Low-volume bowel preparation provides safe and effective pan-enteric capsule endoscopy in suspected or established Crohn’s disease

Abstract

BACKGROUND

Effective pan-enteric capsule endoscopy (PCE) visualization requires thorough bowel preparation. We hypothesized that a novel low-volume protocol (LVP) provides comparable safety and effectiveness to the standard high-volume protocol (HVP).

AIM

To evaluate the safety and effectiveness of a novel low-volume vs standard high-volume preparation protocol for PCE.

METHODS

We conducted an observational study including all prospectively documented PCEs at our tertiary center (2018-2021), comparing preparation adequacy and completion rates between patients receiving LVP vs HVP. Preparation adequacy was evaluated by two blinded expert readers (Cohen’s kappa, correlation tests). Finally, an individual participant data meta-analysis combining our cohort with three published cohorts assessed PCE outcomes with HVP vs LVP. A P value < 0.05 was considered statistically significant in all tests.

RESULTS

Our analysis included 67 patients: 12 (17.9%) underwent PCE with HVP, and 55 (82.1%) with LVP. No post-PCE complications were reported. PCE completion was achieved by 89.1% of the LVP group and 75.0% of the HVP group. Adequate small bowel and colonic preparation rates were comparable between the HVP and LVP groups (90.9% vs 89.1%, and 80.0% vs 90.0%, respectively). In our individual patient data meta-analysis, the combined HVP (n = 257) and the LVP (n = 55) were comparable regarding completion rates (87.2% vs 89.1%) and adequate small bowel preparation rates (90.0% vs 89.1%); however, LVP had higher rates of adequate colonic preparation vs combined HVP (80.0% vs 62.7%, respectively, P = 0.017).

CONCLUSION

LVP is a safe and effective alternative to HVP with similar completion rates, better colonic preparation, and lower volume, making it preferable for patients with suspected or diagnosed Crohn’s disease.

Key Words: Pan-enteric capsule endoscopy; Bowel preparation; Crohn’s disease; Low-volume protocol; Completion rate

Core Tip: This study evaluates a novel low-volume, sulfate-free bowel preparation protocol for pan-enteric capsule endoscopy compared to the standard high-volume protocol. Despite the substantially lower volume, low-volume protocol (LVP) achieved comparable completion rates and adequate colonic cleansing. A pooled analysis across cohorts supports these findings and suggests improved colonic visualization with LVP over high-volume protocol. LVP may therefore represent a safe, effective, and potentially more patient-friendly alternative for pan-enteric capsule endoscopy preparation, particularly in the evaluation of Crohn’s disease.

INTRODUCTION

Video capsule endoscopy (VCE) has revolutionized gastrointestinal (GI) imaging since its introduction in 2000[1]. Initially launched as a single-camera, non-steerable device for small bowel (SB) evaluation, it has evolved into a sophisticated first-line diagnostic tool for a variety of SB disorders, including obscure GI bleeding, iron-deficiency anemia, and suspected inflammatory bowel disease[2,3]. VCE provides a non-invasive, radiation-free means of assessing mucosal pathology without the need for sedation[4]. Technological advancements, including higher-resolution imaging, adaptive frame rates, and software innovations that reduce reading time, have further enhanced its utility[5]. The introduction of patency capsules has also improved safety by minimizing retention risk in patients with suspected strictures[6-9]. More recently, the development of pan-enteric capsules has enabled the evaluation of both the small and large intestine in a single examination, expanding its role in evaluating inflammatory bowel disease[10].

