathy
Edema, linear erythema, erosions, hemorrhage
10
Neoplasm
Polypoid or nonpolypoid lesions according to Paris classification, endoluminal masses
11
Vascular anomaly
Telangiectasia, angiodysplasia
12
Erosion
Superficial mucosal defect, linear or rounded in shape, covered or not with fibrin
13
Ulcer
Break in the digestive wall, linear or rounded in shape, bleeding or not, covered or not with fibrin
Upper Gastrointestinal VEC
The positioning of VEC was performed in the endoscopy unit by a nurse experienced in digestive endoscopy and highly trained for capsule procedures. Immediately before swallowing the capsule, 20 mg of hyoscine butyl-bromide was given intramuscularly to reduce peristalsis and optimize esophageal and gastric views. Therefore, the patients swallowed, with a sip of water, the upper gastrointestinal VEC (PillCam upper gastrointestinal, Medtronics Ltd, Saint Peter’s, Dublin, Ireland). A portable data recorder (PillCam Recorder DR3, Medtronics Ltd, Saint Peter’s, Dublin, Ireland) permitted acquisition of images from a double camera system (one for each capsule tip) with a frame rate per second (FRS) from 18 to 35 pictures. A small monitor in the recorder allowed real-time viewing. Once the capsule was swallowed, patients were placed lying supine on the bed, with the trunk tilted 30 degrees for 5 minutes, to decrease the speed of the esophageal transit. Then, they were discharged and returned the next day for the EGD. At the end of the procedure the recorder was removed, and patient discharged. The data were downloaded on the RAPID workstation (Medtronic Ltd, Saint Peter’s, Dublin, Ireland) by RAPID Reader v. 9 software (Medtronic Ltd, Saint Peter’s, Dublin, Ireland) and then analyzed by an endoscopist skilled in capsule procedures.
EGD
All the EGDs were performed with a standard gastroscope (GIF HQ-190 Olympus, Shinjuku, Tokyo, Japan) by an endoscopist and a nurse both experienced in bariatric endoscopy, in standard sedation (midazolam 5–10 mg and hyoscine butyl-bromide 20 mg IV); when necessary, because of the psychomotor agitation or respiratory problems of the patients, an anesthesiologist was involved for the use of propofol and orotracheal intubation.
Evaluation of discomfort of VEC and EGD
To evaluate the general discomfort evoked during upper gastrointestinal VEC and EGD, a questionnaire was administered after each procedure. Patient discomfort was evaluated on a VAS. The VAS consisted of a 100-mm long horizontal straight line defined by anchors with verbal labels, the left end point (0 mm) indicating no discomfort and the right end point (100 mm) indicating overwhelming discomfort. We determined the VAS score (1 mm = 1) by measuring the distance from the left end point to the mark made by the patient.
Results
Twenty-seven individuals were initially enrolled in our study, but one withdrew consent; therefore, 26 subjects underwent upper gastrointestinal VEC plus EGD examination. The sample consisted of eight men (30.8 %) and 18 women (69.2 %); mean age was 37 ± 8.7 years, with a mean BMI of 46.88 ± 6.56 kg/m2 . Nine patients (34.6 %) had type 2 diabetes mellitus (T2DM). [Table 2 ] summarizes baseline demographics.
Table 2
Demographic and clinical characteristics of the study group.
Female/male, n
18/8
Age (y; mean ± SD)
37 ± 8,7
BMI (± SD)
46,88 (± 6,56)
Comorbidity, n (%)
9 (34,6)
4 (11,1)
Upper gastrointestinal VEC
Complete examination of the second duodenum (D2) was achieved in all patients. The mean (± SD) time of capsule transit in the esophagus, stomach, and duodenum was 29 seconds (± 98), 65 minutes (± 21), and 12 minutes (± 16), respectively. The transit time of the capsule through the esophagus, stomach, and duodenum was irrelevant to the aims of the study, as transit occurred while the patient was discharged. For this reason, the transit time was not calculated. Mean residence time in the endoscopic room was 7.27 minutes (± 1,43) for activation of capsule endoscopy.
The mean reading time for capsule video was 46 minutes (± 15) with standard mode. The long average reading time was due to the use of a new device, specific for upper gastrointestinal assessment, and the readers had experience with the ileum and colon capsule. This was the first study that evaluated bariatric patients with a new, specific device, and for this reason, the reading time was longer. No specific challenges were reported.
Pathologies detected by upper gastrointestinal VEC included 16 hiatal hernias, nine esophagitis, eight gastric erythemas, six gastric erosions, five vascular alterations, and two duodenitis; four examinations were normal ([Table 3 ]). The degree of satisfaction of the readers was complete.
Table 3
Number of upper gastrointestinal pathologies detected by VEC and EGD.
