Randomized Controlled Trial of Remimazolam Compared with Placebo in Japanese Patients Undergoing Colonoscopy: A Phase III, Investigator-Initiated Trial

Colonoscopy is an important procedure for the early detection of colorectal cancer. The resection of adenomas during colonoscopy improves patient prognoses [1, 2]. However, colonoscopy is often painful without sedation. Sedative use can reduce pain during colonoscopy and improve the cecal intubation rate [3]. In Japan, benzodiazepines are often used to achieve sedation during colonoscopies. However, the insertion time during colonoscopy, which is painful, is approximately 5 min. Among the currently used benzodiazepines, midazolam has the shortest half-life of 3–4 h. However, the safety and cost of achieving sedation during colonoscopy remain controversial [4].

Remimazolam, a novel, ultrashort-acting benzodiazepine, has a shorter half-life than other benzodiazepines (approximately 40 min). Therefore, remimazolam may shorten the recovery time after colonoscopy [5‒19]. Considering the critical need for safer and more reliable sedation during gastrointestinal endoscopy procedures, we conducted a phase III investigator-initiated trial to compare the effects of remimazolam with those of a placebo (normal saline) in patients undergoing colonoscopy.

This multicenter, randomized, double-blind, placebo-controlled, phase III investigator-initiated trial was conducted at six facilities in Japan, including two clinics, two university hospitals, one community hospital, and one cancer center. The protocol of this trial was established after consultation with the Pharmaceuticals and Medical Devices Agency (PMDA). At the time the protocol was established, there were no sedatives for endoscopy covered by insurance in Japan, so we decided to compare it with a placebo. Considering the ethical aspects of the patients, the study was discontinued, and the subjects were excluded if they experienced severe pain; the use of sedatives such as midazolam and analgesics was allowed. This study was registered in the Japan Registry of Clinical Trials (Registration No. jRCT2031200360) and conducted in accordance with the Declaration of Helsinki. All patients provided informed consent for participation.

Japanese patients aged ≥20 years with a body mass index (BMI) <30 kg/m² who were scheduled to undergo colonoscopy without analgesics were included in the study. Patients with a BMI ≥30 kg/m² were excluded because administering the same dose may affect the effectiveness of sedation.

Patients who consumed >60 g of pure alcohol per day, used benzodiazepines or analgesics regularly, or had a history of drug or alcohol dependence were excluded from the study as these were factors that possibly affected the effectiveness of sedatives. Patients with hypersensitivity to benzodiazepines or flumazenil, contraindications to these drugs, or a history of remimazolam use were also excluded. Patients were excluded from the study if they had severe mental illness, severe respiratory disease, oxygen saturation <95% before or during drug administration, a Mallampati score ≥ III, essential hypotension, difficulty in communication or walking unassisted, or a history of abdominal surgery. These exclusions were due to differences in procedure time and pain experienced during endoscope insertion compared to the general patient population. Some patients with prior abdominal surgery develop adhesions and report discomfort during colonoscopies. Furthermore, patients with a history of colon cancer surgery may have a shortened intestine. Patients with aspartate transaminase or alanine aminotransferase levels ≥2.5 times the institutional upper limit or a total bilirubin level ≥1.5 times the institutional upper limit were also excluded. Pregnant or lactating women, patients with childbearing potential (amenorrhea for <12 months or no use of permanent contraception), and those who refused to use contraception were also excluded. Additionally, patients who had participated in another clinical trial within 4 weeks of the current trial and those deemed unsuitable for the study by physicians were excluded.

