Current Medical Research & Opinion
Curr Med Res Opin Downloaded from informahealthcare.com by University of Missouri Kansas City UMKC on 03/31/15 For personal use only.
0300-7995 doi:10.1185/03007995.2014.960513
Vol. 30, No. 12, 2014, 2493–2503
Article ST-0173.R2/960513 All rights reserved: reproduction in whole or part not permitted
Original Article Randomized, controlled trial of 2 L polyethylene glycol plus ascorbate components versus sodium phosphate for bowel cleansing prior to colonoscopy for cancer screening
C. Ell Dr Horst Schmidt Kliniken GmbH, Klinik fu¨r Innere Medizin II, Wiesbaden, Germany
W. Fischbach Gastroenterology Department, Klinikum Aschaffenburg, Germany
P. Layer Abt. Innere Medizin, Israelitisches Krankenhaus, Hamburg, Germany
Abstract Background: Colonoscopy requires effective bowel preparation for adequate mucosal visualization. Safety and acceptability of bowel preparation are key components in colorectal cancer screening (CRC) populations. Objective: To compare the efficacy, safety and acceptability of bowel preparation with polyethylene glycol (PEG), ascorbic acid, sodium ascorbate (ascorbate components), sodium sulfate and electrolytes (PEGþAsc) or sodium phosphate (NaP).
M. Halphen Norgine Limited, Harefield, Uxbridge, Middlesex, UK Address for correspondence: Dr Marc Halphen, Norgine Limited, Norgine House, Widewater Place, Moorhall Road, Uxbridge, UB9 6NS, UK. Tel.: þ44 (0)1895 826600; [email protected] Key words: Ascorbate components – Ascorbic acid – Colon cancer screening – Colonoscopy – Polyethylene glycol – Sodium ascorbate – Sodium phosphate Accepted: 22 August 2014; published online: 23 September 2014 Citation: Curr Med Res Opin 2014; 30:2493–503
Methods: Consenting adults undergoing elective out-patient colonoscopy for CRC were randomized to take 2 L PEGþAsc or 90 mL NaP (control) following manufacturer’s instructions. PEGþAsc was taken the evening before and morning of the colonoscopy; NaP was taken the morning and evening before colonoscopy. Participants followed a restricted diet specific to each preparation. Primary endpoint was bowel cleansing success (100% colon mucosa visible) rated by an independent expert panel (all experienced endoscopists) unaware of treatment allocations. Subject reported outcomes about the preparations were elicited. Adverse events were recorded. Clinical trial registration: Clinicaltrials.gov: NCT00427089. Results: Successful bowel cleansing rate was significantly higher in the PEGþAsc (N ¼ 242) than the NaP (N ¼ 114) group (PEGþAsc 93.4% [95% CI 89.5–96.2] versus NaP 22.8% [15.5–31.6%], p50.0001). Subject reported outcomes on acceptability of the two different preparations were not significantly different (p ¼ 0.238). However, taste ratings for PEGþAsc were significantly better versus NaP (mean VAS: 31.2 and 38.1 respectively, p ¼ 0.0111). The proportion of patients prepared to receive the same preparation again was significantly higher in the PEGþAsc group (88.4% vs. 78.1%, p50.0001). Conclusions: PEGþAsc provided superior bowel cleansing to NaP and was well tolerated. Findings for PEGþAsc are aligned with previous similar studies; however, differences observed in NaP cleansing results, especially for the proximal colon segments, may be due to factors including: differences in demographics and population types and the use of the validated Harefield Cleansing Scale as an assessment tool combined with expert reviews, which may have resulted in conservative cleansing assessments.
! 2014 Informa UK Ltd www.cmrojournal.com
2L PEGþAsc versus sodium phosphate for bowel cleansing Ell et al.
2493
Current Medical Research & Opinion Volume 30, Number 12
December 2014
Curr Med Res Opin Downloaded from informahealthcare.com by University of Missouri Kansas City UMKC on 03/31/15 For personal use only.
