Pan-intestinal capsule endoscopy as first-line procedure in patients with suspected mid or lower gastrointestinal bleeding

Abstract

Background Pan-intestinal capsule endoscopy (PCE) evaluates the small bowel and colon noninvasively. This study evaluated diagnostic accuracy and safety of PCE vs. colonoscopy as first-line examination in suspected mid–lower gastrointestinal bleeding (MLGIB).

Methods In this prospective, single-center, single-blinded cohort study, consecutive patients with suspected MLGIB underwent PCE followed by same-day colonoscopy. Diagnostic accuracy for potentially hemorrhagic lesions (PHLs; combined diagnosis by PCE + colonoscopy) and incidence of adverse events were assessed.

Results 100 patients were included (median age 70 [range 18–92] years; 65% female). PHLs were diagnosed in 46 patients, including small-bowel and/or colon angioectasias in 32. PCE correctly identified 54 individuals without PHLs, and 95.7% (44/46) of those with PHLs vs. 50.0% (23/46) for colonoscopy (P<0.01). PHLs were detected by PCE alone in 65.2% (30/46), both examinations in 28.3% (13/46), and colonoscopy alone in 6.5% (3/46). PHLs were diagnosed at the ileocolonic region in 28% of patients, with PCE diagnosing 25/28 cases (89.3%) and colonoscopy diagnosing 23/28 (82.1%; P=0.13). Interventional procedures were performed at colonoscopy in 13/81 patients with iron-deficiency anemia (16.0%) vs. 6/19 patients with overt bleeding (31.6%; P<0.01). No significant adverse events occurred with PCE vs. 2% with colonoscopy.

Conclusions In patients with MLGIB, PCE avoided further invasive procedures in >50% of patients. PCE was safe and more effective than colonoscopy in identifying PHL both in the small bowel and colon. These results support the potential use of PCE as first-line examination in patients with suspected MLGIB.

Introduction

Capsule endoscopy using double-headed capsules permits a noninvasive, safe, and comfortable full endoscopic assessment of the small bowel and colon in one single procedure (pan-intestinal capsule endoscopy, PCE) [11] [22]. Current standard of care involves patients with suspected mid or lower gastrointestinal bleeding (MLGIB; i.e. those with iron-deficiency anemia [IDA] and/or overt bleeding after nondiagnostic esophagogastroduodenoscopy [EGD]) initially undergoing invasive colonoscopy, followed by capsule endoscopy with a single-headed capsule to evaluate the small bowel if the colonoscopy is negative. Recently, the use of PCE has been hypothesized as a first-line diagnostic examination in this setting [33] [44], guiding the subsequent approach. Depending on the presence, type, and location of potentially hemorrhagic lesions (PHLs) detected by PCE, patients may subsequently require a conventional colonoscopy, device-assisted enteroscopy, second look EGD, or no further endoscopic procedures. This innovative concept, reserving invasive colonoscopy or enteroscopy as a second step only for those cases in which the noninvasive diagnostic PCE detected lesions requiring biopsy and/or endoscopic treatment in the colon and/or small bowel, is regarded as potentially more effective and safer than the current standard practice.

We aimed to evaluate the diagnostic accuracy and safety of PCE vs. the current standard of care (conventional colonoscopy) as a first-line test to investigate patients with IDA and/or overt bleeding after a nondiagnostic EGD (suspected MLGIB).

Methods

Study design and patients

We conducted a prospective, single-center, single-blinded study from July 2020 to September 2022, including consecutive patients with suspected MLGIB (IDA; hemoglobin <12.0 g/dL in female patients or <13.0 g/dL in male patients, and/or overt bleeding as melena and/or hematochezia), after nondiagnostic EGD. Patients underwent PCE using a double-headed capsule, followed by same-day conventional colonoscopy, under anesthesiologist-assisted deep sedation and without additional bowel preparation, performed by an independent gastroenterologist who was not involved in reading the PCE (see Fig. 1s in the online-only Supplementary material).

Exclusion criteria

Exclusion criteria were as follows: lack of informed consent; known or suspected intestinal stricture; pregnancy or breastfeeding; suspected lesions in the index EGD possibly related to IDA or digestive bleeding; patients unable to walk for at least short periods, and/or with neurological and/or psychiatric conditions potentially leading to protocol deviations; patients with other presumed etiology for IDA (e.g. vaginal bleeding); allergy or contraindications to any of the drugs or products used in the study.