Crohn’s disease (CD) is a chronic inflammatory bowel disease that can involve any segment of the GI tract; however, it most commonly involves the terminal ileum and proximal colon[11,12]. The disease often follows a progressive course, with up to half of patients developing strictures, fistulas, or other complications within 20 years of diagnosis[13,14]. Although severe complications such as perforation or SB adenocarcinoma are uncommon, the cumulative burden of disease underscores the need for comprehensive and repeated evaluation of the entire GI tract. Solitary SB involvement, particularly of the terminal ileum, occurs in about 30%-40% of patients with CD, whereas combined SB and colonic involvement is observed in a similar proportion[14,15]. Moreover, CD is characterized by discontinuous lesions (“skip lesions”), making full SB evaluation challenging and frequently requiring multiple diagnostic modalities[16,17]. In this context, capsule endoscopy has emerged as a valuable tool for both diagnosis and monitoring of CD. Clinical guidelines already recognize capsule endoscopy as superior to barium studies and computed tomography enterography (CTE) in suspected CD, due to its higher diagnostic yield[18,19]. The pan-enteric capsule was specifically designed for CD activity assessment, enabling evaluation of mucosal healing or inflammation in both the small and large bowel in a single test[10]. Considering follow-up colonoscopies in CD rarely require biopsies, particularly from the SB, this purely observational endoscopic approach is appropriate[20].

As with all endoscopic procedures, however, adequate bowel preparation is essential for diagnostic accuracy. While SB evaluation can often be achieved with dietary modification and fasting, colonic visualization requires the use of purgatives and laxatives[21]. Standard pan-enteric capsule protocols typically employ high-volume, sulfate-based preparations, which can impair tolerability and adherence[22]. To address this, we aimed to evaluate the effectiveness and safety of a simplified, sulfate-free, low-volume protocol (LVP) compared with the standard high-volume protocol (HVP) in patients with suspected or established CD.

MATERIALS AND METHODS

Study design and population

This study had two parts: The first part was a retrospective analysis of our prospectively maintained database of VCEs performed at a tertiary medical center, and the second part was a comparison of the primary outcomes between our cohort and a combination of historical cohorts of patients undergoing VCEs in other centers [individual participant data (IPD) meta-analysis]. In the first part, a retrospective analysis of all consecutive pan-enteric capsule endoscopies (PCEs) documented prospectively and performed in our tertiary referral VCE service between 1 October, 2018 and 30 June, 2021 was performed. This study included only adult patients with CD or suspected CD, as the PCE (PillCam Crohn’s) is specifically indicated for this population. Patients with one of the following criteria were excluded: Age < 18 years, inability to swallow the capsule, significant technical malfunction of the PCE recording, or significant lack of data. The pan-enteric capsule (PillcamCrohn^®, Given Imaging, Yokneam, Israel) and the RAPID 9 software (PillCam™ software v9, Given Imaging, Ltd., Yokneam, Israel) were used in all procedures. Patients were referred to PCE by the treating gastroenterologist due to one of three possible indications: Exacerbation of established CD, evaluation of established CD in clinical remission, and suspected CD with negative ileo-colonoscopy in the 5 years prior to the VCE. In order to reduce potential capsule retention and in accordance with the European Crohn’s and Colitis Organization guideline[23], all patients with established CD were administered the Agile^© Patency Capsule (Given Imaging, Yokneam, Israel) prior to the PCE procedure. Patients with suspected CD and no obstructive symptoms were offered the patency capsule as well, but could choose to decline this preliminary procedure.

VCE preparation protocols

PCE requires special preparation before and after capsule ingestion, and the cohort was divided into two groups according to these protocols elaborated in Table 1. During the implementation period (October 2018 to December 2018), the HVP protocol was administered pre- and post-PCE according to the manufacturer’s recommendations as published in the literature[24] (Table 1). The first PCE protocol published included (in brief): 4 L of polyethylene glycol (PEG) before capsule ingestion and 2 doses of Suprep^® (Braintree Laboratories, Inc., Braintree, MA, United States) or pico-sulfate solution with 2 L of water after ingestion[22,25]. The new LVP was developed at our center and administered starting from January 1, 2019, until June 30, 2021 (Table 1). The LVP protocol (in brief) is based on 2 L PEG with ascorbic acid and lactulose instead of sulfate-based solutions. The study was approved by the local Institutional Review Board (No. 0407-21-TLV).

Table 1 Details of the low and high-volume preparation protocols according to date and time of pan-enteric capsule endoscopy.

