Researches

Inulin-Type Prebiotics

A Review (Part 2)
Greg Kelly, ND

Abstract

This article is part 2 of a two-part review of inulin-type prebiotics. In part 1, inulin-type prebiotics were defined and food applications were explored. Evidence of ability to modulate gut microflora was also examined. Part 2 discusses the clinical research on inulin-type prebiotics. Prebiotics are a category of nutritional compounds grouped together based on ability to promote growth of specific beneficial (probiotic) gut bacteria. According to Roberfroid’s definition, a prebiotic is “a selectively fermented ingredient that allows specific changes, both in the composition and/or activity of the gastrointestinal microflora that confers benefits upon host well-being and health.”1 Many dietary fibers, especially soluble fibers, exhibit some prebiotic activity; however, Roberfroid only identifies two groupings of nutritional compounds that meet his definition. These two groupings or sub-categories can be described as inulin-type prebiotics and galactooligosaccharides (GOS).1 Part 1 focused on inulin-type prebiotics, which include fructooligosaccharides (FOS), oligofructose, and inulin. These inulin-type prebiotics are oligo- or polysaccharide chains comprised primarily of linked fructose molecules. They are considered to be bifidogenic (stimulating the growth of Bifidobacteria species). This grouping of prebiotics was selected for review because they represent the most widely commercially available and the most researched prebiotic compounds. In prebiotic clinical studies, inulin-type prebiotics have been studied as an isolated intervention or combined with other types of prebiotics (primarily GOS) as part of “prebiotic mixtures.” This review focuses on the former; however, a summary of mixed prebiotic research is provided.

Nomenclature

Within the category of inulin-type prebiotics, there are no uniformly accepted and used standards for nomenclature. To provide consistency, the nomenclature outlined in Table 1 will be used within this review. In instances where authors of a research paper specify what they used, the nomenclature outlined in Table 1 will be used even if the authors of the original paper used different terminology. Table 2 reviews the abbreviations originally defined in part 1.

Table 1.Prebiotic Nomenclature

Mixed Prebiotics Overview and Research Summary

The clinical research that has investigated “prebiotic mixtures” has primarily used inulin-type prebiotics combined with galactooligosaccharide prebiotics. GOS, which are oligo- or polysaccharide chains comprised primarily of linked galactose units, are one of the two categories of compounds that meet Roberfroid’s definition of a prebiotic.1 Similar to inulin-type prebiotics, GOS are selectively fermented in the colon and promote specific changes in the composition of the gastrointestinal microflora. Specifically, they stimulate the growth of Bifidobacteria and Lactobacilli species.

Table 2. Degree of Polymerization: Definitions

A specific GOS/inulin-type prebiotic mixture has been extensively investigated. It is comprised of short-chain GOS combined with long-chain inulin (inulin HP) in a 9:1 ratio. This mixture has been researched for infant nutrition applications and is typically added to standard formulas in these trials. The GOS/inulin HP mixture was designed to more closely mimic the oligosaccharide portfolios found in human breast milk than inulin-type prebiotics alone, since the oligosaccharides in breast milk contain relatively high amounts of galactose polymers.6 Studies in preterm and term infants have shown a formula supplemented with this GOS/inulin HP prebiotic mixture results in an intestinal microbiota similar to that found in breast-fed infants. In infants, this GOS/inulin HP combination has been reported to: (1) reduce the incidence of atopic disease, recurrent wheezing, and allergic urticaria in infants with a parental history of atopic disease;5,8,9 (2) result in fewer episodes of upper respiratory tract infections, fever, and antibiotic prescriptions;8,10,11 (3) reduce the incidence of acute diarrhea;11 (4) improve stool consistency, increase stool frequency, and accelerate gastrointestinal transport time;3,12-14 (5) increase fecal secretory IgA;15,16 and (6) significantly lower bilirubin levels during the first 72 hours of life in formula-fed infants.17

Table 3. Summary of Inulin HP Research Findings

Since GOS has prebiotic activity in isolation, and hence presumably some ability to influence physiological responses to supplementation, it is not possible to determine which aspects of the clinical response in the infant nutritional studies of the GOS/inulin HP prebiotic mixture are due to inulin-type prebiotics and which are due to the GOS component of the prebiotic.

Therapeutic Uses of Inulin-type Prebiotics

This section discusses clinical research on inulin-type prebiotics used as stand-alone clinical interventions.

• Blood Sugar Regulation

Several investigators have attempted to determine whether supplementing the diet with inulin-type prebiotics has an effect on blood sugar regulation. Studies in both normoglycemic and hyperglycemic subjects have largely reported no statistically observable differences between inulin-type prebiotics and placebo interventions.

van Dokkum et al recruited 12 healthy male subjects for a trial lasting 84 days during which time the participants received a constant and controlled diet. The trial was divided into four distinct experimental periods in which the same diet was supplemented with: (1) 15 g/day inulin, (2) 15 g/day FOS, (3) 15 g/day GOS, or (4) no added prebiotic. No statistically significant differences in glucose tolerance tests and insulin responses to these tests were observed among four experimental periods.

Letexier et al gave inulin HP or placebo (maltodextrin) to eight healthy subjects in a double-blind, randomized, placebo-controlled (RCT) crossover study for six weeks. The daily dose and duration was 10 g given in two divided doses at breakfast and dinner. No statistically significant changes in plasma glucose or insulin levels were observed.19 Table 3 summarizes this and other inulin HP research.

Boutron-Ruault et al conducted an open multicenter pilot study of adenoma and adenoma-free subjects. All participants received 10 g FOS daily in two divided doses for three months; 74 subjects completed the study. No statistically significant differences were observed in blood glucose or insulin following supplementation with FOS.

Giacco et al conducted a crossover RCT in 30 subjects with mild hyperlipidemia. Participants received either 10.6 g FOS or 15 g placebo (maltodextrin plus aspartame) as powder in packets and were instructed to add it daily to coffee or tea for two months, after which time they were switched to the other intervention for an additional two months. No statistically significant differences were observed in fasting blood glucose or insulin levels. Although postprandial plasma glucose levels were similar after FOS and placebo at every time point, postprandial insulin response was significantly reduced after FOS compared to placebo.

Daubioul et al investigated the effects of 16 g oligofructose daily on blood glucose and insulin in seven subjects with biopsy-confirmed nonalcoholic steatohepatitis (NASH) – (methodology described under section on fatty liver disease). No statistically significant differences were observed for glucose or insulin levels during the oligofructose versus placebo periods.

Luo et al gave 12 healthy, normoglycemic males 20 g FOS or placebo (sucrose) daily for two four-week periods, separated by a two-week washout interval. In this crossover RCT, the intervention was incorporated into cookies that were eaten throughout the day to reach the target intervention levels of active and placebo compounds. No statistically significant differences were observed for insulin-stimulated glucose metabolism. A decrease in basal hepatic glucose production was observed in the group receiving FOS

Luo et al subsequently investigated the effects of FOS in 12 persons with type 2 diabetes. Subjects were maintained on their existing medical therapies (10 on oral hypoglycemic agents and two on antidiabetic dietary management) throughout the study. The study design was identical to that described above as was the daily dose of FOS or placebo (sucrose). The FOS or sucrose was supplied as powder in packets and participants were instructed to add these as sweeteners to food or beverages several times daily to complete a daily dose of 20 g; 10 subjects completed the study period. No statistically significant differences were observed for fasting plasma glucose concentrations. Levels of hemoglobin A1c and fructosamine (markers of long-term blood sugar control) also did not differ during the active and placebo periods. No significant differences were observed for fasting insulin, insulin binding to erythrocytes, and insulin-mediated glucose disposal for the two interventions. Basal hepatic glucose production also did not differ during the FOS and sucrose periods.

Yamashita et al conducted a 14-day uncontrolled study on poorly controlled type 2 diabetics with high glucose levels. Subjects received 8 g FOS daily for 14 days. The intervention resulted in an eight-percent decrease in fasting blood glucose concentrations.

Alles et al investigated the effects of oligofructose on blood glucose in 20 patients with type 2 diabetes. Seventeen of the participants were using oral hypoglycemic medications to assist with blood sugar control. The study was a randomized, single-blind, crossover design with subjects consuming either a placebo (4 g/day glucose) or oligofructose (15 g/day) for 20 days in two divided doses. The dose of oligofructose was gradually increased during the first three days by 5 g/day in an attempt to prevent gastrointestinal side effects. Compared to placebo, oligofructose did not significantly affect fasting blood glucose levels.

Existing evidence, taken as a whole, suggests supplementation with inulin-type prebiotics is unlikely to have a positive effect on blood sugar in normoglycemic individuals or have a clinically relevant benefit in terms of improving metabolic control in hyperglycemic individuals. Table 4 summarizes the clinical studies of FOS for blood sugar management and other conditions.

• Colon Cancer

In multiple experimental animal models inulin-type prebiotics have demonstrated anticarcinogenic properties. In studies that have examined the effects of inulin-type prebiotics on chemically induced pre-neoplastic lesions or tumors in the colon of rats and mice, the most consistent findings have been reductions in aberrant crypt foci, tumor incidence, and metastasis of implanted tumor cells.27 Only a single preliminary study in humans exists.