Technical procedures and reporting

All PCE videos were read by one single experienced investigator (B.R.), who was blinded to the conventional colonoscopy findings and reports. The description of the PHL used the standard terminology [55] for vascular [66] and inflammatory [77] lesions in capsule endoscopy. If the PCE identified any PHL in the upper gastrointestinal (GI) region that was confirmed at the subsequent second look EGD, patients were excluded from the study. The quality of bowel cleansing in PCE was classified using the CC-CLEAR (Colon Capsule CLEansing Assessment and Report) scale [88] for the colon, and the Small Bowel CLEansing Assessment and Report (SB-CLEAR) for the small bowel [99]. We also assessed completeness (capsule expelled or hemorrhoidal pedicles reached before battery exhaustion).

Colonoscopies were performed by one of two experienced gastroenterologists (T.C.G., M.J.M.), who were blinded and not involved in the PCE reading. The quality of colonoscopy preparation was assessed using the Boston Bowel Preparation Scale [1010]. During colonoscopy, all required procedures were performed according to standard clinical practice, including biopsies, hemostatic procedures, or polypectomies as indicated. Ileal intubation was not mandatory and was performed according to individual clinical evaluation and technical feasibility. All colonoscopy videos were fully recorded.

Study end points and analysis of outcomes

All endoscopic findings potentially responsible for bleeding detected by any of the examinations were assessed and compared for diagnostic efficacy analysis. The main outcome was the diagnostic accuracy of both modalities (PCE vs. colonoscopy) for the detection of PHL, such as vascular lesions, ulcers, or tumors. The gold standard was defined by a consensus panel (B.R., T.C.G., M.J.M.), based on the assessment of the results of both examinations. For each examination, PCE and colonoscopy findings were compared, and, in case of disagreement, they were re-discussed until a consensus was reached, with the intervention of an independent reader (J.C.) who was experienced in both techniques and not involved in the primary reading of any of the examinations. As a secondary outcome, we evaluated safety (incidence of adverse events), defined as capsule retention, bleeding, perforation, or cardiopulmonary complications.

Statistics

Assuming a 70% estimated prevalence of PHL in the small bowel and/or colon in patients with IDA and/or overt bleeding (mainly melena) after first nondiagnostic EGD, and estimating a sensitivity of 90% and specificity of 80% for the endoscopic procedures, we determined, for confidence interval of 95%, significance level of 0.05, and 80% statistical power, that 68 patients were needed to meet the study purposes. As the dropout rate could not be anticipated, especially regarding the possible nonadherence to the intensive bowel preparation regimen and the risk of incomplete examinations, we decided to enlarge the sample to 100 consecutive patients.

Categorical variables are described using absolute frequencies and percentages, and continuous variables are described using median and range. Analysis of categorical variables used the chi-squared test, McNemar test, or Fisher’s exact test as appropriate.

Ethics

The study was approved by the Academic Center and the Ethic Committee of Hospital da Senhora da Oliveira, Guimarães, Portugal.

Results

Characterization of study population

We included 100 consecutive patients with suspected MLGIB who presented with IDA (81.0%) or overt bleeding such as melena and/or hematochezia (19.0%). Patients had a mean age of 66.5 years (median age 70 [range 18–92] years), and 65% were female. Mean hemoglobin level was 8.4 g/dL (median 8.5 [range 4.1–12.8] g/dL). Patients with overt bleeding presented with lower hemoglobin levels than patients with IDA (89.5% vs. 66.7% of patients with levels <10 g/dL; P=0.04). Hospital admission was required in 40.0% of cases (28.4% of patients with IDA vs. 89.5% of patients with overt bleeding; P <0.001). Of patients with overt bleeding, 74.0% underwent PCE and colonoscopy within the first 2 weeks after the bleeding episode, with 16.0% undergoing these procedures within the first 72 hours. There were more female patients than male patients in the IDA subgroup (70.4% vs. 29.6%), and more male patients than female patients with overt bleeding (57.9% vs. 42.1%; P=0.02). No cases of overt bleeding occurred in patients under the age of 40 years. Approximately one-third (34.0%) of patients were receiving antiplatelet medication (32.0% of patients with IDA and 42.0% of patients with overt bleeding; P=0.28), and approximately one-quarter of patients (26.0%) were receiving anticoagulant drugs (27.0% of patients with IDA and 21.0% of patients with overt bleeding; P=0.41). Nonsteroidal anti-inflammatory drugs, including aspirin, were used by 37.0% of patients (36.0% of patients with IDA and 42.0% of patients with overt bleeding; P=0.40). [Table 1Table 1] summarizes the baseline characteristics of the population.

Table 1 Table 1 Characterization of study population and procedural details.