Introduction Worldwide diagnosis of colorectal cancer (CRC) in 2012 was reported to be approximately 1.36 million new cases, representing around 10% of all reported cancer cases1. In Europe in 2012, the incidence of new cases of CRC was 446.8 thousand with 214.7 thousand deaths reported in the same year2. Population screening programs in Europe include the use of Fecal Occult Blood tests (FoBt), colonoscopies and to a lesser extent flexible sigmoidoscopy3. Mass screening with FoBt has been shown to result in a reduction of around 16% in the mortality from CRC and has been adopted in several European countries4. In addition, colonoscopy as a primary screening modality for CRC detection has been established in Germany since 20024. Recently, it has been shown that an increase in colonoscopy completion rate (as defined by ileocecal visualization) within a CRC screening population has been associated with a reduction in CRC associated mortality5. Reasons for failure to visualize the ileocecal valve during colonoscopy may include uncontrolled looping, patient discomfort and poor bowel preparation6. Recently it has been shown that several independent factors, including diabetes, chronic health conditions and patient self-activation may influence the quality of bowel preparation7. Any of these factors may affect the diagnostic sensitivity of screening colonoscopies which depend on complete mucosal visualization8. Consequently, effective and well tolerated bowel cleansing is essential before the procedure. There are a number of different bowel lavage preparations available. Traditional polyethylene glycol (PEG) preparations require consumption of a large volume of liquid (4 L) which many patients find unpleasant and some find impossible to drink9. Lower volume preparations such as sodium phosphate (NaP) solutions have been associated with disturbances of electrolyte balance10 and kidney function11–13, sometimes with fatal consequences14–16. The safety and acceptability of bowel preparation is of particular importance when colonoscopy is used for routine screening programs of asymptomatic individuals. A modified 2 L PEG-based cleansing agent, consisting of PEG plus ascorbic acid, sodium ascorbate (ascorbate components), sodium sulfate and electrolytes (PEGþAsc; MOVIPREP) has become an established preparation worldwide17–19 with citations in a number of usage guideline recommendations20,21. A large prospective, randomized study in hospitalized patients undergoing elective colonoscopy showed that the 2 L PEGþAsc solution provides the same cleansing quality as the established 4 L PEG þ electrolytes (PEGþE) solution18. Therefore the inclusion of ascorbate components in the 2 L PEGþAsc formulation allows the volume of the cleansing solution to be reduced with no loss in efficacy. 2494
2L PEGþAsc versus sodium phosphate for bowel cleansing Ell et al.
We report the results from a randomized, controlled trial comparing PEGþAsc and NaP solution (Fleet Phospho-soda) in ambulatory subjects undergoing screening colonoscopy for colon cancer assessment.
Patients and methods Adults undergoing elective, outpatient colonoscopy for CRC screening at 12 centers in Germany were recruited to the study. All participants gave written, informed consent to take part, and the protocol was approved by the appropriate research ethics committees and the competent authorities.
Exclusion criteria The study was explained to all subjects scheduled to undergo a CRC screening colonoscopy at participating centers before noon. Consenting individuals were included unless they had: ileus, intestinal obstruction, toxic megacolon, history of colonic resection, heart failure, hypertension, liver failure or renal insufficiency (moderate or severe). Women were excluded from the study if they were breastfeeding, pregnant, or at risk of becoming pregnant. Patients on permanent medication requiring stable serum concentrations (e.g. neuroleptics) and those who, in the opinion of the investigator, might have been at risk from entering the trial were also excluded.
Bowel cleansing schedules and dietary restrictions Participants received either PEGþAsc (MOVIPREP, Norgine, Harefield, UK) as sachets containing 100 g macrogol 3350, 7.5 g sodium sulfate, 2.7 g sodium chloride, 1 g potassium chloride, 4.7 g ascorbic acid, 5.9 g sodium ascorbate and lemon flavoring to be dissolved in water to make a 1 L solution or the control drug, NaP (Fleet Phospho-soda, E.C. De Witt Ltd, Runcorn, UK) in bottles containing 24.4 g sodium hydrogen phosphate, 10.8 g disodium phosphate dodecahydrate and ginger–lemon flavoring in 45 mL liquid to be diluted with 120 mL water. Those receiving PEGþAsc drank 1 L of solution the evening before colonoscopy and 1 L on the morning of the procedure (at least 1 hour before colonoscopy), followed by at least 0.5 L of clear fluid each time. The NaP group took 45 mL NaP diluted with 120 mL cold water, followed by 240 mL of cold water the morning before the colonoscopy (after breakfast) and again on the evening before the procedure, as per the country manufacturer’s instructions for morning colonoscopy appointments at the time the study was conducted. Colonoscopies in both treatment arms were scheduled to start before 12.00 p.m. On the day before the colonoscopy, the PEGþAsc group were advised to have a normal www.cmrojournal.com ! 2014 Informa UK Ltd
Curr Med Res Opin Downloaded from informahealthcare.com by University of Missouri Kansas City UMKC on 03/31/15 For personal use only.