Patients were randomly assigned to the remimazolam or placebo group at a ratio of 4:1, using the number method. Allocation adjustment factors included age (≥75 or <75 years), body weight (≥45 or <45 kg), and medical institution type. The cutoff for body weight was 45 kg, according to data from the Health and Nutrition Survey published by the Ministry of Health, Labor, and Welfare of Japan (https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/kenkou_iryou/kenkou/eiyou/h29-houkoku.html). Once a patient was registered on the computer, a number was assigned to them. An allocation table based on the assignment of patients into the remimazolam or placebo groups was distributed in advance to medical staff not involved in endoscopy, and decisions were made based on their assigned numbers. Normal saline solution was used as a placebo. The patients and medical staff involved in colonoscopy were blinded to the patient group. Remimazolam was diluted with a 20-mL saline syringe to 1 mg/1 mL to make it indistinguishable from placebo.

A colonic cleanser for colonoscopy was administered in accordance with the standard practice of each medical institution. Colonoscopy was performed only in the outpatients. An initial dose of 3 mg (3 mL) remimazolam or placebo was administered intravenously. Patients aged ≥75 years and weighing <45 kg received a half dose (1.5 mg) at the discretion of the endoscopist. The level of sedation was evaluated using the modified observer’s alertness/sedation score (MOAA/S score) (Table 1) every 2 min after administration of the first dose of remimazolam or placebo [20]. Sedation was defined as a MOAA/S score of ≤4. Colonoscopy was initiated after sedation. If the patient was not sedated (MOAA/S score ≥5), an additional dose of 1 mg was administered. If sedation was not achieved after five additional doses were administered, the sedative was deemed ineffective. Administration of rescue medications including midazolam and propofol was not permitted. The patients’ vital signs, including blood pressure, heart rate, and respiratory rate, oxygen saturation, and MOAA/S scores were measured every 5 min after colonoscopy was initiated. If signs of arousal, such as an MOAA/S score ≥5 or body movement, were observed at intervals other than every 5 min, an additional dose of the assigned drug was administered, particularly if at least 2 min had elapsed since the previous dose. Patients’ vital signs, oxygen saturation, and MOAA/S scores were assessed immediately after colonoscopy, at 5 min post-procedure, and subsequently every 10 min from 10 to 60 min after colonoscopy. Walking ability was evaluated in patients with an MOAA/S score of 5. The presence or absence of re-sedation was observed 60 min after the end of colonoscopy. The presence or absence of adverse events was confirmed 7–14 days after the procedure. All colonoscopies were performed by doctors certified by the Japanese Gastroenterological Endoscopy Society.

The primary endpoint was the sedation success rate during colonoscopy, which was defined as the achievement of sedation (MOAA/S score ≤4) before the initiation of colonoscopy, successful completion of colonoscopy, and use of ≤5 additional doses of the assigned drug per 15 min of the procedure.

Secondary endpoints included the dose needed to achieve sedation, duration from procedure completion to regaining consciousness (assessed by MOAA/S score of 5), time from procedure completion to ambulation (defined as the patient’s ability to walk 5 m in a straight line without staggering), and incidence of adverse events. Physicians and patients completed a questionnaire regarding their satisfaction with the procedure. The endoscopists and patients completed a satisfaction questionnaire as follows: 5, very satisfied; 4, satisfied; 3, normal; 2, somewhat dissatisfied; and 1, very dissatisfied.

Colonoscopy completion was defined as reaching the cecum and sufficiently observing the anus. Loss of consciousness was defined as a MOAA/S score ≤1 at ≥2 consecutive time points. Oxygen was administered (2–5 L/min) when oxygen saturation was <94%. The airway was secured if the respiratory rate fell below five breaths/min. If breathing did not recover after securing the airway, manual ventilation was performed using a back-valve mask. Intravenous flumazenil was administered when there was an urgent need to reverse sedation with the assigned drug. These criteria were determined with reference to the opinions of anesthesiologists who serve as external efficacy and safety evaluation committee members, and more stringent criteria were set in order to sensitively detect respiratory depression in patients. Regarding adverse events, in both groups, investigations were conducted during the endoscopic examination and up to the outpatient visit 1 week after the end of examination. The severity of adverse events was determined according to the Common Terminology Criteria for Adverse Events version 5.0 (https://ctep.cancer.gov/protocolDevelopment/electronic_applications/ctc.htm#ctc_50).

