Comparing the efficacy of mannitol and polyethylene glycol in treating patients with poor bowel preparation: a randomized controlled clinical study.

Introduction
Colorectal cancer (CRC) is one of the most prevalent malignancies worldwide. According to 2020 global cancer statistics, colorectal cancer has an incidence rate of 10% and a mortality rate of 9.4% [1]. The majority of colorectal cancers arise from the transformation of benign lesions, notably colorectal polyps [2]. Colonoscopy and endoscopic colonic polypectomy are recognized as pivotal factors in reducing both the morbidity and mortality associated with CRC [3]. These procedures are now endorsed by numerous national guidelines [4]. However, the effectiveness of colonoscopy is contingent upon the quality of bowel preparation, with up to one-quarter of patients experiencing inadequate bowel preparation [5].
Despite ongoing research into bowel cleansing protocols, such as dietary modifications, laxative agents, and cleansing regimens [6, 7], instances of insufficient bowel preparation persist during colonoscopies [6, 8, 9]. Current guidelines recommend implementing repeat measures in cases of inadequate bowel preparation or rescheduling the endoscopy [10]. However, approximately 30% of patients do not adhere to physicians’ instructions for rescheduling tests effectively [11]. Same-day remedies may prove more effective in enhancing compliance, suggesting that same-day remediation could offer more advantages than rescheduling the examination. Regarding repeat measures, studies have confirmed that orally administered polyethylene glycol (PEG) is superior to enemas for achieving intestinal cleanliness [12, 13]. Nevertheless, the unpalatable taste of PEG and the large dosage required render it unpopular among patients, leading to suboptimal intestinal preparation. Opting for orally administered mannitol is beneficial because its usage requires a small dosage and involves a pleasant taste, potentially increasing patient compliance [14]. To date, no prospective randomized controlled trials have focused on evaluating the efficacy and safety of mannitol for intestinal remediation.
This study was designed to assess and compare the effectiveness and tolerability of orally administered mannitol as a same-day remedial regimen for patients with inadequate bowel preparation that is not inferior to orally administered PEG.

Methods

Study design and population
The research was a prospective, randomized, open-label, blinded end point, noninferiority trial. The ethical review committee of the First Affiliated Hospital of Ningbo University approved the study (2023-066A), and it was registered at ClinicalTrials.gov (NCT05912114). Furthermore, approval was obtained from another collaborating institution, Ninghai Second Hospital, in China. Every patient gave written consent after being informed about the study. The Declaration of Helsinki was followed in the conduct of the study and this study design adhered to CONSORT guidelines.
Patients aged 18–75 years at the First Affiliated Hospital of Ningbo University and Ninghai Second Hospital in China and with a Boston Bowel Preparation Scale (BBPS) score ≤ 1 in any of the three colon segments (right side, transverse, and left side) were enrolled between July and December 2023. Those excluded from the study were pregnant women, patients with severe electrolyte disorders, individuals with a history of colorectal surgery, patients experiencing an acute exacerbation of inflammatory bowel disease, patients with severe medical conditions that made them unsuitable for colonoscopy, patients with a known allergy to mannitol or PEG, and patients who refused to sign the informed consent form. All patients were administered either orally administered mannitol or PEG solution.

Bowel preparation
All patients were recommended a split-dose bowel preparation regimen (1 L PEG solution the night before and 2 L the next day for a bowel preparation) and a low-residue diet 1 day before the time of their first bowel preparation. In the PEG group, we utilized PEG powder, where each packet comprised 0.74 g of potassium chloride, 1.68 g of sodium bicarbonate, 1.46 g of sodium chloride, 5.68 g of sodium sulfate, and 60 g of polyethylene glycol 4000. The solution was prepared by diluting each packet with 1 L of water and was to be consumed entirely 2–3 h before the scheduled repeat colonoscopy. In the mannitol group, a 1000-ml mannitol solution was prepared by combining 250 ml of 20% mannitol (Shanghai Baxter Medical Supplies Co.) with 750 ml of water. The complete solution was administered 2–3 h before the scheduled colonoscopy procedure. Salvage bowel preparation protocols were initiated immediately following completion of the initial colonoscopy and confirmation of patient eligibility.