Current Medical Research & Opinion Volume 30, Number 12
breakfast, a light lunch, and liquid supper (e.g. clear soup or yoghurt). No further food was permitted until after the colonoscopy but clear fluids could be taken. In line with the approved dosing instructions, the NaP group were not permitted any solid food on the day before the colonoscopy. They were advised to have at least one glass of clear fluid for breakfast (e.g. black tea or coffee, fruit juice, soda, soup or water), at least three glasses of clear fluid for lunch and at least one glass of clear fluid in the evening. The NaP group were permitted clear fluids until midnight before the procedure. Subjects receiving NaP (90 mL with additional 240 mL water) had a total additional water intake of at least 1680 mL (240 mL before and after each 45 mL NaP dose diluted in 120 mL water plus 720 mL water for lunch on the day before colonoscopy).
Randomization and treatment allocation Participants were randomized in a 2:1 ratio to receive either PEGþAsc or NaP. Treatments were allocated using a computer-generated, randomized code list with a block size of 6. After obtaining informed consent and ensuring the individual was eligible to enter the study, the investigator assigned the next randomization number and dispensed the appropriate bowel preparation and instruction leaflet.
Assessments The primary outcome measure was the overall quality of colon cleansing assessed by an expert panel of three experienced gastroenterologists who reviewed recordings of the colonoscopies, unaware of the treatment allocation (i.e. blinded assessment). The degree of gut cleansing was also recorded by the endoscopist. As the endoscopists could not be fully blinded the primary outcome measure was based on the consensus assessment of the expert panel. The degree of cleansing was recorded according to the Harefield Cleansing Scale (HCS)18,22. Each bowel segment (rectum, sigmoid colon, descending colon, transverse colon and ascending colon) was assessed individually for 100% mucosal visualization. Each segment was rated using a 5 point scale; using following scores: 4 – colon empty and clean; 3 – presence of clear liquid in the gut; 2 – presence of brown liquid or small amounts of semisolid residual stool, fully removable by suction or displaceable, thus allowing a complete visualization of the underlying mucosa; 1 – presence of semisolid stool, only partially removable with a risk of incomplete underlying mucosal visualization; 0 – presence of semisolid or solid stool, colonoscopy incomplete or had to be stopped. The overall quality of colonic cleansing to achieve 100% mucosal visualization was based on the assessment of the individual segments using the following grading ! 2014 Informa UK Ltd www.cmrojournal.com
December 2014
scale – A: all segments clean (i.e. scores of 3 or 4 in all segments); B: residual brown liquid or easily removable semisolid stool (i.e. a score of 2) in at least one segment; C: partially removable stool preventing complete visualization of mucosa (i.e. a score of 1) in at least one segment; or D: at least one segment could not be examined due to the presence of solid stool (i.e. a score of 0). Grades A or B indicate successful cleansing with complete visualization of 100% of the colon mucosa, whereas Grades C or D indicate insufficient colonic preparation and therefore are regarded as cleansing failures. Participants completed a diary card recording the amount of study preparation taken together with their views on satisfaction with the gut cleansing procedure, the acceptability and taste of the preparation and their willingness to take the same preparation for another colonoscopy. Participants also reported any problems with drinking the solution and the impact of dietary restrictions. Blood was sampled before and after bowel cleansing (at the screening and colonoscopy visits) for hematological and biochemical analysis. Vital signs were monitored during colonoscopy. All adverse events were recorded.
Sample size and statistical analysis The required sample size was calculated to determine a difference in the rate of successful bowel cleansing of at least 15% using a randomization ratio of 2:1. This required 200 evaluable participants in the PEGþAsc group and 100 in the NaP group. The aim was therefore to recruit up to 360 participants. A one-sided Fisher’s exact test was used to compare the primary outcome variable between treatment groups. If the colonoscopy recordings (videotapes/DVDs) were not available, the ratings of the endoscopist who performed the procedure were used instead of those from the expert reviewers. If ratings for the degree of cleansing were missing for any segment, they were assumed to be ‘not clean’ (i.e. the worst case).