	Overall (n=100)	IDA (n=81)	Overt bleeding (n=19)
IDA, iron deficiency anemia; PCE, pan-intestinal capsule endoscopy; N/A, not applicable.
Age, years
Mean	66.5	65.8	69.5
Median (range)	70 (18–92)	70 (18–87)	75 (42–92)
Male sex, n (%)	35 (35.0)	24 (29.6)	11 (57.9)
Clinical presentation, n (%)
Occult (IDA)	81 (81.0)	81 (100)	N/A
Melena	12 (12.0)	N/A	12 (63.2)
Hematochezia	7 (7.0)	N/A	7 (36.8)
Hospital admission, n (%)	40 (40.0)	23 (28.4)	17 (89.5)
Timing (PCE + colonoscopy) if overt bleeding, n (%)
<72 hours	N/A	N/A	3 (15.8)
72 hours – 14 days	N/A	N/A	11 (57.9)
>14 days	N/A	N/A	5 (26.3)
Hemoglobin, g/dL
Mean	8.4	8.6	7.3
Median (range)	8.5 (4.1–12.8)	9 (4.1–12.8)	7.1 (5.1–11.6)
Lowest hemoglobin, n (%)
<7 g/dL	32 (32.0)	23 (28.4)	9 (47.4)
7–10 g/dL	39 (39.0)	31 (38.3)	8 (42.1)
>10 g/dL	29 (29.0)	27 (33.3)	2 (10.5)
Procedural details, n (%)
PCE complete	76 (76.0)	61 (75.3)	15 (78.9)
Colonoscopy complete	95 (95.0)	79 (97.5)	16 (84.2)
Retrograde ileoscopy during colonoscopy	43 (43.0)	35 (43.2)	8 (42.1)
Adequate bowel preparation at PCE	72 (72.0)	63 (77.8)	9 (47.4)
Adequate bowel preparation at colonoscopy	85 (85.0)	73 (90.1)	12 (63.2)

Procedural details

PCE was complete in 76.0% of cases (75.3% of patients with IDA and 78.9% of patients with overt bleeding; P=0.50) compared with a 95.0% completion rate for colonoscopy (97.5% of patients with IDA and 84.2% of patients with overt bleeding; P=0.046) (PCE vs. colonoscopy P <0.001). Retrograde ileoscopy was performed, according to the judgment of the colonoscopist, in 43.0% of cases. Colon cleansing was adequate (CC-CLEAR score ≥6 and no segmental scores <2) in 72.0% of PCE examinations (77.8% of patients with IDA and 47.4% of patients with overt bleeding; P=0.015). In colonoscopies, adequate cleansing (Boston Bowel Preparation Scale score ≥6, with no segmental score <2) was recorded for 85.0% (90.1% of patients with IDA and 63.2% of patients with overt bleeding; P=0.02) (PCE vs. colonoscopy P=0.01) ([Table 1Table 1]). The completion rate of PCE was not associated with inpatient status (87.5% vs. 68.3% for outpatients; P=0.02), but was lower for elderly patients at a cutoff of 80 years old (57.9% vs. 80.2% for patients aged <80 years; P=0.04). The quality of colon cleansing at PCE was not significantly affected by advanced age (≥80 year 84.2% vs. <80 years 85.2%; P=0.12) or inpatient status (inpatients 77.5% vs. outpatient 90%; P=0.23). The small bowel was adequately clean (SB-CLEAR ≥6, with no segmental score <2) in 96.0% of patients.

Analysis of outcomes

Combining the results of PCE and conventional colonoscopy, PHLs were diagnosed in 46% of patients, including angioectasias in more than two-thirds (32/46, 69.6%) of cases. Angioectasias were located in the colon (14%), small bowel (11%), or both (7%). Other types of PHLs detected in the small bowel and/or colon included ulcerated lesions and large polyps or tumors ([Fig. 1Fig. 1]). Overall, small-bowel PHLs were detected in 25% and colonic PHLs in 28% of the entire cohort. In 54% of patients, no PHLs were detected by either of the procedures. A summary of all the PHLs detected by PCE and/or colonoscopy is presented in [Table 2Table 2].

Fig. 1 Fig. 1 Types of potentially hemorrhagic lesions detected by pan-intestinal capsule endoscopy in the small bowel (upper row) and colon (lower row). a Ulcerated stricture. b Large polyp. c Tumor. d Bleeding angioectasia. e Angioectasia. f Large polyp. g,h Angioectasia. i Tumor. j Ulcer with bleeding vessel.

Table 2 Table 2 Main endoscopic findings: potentially hemorrhagic lesions detected by the combination of both examinations (pan-intestinal capsule endoscopy plus colonoscopy).