Boutron-Ruault et al conducted an open multicenter pilot study of colon adenoma and adenoma-free subjects. All participants received 5 g FOS twice daily for three months; 74 subjects completed the study. Cell proliferation at the rectal crypts was not significantly modified by FOS ingestion in either group

While the results observed in animal studies have been promising, human intervention evidence is very limited and currently not very compelling. The question of whether inulin-type prebiotics have an ability to prevent or alter the clinical progression of colon cancer in humans is unanswered and requires future research.

• Fatty Liver Disease

One study has been conducted of the effects of inulin-type prebiotics on the clinical course of fatty liver disease. A beneficial effect was observed on serum aspartate aminotransferase (AST) but not on serum alanine aminotransferase (ALT).

Daubioul et al investigated the effects of oligofructose on liver enzymes in seven subjects with biopsy-confirmed NASH. Each subject received 16 g oligofructose or placebo (maltodextrin) daily in a crossover RCT. Observations were made over two eight-week periods of supplementation with the active or placebo treatments and separated by a washout period of minimum five weeks. Oligofructose decreased AST after eight weeks; however, no statistically significant differences were observed for ALT. Analysis of liver ultrasonography did not reveal any difference in liver size.

While larger and longer studies might produce more positive results, based on this one small study it does not appear inulin-type prebiotics alter the course of fatty liver disease in a clinically significant manner.

Table 4. Summary of FOS Research Findings

• Gastrointestinal Health

Because inulin-type prebiotics are thought to exert therapeutic effect in the gastrointestinal tract, they have been investigated for a wide variety of gastrointestinal conditions.

Bowel Transit, Stool Consistency, and Stool Frequency

Several studies have investigated the effects of inulin-type prebiotics on aspects of bowel transit time, stool consistency, or stool frequency in infants and adults. More research is required before definitive conclusions can be reached; however, existing evidence supports an effect in infants but does not support a significant effect in adults who have no pre-existing functional gastrointestinal problems.

Kapiki et al conducted an RCT in preterm infants. Starting within the first two weeks after birth, 56 bottle-fed preterm infants were randomized to receive a formula supplemented with oligofructose or maltodextrin for 14 days. Supplementation with oligofructose increased daily stool frequency. Stool consistency also differed between the two groups, with infants fed the oligofructose having softer stools

In an RCT of 56 healthy children ages 16-46 weeks, Moore et al supplemented cereals with either FOS or placebo for 28 days. Average daily FOS consumption was 0.74 g/day. FOS supplementation resultedin softer stools and significant increase in the mean number of stools per day compared with placebo. There were also fewer days with skipped stools in the infant group receiving FOS.

In a crossover RCT, Euler et al gave formula-fed, full-term infants (ages 2-6 weeks) a cow’s milk formula with or without oligofructose at a dose of both 1.5 and 3 g/liter; washout weeks preceded and followed a week of oligofructose-supplemented formula feeding. The frequency of stools was greater and the consistency of stools was looser during oligofructose supplementation. These observed differences were dose-dependent, with the higher dose of oligofructose producing a greater statistically significant effect.

In the study by van Dokkum et al (previously discussed in the blood sugar regulation section) aspects of general bowel function were assessed. In this study, 12 healthy male subjects had a control diet supplementedwith: (1) 15 g/day inulin, (2) 15 g/day FOS, (3) 15 g/day GOS, or (4) no added prebiotic. Fecal wet weight, dry weight, and bowel transit times were similar in all four study periods. After log transformation, periods of inulin and GOS supplementation produced statistically significantly greater fecal wet weight than what was observed during FOS supplementation.

Bouhnik et al gave six elderly men and women 8 g FOS daily in two divided doses for four weeks. The subjects did not have any history of gastrointestinal issues. Stool weight, water percentage, and bowel transit time were unchanged following prebiotic supplementation.

Constipation

In subjects with existing constipation, supplementation with inulin-type prebiotics might have a clinical role; however, placebo-controlled trials should be conducted before any definitive conclusions can be drawn.

Kleessen et al gave inulin to a group of 10 elderly females with constipation – defined as one or two bowel movements per week and hard stool consistency. The women received inulin for 19 days. An initial dose of 20 g daily was given from days 1 to 8 and the dose gradually increased to 40 g daily from days 9 to 11. Daily inulin supplementation remained at 40 g from days 12 to 19. Stool consistency was assessed as being softer with inulin supplementation. Inulin supplementation increased stool frequency in 7 of 10 patients from pre-intervention of 1-2 bowel movements per week to 8-9 bowel movements per week. This occurred independent of the total daily amount of inulin ingested. In the other three subjects, the results were mixed with responses in two cases appearing to differ based on the dose of inulin. One woman had an increase in stool frequency to five per week at a dose of 20 g inulin daily and an increase to seven bowel movements per week at the 40-g dose. Another woman had a slightly poorer outcome at the 40-g dose (four bowel movements weekly) compared with the 20-g daily dose (five bowel movements per week). The last woman had no distinct change in stool frequency following supplementation with inulin, but the stool was softer.

Diarrhea Prevention

One study administered inulin-type prebiotics to infants to determine whether prebiotics might help prevent occurrences of acute diarrhea. The findings do not suggest a clinical role for reducing the incidence or severity of diarrhea in infants.

In an RCT Duggan et al compared the effect of cereal with or without added oligofructose (0.55 g/15 g cereal) on diarrhea incidence and severity in 282 infants, ages 6-12 months, for six months. The mean number of days with diarrhea was 10.3 in the non-supplemented cereal group and 9.8 in the oligofructose-supplemented cereal group, a difference not statistically significant.

Inflammatory Bowel Disease

Inulin-type prebiotics have shown therapeutic benefits in animal models of colitis.34 The human studies on inflammatory bowel disease (IBD) have included small numbers of patients and interventions have been for only 2-3 weeks. All three studies reported some degree of functional change subsequent to supplementation with inulin-type prebiotics that suggest the potential to modulate inflammation in subjects with active IBD. Changes in disease activity have been mixed in persons with active disease. No trials have been conducted to determine whether chronic supplementation with inulin-type prebiotics might help prevent disease reoccurrence or sustain periods of clinical remission.

Casellas et al conducted an RCT of inulin-type prebiotics in patients with acute ulcerative colitis. Participants had been previously in remission and presented with a relapse of mild-to-moderate severity. Subjects were treated with the conventional medication mesalazine (at a dose of 3 g/day) and randomly allocatedto receive either 12 g oligofructose-enriched inulin HP (in three divided doses) or placebo (maltodextrin) daily for two weeks; 15 subjects completed the trial. All participants except one (in the placebo group) demonstrated a decline in disease activity, assessed using the Rachmilewitz Index. Based on this index, all subjects in the active and six of eight subjects in the placebo group were considered to be in remission at the end of the two-week intervention period, demonstrating no significant difference in disease activity reduction scores between groups. Scores on a self-assessed dyspeptic symptoms questionnaire decreased significantly with active treatment but not with placebo, suggesting at least a reduction of perceived dyspeptic symptoms with prebiotic supplementation. Concentrations of inflammatory markers interleukin-8 (IL-8) and prostaglandin-2 (PGE-2) in rectal fluid were measured before and after the intervention period. No differences were observed between groups. Concentrations of fecal calprotectin were assessed. Calprotectin is a calcium-binding protein found in neutrophilic granulocytes, which is used as a biomarker of intestinal inflammation and as a predictor of relapse in subjects with IBD. Although it was elevated in both groups at the start of the intervention period, a significant reduction in this marker was observed at the end of the study in the active group but not in the placebo group.

Lindsay et al gave 10 patients with active ileocolonic or colonic Crohn’s disease 15 g oligofructose-enriched inulin daily for three weeks. Supplementation resulted in a statistically significant reduction in the Harvey Bradshaw Index (HBI) from 9.8 to 6.9. Four patients were categorized as achieving clinical remission based on reductions in HBI (with clinical remission defined as a fall in HBI to 5 or less). Mean Crohn’s Disease Activity Index (CDAI) fell from 250.9 to 220.6, although this change was not statistically significant. The researchers used HBI rather than CDAI as the disease activity index, since the HBI places less weight on subjective criteria such as abdominal symptoms and supplementation was expected to and did induce an increase in the severity of flatulence. A statistically significant increase in the percentage of anti-inflammatory interleukin-10 positive CD11c+ dendritic cells was observed following supplementation. Supplementation with the oligofructose-enriched inulin also resulted in statistically significant increases in toll-like receptor 2 (TLR2) and TLR4 expression. These changes suggest some degree of immunomodulatory activity and possibly the initiation of cytoprotective mechanisms in colonic cells.