Pathology
Total number
Pathologies detected by both upper gastrointestinal VEC and EGD
Pathologies detected only by upper gastrointestinal VEC
Pathologies detected only by EGD
Hiatal hernia
18
16
2
Esophagitis
9
9
Gastric erythema
8
8
Gastric erosion
6
4
2
Gastric angioectasia
5
4
1
Gastric polyp
1
1
Hyperplastic polyp
1
1
Duodenitis
2
2
Normal appearance
5
4
1
VEC, video-endoscopic capsule; EGD, esophagogastroduodenoscopy
In our experience, there was a significant interoperator agreement in terms of visibility score both at all sites of the stomach (fundus, body, antrum) (r = 0.69–0.85, P < 0.001) and in all explored tracts (r = 0.75–0.89, P < 0.001).
EGD
There were no adverse events during any examinations; in one case orotracheal intubation was required. In nine patients, the endoscopic intubation was regular, in six patients it was difficult, and eight patients were agitated during the examination despite premedication. In three other patients, anesthetic assistance was requested.
Mean residence time in the endoscopic room was 38.8 minutes (± 5.7), of which, 13.4 minutes (± 1.33) was for endoscopy.
Pathologies detected by EGD included 18 hiatal hernias, nine esophagitis, eight gastric erythemas, four gastric erosions, four vascular alterations, one hyperplastic polyp, and two duodenitis; five examinations were normal ([Table 3 ]).
We found significant interoperator agreement in terms of visibility score at all sites in the stomach (r = 0.63–0.81, P < 0.001). Considering the global evaluation of all explored tracts, the correlations between the two endoscopists were even higher (r = 0.81–0.97, P < 0.001).
Diagnostic yield of upper gastrointestinal VEC vs EGD (primary endpoint)
Considering the primary endpoint, upper gastrointestinal VEC detected 47 upper gastrointestinal pathologies and EGD detected 47 ([Table 3 ]). There was no difference in detection of significant lesions.
EGD and VEC concomitantly identified hiatal hernia (n = 16), esophagitis (n = 9), gastric erythema (n = 8), and erosive duodenitis (n = 4).
VEC alone additionally identified two cases of gastric erosions and one of angioectasia. However, EGD alone identified two additional hiatal hernias and one hyperplastic polyp.
In detail, for the gastric evaluation, fundus, body, and antrum were separately assessed. No gastric fundal lesions were revealed by VEC or EGD; EGD missed erosions in two patients and one vascular lesion detected by VEC in the body. Moreover, EGD revealed one hyperplastic polyp of the body missed by VEC. No differences were noted for antrum evaluation with two techniques.
VEC had sensitivity, specificity, and positive and negative predictive values for identifying significant lesions of 91.3 % (95 % confidence interval [CI], 71.96 %-98.93 %), 83.33 % (95 % CI, 35.88 %-99.58 %), 98.01 % (95 % CI, 89.13 %-99.66 %), and 51.57 % (95 % CI, 21.77 %-80.76 %), respectively, compared with EGD as the standard criterion.
The accuracy was 90.51 % (95 % CI, 73.75 %-98.18 %) and the chi-square statistic was 0.1153 (P = 0.73).
Clinical outcomes of VEC vs EGD (secondary endpoint)
Based on the findings from VEC, all 26 patients were identified as potentially being safe to undergo surgery as well as EGD. No complications were seen with VEC or EGD. Mean residence time in the endoscopic room was 7.27 minutes (± 1,43) and 38.8 minutes (± 5,7) for upper gastrointestinal VEC and EGD, respectively (P < 0.05). All upper gastrointestinal VEC procedures were conducted with only one experienced nurse while EGDs required a mean of 3.2 staff members (1 nurse, 1 endoscopist, and 1 anesthetist in some described cases).
Furthermore, considering patient-reported discomfort, VAS for VEC was significantly lower compared to that of EGD (0.27 ± 0.7 vs 5 ± 1.98, respectively P < 0.001).
Finally, concerning global costs of the procedures, EGD was significantly cheaper than VEC (122.82 ± 4.01 vs 596 €, P < 0.001).
Discussion
The role of routine EGD in preoperative assessment of obese patients selected for bariatric surgery has changed over time, and today, it is still debated [5 ]
[6 ]
[7 ]
[8 ].
Some studies suggest that EGD can identify patients with asymptomatic findings that may change surgical planning [18 ]. Moreover, patients with symptoms of gastroesophageal reflux disease (GERD) or who chronically take antisecretory drugs should undergo EGD [21 ]. Although most patients with abnormalities in these studies were asymptomatic, endoscopic findings definitively modified the surgical approach only 1 % to 9 % of cases [8 ]
[18 ]
[19 ]
[20 ].