		Low-volume protocol	Standard high-volume protocol
Day	Time	Instructions
1 day	All day	Clear liquid diet	Clear liquid diet
	19:00-21:00	1000 mL PEG 3350 with ascorbic acid (MoviPrep©)	2000 mL PEG 3350 (Meroken©)
PCE day	07:00-09:00	1000 mL PEG 3350 with ascorbic acid (MoviPrep©)	2000 mL PEG 3350 (Meroken©)
	10:00	PCE ingestion	PCE ingestion
	11:00	Optional: 10 mg metoclopramide if alert “0’’	Optional: 10 mg metoclopramide if alert “0’’
	Small bowel detection - alert “1”	150 mL lactulose + 1500 mL of water	180 mL oral sulfate solution mixed in 300 mL of water + 1000 mL of water
	3 hours later - alert “2”	90 mL lactulose + 500 mL of water	90 mL oral sulfate solution mixed in 150 mL of water + 1000 mL of water
	2 hours later - alert “3”	10 mg bisacodyl suppository	10 mg bisacodyl suppository
	2 hours later - alert “4”	A light meal	A light meal
Total liquid volume		4240 mL	6720 mL

PCE: Pan-enteric capsule endoscopy; PEG: Polyethylene glycol.

IPD meta-analysis

In order to increase the sample size, we created an IPD meta-analysis based on three published papers[24-27], in which PillcamCrohn^® (Given Imaging, Yokneam, Israel) was used with a preparation protocol identical to the protocol of our HVP group (Table 1). These three studies were joined to our HVP group to create the “combined HVP group”. In this methodology, only categorical variables can be calculated.

Primary and secondary outcomes

Our primary outcomes for both parts of the study were: Bowel preparation effectiveness and safety of VCE.

Bowel preparation effectiveness: The quality of bowel preparation was established by retrospective review of all PCE videos, by two blinded expert PCE readers, scoring three segments in the SB (proximal, middle, and distal) and a general assessment of the colon. Moreover, at the end of every video, the reader answered a yes/no question: “Would you recommend that this patient repeat the PCE due to inadequate preparation?”. All videos were anonymized and coded by the software. The cleansing score was assessed using a 4-point Likert scale[26] as detailed in Table 2. “Poor”/“fair” were defined as “inadequate preparation,” and “good”/“excellent” were defined as “adequate preparation”. A general assessment of the SB was considered adequate if all three SB segments were scored as adequate.

Table 2 Four-point Likert scale for bowel preparation assessment.

Score	Description	Definition of adequacy
1	Excellent	Adequate
2	Good	Adequate
3	Fair	Inadequate
4	Poor	Inadequate

VCE safety: Safety was evaluated by the lack of “true capsule retention”, which was defined as the inability to extract the capsule > 2 weeks after capsule ingestion. A complete PCE study was defined by successful expulsion of the capsule and documentation of the toilet. Early complications (< 24 hours) were assessed by questioning the patient regarding dehydration, bowel obstruction, electrolyte disturbances, emergency room (ER) visit, or hospital admissions, on the follow-up visit the next morning (24 hours after ingestion). Late complications (24 hours to 30 days after ingestion) included ER visits or hospital admission documented in the patients’ medical records.

Data collection

Data were collected from PCE referral letters, medical admissions, and final reports, all documented prospectively in patients’ medical files. CD characteristics, including the Montreal classification[28] and disease duration, were described and analyzed only for patients with established CD and not suspected CD, while typical medications were described for the entire cohort, since they may have been prescribed for other inflammatory conditions (i.e., rheumatoid arthritis). Clinical and demographic details and relevant symptoms were documented according to the patient’s reports on the day of the examination or recovered from the referral letter, no more than 6 months prior to PCE. Laboratory results no more than 6 months before the PCE date were documented. C-reactive protein > 0.5 mg/dL was defined as elevated; hemoglobin level < 13.8 g/dL (males) or < 12.1 g/dL (females) was defined as Anemia, and fecal calprotectin level > 150 μg/mg as elevated. The PCE disease activity score was calculated by using the “Eliakim score” based on the original scores given in the original report of the PCE[26]. The severity was graded on a scale of 1-3, with scores for the most common and severe lesions. Disease extent was expressed as a percentage. Separate scores were assigned to the colon and three parts of the SB and then summed for the total Pan-enteric Crohn’s Capsule Score. In the second part of the study, data regarding CD status, sex, PCE completion rates, and adequacy of SB and colon preparation were collected from the studies by Leighton et al[24], Bruining et al[27], and Eliakim et al[26] as published.