Welters et al gave 20 patients with an ileal pouch-anal anastomosis 24 g inulin (in two divided doses) or placebo daily for three weeks in a crossover RCT. There were no differences in observed clinical symptoms between the inulin and placebo periods. However, compared with the placebo, inulin supplementation did increase butyrate concentrations (beneficial fatty acid and the main fuel of colonocytes), lower colonic pH, decrease numbers of Bacteroides fragilis, and diminish concentrations of secondary bile acids in feces. These changes were accompanied by endoscopically and histologically verified reductions in inflammation of the mucosa of the ileal reservoir.

Table 5. Summary of Oligofructose Research Findings

Given the small numbers of subjects in the studies conducted to date, and the mixed results reported, it is not possible to determine whether inulin-type prebiotics can reduce disease activity in individuals with acute inflammation. Larger trials are required before definitive conclusions can be drawn. Whether inulin-type prebiotics would have a role in keeping patients with IBD in remission between bouts of active disease has not been explored.

Irritable Bowel Syndrome

Two studies have investigated whether inulin-type prebiotics have a role in irritable bowel syndrome (IBS). One study observed no statistical benefit while the other reported an improvement in symptom scores following supplementation with an inulin-type prebiotic.

Olesen et al conducted a parallel group RCT. The study consisted of a two-week, single-blind run-in phase and a 12-week, double-blind comparative phase. Participants were randomly assigned to receive oligofructose or a placebo (dried and powdered glucose syrup). The dose used was 10 g/day for the first two weeks and 20 g/day for the following 10 weeks. Seventy-five patients (38 in the oligofructose group and 37 in the placebo group) completed the trial. In terms of clinical changes (as assessed by changes in symptom scores), no significant differences were observed between the two groups after 12 weeks of supplementation. IBS symptoms improved in 58 percent and worsened in eight percent of the participants receiving oligofructose, while among participants receiving the placebo 65 percent improved and 13 percent worsened. A statistically significant increase in defecation frequency was observed in participants receiving the oligofructose compared to placebo at weeks 4 and 6; however, no statistically significant difference persisted by the end of the 12-week intervention

Paineau et al investigated the use of FOS in 105 patients diagnosed by Rome II criteria with a minor degree of functional bowel disorder (FBD). As defined in the article, FBD includes symptoms of abdominal bloating, rumbling, transit disorders (occasional constipation and/or diarrhea, possibly alternating), abdominal pains, and flatulence. Thus, FBD is symptomatically the same as IBS. Participants were randomized to receive either 5 g FOS or placebo (50% sucrose and 50% maltodextrins) daily in divided doses for six weeks; 97 participants completed the study. At baseline, overall scores on the Functional Digestive Disorders Quality of Life questionnaire were similar between groups. Following six weeks of supplementation a statistically significant decrease in symptom scores was observed in the group taking FOS compared to placebo.

The response to inulin-type prebiotics in persons with IBS-like symptoms has been mixed. This might be due to the differences in what was supplemented (FOS in one case and oligofructose in the other), differences in the length of the supplementation periods, or other factors. Additional research might determine whether inulin-type prebiotics have any clinical role in managing IBS symptoms; however, since, in the longer study, some of the benefits observed early (an increase in stool frequency at 4 and 6 weeks) did not appear to persist, future studies should be of sufficient duration to determine occurrence and persistence of benefits.

Table 5. summarizes the clinical research of oligofructose for gastrointestinal health and other clinical situations.

• Immunity

Type prebiotics in infants. In this RCT, Firmansyah et al supplemented the diet of 50 infants (ages 7-9 months) with oligofructose-enriched inulin or placebo for four weeks prior to measles vaccination. The average daily intake of the inulin-type prebiotic was 0.2 g/kg body weight. A significantly greater rise in post-vaccination antimeasles IgG was observed in infants supplemented with the oligofructose-enriched inulin.

While this study suggests the potential for inulin-type prebiotics to influence aspects of immune response in infants, more research is required to determine to what extent supplementation influences other populations and other aspects of immunity.

• Lipids

The effect of inulin-type prebiotics on lipid levels has been investigated in a variety of studies. In individuals with normal lipid levels, the most consistent observation is that inulin-type prebiotics have no statistically significant effect on lipid levels. In persons with elevated lipid levels findings have been mixed, with some studies reporting improvements in lipid levels subsequent to supplementation and other studies reporting no effect of supplementation.

Luo et al reported that 20 g FOS daily had no effect on lipids in healthy, normoglycemic males (methodology described in the section on blood sugarregulation).23 Similarly, in the study van Dokkum et al conducted for 84 days in 12 healthy male subjects (methodology described in the section on blood sugar regulation) neither inulin nor FOS produced statistically significant differences in total cholesterol, LDL-cholesterol, total HDL-cholesterol, HDL-2 and HDL-3 concentrations, apolipoprotein A-1 and B, or triglycerides.

In the study by Letexier et al in which eight healthy subjects were given 10 g inulin HP or placebo daily (methodology described in the section on blood sugar regulation), total cholesterol, HDL-cholesterol, and LDL-cholesterol concentrations did not change significantly. However, a statistically significant decrease in plasma triglyceride concentrations was observed after inulin HP ingestion.

Pedersen et al also studied the effect of 14 g inulin daily on blood lipids in 64 healthy, normolipidemic females in a crossover RCT. Inulin produced no statistically significant changes in total cholesterol, HDL-cholesterol, LDL-cholesterol, or triglycerides.

In the study by Olesen et al (methodology described in the section on IBS) of individuals with IBS, lipids were also assessed. Oligofructose produced no significant changes compared to placebo in these normolipidemic individuals.

In the study by Boutron-Ruault et al (described in the colon cancer section) lipid levels were assessed before and after three months of FOS supplementation. No statistically significant differences were observed in total cholesterol or HDL-cholesterol following supplementation with FOS in these normolipidemic individuals.

Daubioul et al investigated the effect of 16 g oligofructose daily on blood glucose, insulin, and lipids in seven subjects with biopsy confirmed NASH (methodology described in fatty liver disease section). As a group, baseline total cholesterol was below 200 mg/dL and triglycerides values were in the normal range. No statistically significant differences were observed for total cholesterol, HDL- or LDL-cholesterol, or triglycerides between the oligofructose and placebo periods.

Yamashita et al conducted an uncontrolled 14-day clinical study in type 2 diabetics with elevated lipid levels. FOS was given at a daily dose of 8 g. A six-percent decrease in total cholesterol and a 10-percent decrease in LDL-cholesterol were reported.

Balcazar-Munoz et al conducted an RCT in 12 obese individuals with high triglycerides and cholesterol. Subjects were randomized to receive 7 g/day inulin or placebo for four weeks. Inulin administration produced a significant reduction in total cholesterol (248.7 to 194.3 mg/dL), LDL-cholesterol (136.0 to 113.0 mg/dL), VLDL-cholesterol (45.9 to 31.6 mg/dL), and triglyceride (235.5 to 171.1 mg/dL) concentrations.

Alles et al investigated the effects of 15 g/day oligofructose on lipids in 20 patients with type 2 diabetes (methodology described in the section on blood sugar regulation). Subjects had mild hyperlipidemia, with baseline total cholesterol of 216 mg/dL for men and 242 mg/dL for women. Baseline serum triglycerides were 235 mg/dL for men and 261 mg/dL for women. Compared to placebo, oligofructose had no significant effect on total cholesterol, HDL-cholesterol, LDL-cholesterol, or triglycerides.

Giacco et al conducted a crossover RCT in 30 subjects with mild hyperlipidemia (methodology described in the section on blood sugar regulation). To be eligible for inclusion subjects had to have plasma cholesterol greater than 200 and less than 300 mg/dL and plasma triglycerides less than 305 mg/dL. A daily dose of 10.6 g FOS or placebo (maltodextrin plus aspartame) was given for two months. No statistically significant differences were observed in fasting total cholesterol, VLDL-cholesterol, LDL-cholesterol, HDL-cholesterol, apolipoprotein A-1, or triglycerides. A statistically significant increase in fasting plasma lipoprotein(a) (a cardiac risk factor) was observed with FOS (an increase from 33 to 37 mg/dL).

Luo et al investigated the effect on lipids of 20 g FOS or placebo daily in 10 subjects with type 2 diabetes (methodology described in the section on blood sugar regulation). At baseline average triglyceride levels, total cholesterol, and HDL-cholesterol were 133, 207, and 38 mg/dL, respectively. No statistically significant differences were observed subsequent to the interventions for fasting levels of triglycerides, total or HDL-cholesterol, apolipoproteins A-1 or B, or lipoprotein(a).

In normolipidemic subjects, inulin-type prebiotics do not appear to positively modify lipid levels. The one positive reported result was a reduction in triglycerides with inulin HP (studies with FOS, oligofructose, and inulin demonstrated no statistically significant changes). More research is required before conclusions can be drawn based on this preliminary finding; however, it is theoretically possible a long-chain inulin-type prebiotic (inulin HP) might produce different responses than inulin-type prebiotics with short-chain fructans.