Based on previous reports of a sedation success rate of 91.3% in the remimazolam group and 1.7% in the placebo group, the success rates of sedation during colonoscopy were predicted to be 90% and 5% in the remimazolam and placebo groups, respectively. Twenty-five patients (20 and 5 in the remimazolam and placebo groups, respectively) were required to ensure a power of ≥90%, with a two-sided significance level of 5%. After discussions with the PMDA regarding safety concerns, the sample size was increased to 45 individuals (36 patients in the remimazolam group and 9 in the placebo group) to thoroughly evaluate the efficacy and safety of remimazolam, especially concerning adverse events. Based on an estimated dropout rate of 10%, the final target sample size was 50 patients (remimazolam group, n = 40; placebo group, n = 10).

Continuous variables are presented as medians and interquartile ranges, while categorical variables are expressed as counts and percentages. Fisher’s exact test and Wilcoxon rank-sum test were used to compare categorical and continuous variables between groups, respectively. Physician and patient satisfaction were categorized into three levels: 1 or 2 (not satisfied), 3 (neutral), and 4 or 5 (satisfied). The comparison of satisfaction rates between groups was conducted using the Wilcoxon rank-sum test. Statistical significance was defined as p < 0.05. Statistical analyses were performed using JMP software (version 13.0.0; SAS Institute, Cary, NC, USA).

Fifty-two patients were enrolled in this study between September and December 2021 (Fig. 1). One patient in the placebo group was excluded owing to withdrawal of consent. Therefore, the final analyses included 40 patients in the remimazolam group and 11 patients in the placebo group.

Baseline characteristics such as age, sex, height, weight, and BMI, and American Society of Anesthesiologists classification did not differ significantly between groups (Table 2).

The successful sedation rate was 95.0% (38/40 patients) in the remimazolam group and 0.0% (0/11 patients) in the placebo group (p < 0.01; Table 3). Despite the administration of five additional doses, sedation was not achieved in 1 patient in the remimazolam group. Moreover, over five additional doses of the assigned drug were administered to 1 patient (2.5%) after colonoscopy was initiated. Examination could not be completed in 1 patient (9.1%) in the placebo group.

Patients in the remimazolam group required significantly fewer additional and total doses to achieve sedation before the initiation of colonoscopy than those in the placebo group (p < 0.01; Table 4). The time from the end of endoscopy to regaining consciousness was 0.0 min in both groups. The time from completion of endoscopy to ambulation was 5.0 min in the remimazolam group and 0.0 min in the placebo group (p = 0.02).

In the remimazolam group, 2 patients had SpO₂ levels <94%, necessitating oxygen administration. However, no other adverse events related to vital signs, such as hypotension or bradycardia, requiring medical intervention, were observed. Additionally, none of the patients in this group required flumazenil administration, manual ventilation, respiratory depression, or experienced hypotension intraoperatively. One (2.5%) and 5 (12.5%) patients in the remimazolam group experienced moderate and mild adverse events, respectively, after leaving the endoscopy room. In contrast, none of the patients in the placebo group experienced any adverse events (Tables 5, 6). Physician (Table 7) and patient (Table 8) satisfaction rates regarding the time to sedation, depth of sedation, and pain relief were higher in the remimazolam group than in the placebo and midazolam groups (p < 0.01).

This study evaluated the efficacy and safety of remimazolam in Japanese patients undergoing colonoscopy and demonstrated a significantly higher successful sedation rate with remimazolam than with a placebo. Although previous comparative studies have reported the usefulness and safety of remimazolam for sedation during colonoscopy [8, 13‒17], this is the first placebo-controlled trial. Moreover, these previous studies used additional sedatives and analgesics such as midazolam and propofol, and none evaluated the effects of remimazolam alone. This paucity of data is exacerbated by the fact that remimazolam for gastrointestinal endoscopy is not covered by health insurance in Japan. In this study, following discussions with PMDA, remimazolam was compared with placebo rather than with benzodiazepine-like midazolam.