Randomization
A simple randomization strategy was employed, and a researcher, independent of the trial, created sealed envelopes in numerical order using random numbers generated by a computer. As soon as informed consent was obtained, patients were randomly assigned to either the mannitol group or the PEG group at a 1:1 ratio using the envelopes. The envelopes were sealed and numbered, and assignments were opened by an independent secretary before the repeat colonoscopy. Before the initial colonoscopy, a researcher administered a questionnaire about sex, age, body mass index (BMI), history of prior abdominal surgery, history of hypertension, and diabetes, and BBPS scores were collected after the initial colonoscopy by this researcher. A blinded independent investigator gathered data during the patient’s repeat colonoscopy and recorded the cecal intubation time, BBPS score, and potential negative reactions such as nausea, abdominal pain, and bloating. The assessment also included information on drug taste, patient satisfaction, and patient receptiveness to continue medication for bowel preparation before the next colonoscopy. All assessors and endoscopists in the remediation process were blinded to participant allocation and intervention, but all the patients were not blinded in this study.

Study procedures
After randomization, all eligible patients were instructed to consume either 1 L of mannitol solution or 1 L of PEG solution orally. Subsequently, a repeat colonoscopy was scheduled with a 2- to 3-h interval. Before starting the repeat colonoscopy, the researchers inquired about the patients’ adherence to consuming the prescribed 1 L of solution and the characteristics of the final rectal effluent. Both the initial and repeat colonoscopies utilized adult video colonoscopes, specifically the Olympus CF-HQ/H-290/260I video colonoscope, without incorporating a magnetic endoscopic imager. Air insufflation was used for all the colonoscopies in this study. No anesthesia was administered to any patients undergoing repeat measures.

Outcomes
The primary aim of this study was to evaluate and compare the efficacy of orally administered mannitol in repeat colonoscopy to whether it is non-inferior to the PEG, where a BBPS score of ≥ 2 for all three colon segments (right side, transverse, and left side colons) was considered adequate.
The secondary outcomes included the adverse effects of drug remediation (such as nausea, bloating, and abdominal pain) as well as the taste and satisfaction (measured on a 4-point scale). Additionally, the study examined patient inclination to reuse the same medication for bowel preparation before colonoscopy, the rate of successful insertion into the cecum, and verification of cecum intubation through identification of the ileocecal flap on endoscopic video. Withdrawal time was defined as the duration from cecal identification to removal of the colonoscope, excluding the time required for biopsy and polyp removal [15]. Satisfaction was measured using a numeric rating scale from 1 (very satisfied) to 4 (very unsatisfied).

Statistical analysis
Orally administered PEG solutions achieve sufficient bowel preparation in approximately 88.4% of patients [16]. Another study reported an oral preparation adequacy of approximately 75% with 1 L of mannitol solution [14]. Therefore, leveraging existing research findings, we assumed that the adequacy rates for bowel preparation are 90% in both the control and experimental groups, with a noninferiority margin of 15%. This ensured that a minimum of 75% adequacy was maintained in the experimental group. Employing the noninferiority test method, with α = 0.025 and β = 0.2, the sample size was calculated as 63 patients in each group. Assuming a 5% loss to follow-up rate, a total of 67 patients were included in each group, resulting in a total of 134 patients. The sample size calculation was performed using PASS 15.0.5.
All categorical or proportional variables in the results were compared using chi-square or Fisher’s exact tests, and the P values were calculated. In addition, we tested each continuous variable for normality and compared it with the t test. In addition to expressing the results as the mean (standard deviation, SD), we assessed the scores of the questions on a 4-point scale using Wilcoxon rank-sum tests, and the P value was also calculated. All the statistical analyses in this study were conducted using SPSS version 21. Primary outcome indicators were calculated using both intention-to-treat (ITT) and per-protocol (PP) analyses, while secondary outcome indicators were calculated using only ITT analyses. P values < 0.05 were considered indicative of statistical significance.