In individuals with elevated lipids the results have been mixed, although the preponderance of studies report no benefit with supplementation. The differences in reported outcomes might be due to differences in the study populations, the duration of the studies (although all have been relatively short-term – lasting for one month or less), or differences in the type of inulin-type prebiotic used. The one positive trial using FOS was not placebo-controlled and lasted only two weeks. In the placebo-controlled trials of FOS, which were also short-term in nature, no significant differences were observed. In the trial that administered oligofructose, no statistically significant benefits were reported. In the single inulin trial positive results were reported. Whether this was because inulin was used rather than exclusively short-chain inulin-type prebiotic supplements (FOS and oligofructose), differences in the study population (obese subjects with high cholesterol and triglycerides), or some other factor cannot be determined. While it is possible that some forms of inulin-type prebiotics might positively influence lipids in some subsets of the population, until more definitive research is available and longer-term studies are conducted, these prebiotics should not be expected to significantly modify lipid levels.

• Mineral Metabolism and Bone Remodeling

In animals, inulin-type prebiotics have consistently shown a positive impact on aspects of calcium and magnesium absorption. In humans the effects have been somewhat less consistent for absorption of these and other minerals. In a review of this subject, Scholz-Ahrens et al suggested the inconsistent results might be caused by differences in experimental design.43 Experimental design differences have included: (1) study populations, (2) minerals assessed, (3) assessment techniques used to gauge changes in mineral metabolism, and (4) dose, type of prebiotic, and duration of intervention.

The effect of inulin-type prebiotics on calcium absorption has been the most researched, while the absorption of copper, iron, magnesium, selenium, and zinc has been studied less intensively. Although some studies have been long-term, most have been short-term. Several of the studies have reported inulin-type prebiotics enhance intestinal calcium absorption. Roberfroid notes the most convincing data has been with oligofructose-enriched inulin HP in adolescents and postmenopausal women.1 The studies on other minerals have been inconsistent.

Yap et al investigated the effect of inulin on mineral absorption in 36 healthy, formula-fed infants. Significant increases in iron absorption and retention, magnesium retention, and zinc absorption and retention were observed in infants supplemented with inulin. Calcium and copper absorption and retention were not influenced by supplementation with inulin.

Teuri et al investigated the effect of inulin on calcium metabolism using a randomized, two-period crossover design (each period consisting of one test day) in 15 young healthy women (mean age 23.7). The volunteers were given cheese containing 210 mg of calcium with or without added inulin (15 g/dose). The authors conclude that inulin did not acutely affect markers of calcium metabolism compared to a corresponding breakfast without inulin.

In a crossover study, Coudray et al investigated the effect of inulin on absorption and balance of calcium, magnesium, iron, and zinc in nine healthy young men (mean age 25.5). Active treatment lasted 26 days and participants received up to 40 g inulin daily. The authors reported inulin increased calcium absorption and calcium balance but had no effect on the other minerals.

van den Heuvel et al investigated the effect of inulin or oligofructose on calcium and iron absorption in 12 healthy males (ages 20-30) using a randomized, crossover design. The dose studied was 15 g daily for 21 days and the effects on calcium absorption were compared to a control period with no supplementation. No significant effect on calcium or iron absorption was observed.

In a second study, the same authors investigated the effect of 15 g/day oligofructose compared with placebo (sucrose) on calcium absorption in a crossover experiment in a younger study group – 12 healthy, 14- to 16-year-old boys for nine days per treatment period. The authors reported oligofructose significantly increased fractional calcium absorption compared to the placebo periods.

Griffin et al reported an oligofructose-enriched inulin HP prebiotic significantly increased calcium absorption in girls. A group of 59 adolescent girls at or near menarche was supplemented with either 8 g oligofructose, oligofructose-enriched inulin HP, or placebo (sucrose) daily in a randomized, crossover design. Each intervention was given for three weeks with two-week washout periods between interventions. Throughout the study, subjects consumed approximately 1,500 mg/day dietary calcium. The authors reported calcium absorption was significantly higher in the group receiving the oligofructose-enriched inulin HP compared to the placebo group. Unlike the study by van den Heuvel et al of adolescent boys, in this study oligofructose, compared to placebo, did not produce a statistically significant change in calcium absorption.

Abrams et al conducted the first long-term study of inulin-type prebiotics and mineral absorption. The objective was to assess the effect on calcium absorption and bone mineral accretion after eight weeks and one year of supplementation in adolescents ages 9-13. One hundred subjects were randomly assigned to receive 8 g/day oligofructose-enriched inulin HP or a placebo (maltodextrin). At the end of the study, 92 subjects were available for calcium absorption measurement, which was significantly greater in the active group than in the control group at both eight weeks and one year.

Holloway et al investigated the effects of an oligofructose-enriched inulin HP prebiotic in 15 postmenopausal women in a six-week RCT. Fractional absorption of calcium and magnesium increased following active treatment compared to placebo.

In a three-month study, Kim et al reported that inulin (8 g/day) compared to placebo (maltodextrin/sucrose mixture) increased calcium absorption in postmenopausal women (mean age 60-61) who were not receiving hormonal replacement therapy.

Tahiri et al gave 12 postmenopausal women a daily dose of 10 g FOS or placebo (sucrose) for five weeks using a crossover design. Mean intestinal calcium absorption was not significantly different between the two interventions nor were plasma parathyroid hormone or 1,25-dihydroxyvitamin D concentrations.

Tahiri et al also investigated the effect of FOS supplementation on magnesium absorption in 11 postmenopausal women. Participants received 10 g FOS or placebo for five weeks in a crossover trial separated by a washout period of three weeks. Magnesium absorption increased with FOS supplementation.

Ducros et al investigated the effect of FOS on copper, selenium, and zinc absorption in 11 postmenopausal women. Participants were given either 10 g/day FOS or placebo for five weeks in random order, followed by a washout period of three weeks. Copper absorption was significantly enhanced with FOS supplementation; however, no differences in selenium or zinc absorption were observed.

Several studies have looked beyond mineral absorption to investigate the effect of inulin-type prebioticson biomarkers of bone remodeling. The results have been inconsistent for the entire category of inulin-type prebiotics; however, the studies using an oligofructose-enriched inulin HP prebiotic mixture have reported positive results.

Tahiri et al (described above) reported that 10 g FOS daily for five weeks had no effect on bone turnover biomarkers in postmenopausal women. Plasma osteocalcin concentrations and urinary deoxypyridinoline excretion were monitored, and no significant differences were observed when active and placebo periods were compared.

In the Kim et al study (described above) the effect of inulin on serum bone parameters related to bone turnover and bone mineral density was assessed. After three months of supplementation, the level of serum alkaline phosphatase was significantly lower in the inulin. group compared to placebo, and a non-significant trend toward a slight reduction in urinary deoxypyridinoline was also observed, both indicating reductions in bone turnover. No significant effect on bone mineral density was observed.

Dahl et al gave institutionalized adults thickened beverages fortified with inulin (13 g/day) or isocaloric standard modified starch-thickened beverages. The three-week trial was double-blind and crossover in design. The authors reported the inulin-fortified beveragefailed to induce any change in bone resorption rate, as assessed by the measurement of cross-linked N-telopeptides.

While these three studies reported no positive effects, studies using oligofructose-enriched inulin HP have reported positive results. Holloway et al (described above) reported that, compared to placebo, bone resorption (assessed by urinary deoxypyridinoline crosslinks) was greater and bone formation (assessed by serum osteocalcin) showed an increase after six weeks of treatment. A greater response in bone turnover biomarkers was observed in postmenopausal women with lower initial spine bone mineral density. Positive changes in these parameters occurred only in volunteers who increased mineral absorption in response to supplementation with oligofructose-enriched inulin HP.

Abrams et al (described above) assessed whole-body bone mineral content and whole-body bone mineral density after one year of supplementation with 8 g/day oligofructose-enriched inulin HP. Of 100 adolescents who began the study, 95 were available for the bone mineral measurements. Compared to placebo, whole-body bone mineral content and whole-body bone mineral density were significantly greater in the active group.

A few studies have attempted to determine whether predictable factors might impact the effect inulin-type prebiotics have on calcium absorption. Griffin et al conducted a study on 54 adolescent subjects to determine which subject characteristics were associated with a beneficial effect of oligofructose-enriched inulin HP. In a crossover trial, subjects were given 8 g oligofructose-enriched inulin HP or placebo (sucrose) daily for three weeks. In the group as a whole the active treatment increased calcium absorption. The most consistent identifiable determinant of a beneficial response to the prebiotic was the fractional calcium absorption during the placebo period. Individuals with lower calcium absorption during the placebo period showed the greatest benefit in terms of improved calcium absorption when given oligofructose-enriched inulin HP.

Roberfroid commented on these findings: “An interesting conclusion of the studies in adolescence is the inverse correlation between the relative increase in absorption caused by inulin-type fructans and the basal absorption capacity as measured before the intervention. The same correlation was demonstrated in analyses of the animal data. That would indicate that, with regard to mineral absorption, consuming inulin-type fructans would benefit more the adolescents who have a low basal level.