In 2005, guidelines from the European Association for Endoscopic Surgery (EAES), strongly recommended EGD for all bariatric procedures [7 ]. However, in the 2020 update of these guidelines [8 ], EAES recognized that endoscopy might be more appropriate only in selected patients with upper gastrointestinal symptoms, stating that the desirable effects of adhering to the recommendation probably outweighed the undesirable effects, even if there was a lack of confidence about these trade-offs.
Similarly, the American Society of Gastrointestinal Endoscopy and the American Society for Metabolic and Bariatric Surgery recommend that EGD be performed in patients scheduled for bariatric surgery only after a thorough discussion with the surgeon about the type of bariatric procedure and that individual patient’s symptoms [6 ]
[21 ]
[22 ].
In contrast, in Japan, because of the very high incidence of gastric cancer, atrophic gastritis (precancerous status), and Helicobacter pylori infection [23 ], routine EGD is considered necessary before bariatric surgery [24 ]. In a Japanese series [25 ], 66 % of preoperative EGDs revealed abnormal findings and, subsequently, perioperative management was changed in more than 10 % of patients.
The main reason for the more prudent attitude and individualization of indications for each patient with a trend toward preoperative EGD in Western countries is the poor respiratory performance of obese patients during upper endoscopy; macroglossia, sleep apnea, reduced oropharyngeal space by the endoscope, and impaired lung expansion often evolve into dangerous oxygen desaturation [25 ]
[26 ]. Moreover, because of their low compliance, morbidly obese patients frequently need higher doses of sedation managed by the anesthesiologist, which is a high technical risk, with increased time and costs for endoscopic procedures [23 ]
[24 ]
[26 ]
[27 ]
[28 ].
From the perspective of standardizing preoperative work-up, new upper gastrointestinal VEC could be a valid alternative as a first-line test for endoscopic assessment of obese patients prior to bariatric surgery, with EGD reserved only for cases in which surgery may be contraindicated or delayed after appropriate therapy.
The diagnostic advantages of the new upper gastrointestinal VEC already have been validated in several papers and its technical features make it suitable studying different upper digestive diseases. Among possible preoperative findings, those of particular interest in patients with morbid obesity selected for bariatric surgery are esophagitis, Barrett’s esophagus (BE), hiatal hernia, and esophagogastric varices [29 ]
[30 ].
Previous experience has shown that esophageal transit time was just 2 seconds and the gastroesophageal junction (GEJ) was visualized in only 24 % of patients using old-generation, single-headed, small-bowel capsules, with two FRS [31 ]. Obviously, this kind of capsule is not suitable for esophageal imaging [32 ].
A first-generation, twin-headed, dual-camera esophageal video capsule with an FRS of four pictures was developed specifically for visualization of the GEJ and showed good results in identification of esophagitis, but only moderate or barely acceptable sensitivity and specificity for identifying BE [33 ]
[34 ]
[35 ]
[36 ]. Subsequently, a second-generation esophageal capsule with a higher FRS of 14 pictures showed very high-quality esophageal and GEJ images, excellent visualization of the Z line, and good results for diagnosing erosive esophagitis and suspected BE [37 ]
[38 ]
[39 ]
[40 ].
The new upper gastrointestinal VEC system features innovative technology, operating at 35 FRS for the first 10 minutes of the procedure and 18 FRS for the last 80 minutes, with excellent performance in identifying esophagitis and suspicion of BE [31 ]
[39 ].
Identification of hiatal hernia often leads to modification of the surgical bariatric strategy, given the risk of postoperative development or worsening of preoperative symptoms [40 ].
The competence of the mechanical anti-reflux barrier, as an expression of a hiatal hernia, can be evaluated endoscopically in two ways. The most often used is measurement of the axial length between the hiatus and the GEJ. During EGD, the hiatus is represented by the diaphragmatic pinch and the GEJ is defined by the proximal margin of the gastric folds [41 ]. Given the physiologic dynamics in this area, it can be very difficult to measure the length of a hiatal hernia [42 ] and interobserver agreement about endoscopic measurement is poor even under ideal conditions [43 ].
The second way to assess a hiatal hernia is by grading the gastroesophageal flap valve using the Hill classification [44 ], in retroversion position. Many studies show an association between higher Hill grades with GERD [45 ]
[46 ], low levels of lower esophageal sphincter pressure, and increased prevalence of hiatal hernia [47 ]
[48 ]. The Hill classification is reproducible and provides useful information for evaluation of patients with suspected GERD who are undergoing endoscopy [45 ].
A paper comparing these two methods concluded that Hill classification is superior to measurement of the axial length for diagnosis of hiatal hernia [49 ]. From that perspective, upper gastrointestinal VEC (with its twin-headed, dual cameras, and high FRS) can generate pictures simulating the endoscopic intragastric retroversion visualization according to the Hill classification, with optimal reliability for diagnosing a hiatal hernia.