Results Of the 363 patients randomized, 356 drank at least some of the study medication and therefore comprised the ‘intention to treat’ (ITT) population for efficacy and safety analysis (Figure 1). Seven patients did not receive any study medication because of: a serious adverse event in the screening phase, administrative reasons (two cases), withdrawal of consent, withdrawal because of relative’s illness, and withdrawal for unspecified reasons (two cases). The per protocol (PP) population comprised 332 cases based on 2L PEGþAsc versus sodium phosphate for bowel cleansing Ell et al.
2495
Current Medical Research & Opinion Volume 30, Number 12
December 2014
(A)
Volume of cleansing solution ingested (in mL±SD)
Subjects Screened N=363
2000
1,998.10
1500 1000 500
No Study Medication N=7
326.05 0 (B)
110 100
Safety Population N=356
90
99.40
98.8
Compliance (in %±SD)
80
ITT Population N=356 PEG+Asc, N=242 NaP, N=114
PP Population N=332 PEG+Asc, N=226 NaP, N=106
60 50 40 20 10 0
(C)
4500 4000
Figure 1. CONSORT diagram showing patient disposition. Table 1. Demographic characteristics of study subjects undergoing screening colonoscopy for colon cancer. Participant characteristics
70
30
Protocol Violators N=24 PEG+Asc, N=16 NaP, N=8
PEGþAsc N ¼ 242
NaP N ¼ 114
All Subjects N ¼ 356
60.1 9.1 77.9 13.6
60.3 10.6 76.1 12.5
60.2 9.6 77.3 13.3
127 (52.5) 115 (47.5)
56 (49.1) 58 (50.9)
183 (51.4) 173 (48.6)
Volume of additional fluid intake drunk (in mL±SD)
Curr Med Res Opin Downloaded from informahealthcare.com by University of Missouri Kansas City UMKC on 03/31/15 For personal use only.
Subjects Randomized N=363
2500
3500 3000 2,900.48
2500 2000 1500
1,906.21
1000 500 0 PEG+Asc
Age (mean SD in years) Weight (mean SD in kg) Gender [N (%)] Female Male
PEGþAsc ¼ polyethylene glycol plus ascorbic acid, sodium ascorbate (ascorbate components), sodium sulfate and electrolytes; NaP ¼ sodium phosphate; SD ¼ standard deviation.
the criterion of absence of major protocol violations (N ¼ 24). The mean age was 60.2 9.6 (range 18–83) years, mean weight was 77.3 13.3 (range 46–120) kg, and 183 of the participants (51.4%) were women (Table 1). Compliance with bowel preparation was good, with 99% of participants in both treatment groups consuming at least 80% of the required cleansing solution (Figure 2).
Efficacy of gut cleansing Rates of successful gut cleansing (Grade A or B) were statistically different between the two groups 93.4% (95% CI 2496
2L PEGþAsc versus sodium phosphate for bowel cleansing Ell et al.
NaP
Figure 2. Participants’ compliance with bowel cleansing solutions comparing PEGþAsc with NaP. (A) Mean volume ingested. (B) Percentage of participants rated as compliant (i.e. consuming at least 80% of the required cleansing solution). (C) Mean volume of additional fluid ingested.
89.5 to 96.2) for the PEGþAsc group and 22.8% (95% CI 15.5 to 31.6%) for the NaP group according to the blinded expert panel (Figure 3) (p50.0001). According to the endoscopist’s assessment, the rates of successful cleansing were 95.0% (91.5 to 97.4%) for PEGþAsc and 45.6% (36.3 to 55.2%) for NaP (Fisher’s exact test p50.0001; Figure 3). These findings were unchanged when protocol violators or participants who had their colonoscopy after midday (N ¼ 34) were excluded from the analysis (Figure 3). The cleansing of all assessed colon segments was better in the subjects receiving PEGþAsc than for those taking NaP (Figure 4, A and B). The total difference in the mean segmental cleansing scores between the two treatment groups was statistically significant in favor of PEG þ Asc (p50.001). Furthermore, there was a www.cmrojournal.com ! 2014 Informa UK Ltd
Current Medical Research & Opinion Volume 30, Number 12
ITT Population p
Curr Med Res Opin Downloaded from informahealthcare.com by University of Missouri Kansas City UMKC on 03/31/15 For personal use only.