In the study of 100 adolescents (described above) by Abrams et al, polymorphisms of the Fok1 vitamin D receptor gene were determined. A significant interaction of genotype with prebiotic supplementation was observed following eight weeks of supplementation but not at one year. Participants with the ff genotype had the least initial response to prebiotic supplementation, while the FF and Ff genotypes were associated with higher calcium absorption. In the active treatment group, the percentage of children with an increase in calcium absorption of greater than or equal to three percent after eight weeks of inulin-type prebiotic supplementation were 92 percent, 62 percent, and 50 percent for FF, Ff and ff subjects, respectively. The authors suggested that, “…at least initially, the magnitude of the benefit was affected by genetic modifiers of calcium absorption, including polymorphisms of the Fok1 gene.”

Using the same group of adolescents, Abrams et al identified 32 of the 48 participants receiving the active treatment as “responders.” To be identified as “responders,” participants had a three-percent increase in calcium absorption after eight weeks of inulin-type prebiotic supplementation. Compared to the non-responder and placebo groups, the responders had significantly greater retention of calcium and accretion of this calcium in the skeleton over a year based on whole-body, dual-energy x-ray absorptiometry data and increases in whole-body bone mineral content.

In terms of a mechanism of action, it has been speculated that inulin-type prebiotics optimize passive calcium absorption and this optimization occurs primarily in the colon.59 A study by Abrams et al lends support to this hypothesis. In the study, the authors report the effects of inulin-type prebiotics on calcium absorption occur principally in the colon, with approximately 70 percent of the increase attributed to colonic absorption. This conclusion was based on a study with 13 participants, ages 18-27 years, who took 8 g oligofructose-enriched inulin HP daily for eight weeks.

the proximal colon, while longer-chain polymers can reach the distal colon where they are metabolized. This might explain why shorter-chain polymers like FOS and oligofructose have had less consistent effects on calcium absorption than the oligofructose-enriched inulin HP (which contains a mix of short- and long-chain polymers) and why the latter are presumably better able to influence passive calcium absorption throughout the entirety of the colon.

Table 6. Summary of Inulin Research Findings

Table 7. Summary of Oligofructose-enriched Inulin/Inulin HP Findings

While evidence on inulin-type prebiotics and mineral absorption is mixed, existing evidence suggests positive results are more likely to occur in adolescents and postmenopausal women. Even in these subsets of the population, there is a variable response to supplementation. The variability in response appears to be influenced by baseline calcium absorption – with lower baseline absorption expected to produce better response to inulin-type prebiotics – and possibly genetics. More research in this area is warranted.

In their review, Scholz-Ahrens concludes that, “Certain experimental or dietary conditions and the physiological characteristics of the target group studied might favor a positive outcome of a study.”43 This conclusion appears to be correct. Rather than inulin-type prebiotics having a uniform effect on mineral absorption, the specific type of inulin-type prebiotic used, the duration of use, and individual characteristics (genetics, age, baseline calcium status, etc.) appear to play roles in terms of response.

With respect to improving calcium absorption and producing a long-term positive effect on bone health biomarkers, the best choice for supplementation appears to be oligofructose-enriched inulin HP. Current evidence suggests inulin-type prebiotics would not be expected to have a significant impact on absorption of iron, selenium, and zinc, and might have a positive effect on copper and magnesium in some population subsets.

Table 6 summarizes the clinical research on inulin for bone metabolism and other clinical applications. Table 7 summarizes the clinical research on inulin HP/oligofructose-enriched inulin on various clinical conditions.

• Weight

Inulin-type fructans might have an influence on weight regulation; however, research in this area is very limited.

Abrams et al randomized 97 adolescents (ages 9-13) to receive 8 g oligofructose-enriched inulin HP or placebo (maltodextrin) daily for one year. Eighty-nine of the subjects who completed the one-year intervention were available for follow-up at two years. The subjects who received the prebiotic demonstrated a statistically significantly smaller increase in body mass index (BMI) compared to the control group. The average difference in BMI increase between the two groups was 0.52 kg/m2. Increases in body fat mass were also less in the group receiving the prebiotic at one year, the average gain in body fat being 0.84 kg greater in the placebo group. These observed differences tended to be maintained one year after supplementation was stopped.

Safety/Toxicity

In 1992 a committee of U.S. experts declared inulin-type prebiotics as Generally Recognized as Safe (GRAS).

In animal experiments, the LD50 for FOS is more than 9 g/kg for acute dosing. No treatment-related chronic toxicity has been reported for oral doses of 4.5 g/kg for six weeks. In animal experiments, FOS shows no toxicity compared with existing sugars commonly used in the food supply and no observable negative effects on pregnant rats or the development of fetuses and newborns.62 Presumably, these FOS results are applicable to all inulin-type prebiotics.

Potential Gastrointestinal Side Effects

The primary side effects of inulin-type prebiotics are gastrointestinal and can include osmotic diarrhea, abdominal rumbling, bloating, cramping, and excessive flatulence. These side effects are similar to the effects produced by lactose in people with lactose maldigestion.

Because of the configuration of the bond between fructose monomers (as described in part 1), inulin-type prebiotics are not broken down by intestinal enzymes, which is the probable cause of osmotic diarrhea. Although daily doses of 40-50 g can cause an osmotic effect, doses greater than 50 g would be expected to produce osmotic diarrhea in a large percent of the population.63 While osmotic diarrhea is typically expected only with high doses, in one study abdominal distension was reported at daily doses of 10.6 g FOS.

Although doses greater than 40 g/day are generally necessary to produce abdominal rumbling and bloating, and doses greater than 50 g/day to cause abdominal cramping,63 bloating has been reported with daily doses as low as 2.5-5 g64 and abdominal pain has been reported at doses as low as 10 g daily.

Excessive flatulence is the most well-established side effect and can occur at daily doses as low as 1-2 g in sensitive individuals.

As noted in part 1, Rumessen and Gudmand-Hoyer concluded, after comparing the response to inulin-type prebiotics of different chain lengths, that chain length influences abdominal side effects; shorter-chain inulin-type prebiotics produce more abdominal side effects than longer-chain ones.65 Other researchers suggest a similar correlation, with chain length influencing adverse responses.

Based on these observations, gastrointestinal tolerance to inulin-type prebiotics would be expected to decrease (and side effects increase) when FOS or oligofructose – both of which contain exclusively short-chain inulin polymers – are used as opposed to when inulin (a mix of short- and long-chain inulins) is supplemented. In theory, inulin HP – consisting entirely of long-chain inulin polymers – would be the best tolerated inulin-type prebiotic.

This differential response, in terms of gastrointestinal side effects, is theorized to be due to shorter-chain inulin polymers being metabolized primarily in the proximal colon; whereas, longer-chain polymers reach the distal colon before being fermented. The result is shorter-chain polymers are more rapidly fermented than longer-chain polymers, and thus seem to be more poorly tolerated.

Inulin-type prebiotics contain free sugars (fructose, glucose, and sucrose) unless removed by additional processing. The potential contribution of free sugar content of inulin-type prebiotics to the incidence of abdominal side effects has not been explored.

Dosing Prebiotics

For the promotion of healthy bacterial flora, the usual recommendation for supplementation of inulin-type prebiotics is a daily dose of 2.5-10 g. As detailed in part 1, 2.5-5 g daily is the low end for bifidogenic effects, which are typically dose-dependent. Bouhnik et al reported the optimal FOS dose for producing a bifidogenic effect, while remaining relatively well tolerated, is 10 g FOS daily.

In studies of inulin-type prebiotics, two supplementation approaches have been used to minimize gastrointestinal side effects: dose in two or more divided doses and start with a lower dose and increase after a week or more of supplementation. One or both strategies were used in many of the studies and are potential approaches for lessening the likelihood of side effects.

In subjects who complain of side effects, several options exist. One option is to reduce the dose. A second option is to switch to a product consisting of either a lower percent of short-chain polymers or entirely of long-chain polymers (i.e., inulin or inulin HP instead of FOS or oligofructose).

In patients who complain of abdominal side effects it would be useful to inquire whether they are consuming foods or beverages that have added inulin-type prebiotics as ingredients. Since abdominal side effects tend to be dose dependent, dietary sources would be expected to contribute in an additive way to those prescribed as a supplement.

Conclusions

For infant nutritional applications, the majority of clinical evidence has been obtained from a combination of inulin HP and GOS. This specific mixture has been added to formulas for term and preterm infants.

For non-infant nutrition applications, inulin-type prebiotics have been investigated for a range of clinical uses, with the majority falling into four categories:

• Blood sugar regulation
• Blood lipids
• Gastrointestinal health
• Mineral absorption and biomarkers of bone health

A variety of studies have assessed biomarkers of blood sugar regulation in normo- and hyperglycemic subjects. To date, no form of inulin-type prebiotic has shown a consistent beneficial effect.

Lipids have been investigated in a variety of studies. In persons with normal lipid levels, the most consistent observation is inulin-type prebiotics have no statistically significant effect on lipid levels. In individuals with elevated lipids the results have been mixed, although the preponderance of studies report no benefit from supplementation. Positive results were reported in one trial of inulin. While definitive conclusions should wait for further research clarification, current evidence suggests that inulin, but not FOS or oligofructose, might positively influence lipids in hyperlipidemic individuals.