Because bariatric surgery in obese patients with cirrhosis should only be performed on those with compensated disease, after careful evaluation, by an experienced surgeon at a high-volume bariatric center, preoperative identification of esophagogastric varices represents an important goal [50 ]. A systematic review with meta-analysis confirms the feasibility, safety, and tolerance of capsule endoscopy in diagnosis and grading of esophageal varices, with a diagnostic accuracy of 92 % [51 ]. Another multicenter study showed high sensitivity and specificity for detection and grading of esophageal varices and detection of hypertension gastropathy [52 ].
Our experience confirms the data in the literature and underscores the reliability of upper gastrointestinal VEC for diagnosing esophagitis and hiatal hernia with high sensitivity and specificity. No cases suspected of BE or esophago-gastric/hypertension gastropathy varices occurred in our series. [Fig. 1 ] shows a comparison of upper gastrointestinal VEC images and those from EGD.
Fig. 1 Comparison of upper gastrointestinal VEC images with those for EGD.
All our statistical evaluations showed diagnostic noninferiority of upper gastrointestinal VEC compared with EGD and our results confirmed the advantages of VEC in terms of operative time (with quicker availability of the endoscopic room for other procedures), number of health workers involved [9 ]
[50 ], and total amount of needed drugs (saving time and money and reducing the rate of AEs to the drugs).
Upper gastrointestinal VEC was associated with a lower rate of technical problems during the procedure and also was well accepted by our cohort of patients, with a lower average score of “no discomfort” on the VAS, compared with EGD, and a greater willingness to repeat the procedure, if necessary. All these data are summarized in [Fig. 2 ].
Fig. 2 Results of secondary endpoints of the study.
Our study also underscores that patients undergoing upper gastrointestinal VEC entered and exited the endoscopy unit on their own. Conversely, with EGD, patients they were transported in an ambulance by a stretcher-bearer because of the premedication that was required. Therefore, that aspect also favors the upper gastrointestinal VEC.
The cost issue, in our experience, is the main disadvantage of the capsule, but we must make a few comments about it. First, upper gastrointestinal VEC is a new device, just recently introduced, with a price not yet corresponding to an established market. As shown in [Fig. 3 ], in some countries (Italy for example), EGD is cheaper than VEC. On the other hand, in other countries where bariatric surgery is widely performed, the prices are similar or EGD is more expensive than VEC [53 ]
[54 ]
[55 ]
[56 ]
[57 ]. At present, it seems difficult to unequivocal make pharmaco-economic determinations. Moreover, we must consider that in this particular patient population, use of upper gastrointestinal VEC offers undeniable clinical and organizational advantages, regardless of the costs.
Fig. 3 Comparison of costs for upper gastrointestinal VEC versus EGD in some Western countries (in €).
Finally, the COVID-19 pandemic has brought attention to another concern regarding preoperative upper endoscopy, which is airborne transmission of SARS-CoV-2 infection. Dispersion of air droplets is well known during EGD, and therefore, it is considered a high-risk procedure for healthcare workers [58 ]. Several scientific societies have released guidelines on how to reduce risks in the endoscopy room [9 ]
[59 ], but it is undeniable that upper gastrointestinal VEC carries lower risks than traditional EGD. In 2020, given the fear of catching COVID-19, many patients cancelled their endoscopy appointments, which resulted in significant and worrying delay in endoscopic diagnosis of gastrointestinal cancer [60 ].
Although our pilot study was launched before the COVID-19 pandemic, we found our experience particularly important for safe endoscopy in a COVID-19 setting.
Strengths and limitations
The strengths of our study are that both EGD and upper gastrointestinal VEC were blindly reviewed by two separate endoscopists and this was the first pilot study investigating the usefulness of upper gastrointestinal VEC in patients with morbid obesity undergoing bariatric surgery.
The main limitation of our study is that it was carried out in a single center with a relatively small sample. However, as has been noted previously, our sample size guarantees a statistical power of 80 %.
Conclusions
In conclusion, our report seems to confirm that upper gastrointestinal VEC, compared with EGD, shows diagnostic noninferiority in detecting significative lesions in preoperative work-up of obese patients prior to bariatric surgery. These results lead us to propose that upper gastrointestinal VEC could be the real first-line endoscopic diagnostic test in obese asymptomatic patients selected for bariatric surgery. From that perspective, EGD could be reserved only for patients suspected to have specific pathologies for whom surgery could be contraindicated or delayed after appropriate therapy, reducing the number of EGDs performed. This point is very important, particularly during the COVID-19 pandemic, because EGD is a high-risk contamination procedure [9 ]
[10 ]
[61 ]. A multicenter experience, in a larger sample of patients, is required to confirm our preliminary results.