A previous phase II trial indicated that the administration of initial and additional doses of 3 mg and 1 mg, respectively, was optimal for remimazolam during colonoscopy [4]. In our study, the number of additional doses required before colonoscopy was zero, which confirmed the optimal dose of remimazolam. Additionally, 95% of patients in the remimazolam group in this study completed colonoscopy with one additional dose. In contrast, sedation was not achieved prior to the start of colonoscopy in any patient in the placebo group, and 1 patient was unable to complete colonoscopy due to severe pain.

A phase IIb trial conducted in the USA reported sedation success rates during colonoscopy of 92.5%, 95.0%, and 97.5% when remimazolam was administered at initial and additional doses of 8.0 and 3.0 mg, 7.0 and 2.0 mg, and 5.0 and 2.0 mg, respectively [15]. Moreover, a phase III study reported that the sedation success rate during endoscopy was 91.3% with initial and additional doses of remimazolam of 5.0 and 2.0 mg, respectively [14]. Another phase III trial conducted in China reported a sedation success rate during colonoscopy of 98.9% when patients were sedated using initial and additional doses of remimazolam of 7.0 and 2.5 mg, respectively [17].

In this study, the time from the end of endoscopy to regaining consciousness did not differ significantly between both groups. Moreover, the recovery time was short, and the time to ambulation ranged from 0 to 10 min. In some cases, the sedative effects were worn off by the end of the procedure; however, during colonoscopy, patients typically experienced discomfort, mainly when the colonoscope was inserted. Therefore, we believe that there is no disadvantage for the patient to be awake from the middle to the end of colonoscopy.

In a phase III study conducted in the USA, the time from the end of colonoscopy to regaining consciousness was 7.35, 21.95, and 15.84 min in the remimazolam, placebo, and midazolam groups, respectively. The times from the end of colonoscopy to ambulation were 42.65, 53.18, and 47.92 min, respectively. Notably, patients in the placebo group of this study were administered midazolam as rescue medication [14]. Moreover, a Chinese phase III study reported the time from the end of colonoscopy to regaining consciousness as 8.14 ± 2.87 and 7.74 ± 2.93 min in the remimazolam and propofol groups, and the time to discharge was 18.30 ± 7.35 and 17.84 ± 6.05 min, respectively [8]. The recovery times were longer in these trials than in the current study due to the higher initial dose of remimazolam and concomitant use of fentanyl as an analgesic.

In previous studies, a MOAA/S score ≤3 was used to define sedation; therefore, the patients in those studies were likely to be in a deeper state of sedation than those in our study. A MOAA/S score ≤4 was used to define sedation in this study, as conscious sedation is recommended for colonoscopy to ensure patient safety [4].

In this study, a few moderate and mild adverse events occurred, with no reports of severe adverse events. None of the patients experienced hypotension or respiratory depression or required mechanical ventilation or flumazenil during colonoscopy. In a previous study comparing sedation with remimazolam (initial dose, 5 mg; additional dose, 2.5 mg) and midazolam (initial dose, 1.75 mg; additional dose, 1 mg) during colonoscopy, hypotension, and hypoxemia occurred in 38.9% and 1.0% of patients in the remimazolam group, respectively [14]. Another study reported a 10.6% decrease in blood pressure and a 1.1% decrease in respiratory depression in patients administered remimazolam [17]. The rates of these adverse events have been high in previous studies. The efficacy and safety of sedatives depend on the dose of the drug and vary greatly. In this study, an initial remimazolam dose of 3 mg and an additional dose of 1 mg were effective and safe during colonoscopy, without analgesics.