	Total (n=100)	IDA (n=81)	Overt bleeding (n=19)
IDA, iron deficiency anemia; PHL, potentially hemorrhagic lesion. 1Dieulafoy’s lesion; blue rubber bleb nevus.
Findings, n (%)
PHLs	46 (46.0)	37 (45.7)	9 (47.4)
Angioectasias	32 (32.0)	25 (30.9)	7 (36.8)
Colon	14 (14.0)	11 (13.6)	3 (15.8)
Small bowel	11 (11.0)	10 (12.3)	1 (5.3)
Small bowel + colon	7 (7.0)	4 (4.9)	3 (15.8)
Ulcers/erosions	7 (7.0)	7 (8.6)	0
Small-bowel	5 (5.0)	5 (6.2)	0
Colon	2 (2.0)	2 (2.5)	0
Polyps/tumors	5 (5.0)	4 (4.9)	1 (5.3)
Small-bowel	2 (2.0)	2 (2.5)	0
Colon	3 (3.0)	2 (2.5)	1 (5.3)
Colon vascular lesions1	2 (2.0)	1 (1.2)	1 (5.3)
None	54 (54.0)	44 (54.3)	10 (52.6)

PCE detected PHLs in 44% of the entire cohort, corresponding to 95.7% (44/46) of all PHLs according to the consensus panel. Colonoscopy detected PHLs in 23% of the entire cohort, corresponding to 50.0% (23/46) of all patients with PHLs (P <0.01) ([Fig. 2Fig. 2]). Moreover, considering the ability to detect all PHLs in patients with multiple PHLs, a fully correct diagnosis was achieved by PCE in 43/46 patients (93.5%; 95%CI 82.5%–97.8%) compared with 16/46 (34.8%; 95%CI 22.7%–49.2%) for colonoscopy (P <0.01). Lesions responsible for bleeding were detected by PCE alone in 30/46 positive cases (65.2%), by both examinations in 13/46 (28.3%), and by colonoscopy alone in 3/46 (6.5%). Fig. 2s illustrates some examples of the same lesions detected by both PCE and conventional colonoscopy.

Fig. 2 Fig. 2 Proportions of potentially hemorrhagic lesions detected by pan-intestinal capsule endoscopy and colonoscopy in patients with suspected mid–lower gastrointestinal bleeding. IDA, iron deficiency anemia; PCE, pan-intestinal capsule endoscopy.

Among patients with IDA, PCE detected PHLs in 35/81 patients (43.2%), corresponding to 94.6% (35/37) of all IDA patients with PHLs, whereas colonoscopy detected PHLs in 16/81 patients (19.8%), corresponding to 43.2% (16/37) of all IDA patients with PHLs (P <0.01). Among patients with overt bleeding, PCE detected PHLs in 9/19 (47.4%), corresponding to 100% (9/9) of all overt bleeding patients with PHLs, whereas colonoscopy detected PHLs in 7/19 patients (36.8%), corresponding to 77.8% (7/9) of all overt bleeding patients with PHLs (P=0.02) ([Fig. 2Fig. 2]).

PCE had a sensitivity of 95.7% (44/46), specificity of 100% (54/54), negative predictive value (NPV) of 96.4% (54/56), and positive predictive value (PPV) of 100% (44/44). Colonoscopy had a sensitivity of 50.0% (23/46), specificity of 100% (54/54), NPV of 70.1% (54/77), and PPV of 100% (23/23). [Table 3Table 3] further describes the diagnostic accuracy of PCE vs. colonoscopy in the subgroups of patients with IDA and overt bleeding. Absolute numbers are presented in Table 1s.

Table 3 Table 3 Diagnostic accuracy of pan-intestinal capsule endoscopy versus colonoscopy for the detection of potentially hemorrhagic lesions.

Patients	Total (n=100)		IDA (n=81)		Overt bleeding (n=19)
IDA, iron deficiency anemia; PCE, pan-intestinal capsule endoscopy; PHL, potentially hemorrhagic lesions; NPV, negative predictive value; PPV, positive predictive value.
PHLs detected	All findings (n=46)	Colonic findings (n=28)	All findings (n=37)	Colonic findings (n=20)	All findings (n=9)	Colonic (n=8)
PCE, %
Sensitivity	95.7	89.3	94.6	85.0	100	100
Specificity	100	100	100	100	100	100
NPV	96.4	96.0	95.7	95.3	100	100
PPV	100	100	100	100	100	100
Colonoscopy, %
Sensitivity	50.0	82.1	43.2	85.0	77.8	75.0
Specificity	100	100	100	100	100	100
NPV	70.1	93.5	67.7	95.3	83.3	84.6
PPV	100	100	100	100	100	100