Because of a bifidogenic effect, the uses of inulin-type prebiotics for improvement of gastrointestinal function and as a potential therapy in gastrointestinal clinical conditions have been investigated. Several studies have investigated bowel transit time, stool consistency, and stool frequency. Existing evidence supports an effect in infants but not in adults without pre-existing functional problems. In subjects with existing constipation, one uncontrolled trial found supplementation with inulin improved stool frequency. In infants, inulin-type prebiotics as a stand-alone intervention do not appear to prevent diarrhea. Three studies report some degree of functional change subsequent to supplementation in persons with active inflammatory bowel disease. These studies were mixed with respect to the active intervention reducing disease activity compared to placebo. No trials have been conducted to determine whether chronic supplementation might help prevent disease reoccurrence or sustain periods of clinical remission. The two studies on IBS also had mixed findings.

Mineral absorption and biomarkers of bone health have received a significant degree of research attention. Current evidence suggests inulin-type prebioticsdo not have a significant impact on absorption of iron, selenium, and zinc, but might have a positive effect on copper and magnesium in some population subsets.

There appears to be significant individual variability in response to inulin-type prebiotics and effects on calcium absorption. Adolescents and perimenopausalwomen have responded positively more consistently than young adults. However, even in these population subsets, some individuals respond better than others. Baseline calcium absorption and possibly genetics appear to influence response. The best choice for prebiotic supplementation to positively impact calcium absorption and bone health biomarkers appears to be oligofructose-enriched inulin HP, which has produced the most consistent results.

Based on available research, it appears inulin HP is the best prebiotic to reduce the likelihood of gastrointestinal side effects. FOS and oligofructose are considered the forms most likely to produce side effects.

Other studies

The efficacy of alverine citrate ⁄ simeticone combination on abdominal pain ⁄ discomfort in irritable bowel syndrome- a randomized, double-blind, placebo-controlled study

T. WITTMANN*, L. PARADOWSKI, P. DUCROTTE´ , L. BUENO§ & M.-C. ANDRO DELESTRAIN

Summary

• Background

Alverine citrate and simeticone combination has been used for almost 20 years in irritable bowel syndrome (IBS), but supportive scientific evidence of efficacy was limited.

• Aim

To evaluate the efficacy of alverine citrate and simeticone combination in patients with IBS-related abdominal pain ⁄ discomfort.

• Methods

total of 412 IBS patients meeting ROME III criteria were included in this double-blind randomized placebo-controlled study if their abdominal pain ⁄ discomfort intensity was at least 60 mm on a 0–100 mm visual analogue scale (VAS) during a 2-week run-in treatment-free period. Patients were randomly assigned through the use of Interactive Voice Response System to receive either alverine citrate 60 mg with simeticone 300 mg three times daily or matching placebo for 4 weeks.

• Results

The full analysis set included 409 patients (71.4% female: mean age: 46.2  13.9 years). At week 4, alverine citrate and simeticone group had lower VAS scores of abdominal pain ⁄ discomfort (median: 40 mm vs. 50 mm, P = 0.047) and higher responder rate (46.8% vs. 34.3%, OR = 1.3; P = 0.01) as compared with placebo group. Patient receiving alverine citrate and simeticone reported greater global symptom improvement compared with those receiving placebo (P = 0.0001). Reported adverse events were similar in both groups.

• Conclusion

Alverine citrate ⁄ simeticone combination was significantly more effective than placebo in relieving abdominal pain ⁄ discomfort in patients with IBS.

Introduction

Irritable bowel syndrome (IBS) is a common, nonlifethreatening condition characterized by abdominal pain and ⁄ or discomfort associated with altered bowel habits (constipation, diarrhoea or both) and very often bloating, which are not explained by bowel anatomical anomalies or biochemical abnormalities. Depending on the criteria, the prevalence of IBS is estimated to rangebetween 8% and 15% in North America and Europe. When using Rome III criteria,4 a homogeneous IBS population may be obtained. Symptoms tend to recur at highly variable intervals, worsen during flares and significantly impact on quality of life in a large subset of patients.

IBS has been linked to several pathophysiological mechanisms: visceral hypersensitivity, digestive motor disturbances, brain-gut axis dysfunction, intestinal dysbiosis and micro-inflammatory changes in the gut wall.7–10 Moreover, a subgroup of IBS patients reveals psychological disturbances which may interfere with the symptoms are dealt with.4, 11 Although all these causes probably overlap and vary in importance from one patient to another, visceral hypersensitivity appears to be a key pathophysiological factor.

A recent review of available IBS treatments in Europe concluded that the efficacy of most is limited by a low level of evidence2 for several reasons: a multifactorial pathophysiology resulting in heterogeneous patients groups, a variety of symptoms leading to very diverse expectations, a strong placebo effect13, 14 and a lack of tools with sufficient validity, reliability and sensitivity to change to allow for proper symptom assessment.

The alverine citrate ⁄ simeticone (ACS) combination (Meteospasmyl; Mayoly-Spindler, France) has been available in Europe since 1990 for the treatment of symptoms related to functional bowel disorders (FBDs). It combines 60 mg of alverine citrate, an active substance derived from papaverine with 300 mg of simeticone (dimeticone enriched with silicon dioxide) in a soft capsule. In pharmacological studies, alverine citrate has been shown to exert an effect on intestinal motility and intestinal sensitivity, two factors recognized as involved in the onset of FBD, without exhibiting the potential drawbacks of a medication with systemic effects.15–19 Alverine citrate affects basal and stimulated motility via a calcium-dependent and -independent inhibition of neuronal excitability15 as well as direct inactivation of L-type Ca2+ of smooth muscle cells.19 Experimental findings support an antinociceptive action by selective receptor-mediated mechanisms. Alverine has been shown to bind to 5-HT1A receptors thereby acting as an antagonist that reduces the visceral pronociceptive effect of 5-HT.18 This mechanism of action may account for its antinociceptive effects in postinflammatory visceral hypersensitivity.16 Simeticone is an inert substance with antifoaming activity. Additionnally, simeticone is able to reduce stress-induced increase in colonic permeability in animals (unpublished data). By limiting mucosal entry of immune stimulating substances, simeticone is likely to reduce the sensitization of primary afferent nerve endings.

In previous double-blind, clinical studies, ACS has been shown to have a therapeutic effect in patients with FBDs (cited by Coelho et al.18) when compared with other antispasmodics. However, these studies were performed in patients suffering from FBD not strictly defined as IBS. Moreover, these studies were not consistent with current recommendations for IBS clinical studies.

Our double-blind study was designed to assess the symptomatic efficacy of ACS, when administered three times daily for 4 weeks, for the relief of abdominal pain ⁄ discomfort in any subgroup of IBS patients meeting Rome III criteria.

Methods

The design and methods of this study were consistent with the recommendations of Rome III, EMEA and expert groups.

Study design

This multicenter, randomized, double-blind, placebocontrolled, parallel-group study was conducted in 17 gastroenterological sites in Hungary and Poland from July 2007 to July 2008. The final protocol and amendments were approved by each country’s central independent ethics committee and the study was conducted in accordance with the International Conference on Harmonisation guidelines for Good Clinical Practice, the Declaration of Helsinki and applicable local regulations. Before entry, patients received detailed information on the study and signed a written consent form.

After a 2-week run-in treatment-free period, patients satisfying the eligibility criteria were randomized to either ACS combination or matching placebo, in a 1:1 ratio. Patients were treated for 4 weeks.

Randomization, stratified by country, was centralized through an Interactive Voice Response System.

Patient selection

Patients were eligible for inclusion if they met the following criteria: adult patient (aged 18–75 years), suffering from IBS as defined by Rome III criteria24 and for no longer than 5 years. At the end of the run-in period, IBS patients were randomized into one of the two treatment groups if they exhibited:

• Abdominal pain ⁄ discomfort for at least 2 days ⁄ week during the run-in period
• Intensity of abdominal pain ⁄ discomfort ‡60 mm on a visual analogue scale (VAS) at randomization (based on the previous 7 days).

Patients were excluded from participation for the following reasons: if their diagnosis was not IBS, if they presented alarming signs (anaemia, rectal bleeding, unexplained weight loss, general health status impairment) or in case of any underlying cause raising doubt as to the IBS diagnosis (diabetes, thyroid dysfunction, biliary or pancreatic disorders or infectious diarrhoea), history of gastro-intestinal cancer or significant gastro-intestinal surgery, acute ⁄ uncontrolled systemic pathology or liver function tests ‡3 times the upper normal limits. Patients with known intolerance to ACS combination or one of its components were also excluded from entry, as were patients with regular use of ACS combination during the 6 months prior to the study.

Drug treatment

During the active treatment period, patients took either ACS (alverine citrate 60 mg + simeticone 300 mg) or matching placebo. ACS and placebo were identical in appearance and taste and administered orally as soft capsules three times a day, prior to meals for 4 weeks. No study drug was administered during the run-in period.