0300-7995 doi:10.1185/03007995.2014.960513
Vol. 30, No. 12, 2014, 2493–2503
Article ST-0173.R2/960513 All rights reserved: reproduction in whole or part not permitted
Original Article Randomized, controlled trial of 2 L polyethylene glycol plus ascorbate components versus sodium phosphate for bowel cleansing prior to colonoscopy for cancer screening
C. Ell Dr Horst Schmidt Kliniken GmbH, Klinik fu¨r Innere Medizin II, Wiesbaden, Germany
W. Fischbach Gastroenterology Department, Klinikum Aschaffenburg, Germany
P. Layer Abt. Innere Medizin, Israelitisches Krankenhaus, Hamburg, Germany
Abstract Background: Colonoscopy requires effective bowel preparation for adequate mucosal visualization. Safety and acceptability of bowel preparation are key components in colorectal cancer screening (CRC) populations. Objective: To compare the efficacy, safety and acceptability of bowel preparation with polyethylene glycol (PEG), ascorbic acid, sodium ascorbate (ascorbate components), sodium sulfate and electrolytes (PEGþAsc) or sodium phosphate (NaP).
M. Halphen Norgine Limited, Harefield, Uxbridge, Middlesex, UK Address for correspondence: Dr Marc Halphen, Norgine Limited, Norgine House, Widewater Place, Moorhall Road, Uxbridge, UB9 6NS, UK. Tel.: þ44 (0)1895 826600; [email protected] Key words: Ascorbate components – Ascorbic acid – Colon cancer screening – Colonoscopy – Polyethylene glycol – Sodium ascorbate – Sodium phosphate Accepted: 22 August 2014; published online: 23 September 2014 Citation: Curr Med Res Opin 2014; 30:2493–503
Methods: Consenting adults undergoing elective out-patient colonoscopy for CRC were randomized to take 2 L PEGþAsc or 90 mL NaP (control) following manufacturer’s instructions. PEGþAsc was taken the evening before and morning of the colonoscopy; NaP was taken the morning and evening before colonoscopy. Participants followed a restricted diet specific to each preparation. Primary endpoint was bowel cleansing success (100% colon mucosa visible) rated by an independent expert panel (all experienced endoscopists) unaware of treatment allocations. Subject reported outcomes about the preparations were elicited. Adverse events were recorded. Clinical trial registration: Clinicaltrials.gov: NCT00427089. Results: Successful bowel cleansing rate was significantly higher in the PEGþAsc (N ¼ 242) than the NaP (N ¼ 114) group (PEGþAsc 93.4% [95% CI 89.5–96.2] versus NaP 22.8% [15.5–31.6%], p50.0001). Subject reported outcomes on acceptability of the two different preparations were not significantly different (p ¼ 0.238). However, taste ratings for PEGþAsc were significantly better versus NaP (mean VAS: 31.2 and 38.1 respectively, p ¼ 0.0111). The proportion of patients prepared to receive the same preparation again was significantly higher in the PEGþAsc group (88.4% vs. 78.1%, p50.0001). Conclusions: PEGþAsc provided superior bowel cleansing to NaP and was well tolerated. Findings for PEGþAsc are aligned with previous similar studies; however, differences observed in NaP cleansing results, especially for the proximal colon segments, may be due to factors including: differences in demographics and population types and the use of the validated Harefield Cleansing Scale as an assessment tool combined with expert reviews, which may have resulted in conservative cleansing assessments.
! 2014 Informa UK Ltd www.cmrojournal.com
2L PEGþAsc versus sodium phosphate for bowel cleansing Ell et al.
2493
Current Medical Research & Opinion Volume 30, Number 12
December 2014
Curr Med Res Opin Downloaded from informahealthcare.com by University of Missouri Kansas City UMKC on 03/31/15 For personal use only.