Less than one-third (28.0%) of patients had PHLs in the ileocolonic region, including 20/81 patients (24.7%) with IDA and 8/19 patients (42.1%) with overt bleeding (P=0.03); in these cases, PCE made the diagnosis in 25/28 cases (89.3%) and colonoscopy diagnosed 23/28 (82.1%) (P=0.13) ([Fig. 2Fig. 2]). Additional procedures such as argon plasma coagulation with or without clips (n=14), biopsies (n=3), or polypectomy (n=2) were performed at the index colonoscopy in 13/81 patients with IDA (16.0%) vs. 6/19 patients with overt bleeding (31.6%; P <0.01). Moreover, incidental polyps unrelated to the clinical condition and with negligible bleeding potential were detected and removed by standard polypectomy technique in 16 patients. Other reported colonic findings unrelated to bleeding included diverticula (28 patients) and subepithelial lesions (6 patients).

Regarding safety, no significant adverse events occurred with PCE compared with two cases (2.0%) with colonoscopy (one case of hypoventilation with respiratory depression and one case of persistent bradycardia, both of which were unresponsive to standard supportive measures and required the procedure to be aborted) (P=0.30).

For patients with no lesions detected by any of the procedures (54.0% of the entire cohort), after a mean follow-up of 18.8 (SD 6.6) months (range 8–34 months), the etiology of IDA was attributed to other clinical entities: autoimmune disease, namely systemic lupus erythematosus or vasculitis (n=7); upper GI causes such as subtotal gastrectomy, autoimmune or atrophic Helicobacter pylori-associated gastritis (n=6); liver cirrhosis (n=5); gynecological pathology (n=4); non-GI neoplasia (n=3); chronic kidney disease (n=1). A total of 28 patients remained with an unclear cause of IDA; a large proportion of these patients (20/28, 71.4%) were receiving antiplatelet and/or anticoagulation therapy, mainly due to complex cardiac, cerebrovascular, or peripheral vascular conditions.

Discussion

In our cohort, PHLs were diagnosed in less than half of patients presenting with IDA or overt bleeding, and were mainly small-bowel and/or colon angioectasias. Interestingly, approximately half of patients had entirely normal examinations (no PHLs) and this occurred in both IDA and overt bleeding subgroups. In almost two-thirds of cases, the PHLs were only detectable by PCE. The incremental yield of PCE was observed in both groups of patients, although when considering only the lesions at the ileocolonic region, the diagnostic yield was not significantly different from that of colonoscopy (89.3% vs. 82.1%; P=0.13). We acknowledge that, as we were comparing two procedures that have different characteristics (not entirely investigating the same bowel segments), different outcomes would be expected from the beginning. This is why we considered the study outcome for colonoscopy as the presence of lesions in the colon, and for PCE the presence of lesions in the colon and/or small intestine (i.e. the maximum information that each test could provide). With this methodology, we aimed to compare two different strategies for approaching a specific clinical problem: what can we diagnose with a capsule-first strategy compared with the current standard of care?

Less than one-third of patients with suspected MLGIB had colorectal PHLs. Patients presenting with IDA had an even lower prevalence of PHLs located at the ileocolonic region, compared with patients with overt bleeding (approximately 25% vs. 40%), and nearly half the need for interventional hemostatic procedures at colonoscopy (approximately 16% vs. 32%, respectively). In addition, more than half of patients in our cohort did not have any PHLs either in the small bowel or the colon, and therefore would not benefit from any further invasive procedures after a negative PCE. This means that most patients, according to the current standard of care, are often being subjected to unnecessary invasive colonoscopies, which in our series were associated with a 2% risk of significant adverse events. Furthermore, we demonstrated that even those patients with PHLs at colonoscopy often had other synchronous lesions (mainly angioectasias) in the small bowel (one-third [7/21] of patients with colonic angioectasias had synchronous small-bowel angioectasias), so capsule endoscopy is still required to establish adequate management.

The use of PCE as the first diagnostic examination after nondiagnostic EGD in patients with suspected bleeding has been previously addressed. In 2021, Mussetto et al. evaluated PCE in 128 patients presenting with melena and negative EGD [33]. PCE identified the bleeding site in 83% of patients, which led to small-bowel therapeutic interventions in near half the patients, avoiding standard colonoscopy. Another study conducted by Carretero et al. included 100 consecutive patients with GI bleeding undergoing PCE, which detected small-bowel lesions in 68% and colonic findings in 81% of patients [44]. No further endoscopic procedures were needed in nearly 65% of the patients with negative gastroscopy.