Randomization and blinding

The randomization scheme was performed by a specific contractor (Cardinal Systems, Paris, France), who was independent from the Contract Research Organization (I3 Research) in charge of the study conduct and analysis. Cardinal Systems generated the randomization scheme (stratified by country) using SAS program (SAS Institute Inc., Cary, NC, USA). Codes were held by their statistician. Patients were equally allocated to treatment in block sizes of 4. When a patient was eligible for randomization, investigator contacted the randomization centre through the Interactive Voice Response System. The IVRS prompted the user for key information: VAS value in millimeters, number of pain/discomfort episode during the first and second week of run-in period. If randomization criteria were met, randomization was confirmed and IVRS allocated the appropriate treatment kit number and a notification was sent by fax to the investigator to confirm treatment kit number to be dispensed. All study personnel and participants were blinded to the true identity of the treatment assigned until database was locked.

Prohibited ⁄ authorized treatment

Prohibited treatments during the study were those likely to jeopardize study drug evaluation, i.e. antispasmodics, antidiarrhoea drugs and laxatives. Antidepressants and anxiolytics were allowed, if they were started prior to the study and had been taken at a stable dosage over the last 3 and 1 months respectively. Their dosage had to be stable during the 4-week study. Patients were advised not to change their diet throughout the study.

Data collection

Visits were scheduled at inclusion (week )2), randomization (week 0), and then at weeks 1, 2 and 4. Throughout the duration of the study (from weeks )2 to 4), patients had to report in a paper diary on a daily basis: abdominal pain ⁄ discomfort, bloating, any other symptoms, number and form of stools, medications used and study drug intake. At each visit, patients rated their abdominal pain ⁄ discomfort on a 0–100 mm VAS relating to the previous week. Discomfort was defined as an uncomfortable, but not painful sensation. In IBS patients, the 0–100 mm VAS of abdominal pain was shown to be a valid and a reliable tool,25 with discriminatory power.26 At randomization and end of treatment (week 4), patients quantified the impact of IBS on their daily life; they assessed overall treatment efficacy at week 4.

Efficacy endpoints

Primary efficacy endpoint. The primary efficacy endpoint was the difference in the magnitude of change in abdominal pain ⁄ discomfort VAS scores from weeks 0 to 4 between treatment groups. In addition, a patient was defined as responder if abdominal pain ⁄ discomfort VAS score decreased by at least 50% at week 4 compared to week 0. This definition of responder is considered accurate.21, 27 In the statistical analysis plan, the difference in responder rates between both groups was taken into account in the secondary analysis of the primary efficacy endpoint.

Secondary efficacy criteria. Secondary end-points included intensity of abdominal pain assessed by VAS at weeks 1 and 2, overall patient assessment regarding symptom relief, IBS impact on daily life and changes in remaining IBS symptoms as compared to week 0.

Patients were asked to assess the outcome of their IBS symptoms at the end of week 4 by responding to the following statement, ‘The treatment helped to improve my bowel problems’ using a 5-point Likert scale: ‘strongly disagree’, ‘disagree’, ‘neither agree nor disagree’, ‘agree’ and ‘strongly agree’.

At randomization (week 0) and end of treatment (week 4), patients were also asked to grade the impact of IBS on their daily life by responding to the following statement: ‘My bowel problems limit my life and everyday activities’. A 5-point Likert scale was used, with scores ranging from 1 to 5: ‘extremely’, ‘quite a bit’, ‘moderately’, ‘a little’ or ‘not at all’.

Each day, patients recorded the frequency and form of stools in a paper diary. At each visit the investigator assessed the weekly average frequency and the most frequent type of stools by using the Bristol Stool Form Scale (BSFS). The presence of other IBS symptoms (bloating, straining, urgency and feeling of incomplete defecation) was also noted.

Concomitant factors

Although psychological factors are not required for IBS diagnosis, they may influence GI symptoms. 11, 28, 29 The Hamilton Rating Scales for Depression (HAM-D) and Anxiety (HAM-A) were used to detect symptoms of depressive or anxiety disorders and if present to explore their potential impact on the treatment’s effect. Trained and certified raters conducted patient’s interviews and filled in the HAM-D and HAM-A questionnaires at weeks 0 and 4.

Safety

All adverse events (AEs) were actively sought at each visit. Standard laboratory tests for haematology and biochemistry were performed at inclusion (week )2) and at end of treatment (week 4). Haematology included haemoglobin, total white blood cell count, platelet count and a differential count including neutrophiles, lymphocytes, monocytes, eosinophiles and basophiles. Biochemistry included serum creatinine, alkaline phosphatases, aspartate amino transferase, alanine amino transferase and total bilirubin.

Statistical analysis

Analyses were performed on three sets: safety analysis set, full analysis set (FAS) and per protocol set. The safety analysis set included all randomized patients who received at least one dose of study treatment. The FAS included all randomized patients who took at least one dose of study treatment and for whom at least one on-treatment main criterion measure was available. The FAS population corresponded to the recommended intention-to-treat (ITT) population. The per-protocol set (PP) included a subgroup of the FAS population fulfilling the following criteria: minimal time of exposure to treatment (‡3 weeks), minimal treatment compliance during minimal study drug exposure time (‡66%) and evaluation of the main criterion (abdominal pain ⁄ discomfort VAS score at week 4).

Primary efficacy criterion was assessed using two analyses. The primary analysis focused on the betweengroup difference in the degree of change in abdominal pain ⁄ discomfort VAS scores from weeks 0 to 4. Due to the non-normality of the VAS score change distribution, a rank-based nonparametric analysis of covariance (ANCOVA, Quade’s analysis)30 was performed on week 4 VAS values in the ITT population. This was conducted after imputing missing values by last-observation- carried-forward (LOCF). The nonparametric analysis of the week 4 VAS values adjusted for week 0 VAS values is equivalent to the nonparametric analysis of the VAS score changes adjusted for week 0 value. The secondary analysis of the primary efficacy criterion focused on responder rates using a Cochran-Mantel- Haenszel Test adjusted for country. For consistency, responder rates were calculated, after imputing missing values using the LOCF-based method. This analysis was also performed with a 60% cut-off value. A logistic regression model adjusting for treatment and country was used and the odds ratio for the treatment effect with 95% confidence interval was calculated. The efficacy analyses were primarily performed on the ITT population and secondarily on the PP population. The safety analysis was performed on the safety population.

The study was planned to assess superiority of ACS combination over placebo. The sample size was calculated based on the primary end point. Notably, there are no guidelines specifying the threshold considered clinically significant for IBS. Considering a conservative expected standard deviation of 25 mm of VAS score change from weeks 0 to 4, and a conventional two-sided type-1 error of 5%, the sample size was established at 200 patients in each group, enabling the detection of a minimal clinically significant effect of 7 mm with a guaranteed 80% power. All statistical analyses were performed using the SAS software version 9. As no direct action on any specific subtype was expected, no subgroup analysis on bowel habit subtypes was performed and the different IBS subtypes were not taken into account for the calculation of the study sample size.

Results

Baseline patient characteristics

Between July 2007 and June 2008, 429 patients were selected and entered the 2-week treatment-free run-in period. Among these patients, 17 (4.0%) were not randomized for the following reasons: VAS <60 mm (five patients), consent withdrawal (10 patients), serious AE (one patient) and lost to follow-up (one patient). Among the remaining 412 patients, 207 were randomly assigned to ACS, and 205 to placebo. A total of 399 (97.0%) patients completed the study, whereas 13 (3.0%) discontinued the study. The most common reasons for study discontinuation were: lack of efficacy in five patients (ACS: one patient; placebo: four patients) and AEs in three patients (ACS: one patient; placebo: two patients). One patient in each group was lost to follow-up. The flow chart of the study is shown in Figure 1. Of the randomized patients, all 412 comprised the safety set (207 in ACS and 205 in placebo groups), 409 formed the ITT set (205 in the ACS and 204 in the placebo groups) and 386 comprised the PP set (194 in the ACS and 192 in the placebo groups).

Baseline patient characteristics of ITT population are displayed in Tables 1 and 2. Mean age was 46.2  13.9 years (mean  s.d.). No differences were observed between both groups regarding demographic characteristics, abdominal pain ⁄ discomfort score, occurrence of other gastrointestinal symptoms, bowel habit disorders, and anxiety (HAM-A) and depression (HAM-D) scores. Concomitant medical conditions were mainly hypertension (27.7%), gastroesophageal reflux disease (18.0%) and psychological disorders (14.3%). The most common concomitant drugs were: cardiovascular agents (33.3%) antiacids, including mainly proton pump inhibitors (26.7%), anxiolytics (10.7%) and antidepressants (8%). Concomitant pathologies and drugs were comparable between both groups.