Introduction Worldwide diagnosis of colorectal cancer (CRC) in 2012 was reported to be approximately 1.36 million new cases, representing around 10% of all reported cancer cases1. In Europe in 2012, the incidence of new cases of CRC was 446.8 thousand with 214.7 thousand deaths reported in the same year2. Population screening programs in Europe include the use of Fecal Occult Blood tests (FoBt), colonoscopies and to a lesser extent flexible sigmoidoscopy3. Mass screening with FoBt has been shown to result in a reduction of around 16% in the mortality from CRC and has been adopted in several European countries4. In addition, colonoscopy as a primary screening modality for CRC detection has been established in Germany since 20024. Recently, it has been shown that an increase in colonoscopy completion rate (as defined by ileocecal visualization) within a CRC screening population has been associated with a reduction in CRC associated mortality5. Reasons for failure to visualize the ileocecal valve during colonoscopy may include uncontrolled looping, patient discomfort and poor bowel preparation6. Recently it has been shown that several independent factors, including diabetes, chronic health conditions and patient self-activation may influence the quality of bowel preparation7. Any of these factors may affect the diagnostic sensitivity of screening colonoscopies which depend on complete mucosal visualization8. Consequently, effective and well tolerated bowel cleansing is essential before the procedure. There are a number of different bowel lavage preparations available. Traditional polyethylene glycol (PEG) preparations require consumption of a large volume of liquid (4 L) which many patients find unpleasant and some find impossible to drink9. Lower volume preparations such as sodium phosphate (NaP) solutions have been associated with disturbances of electrolyte balance10 and kidney function11–13, sometimes with fatal consequences14–16. The safety and acceptability of bowel preparation is of particular importance when colonoscopy is used for routine screening programs of asymptomatic individuals. A modified 2 L PEG-based cleansing agent, consisting of PEG plus ascorbic acid, sodium ascorbate (ascorbate components), sodium sulfate and electrolytes (PEGþAsc; MOVIPREP) has become an established preparation worldwide17–19 with citations in a number of usage guideline recommendations20,21. A large prospective, randomized study in hospitalized patients undergoing elective colonoscopy showed that the 2 L PEGþAsc solution provides the same cleansing quality as the established 4 L PEG þ electrolytes (PEGþE) solution18. Therefore the inclusion of ascorbate components in the 2 L PEGþAsc formulation allows the volume of the cleansing solution to be reduced with no loss in efficacy. 2494
2L PEGþAsc versus sodium phosphate for bowel cleansing Ell et al.
We report the results from a randomized, controlled trial comparing PEGþAsc and NaP solution (Fleet Phospho-soda) in ambulatory subjects undergoing screening colonoscopy for colon cancer assessment.
Patients and methods Adults undergoing elective, outpatient colonoscopy for CRC screening at 12 centers in Germany were recruited to the study. All participants gave written, informed consent to take part, and the protocol was approved by the appropriate research ethics committees and the competent authorities.
Exclusion criteria The study was explained to all subjects scheduled to undergo a CRC screening colonoscopy at participating centers before noon. Consenting individuals were included unless they had: ileus, intestinal obstruction, toxic megacolon, history of colonic resection, heart failure, hypertension, liver failure or renal insufficiency (moderate or severe). Women were excluded from the study if they were breastfeeding, pregnant, or at risk of becoming pregnant. Patients on permanent medication requiring stable serum concentrations (e.g. neuroleptics) and those who, in the opinion of the investigator, might have been at risk from entering the trial were also excluded.
Bowel cleansing schedules and dietary restrictions Participants received either PEGþAsc (MOVIPREP, Norgine, Harefield, UK) as sachets containing 100 g macrogol 3350, 7.5 g sodium sulfate, 2.7 g sodium chloride, 1 g potassium chloride, 4.7 g ascorbic acid, 5.9 g sodium ascorbate and lemon flavoring to be dissolved in water to make a 1 L solution or the control drug, NaP (Fleet Phospho-soda, E.C. De Witt Ltd, Runcorn, UK) in bottles containing 24.4 g sodium hydrogen phosphate, 10.8 g disodium phosphate dodecahydrate and ginger–lemon flavoring in 45 mL liquid to be diluted with 120 mL water. Those receiving PEGþAsc drank 1 L of solution the evening before colonoscopy and 1 L on the morning of the procedure (at least 1 hour before colonoscopy), followed by at least 0.5 L of clear fluid each time. The NaP group took 45 mL NaP diluted with 120 mL cold water, followed by 240 mL of cold water the morning before the colonoscopy (after breakfast) and again on the evening before the procedure, as per the country manufacturer’s instructions for morning colonoscopy appointments at the time the study was conducted. Colonoscopies in both treatment arms were scheduled to start before 12.00 p.m. On the day before the colonoscopy, the PEGþAsc group were advised to have a normal www.cmrojournal.com ! 2014 Informa UK Ltd
Curr Med Res Opin Downloaded from informahealthcare.com by University of Missouri Kansas City UMKC on 03/31/15 For personal use only.