This “PCE-first” management strategy should be followed by: a) conventional colonoscopy (if lesions accessible to colonoscopy require endoscopic hemostasis, biopsies, and/or follow-up of potentially self-limited lesions); b) device-assisted enteroscopy if lesions are detected in the jejunum and/or ileum; c) second-look EGD if the capsule detects upper GI lesions that could have been missed at the index EGD [1111]; d) no additional procedures, in cases of normal examination or diagnosis of nonsignificant or self-limited conditions, given the high NPV of capsule endoscopy, previously demonstrated in the setting of small-bowel bleeding [1212]. [Fig. 3Fig. 3] illustrates a proposed new algorithm for the approach of patients with suspected MLGIB based on this concept, which is supported by our study results. This new concept may position PCE in the center of a new paradigm for the approach of patients with digestive bleeding disorders. Moreover, the ongoing advances in the field of magnetically controlled capsule gastroscopy may, in the future, permit a full pan-endoscopic noninvasive examination that also includes the esophagus and the stomach, possibly also avoiding the need for EGD in selected populations, if the encouraging published preliminary results are supported by further studies [1313] [1414] [1515] [1616].

Fig. 3 Fig. 3 Proposed algorithm for the approach of patients with suspected mid–lower gastrointestinal bleeding. IDA, iron deficiency anemia; GI, gastrointestinal; EGD, esophagogastroduodenoscopy; PCE, pan-intestinal capsule endoscopy; PHL, potentially hemorrhagic lesion.

A few limitations of our study are to be acknowledged. First, as it was a single-center study, an external validation of our observations in a multicenter setting is warranted. Second, we included patients with two different types of clinical presentation, namely IDA and overt bleeding; to minimize possible selection bias we performed a subgroup analysis for the two groups, and also performed a subanalysis considering only the lesions located at the ileocolonic region, with the purpose of evaluating the diagnostic accuracy of PCE vs. colonoscopy when specifically looking at the same accessible segments. Third, we observed a significant rate of inconclusive PCE due to poor cleansing and/or incomplete examinations. In a recently published systematic review [1717], the overall meta-analytical adequate cleansing rate was 72.5%, and the completion rate was 83%. Sodium phosphate is still currently recommended as the preferred booster for capsule endoscopy and PCE. Although security alerts have been issued regarding the use of sodium phosphate for bowel cleansing before conventional colonoscopy, its safety profile when used as a booster for the colon capsule (where only half the dose is administered, in two separate doses and with an additional amount of water) has been consistently demonstrated in large series and recently published meta-analyses [1717] [1818]. In addition, in a study that evaluated 67 fragile patients at higher risk for complications, mainly due to cardiac or renal disease, who underwent PCE using sodium phosphate as a booster, this protocol was safe and did not have a negative impact on kidney function [44]. We observed four cases of incomplete examinations due to ulcerated ileocolonic anastomosis (n=2), blue rubber bleb nevus syndrome (n=1), and small-bowel ulcerated stricture (n=1); the remaining (20%) incomplete examinations were due to “functional” delayed passage of the capsule. This may ultimately be related to the study protocol itself, as PCE was interrupted at the end of the day per protocol in order to perform conventional colonoscopy rather than waiting until the capsule battery life was fully exhausted. Other possible contributing factors could be our patient’s characteristics, such as advanced age [44] [1919] [2020] [2121]. Interestingly, inpatient status was not associated with a higher rate of incomplete examinations in our series. In those patients with incomplete PCE examinations, no lesions were present in the left colon at the subsequent conventional colonoscopy, which is consistent with the higher representation of angioectasias, commonly found in the right colon, as the dominant type of PHL in our cohort. The fact that PCE and colonoscopy were performed on the same day may also be perceived as a limitation, as it is impractical in daily routine. However, for real-life settings, we propose that PCE is performed initially, and only those patients with positive findings are selected for an elective invasive procedure as indicated, to be scheduled as soon as possible.