Statistical analysis

All statistical analyses were performed using SPSS version 27 (SPSS Inc., Chicago, IL, United States). Categorical variables were presented as a percentage (%), and to examine the difference in prevalence between the groups, a χ² test was used. Continuous variables were presented as medians (minimum-maximum), and differences were examined by the Mann-Whitney test. In the second part of the study, continuous variables could not be analyzed. Data collected from the three previous PCE studies[24-27] included only dichotomous variables and were entered into the database according to the proportions published. The χ² test was used to examine the difference in proportions between the “combined HVP” to the LVP group. To determine the degree of agreement between two independent readers, Cohen’s kappa test and correlation tests were used, where r ≤ 0.3 was considered a weak correlation and r ≥ 0.7 was considered a strong correlation. A P value < 0.05 was used as a threshold value for determining statistical significance in all tests. The statistical methods of this study were reviewed and approved by Leshno M and Fliss-Isakov N.

RESULTS

Study population

During the study period (October 2018 to June 2021), 69 PCEs were performed in our center. Of these examinations, 2 PCEs were excluded: One due to age < 18 years and one due to insufficient data regarding CD characteristics. Subsequently, a total of 67 patients were included in the study with a median age of 35.1 years (range: 19.4-90.2 years). Twenty-eight patients (41.8%) were males. Twelve patients (17.9%) underwent PCE with standard HVP, and 55 (82.1%) patients with LVP (Figure 1). Demographic and clinical characteristics of the study groups were comparable (Table 3). All 67 patients included in the study had CD or were referred for suspected CD. Specifically, 17 patients (25.4%) were referred with suspected CD, while 50 patients had an established CD diagnosis: 27 during an exacerbation (40.3%) and 23 in clinical remission (34.3%). The reason for VCE referral was equally distributed between the groups (Figure 2A). The characteristics of CD, as well as clinical and laboratory characteristics, were similar between groups and are elaborated in Table 4.

Figure 1 Study flow diagram. Between October 2018 and June 2021, 69 patients underwent pan-enteric capsule endoscopy. Two were excluded, leaving 67 adults with suspected or established Crohn’s disease. Initially, 12 patients were examined using a high-volume preparation (October 2018 to December 2018). Thereafter, the protocol was replaced by a low-volume preparation, applied in 55 patients (January 2019 to June 2021). PCE: Pan-enteric capsule endoscopy; CD: Crohn’s disease; LVP: Low-volume protocol; HVP: High-volume protocol.

Figure 2 Comparison of study outcomes. A-C: Comparison of study outcomes between the low-volume protocol group and the high-volume protocol group. Distribution of the three possible indications for pan-enteric capsule endoscopy referral within each group (A); small bowel transit time and colonic transit time among the low-volume protocol group (LVP) and the high-volume protocol (HVP) group (B); comparable adequate preparation rates between the LVP and the HVP group; agreement between 2 blinded expert readers (C); D and E: Comparison of study outcomes between the LVP and the combined HVP group. Location of the capsule at the end of the study (maximal study duration was approximately 12 hours) (D); adequate preparation rates between the LVP and the combined HVP group (E). P < 0.05 vs combined high-volume protocol. LVP: Low-volume protocol; HVP: High-volume protocol; CD: Crohn’s disease; CTT: Colonic transit time; SBTT: Small bowel transit time.

Table 3 Clinical and demographic characteristics, n (%).