Current Medical Research & Opinion Volume 30, Number 12
breakfast, a light lunch, and liquid supper (e.g. clear soup or yoghurt). No further food was permitted until after the colonoscopy but clear fluids could be taken. In line with the approved dosing instructions, the NaP group were not permitted any solid food on the day before the colonoscopy. They were advised to have at least one glass of clear fluid for breakfast (e.g. black tea or coffee, fruit juice, soda, soup or water), at least three glasses of clear fluid for lunch and at least one glass of clear fluid in the evening. The NaP group were permitted clear fluids until midnight before the procedure. Subjects receiving NaP (90 mL with additional 240 mL water) had a total additional water intake of at least 1680 mL (240 mL before and after each 45 mL NaP dose diluted in 120 mL water plus 720 mL water for lunch on the day before colonoscopy).
Randomization and treatment allocation Participants were randomized in a 2:1 ratio to receive either PEGþAsc or NaP. Treatments were allocated using a computer-generated, randomized code list with a block size of 6. After obtaining informed consent and ensuring the individual was eligible to enter the study, the investigator assigned the next randomization number and dispensed the appropriate bowel preparation and instruction leaflet.
Assessments The primary outcome measure was the overall quality of colon cleansing assessed by an expert panel of three experienced gastroenterologists who reviewed recordings of the colonoscopies, unaware of the treatment allocation (i.e. blinded assessment). The degree of gut cleansing was also recorded by the endoscopist. As the endoscopists could not be fully blinded the primary outcome measure was based on the consensus assessment of the expert panel. The degree of cleansing was recorded according to the Harefield Cleansing Scale (HCS)18,22. Each bowel segment (rectum, sigmoid colon, descending colon, transverse colon and ascending colon) was assessed individually for 100% mucosal visualization. Each segment was rated using a 5 point scale; using following scores: 4 – colon empty and clean; 3 – presence of clear liquid in the gut; 2 – presence of brown liquid or small amounts of semisolid residual stool, fully removable by suction or displaceable, thus allowing a complete visualization of the underlying mucosa; 1 – presence of semisolid stool, only partially removable with a risk of incomplete underlying mucosal visualization; 0 – presence of semisolid or solid stool, colonoscopy incomplete or had to be stopped. The overall quality of colonic cleansing to achieve 100% mucosal visualization was based on the assessment of the individual segments using the following grading ! 2014 Informa UK Ltd www.cmrojournal.com
December 2014
scale – A: all segments clean (i.e. scores of 3 or 4 in all segments); B: residual brown liquid or easily removable semisolid stool (i.e. a score of 2) in at least one segment; C: partially removable stool preventing complete visualization of mucosa (i.e. a score of 1) in at least one segment; or D: at least one segment could not be examined due to the presence of solid stool (i.e. a score of 0). Grades A or B indicate successful cleansing with complete visualization of 100% of the colon mucosa, whereas Grades C or D indicate insufficient colonic preparation and therefore are regarded as cleansing failures. Participants completed a diary card recording the amount of study preparation taken together with their views on satisfaction with the gut cleansing procedure, the acceptability and taste of the preparation and their willingness to take the same preparation for another colonoscopy. Participants also reported any problems with drinking the solution and the impact of dietary restrictions. Blood was sampled before and after bowel cleansing (at the screening and colonoscopy visits) for hematological and biochemical analysis. Vital signs were monitored during colonoscopy. All adverse events were recorded.
Sample size and statistical analysis The required sample size was calculated to determine a difference in the rate of successful bowel cleansing of at least 15% using a randomization ratio of 2:1. This required 200 evaluable participants in the PEGþAsc group and 100 in the NaP group. The aim was therefore to recruit up to 360 participants. A one-sided Fisher’s exact test was used to compare the primary outcome variable between treatment groups. If the colonoscopy recordings (videotapes/DVDs) were not available, the ratings of the endoscopist who performed the procedure were used instead of those from the expert reviewers. If ratings for the degree of cleansing were missing for any segment, they were assumed to be ‘not clean’ (i.e. the worst case).