We acknowledge that wide implementation of this PCE-first strategy may be limited by numerous factors such as the available resources, possible logistic constraints, and the need to expand training for a larger pool of available colon capsule readers. Regarding the cost-effectiveness of a PCE-first approach, although we should stress that this study was not formally designed to evaluate such an outcome, a few conceptual observations may be considered. For our cohort of 100 patients, following the current standard of care, all of them would have undergone conventional colonoscopy. As only 23% of all patients had colorectal PHLs detected by conventional colonoscopy (and this value is only 16% of all patients if detection of all PHLs in patients with multiple PHLs is considered), up to 84% of patients would still require a capsule endoscopy (using a small-bowel capsule, with a cost that is similar or only marginally lower than PCE), after the index conventional colonoscopy. In other words, we would only have saved 16 capsules following the current standard of care, when compared with a PCE-first approach. Conversely, with a PCE-first approach, conventional colonoscopy would only be justified in less than one-third of patients (28%), being those with colorectal PHLs requiring additional procedures, namely hemostasis, polypectomy, and/or biopsies of lesions previously identified by PCE. Therefore, based on our observations, we could infer that in a cohort of 100 patients with suspected MLGIB, a PCE-first strategy would cost 16 more capsules but save 72 invasive colonoscopies. However, we would also have to take into consideration the relevant issues of incomplete and/or poorly cleansed PCEs. In this case, patients would still require conventional colonoscopy after PCE, even if no colonic PHL had been detected. In our cohort, considering all patients with colonic PHLs detected by PCE plus those with no PHL detected but having incomplete and/or poorly cleansed PCEs, an invasive colonoscopy would still be needed in 57 patients. Therefore, in a real-life clinical setting that considers the suboptimal efficacy of the currently available protocols for PCE bowel preparation, it seems more realistic to expect a 43% instead of a 72% decrease in the demand for invasive colonoscopies after a PCE-first approach in patients with suspected MLGIB. Another aspect, recently addressed in a position paper from the British Society of Gastroenterology on the topic of green endoscopy, is the potential role of capsule endoscopy as a more sustainable alternative to conventional diagnostic endoscopy [2222]. However, PCE requires further evaluation regarding cost-effectiveness and its environmental impact versus the alternative endoscopic procedures [2323].

To conclude, PCE was safe and more effective than conventional colonoscopy in identifying lesions with bleeding potential both in the small bowel and the colon, and therefore should be considered as a possible first-line examination in patients with suspected MLGIB, guiding subsequent management.

Conflict of Interest

B. Rosa has received consultancy fees from Medtronic, unrelated to the current work. T. Cúrdia Gonçalves, M.J. Moreira, F. Dias de Castro, B. Sousa-Pinto, M. Dinis-Ribeiro, and J. Cotter declare that they have no conflict of interest.

Acknowledgement

This study was supported by the Medtronic External Research Program, which granted the double-headed capsules used for PCE procedures.