During colonoscopy, sedatives are primarily used to reduce patient distress. The degree of pain reduction and physician and patient satisfaction were measured by directly comparing the effects of remimazolam with those of a placebo. Both physicians and patients reported that remimazolam resulted in less pain than the placebo. Owing to pain and lack of sedative use, patients may avoid colonoscopy. An increase in patient satisfaction increases the rate of reexamination requests, which facilitates the early detection of gastrointestinal cancer. In the present study, rescue medication was not permitted when sedation was not achieved. Therefore, pain was more intense in the placebo group, and the procedures were challenging to perform, which may account for the low patient and physician satisfaction rates in the placebo group.

This study has several strengths. First, it was a multicenter, double-blind, placebo-controlled trial. In Japan, where there is a growing demand for sedation during colonoscopy, few sedatives are covered by national health insurance for this purpose, making this study particularly relevant. The placebo-controlled design was developed in collaboration with the PMDA and conducted across hospitals of various capacities, demonstrating the possibility of using remimazolam across various clinical settings. Second, patients did not receive analgesics or rescue medications, eliminating potential interactions with other drugs.

Nevertheless, this study had certain limitations. First, the effects of patient ethnicity were not considered. The use of remimazolam has been reported in European and American studies; however, this study included only Japanese patients, as it was conducted to influence the coverage of the drug by national health insurance in Japan. Second, patients with alcoholism and sleep apnea were excluded. The study results may differ between patient populations of different races and between those with different comorbidities. Third, all endoscopies were conducted by expert doctors certified by the Japanese Gastroenterological Endoscopy Society. Fourth, complete anonymization could not be achieved since patients who are sedated are known to the endoscopist to be in the remimazolam group, patients who are not sedated are known to the endoscopist to be in the remimazolam group, and patients who are not sedated are known to the endoscopist to some extent to be in the placebo group before the procedure begins. Moreover, patients who were sedated were not blinded to the postoperative questionnaire because they were known to be in the remimazolam group. Finally, the MOAA/S score, which was used to evaluate the sedation level in this study, is a subjective score.

Remimazolam significantly increased the sedation success rate during colonoscopy compared with placebo. Further investigation to assess the effectiveness of remimazolam compared to other sedative agents is warranted.

We sincerely thank Sho Suzuki, Tomomi Sugita, Toshiki Horii, Noriko Asamizu, Chizuru Naito, and Eri Maruyama from the Nihon University School of Medicine and Chika Kusano from the Kitasato University for their valuable assistance in conducting this study. We also thank Oda Ichiro from the Kawasaki Rinko Hospital and Soichiro Inoue from the St. Marianna University School of Medicine for conducting the external efficacy and safety committee. We thank Editage (www.editage.com) for English language editing.

This multicenter, RCT study was conducted in accordance with the guidelines of the Declaration of Helsinki. The study protocol was approved by the Institutional Review Board of Nihon University School of Medicine, the Nagata Surgery and Gastroenterological Clinic, and the Kiriyama Clinic on March 25, 2021, October 1, 2021, and October 1, 2021, respectively. Additionally, the protocol was approved by the Institutional Review Board of Shizuoka Cancer Centre on September 21, 2021; the Ethics Review Committee of the National Hospital Organization Ureshino Medical Center on June 28, 2021; the Kyushu University Ethical Review Board on August 30, 2021; and the Institutional Review Board of Tonan Hospital on July 29, 2021. This was an investigator-initiated clinical trial, so there was no approval number. All patient enrollment at each facility was conducted after approval by each Institution’s Review Board. Written informed consent was obtained from all patients prior to enrollment. This study was registered in the Japan Registry of Clinical Trials (Registration No. jRCT2031200360) on February 15, 2021.

There are no conflicts of interest to declare for this article.

This study was funded by Mundipharma Deutschland GmbH & Co. KG.

R.I. and H.I. planned and designed the study and wrote the manuscript. R.I., H.I., H.O., K.H., D.Y., M.E., Y.M., Y.N., K.O., S.K., T.S., and Y.K. were involved in the critical revision of the article for important intellectual content. All authors have read and approved the final manuscript.

The data that support the findings of this study are not publicly available as they contain information that could compromise the privacy of research participants but are available from the corresponding author H.I. upon reasonable request.