	LVP (n = 55)	HVP (n = 12)	P value
Age (years)1	35.7 (19.3-90.1)	34.2 (23-58.4)	0.423
Male sex	22 (40.0)	6 (50.0)	0.538
Any GI surgery	5 (9.1)	1 (8.3)	1.000
BMI (kg/m2)1	24.8 (18.2-37)	23.6 (15.9-24.9)	0.235
Smoking
Never	21 (56.8)	10 (90.9)	0.095
Current	7 (18.9)	1 (9.1)
Past	9 (24.3)	0 (0.0)
Medications associated with bowel ulcers
NSAIDS	2 (3.6)	0 (0.0)	0.502
Micropirin	3 (5.5)	0 (0.0)	0.408

¹Continues variables are presented as median (range).

LVP: Low-volume protocol; HVP: High-volume protocol; GI: Gastrointestinal; BMI: Body mass index; NSAIDS: Non-steroidal anti-inflammatory drugs.

Table 4 Crohn’s disease characteristics, n (%).

	LVP	HVP	P value
Disease duration, years, median (range)1	6.3 (0.03-17.6)	4.6 (0.3-16.4)	0.492
Age at diagnosis, years, median (range)1	25 (16-86)	22 (17-58)	0.422
Montreal classification1
Age at diagnosis
> 16	2 (4.8)	0 (0.0)	0.728
17-40	30 (73.2)	6 (85.7)
> 40	9 (22.0)	1 (14.3)
Location
L1	24 (57.1)	4 (50.0)	0.608
L2	3 (7.1)	0 (0.0)
L3	15 (35.8)	4 (50.0)
L4	0 (0.0)	0 (0.0)
Behavior
Inflammatory	32 (76.2)	6 (75.0)	0.660
Structuring	7 (16.7)	2 (25.0)
Penetrate	3 (7.1)	0 (0.0)
Peri-anal disease	10 (23.8)	0 (0.0)	0.184
Symptoms2
Abdominal pain	26 (47.3)	8 (66.7)	0.340
Diarrhea	20 (36.4)	6 (50.0)	0.515
Rectal bleeding	5 (9.1)	1 (8.3)	1.000
Extra-intestinal manifestation	12 (21.8)	4 (33.3)	0.444
Asymptomatic	19 (34.5)	4 (33.3)	1.000
Current treatment2
No treatment	18 (32.7)	3 (25.0)	0.740
5-ASA	8 (14.5)	2 (16.7)	1.000
Thiopurines	6 (10.9)	1 (8.3)	1.000
Methotrexate	0 (0.0)	1 (8.3)	0.179
Biologic	21 (38.2)	3 (25.0)	0.515
Steroids	2 (3.6)	1 (8.3)	0.452
Elevated CRP2	27 (49.0)	9 (75.0)	0.152
Elevated fecal calprotectin2	19 (34.5)	4 (33.3)	1.000
Anemia2	10 (18.1)	3 (25.0)	0.633

¹Data was analyzed only among patients with established Crohn’s disease (low-volume protocol, n = 42; high-volume protocol, n = 8).

²Data was analyzed for the entire cohort (low-volume protocol, n = 55; high-volume protocol, n = 12).

LVP: Low-volume protocol; HVP: High-volume protocol; 5-ASA: 5-aminosalicylic acid; CRP: C-reactive protein.

PCE study

Pre-PCE patency tests were performed in 60/67 patients (89.6%). Out of the seven patients who did not perform the patency tests, six patients were referred due to suspected CD and declined the test, and one patient with established CD was in clinical and laboratory remission and preferred not to do the patency test. Most patients (58 patients, 86.6%) succeeded in expelling the capsule from the body and completed the test within 12 hours. The completion rate among the LVP group was 89.1% while the completion rate among the HVP group was 75.0%; however, this difference was not statistically significant (P = 0.345). One patient (HVP group) did not have cecal documentation due to SB strictures not detected by CTE and, as a result, had delayed SB transit time. However, he reported capsule expulsion on the next day and did not require any further investigations, such as an abdominal X-ray. Additionally, post-PCE complications such as true capsule retention, ER visit, or hospital admission were not recorded in any patient in either group.