Results

Patients’ baseline characteristics
A total of 156 patients were initially screened. After 22 individuals who did not meet the inclusion criteria were excluded, 134 participants from two hospital centers were ultimately enrolled. All participants were recruited from elective patients in outpatient clinics or inpatient departments, excluding those requiring emergency colonoscopy procedures. Importantly, none of the enrolled patients presented with obstructive symptoms at the time of enrollment. As shown in Fig. 1, 134 patients were included in the ITT analysis, while PP analysis comprised 127 patients after excluding 7 cases (3 in the mannitol group, 4 in the PEG group) as a result of failed cecal intubation. Technical failures resulted from persistent solid stool residue in the left colon (n = 6) and severe colonic edema of unidentified etiology (n = 1).
The mean age of the study participants was 56.95 (SD 12.718) years, 54.4% (73/134) were male, and the mean BMI was 23.4 kg/m2 (SD 3.1). The mean BBPS score during the initial colonoscopy was 3.28 (SD 1.5). The most frequent indication for colonoscopy was screening (78.3%). In addition, 37 (27.6%) patients had a history of abdominal surgery. In general, there was no notable difference in baseline information between the two groups (P > 0.05, Table 1).

Quality of bowel preparation
Both ITT and PP analyses were conducted to assess the quality of bowel preparation. The ITT analysis revealed that the rate of adequate bowel preparation was 95.5% (64/67) in the mannitol group and 92.5% (62/67) in the PEG group. The difference in proportions (3.0%, 95% CI − 5.0, 11.0) was greater than the prespecified margin of non-inferiority (− 15.0%), indicating the non-inferiority of mannitol compared with PEG. The results of the PP analysis were consistent with those of the ITT analysis (Fig. 2).

Feasibility, safety, and tolerability
All patients completed repeat bowel preparation and the occurrence of adverse events, nausea, bloating, or abdominal pain did not significantly differ between the two remediation programs, suggesting comparable safety profiles. No significant difference was observed between the groups in terms of taste (19.4% vs. 29.9%, P = 0.160). Participant satisfaction, assessed using a numeric rating scale, was greater in the mannitol group than in the PEG group (1.0 vs. 2.0, P < 0.001). Furthermore, patients tended to use mannitol for subsequent bowel preparation (89.6% vs. 47.8%, P < 0.001; Table 2).

BBPS score for intestinal cleanliness
Subsequent subgroup analysis was conducted on the cleanliness scores of three distinct segments of the colon, revealing no significant differences between the mannitol and PEG groups in terms of cleanliness in any of the colon segments. Overall, the intestinal cleansing efficacy of orally administered mannitol solution for colonic preparation was found to be non-inferior to that of orally administered PEG solution. Moreover, the mannitol group exhibited a favorable trend (Table 3).