Results Of the 363 patients randomized, 356 drank at least some of the study medication and therefore comprised the ‘intention to treat’ (ITT) population for efficacy and safety analysis (Figure 1). Seven patients did not receive any study medication because of: a serious adverse event in the screening phase, administrative reasons (two cases), withdrawal of consent, withdrawal because of relative’s illness, and withdrawal for unspecified reasons (two cases). The per protocol (PP) population comprised 332 cases based on 2L PEGþAsc versus sodium phosphate for bowel cleansing Ell et al.
2495
Current Medical Research & Opinion Volume 30, Number 12
December 2014
(A)
Volume of cleansing solution ingested (in mL±SD)
Subjects Screened N=363
2000
1,998.10
1500 1000 500
No Study Medication N=7
326.05 0 (B)
110 100
Safety Population N=356
90
99.40
98.8
Compliance (in %±SD)
80
ITT Population N=356 PEG+Asc, N=242 NaP, N=114
PP Population N=332 PEG+Asc, N=226 NaP, N=106
60 50 40 20 10 0
(C)
4500 4000
Figure 1. CONSORT diagram showing patient disposition. Table 1. Demographic characteristics of study subjects undergoing screening colonoscopy for colon cancer. Participant characteristics
70
30
Protocol Violators N=24 PEG+Asc, N=16 NaP, N=8
PEGþAsc N ¼ 242
NaP N ¼ 114
All Subjects N ¼ 356
60.1 9.1 77.9 13.6
60.3 10.6 76.1 12.5
60.2 9.6 77.3 13.3
127 (52.5) 115 (47.5)
56 (49.1) 58 (50.9)
183 (51.4) 173 (48.6)
Volume of additional fluid intake drunk (in mL±SD)
Curr Med Res Opin Downloaded from informahealthcare.com by University of Missouri Kansas City UMKC on 03/31/15 For personal use only.
Subjects Randomized N=363
2500
3500 3000 2,900.48
2500 2000 1500
1,906.21
1000 500 0 PEG+Asc
Age (mean SD in years) Weight (mean SD in kg) Gender [N (%)] Female Male
PEGþAsc ¼ polyethylene glycol plus ascorbic acid, sodium ascorbate (ascorbate components), sodium sulfate and electrolytes; NaP ¼ sodium phosphate; SD ¼ standard deviation.
the criterion of absence of major protocol violations (N ¼ 24). The mean age was 60.2 9.6 (range 18–83) years, mean weight was 77.3 13.3 (range 46–120) kg, and 183 of the participants (51.4%) were women (Table 1). Compliance with bowel preparation was good, with 99% of participants in both treatment groups consuming at least 80% of the required cleansing solution (Figure 2).
Efficacy of gut cleansing Rates of successful gut cleansing (Grade A or B) were statistically different between the two groups 93.4% (95% CI 2496
2L PEGþAsc versus sodium phosphate for bowel cleansing Ell et al.
NaP
Figure 2. Participants’ compliance with bowel cleansing solutions comparing PEGþAsc with NaP. (A) Mean volume ingested. (B) Percentage of participants rated as compliant (i.e. consuming at least 80% of the required cleansing solution). (C) Mean volume of additional fluid ingested.
89.5 to 96.2) for the PEGþAsc group and 22.8% (95% CI 15.5 to 31.6%) for the NaP group according to the blinded expert panel (Figure 3) (p50.0001). According to the endoscopist’s assessment, the rates of successful cleansing were 95.0% (91.5 to 97.4%) for PEGþAsc and 45.6% (36.3 to 55.2%) for NaP (Fisher’s exact test p50.0001; Figure 3). These findings were unchanged when protocol violators or participants who had their colonoscopy after midday (N ¼ 34) were excluded from the analysis (Figure 3). The cleansing of all assessed colon segments was better in the subjects receiving PEGþAsc than for those taking NaP (Figure 4, A and B). The total difference in the mean segmental cleansing scores between the two treatment groups was statistically significant in favor of PEG þ Asc (p50.001). Furthermore, there was a www.cmrojournal.com ! 2014 Informa UK Ltd
Current Medical Research & Opinion Volume 30, Number 12
ITT Population p