Median SB transit time for patients in LVP and HVP groups was similar [138.78 minutes (range: 44.27-514.48 minutes) vs 119.91 minutes (range: 64.60-214.12 minutes), respectively, P = 0.331]. Median colonic transit time was comparable as well [LVP: 219.5 minutes (range: 13.72-653.7 minutes) vs HVP: 210.13 minutes (range: 8.32-385.77 minutes), P = 0.743] (Figure 2B). The median values of the endoscopic scores for total pan-enteric Crohn’s capsule score were 2.0 (range: 0.0-53.0) in the LVP group and 0.0 (range: 0.0-8.0) in the HVP group, P = 0.081. Of patients with established CD, three were defined as L1 (only ileal involvement) pre-PCE and changed to L3 (SB and colon) post-PCE. In two patients who were defined as L2 (colon only) pre-PCE, the definition post-VCE was changed to L3. Out of the 17 patients referred for suspected CD, none had findings compatible with CD.

Cleansing scores and adequacy of preparation

There was a statistically significant positive correlation between the cleansing scores (Likert 1-4) of two blinded expert readers in all 4 domains (proximal SB: r = 0.341, P = 0.005; middle SB: r = 0.581, P < 0.001, distal SB: r = 0.666, P < 0.001; colon: r = 0.685, P < 0.001). As for adequacy of preparation, the agreement between the readers was strong regarding the general assessment of the SB (Kappa coefficient: 0.704, P < 0.001) and medium regarding the general assessment of the colon (Kappa coefficient: 0.575, P < 0.001). Strong agreement was also noted on the recommendation of the patient to repeat the test (Kappa coefficient: 0.792, P < 0.001) (Figure 2C).

The rates of adequate SB preparation were similar between the groups (LVP: 89.1% vs HVP: 90.9%, P = 1.000). The inadequacy in all the remaining PCEs was based on inadequate preparation, specifically in the distal SB segment. The rates of adequate colonic preparation were similar between the groups (LVP: 80.0% vs HVP: 90.1%, P = 0.282) (Figure 2C). However, most of the inadequate preparations in the LVP group were scored as “fair” (9/55, 16.4%), and only 2 patients were scored as “poor” (3.6%). Referral for re-examination was recommended in 2 patients, one from each group: The patient in the LVP group due to delayed gastric transit time, and the patient in the HVP group due to inadequate “poor” preparation.

Individual participant data meta-analysis

An IPD meta-analysis was performed by gathering data on PCE procedures previously reported by other groups[24]. The characteristics of these studies are elaborated in Table 5. We created the “combined HVP” group of 257 patients (245 patients from the historical cohorts and 12 HVP patients from our study) and compared them to our LVP group (55 patients). Patient characteristics are described in Table 6. The PCE reached the cecum in 100% of the patients in the LVP group and in 95.9% of patients in the “combined HVP” group (P = 0.217). A complete PCE study was achieved by 89.1% of the LVP group and 87.2% in the “combined HVP” group (P = 0.824) (Figure 2D). Adequate SB preparation was equally prevalent between the LVP and “combined HVP” groups (89.1% vs 90.0%, respectively, P = 0.807) while adequate colonic preparation was more prevalent among the LVP group (80.0% vs 62.7%, respectively, P = 0.017) (Figure 2E).

Table 5 Characteristics of three historical cohorts using the high-volume preparation protocol.

Ref.	CD status	n	Sex	VCE completion rate	Location at the end of the incomplete study (n)	True retention	SB adequate preparation rate	Colonic adequate preparation rate
Leighton et al[24], 2017	Active CD	721	Male 306%; female 611%; unknown 8.3%	91.7%	SB (5); stomach (1)	0	87.9%	39.4%-62.9%
Bruining et al[27], 2020	Active CD	119	Male 40%; female 60%	85.6%	Requested removal (1); SB (1); colon (16)	13	79%-90%	64%
Eliakim et al[26], 2020	CD in remission	542	Male 500%; female 259%; unknown 24.1%	92.6%	SB (2); colon (1)	0	94.4	75%

¹In the original study, 6/72 patients were excluded due to incomplete pan-enteric capsule endoscopy.