Discussion
To our knowledge, this is the first randomized controlled clinical trial to investigate low-dose mannitol as a salvage regimen for bowel preparation. In this study, we show that compared with PEG, mannitol, which is used as an adjunct for bowel preparation, is noninferior in terms of intestinal cleanliness. Both approaches are considered technically feasible; however, mannitol outperforms PEG in terms of patient satisfaction and the receptiveness to choose this regimen again.
PEG solution, commonly employed for bowel preparation, is widely used by scholars as a salvage agent [13]. Despite its prevalence, some reports suggest that patients may exhibit intolerance to PEG solution due to high fluid intake requirements and an unpleasant taste [17]. Mannitol, a derivative and common laxative, is characterized by poor intestinal absorption, palatability, and good tolerability [18, 19]. However, some authors have raised concerns about the potential for intestinal fermentation upon oral ingestion of mannitol, which could lead to colonic gas explosions caused by methane and hydrogen [20]. Contrary to these concerns, studies by Paulo and Carnovali [21, 22], among others, indicated that the mannitol group did not exhibit higher levels of CH4 after gas exchange. Following standardized washing and aeration procedures, mannitol did not increase the potential levels of H2 or CH4. These findings suggested that orally administered mannitol was as safe as other bowel preparation strategies. Taking these factors into account, we chose mannitol as a control for comparison with traditional salvage regimens.
In our study, the majority of patients achieved adequate bowel preparation, aligning with both ITT and PP analyses. Although there was no difference in the adequacy of bowel preparation between the two groups in each colon segment, the mannitol group demonstrated greater BBPS scores than did the PEG group (7.8 ± 1.1 vs. 7.3 ± 1.1, P = 0.007), indicating a more favorable trend. Notably, no patients withdrew from the salvage protocol, potentially because of the simplicity and feasibility of our bowel preparation intervention. Unlike other studies, we did not employ high doses of medications; instead, both the control and experimental groups underwent a 1-L salvage regimen. This approach enhanced patient compliance, and the cleanliness outcomes surpassed the recommended 85% threshold as outlined in the guidelines [23]. Therefore, on the basis of our findings, mannitol can be considered a viable remedy for suboptimal bowel preparation.
According to the results of a clinical study that included 140 patients, administering oral solutions of 4 L of PEG and 2 L of 10% mannitol did not significantly improve intestinal cleanliness (P = 0.75) [24]. On the basis of the findings of another study conducted by Vieira et al., patients were categorized into two groups according to whether they received 1 L of 10% mannitol solution or 2 L of PEG solution for bowel preparation. Patients in both groups achieved high levels of intestinal cleanliness, but no significant differences were observed [25]. Our results deviate from those of the aforementioned studies, possibly because of variations in liquid volumes and drug concentrations. The solution volume used in both groups was 1 L to alleviate the burden on patients during remediation and facilitate quicker medication intake. Further research is necessary to investigate the impact of different drug doses on intestinal remediation cleanliness.
The occurrence of adverse reactions in our study was lower than that in other similar studies [13]. There were no significant differences between the two groups in terms of the incidence of nausea, bloating, abdominal pain, or overall adverse reactions, possibly owing to the lower dosage employed. No adverse events requiring clinical intervention were observed in any of the patients. Although there was no statistically significant difference in taste perception between the two groups, the mannitol group exhibited a greater trend, which could be associated with the relatively small sample size in our study. Consistent with previous research, the mannitol group demonstrated greater advantages in terms of satisfaction and receptiveness to choose the same regimen for future bowel preparation [24].
There are several limitations in this study. Primarily, the number of participants was insufficient, and experienced doctors performed the endoscopy, which might affect the generalizability of the findings to novice medical professionals. Moreover, despite 2 h of remediation, mannitol and PEG still left a large amount of fluid in most patients’ intestines. Indirectly, this residual fluid interfered with the endoscopic visualization and needed to be aspirated, prolonging the endoscopy process time. Too short a drug remediation time would tend to cause this. For excessive fluid accumulation in the intestinal cavity, the previously established 2–3 h preparatory interval could be extended appropriately. Notably, we did not monitor the patients’ electrolyte levels, as this would have required obtaining blood samples, markedly reducing patient receptiveness to participate. However, we conducted follow-up assessments and found no adverse events that required clinical intervention. The low dosage employed in our study further decreased the probability of electrolyte imbalances occurring. Additionally, while our study encompassed patients with inadequate bowel preparation, we did not stratify high-risk subgroups (e.g., hospitalized individuals or those with restricted mobility) who experienced unsuccessful cleansing. This underscores the need for future investigations to determine whether these populations derive greater benefit from salvage bowel preparation protocols. Lastly, while ADR data were not collected, we posit that bowel preparation quality directly correlates with ADR, as optimal mucosal visualization enhances lesion identification.
In summary, the results of this study confirm that orally administered mannitol is not inferior to orally administered PEG. Meanwhile, orally administered mannitol is known for its superior taste and overall lower burden than orally administered PEG solution. This approach introduces a novel method for rehabilitating patients with insufficient bowel preparation.