• References

• 1 Eliakim R, Spada C, Lapidus A. et al. Evaluation of a new pan-enteric video capsule endoscopy system in patients with suspected or established inflammatory bowel disease – feasibility study. Endosc Int Open 2018; 6: E1235-E1246 DOI: 10.1055/a-0677-170. (PMID: 30302381)
• 2 D’Haens G, Lowenberg M, Samaan MA. et al. Safety and feasibility of using the second-generation pillcam colon capsule to assess active colonic Crohn’s disease. Clin Gastroenterol Hepatol 2015; 13: 1480-1486.e1483
• 3 Mussetto A, Arena R, Fuccio L. et al. A new panenteric capsule endoscopy-based strategy in patients with melena and a negative upper gastrointestinal endoscopy: a prospective feasibility study. Eur J Gastroenterol Hepatol 2021; 33: 686-690
• 4 Carretero C, Prieto de Frias C, Angos R. et al. Pan-enteric capsule for bleeding high-risk patients. Can we limit endoscopies?. Rev Esp Enferm Dig 2021; 113: 580-584
• 5 Delvaux M, Friedman S, Keuchel M. et al. Structured terminology for capsule endoscopy: results of retrospective testing and validation in 766 small-bowel investigations. Endoscopy 2005; 37: 945-950 DOI: 10.1055/s-2005-870266. (PMID: 16189766)
• 6 Leenhardt R, Li C, Koulaouzidis A. et al. Nomenclature and semantic description of vascular lesions in small bowel capsule endoscopy: an international Delphi consensus statement. Endosc Int Open 2019; 7: E372-E379 DOI: 10.1055/a-0761-9742. (PMID: 30834297)
• 7 Leenhardt R, Buisson A, Bourreille A. et al. Nomenclature and semantic descriptions of ulcerative and inflammatory lesions seen in Crohn’s disease in small bowel capsule endoscopy: an international Delphi consensus statement. United European Gastroenterol J 2020; 8: 99-107
• 8 de Sousa Magalhaes R, Arieira C, Boal Carvalho P. et al. Colon Capsule CLEansing Assessment and Report (CC-CLEAR): a new approach for evaluation of the quality of bowel preparation in capsule colonoscopy. Gastrointest Endosc 2021; 93: 212-223
• 9 Macedo Silva V, Lima Capela T, Freitas M. et al. Small Bowel CLEansing Assessment and Report (SB-CLEAR): standardizing bowel preparation report in capsule endoscopy. J Gastroenterol Hepatol 2023; 38: 747-751
• 10 Lai EJ, Calderwood AH, Doros G. et al. The Boston Bowel Preparation Scale: a valid and reliable instrument for colonoscopy-oriented research. Gastrointest Endosc 2009; 69: 620-625 DOI: 10.1016/j.gie.2008.05.057. (PMID: 19136102)
• 11 Juanmartinena Fernandez JF, Fernandez-Urien I Sainz, Zabalza Ollo B. et al. Gastroduodenal lesions detected during small bowel capsule endoscopy: incidence, diagnostic and therapeutic impact. Rev Esp Enferm Dig 2018; 110: 102-108
• 12 Pennazio M, Rondonotti E, Despott EJ. et al. Small-bowel capsule endoscopy and device-assisted enteroscopy for diagnosis and treatment of small-bowel disorders: European Society of Gastrointestinal Endoscopy (ESGE) Guideline – Update 2022. Endoscopy 2023; 55: 58-95 DOI: 10.1055/a-1973-3796. (PMID: 36423618)
• 13 Jiang B, Pan J, Qian YY. et al. Clinical guideline on magnetically controlled capsule gastroscopy (2021 edition). J Dig Dis 2023; 24: 70-84 DOI: 10.1111/1751-2980.13173. (PMID: 37220999)
• 14 Jiang X, Pan J, Li ZS. et al. Standardized examination procedure of magnetically controlled capsule endoscopy. VideoGIE 2019; 4: 239-243 DOI: 10.1016/j.vgie.2019.03.003. (PMID: 31194028)
• 15 Szalai M, Helle K, Lovasz BD. et al. First prospective European study for the feasibility and safety of magnetically controlled capsule endoscopy in gastric mucosal abnormalities. World J Gastroenterol 2022; 28: 2227-2242 DOI: 10.3748/wjg.v28.i20.2227. (PMID: 35721886)
• 16 Liu YW, Wang YC, Zhu JH. et al. Magnetically controlled capsule endoscopy in one-time gastro-small intestinal joint examination: a two-centre experience. BMC Gastroenterol 2022; 22: 222 DOI: 10.1186/s12876-022-02302-0. (PMID: 35509022)
• 17 Rosa B, Donato H, Curdia Goncalves T. et al. What is the optimal bowel preparation for capsule colonoscopy and pan-intestinal capsule endoscopy? A systematic review and meta-analysis. Dig Dis Sci 2023; 68: 4418-4431
• 18 Bjoersum-Meyer T, Skonieczna-Zydecka K, Cortegoso Valdivia P. et al. Efficacy of bowel preparation regimens for colon capsule endoscopy: a systematic review and meta-analysis. Endosc Int Open 2021; 9: E1658-E1673
• 19 Gimeno-Garcia AZ, Gonzalez-Suarez B, Ganzo ZA. et al. Predictive factors for inadequate bowel cleansing in colon capsule endoscopy. Gastroenterol Hepatol 2022; 45: 605-613 DOI: 10.1016/j.gastrohep.2022.01.003. (PMID: 35065169)
• 20 de Sousa Magalhaes R, Boal Carvalho P, Rosa B. et al. the prediction of inadequate colon capsule cleansing: a cohort selection guided by CC-CLEAR. GE Port J Gastroenterol 2022; 29: 311-321
• 21 Deane C, Walker C, Ryan B. et al. High diagnostic yield despite a lower completion rates for inpatient versus outpatient colon and pan-intestinal capsule endoscopy: a nested case-control study. BMC Gastroenterol 2023; 23: 61 DOI: 10.1186/s12876-022-02561-x. (PMID: 36894909)
• 22 Sebastian S, Dhar A, Baddeley R. et al. Green endoscopy: British Society of Gastroenterology (BSG), Joint Accreditation Group (JAG) and Centre for Sustainable Health (CSH) joint consensus on practical measures for environmental sustainability in endoscopy. Gut 2023; 72: 12-26 DOI: 10.1136/gutjnl-2022-328460. (PMID: 36229172)
• 23 Lacroute J, Marcantoni J, Petitot S. et al. The carbon footprint of ambulatory gastrointestinal endoscopy. Endoscopy 2023; 55: 918-926 DOI: 10.1055/a-2088-4062. (PMID: 37156511)

Publication History

Received: 05 November 2023

Accepted after revision: 16 February 2024

Accepted Manuscript online:
16 February 2024

Article published online:
14 March 2024