²There were 54 pan-enteric capsule endoscopies performed in 41 patients.

³1 patient - capsule was retrieved by dilation of the small bowel stricture through ileo-colonoscopy.

PCE: Pan-enteric capsule endoscopy; CD: Crohn’s disease; SB: Small bowel; VCE: Video capsule endoscopy.

Table 6 Clinical and demographic characteristics of the individual participant data meta-analysis, n (%).

	LVP (n = 55)	Combined HVP (n = 257)	P value
Sex
Male	22/55 (40.0)	103/257 (40.1)	0.104
Female	33/55 (60.0)	135/257 (52.5)
Unknown	0/55 (0.0)	19/257 (7.4)
CD status1
Active CD	23 (41.8)	195 (75.9)	< 0.001b
CD in remission	19 (34.5)	58 (22.6)
Suspected CD	13 (23.6)	4 (1.6)
Montreal classification1
Age at diagnosis
> 16	2/41 (4.9)	10/126 (7.9)	0.513
17-40	30/41 (73.2)	97/126 (77.0)
> 40	9/41 (22.0)	19/126 (15.1)
Location
L1	24/42 (57.1)	73/168 (43.5)	0.269
L2	3/42 (7.1)	19/168 (11.3)
L3	15/42 (35.7)	76/168 (45.2)
L4	0/42 (0.0)	0/168 (0.0)
Behavior1
Inflammatory	32/42 (76.2)	121/168 (72.0)	0.840
Structuring	7/42 (16.7)	31/168 (18.5)
Penetrating	3/42 (7.1)	16/168 (9.0)
Current treatment1
5-ASA	8/55 (14.5)	21/172 (12.2)	0.651
Immunomodulatory	6/55 (10.9)	44/172 (25.6)	0.025a
Biologic	21/55 (38.2)	98/172 (57.0)	0.020a

^aP < 0.05.

^bP < 0.01.

¹Data was analyzed only among patients with established Crohn’s disease.

LVP: Low-volume protocol; HVP: High-volume protocol; 5-ASA: 5-aminosalicylic acid; CD: Crohn’s disease.

DISCUSSION

The aim of our study was to evaluate the effectiveness and safety of the novel LVP for PCE, developed at a single tertiary center, compared to the current standard high-volume preparation protocol. We have shown similar adequate preparation rates of the SB and possibly higher adequate preparation rates of the large bowel achieved by a protocol with substantially reduced volume and more palatable ingredients without compromising and maybe even improving the completion rates of the procedure. The standard preparation protocol for PCE includes ingestion of high-volume purgatives[24]. Studies on compliance with pre-colonoscopy preparation protocol demonstrate that between 5%-40% of patients do not consume the entire quantity of preparation due to inability to consume a large volume in a short time and the rather unpleasant taste of the purgatives[29,30]. Nevertheless, the quality of preparation in colonoscopy is critical for the proper identification and treatment of focal lesions[31]. While PCE preparation should be more intensive than for colonoscopy due to the inability to wash out content during examination, its main purpose is to assess overall inflammation, not to detect focal lesions. In terms of patient engagement, several studies found that lower volumes of bowel preparation, split doses, and palatability of the preparation improved both the patient experience and efficacy of preparation for colonoscopy[30,32,33]. Moreover, patients with impaired renal function have contraindications for sulfate-based solution consumption[29,34].

Therefore, we have developed the LVP based on 2 L-PEG-ascorbic acid and lactulose instead of sulfate-based solutions. PEG with ascorbic acid (2 L) has been previously proven to be as effective as 4 L PEG for colonic preparation[22,25], and lactulose is an osmotic laxative based on a non-absorbable sugar with a likable taste profile and is generally well-tolerated by patients[35]. In addition, this sulfate-free LVP offers substantial safety advantages. Sulfate-based solutions should be used with caution in patients with clinically significant renal, hepatic, or cardiac impairment, particularly those with cardiovascular pathology who frequently use medications such as diuretics or angiotensin-converting enzyme inhibitors[29]. The LVP in this study does not cause significant hemodynamic changes or electrolyte imbalances and can be safely prescribed for these vulnerable populations. Although our study did not specifically analyze this subgroup, the LVP’s composition makes it a theoretically preferable option.

In our study, adequate SB preparation rates were around 90% in both the LVP and HVP groups, with similar results when compared to the historical “combined HVP” group. For colonic preparation, the LVP group had higher adequate preparation rates (80%) than the “combined HVP” group (62.7%, P = 0.017), despite a reduction of > 2 L in total volume. Since the duration of each PCE study is dependent on the 12-hour lifetime of the battery, the purpose of the preparation protocol is not only cleansing, but also providing a booster for successful capsule expulsion out of the body before the end of the study. Although not statistically significant, more patients in the LVP group achieved a complete PCE study compared with those in the HVP group (89.1% vs 75%, P = 0.345). Two patients experienced a delay in the expulsion of the capsule despite the protocol booster: One due to delayed gastric emptying (LVP group), and one patient had SB strictures, which were not detected by recent CTE, leading to delayed SB transit time (HVP group). It is well known that SB strictures are not always symptomatic and may not necessarily be detected by imaging techniques[36,37]. Therefore, in order to prevent capsule retention, it is recommended to evaluate pre-PCE patency before swallowing VCE[23,38]. In our study, 9 patients had the stricturing phenotype B2 (LVP: 16.7% vs HVP: 25.0%), all of them had successfully passed a pre-PCE patency test, and none demonstrated post-PCE complications, including true capsule retention, hospitalization, or visits to the ER.

Another aspect of the “booster effect” is to allow capsule expulsion but without “over-accelerating” the capsule and having a too short SB transit time. Our results showed that the median SB transit time was similar in both groups, LVP 138.78 minutes (range: 44.27-514.48 minutes) vs HVP 119.91 minutes (range: 64.60-214.12 minutes), but shorter than previously described and recommended for SB VCE, which is generally more than 2 hours[39,40]. However, it is important to remember that PCE aims to evaluate disease load, not focal lesions, and the standardized small bowel transit time should be assessed specifically for PCE. Finally, economic aspects also differ between regimens, with modest cost variations across regions and depending on adjunctive agents as well (lactulose vs pico-sulfate). Although small relative to the overall cost of colonoscopy, these differences may influence system-level expenditures and patient access, particularly where reimbursement is limited. As recently noted by the United States Multi-Society Task Force, regimen choice should also consider safety, comorbidities, patient preference, and associated additional costs[41].

Our study has several limitations. First, the retrospective design may introduce bias due to missing data, although most data were prospectively documented during the PCE procedure. Second, the interpretation of PCE results is subjective, though two independent blinded expert observers using predefined criteria achieved a statistically significant correlation in cleansing scores. Third, the non-randomized allocation to study groups may result in confounding variables, although demographic and clinical characteristics were similar between both cohorts, and clinical guidelines for VCE referrals remained unchanged during the study periods. Furthermore, the observational nature of the study allowed us to examine PCE performance in a “real-life” setting and include patients that were not previously investigated, such as patients with suspected CD. Fourth, the LVP data were collected from a single center, but were comparable to the combined HVP group across three other centers. Lastly, the small LVP sample size limits statistical power, even in the IPD meta-analysis.

Despite these limitations, our study has several key strengths. Data were collected from a prospectively maintained database, ensuring high data quality, and the interpretation of PCE results was strengthened by using two independent, blinded expert observers, who achieved a statistically significant correlation in their cleansing scores. Furthermore, we performed an IPD meta-analysis, combining our cohort with three published cohorts. This substantially increased the sample size and enhanced the statistical power of our analysis, strengthening our conclusions regarding the superior colonic preparation of the LVP.

CONCLUSION

In conclusion, our study shows that a protocol with reduced volume and more palatable ingredients achieves adequate small and large bowel preparation without compromising procedure completion. LVP is similarly safe and effective as HVP. However, due to the small sample size and retrospective design, these findings should be validated in future prospective, randomized trials.