Primary efficacy criterion

ntention-to-treat population: A greater effect on abdominal pain ⁄ discomfort VAS scores from weeks 0 to 4 was observed following ACS compared to placebo, with a week 4 median of: 40.0 mm (range: 0–95) and 50.0 mm (range: 0–100) respectively. The VAS score analysis at week 4 showed a statistically significant difference in favour of ACS combination (P = 0.047) (Figure 2). The responder rates at week 4 were significantly higher with ACS than placebo: 46.8% vs. 34.3% (OR: 1.30; 95% CI: 1.06–1.59; P = 0.01). Sensitivity analyses of responders with a 60% cut-off value showed a statistically significant difference (Figure 3). Per-protocol population A greater reduction in abdominal pain ⁄ discomfort VAS score was observed following ACS [median at week 4: 39.0 mm (range: 3–95)] than placebo [median at week 4: 49.5 mm (range: 0–90)], the difference being statistically significant (P = 0.036). Responder rates at week 4 were significantly higher with ACS than placebo: 49.0% vs. 35.9% (OR: 1.31; 95% CI: 1.07–1.60; P = 0.01).

Secondary efficacy criteria

Abdominal pain ⁄ discomfort at weeks 1 and 2. At week 1, VAS scores reduction and responder rates were greater with ACS group than with placebo, although the difference was not statistically significant. At week 2, abdominal pain ⁄ discomfort VAS scores were significantly (P = 0.02) in favour of ACS [median: 51.0 mm (range: 2–93)] as compared with placebo

[median: 59.0 mm (range: 0–100)] (Figure 2). A statistically significant difference in favour of ACS was also observed on the responder rates (27.9% vs. 17.2%; P = 0.01).

Overall treatment assessment. Positive overall treatment assessments were reported more frequently in the ACS-treated group, while negative assessments were more common in the placebo group, the difference being highly significant (P = 0.0001; Table 3).

IBS life impact score. From weeks 0 to 4, there was a trend towards greater improvements in IBS life impact scores following ACS as compared with placebo (0.97 vs. 0.76), although the difference was not statistically significant (P = 0.08) (Table 4).

Stool assessment and other IBS symptoms. The number and rate of patients who reported constipation as the most frequent stool type during the last 7 days (score 1 or 2) decreased from weeks 0 to 4, from 45 (21.9%) to 29 (14.1%) in the ACS group and from 51 (25%) to 23 (11.3%) in the placebo group. The number and rate of patients who reported diarrhoea (score 6 or 7) decreased from weeks 0 to 4, from 45 (21.9%) to 18 (8.8%) in the ACS group and from 47 (23%) to 32 (15.7%) in the placebo group. No significant difference between the groups was observed with regard to the progression of bowel habit disorders or other IBS symptoms.

Concomitant factors. Reductions in HAM-A and HAM-D scores from weeks 0 to 4 were observed in both groups, although the difference was not statistically significant.

Safety. The incidence of AEs was similar in both groups with 17.9% and 24.4% of patients reporting at least one treatment emergent adverse events (TEAE) under ACS and placebo respectively. TEAEs reported by at least 2% of patients are presented in Table 5. Seven (3.4%) patients in the ACS group and 12 (5.9%) in the placebo group reported treatment related TEAEs, as considered by the investigator. There were no deaths or other drug-related serious adverse events (SAE) in this study. Only one (0.2%) patient (ACS group) reported a SAE (traumatic tendon rupture), which was not considered drug-related. Three patients (one in the ACS group and two in the placebo group) withdrew from the study due to TEAEs, namely eye swelling in the ACS patient, and dizziness and pain in the extremities in the placebo patients.

Discussion

This multicentre, double blind, placebo-controlled, randomized study was the first to assess an antispasmodic drug in IBS patients using Rome III criteria. Patients from hospital databases were prescreened for eligibility. About 50% of these patients were not selected as they did not meet eligibility criteria, in particular, absence of ACS administration in the 6 months prior to inclusion. Furthermore, many patients did not accept to be referred to a psychiatrist for anxiety and depression symptoms rating as they denied any psychological component in their IBS symptoms. This strict screening may account for the low screen failure rate. In the study population, at the end of a 4-week treatment period, the relief of abdominal pain ⁄ discomfort was greater with ACS than placebo. This superior efficacy was associated with significantly higher responder rates in the ACS group (46.8%) compared with placebo (34.3%) at week 4. This resulted in a number needed to treat (NNT) of 8, which is close to the NNT related to pain relief (8.3) previously determined by a meta-analysis of other smooth muscle relaxant studies meta analysis.31, 32 The overall treatment assessment by patients showed a higher rate of symptoms improvement with ACS compared with placebo. However, because of the lack of severity measurement of symptoms other than pain, particularly bloating, no further evidence could be collected. Because of the lack of comparative data between ACS combination and alverine citrate or simeticone, the ACS effect cannot be attributed to one particular component or to a combination. This issue should be addressed in future clinical studies including both pain ⁄ discomfort assessment and objective bloating measurements.

At baseline, bowel habit disorders were self-reported by 84% patients. IBS subtype estimation done by investigators based on patient diaries and using BSFS revealed that 23% of patients were IBS-C and 22% IBSD. Other patients reported stool types 3, 4 and 5 as the most frequent during the 7 days preceding estimation. The discrepancy between the percentages may be explained by the difficulty in assessing bowel habit disorders as perceived by patients, as previously reported by Hungin et al.33 As our patients mainly complained of severe pain ⁄ discomfort, and patients with IBS-M pattern have been reported to exhibit greater pain ⁄ discomfort severity than patients with other stool patterns,34 it may be hypothesized that these patients suffered from mixed type (IBS-M) or unsubtyped IBS, although our data do not allow us to draw conclusions on this issue. The number of patients meeting the BSFS definition of diarrhoea or constipation decreased to 30%. As the decrease was comparable in both the ACS and the placebo groups, it is unlikely to be related to drug itself. In our view, the decrease is more likely related to the natural fluctuations of IBS as reported by Garrigues et al.35 However, it cannot be excluded that some patients changed from one subgroup to another, although changes from constipation to diarrhoea subgroups and vice versa are uncommon.

Patients included in the study had been suffering from IBS for less than 5 years and presented abdominal pain ⁄ discomfort ‡60 mmon VAS. We decided to exclude patients with IBS history longer than 5 years, as they are a subgroup of patients with a more complex clinical profile, and have often used multiple therapies.14 Moreover, the selected population was thought to be representative of IBS patients seen by general practitioners or gastroenterologists in primary and secondary care. Large epidemiological surveys have reported symptom duration of less than 5 years in a large subset of IBS patients, particularly among those seeking a medical advice for the first time.33, 36 This also appears to apply to our patients, as more than three-quarters did not receive any drug for IBS within the 6 months preceding enrolment. For inclusion, pain ⁄ discomfort had to be severe enough to detect a difference between the two groups. Even among patients likely to experience positive changes, the severity of IBS symptoms had to be sufficiently marked to distinguish between treatment and placebo. Previous randomized controlled studies have revealed a large positive placebo response in IBS patients, ranging from 30% to 40%.13 In our study, the results obtained in the placebo group (which displayed a high 50% responder rate (34.3%) on abdominal pain ⁄ discomfort at week 4) are in accordance with previously published results. Hence, the positive overall results of our study cannot be explained by a lower placebo response.

Our selection criteria likely account for the baseline characteristics of the IBS population included in this trial. In our study population, IBS had a moderate impact on quality of life and approximately 10% of patients experienced severe anxiety or depression symptoms. This figure is close to the lower range of psychological co-morbidity reported for IBS.11 Consequently, we acknowledge that our results may not be applicable to long-standing sufferers. In conclusion, ACS combination administered orally three times daily for 4 weeks, significantly improves abdominal pain ⁄ discomfort in IBS patients irrespective of the IBS subgroup. Our results support the conclusion that ACS is a the rapeutic option for IBS patientsseen in primary and secondary care.

Other studies

A Comparative Trial of Zolmitriptan and Sumatriptan for the Acute Oral Treatment of Migraine

R. Michael Gallagher, DO; George Dennish, MD; Egilius L.H. Spierings, MD, PhD; Rohini Chitra, PhD

Introduction

An estimated 10% to 15% of people worldwide suffer from migraine, 1 and more than 80% of these sufferers experience some disability directly related to their migraines. 2 The introduction of selective serotonin (5-HT) receptor agonists provided effective choices for treating acute migraine attacks. Sumatriptan (Imitrex), the first 5-HT 1B/1D -specific receptor agonist available for the acute treatment of migraine, was an improvement over the older antimigraine medications, such as ergotamine. Reported efficacy rates for oral sumatriptan, 50 mg, range from 54% (2 hours post dose) to 77% (4 hours post dose). 3–5 Experimental studies in animals, however, suggest that sumatriptan does not access the CNS unless the blood-brain barrier is disrupted experimentally, which is not the case for the newer generation of triptans. 6 Sites within the CNS may provide additional targets for antimigraine actions, such as the trigeminal nucleus caudalis.7

Zolmitriptan (Zomig), a relatively new 5-HT 1 receptor agonist, has a different pharmacological 7 and pharmacokinetic profile 8 from oral sumatriptan. Zolmitriptan has dual peripheral and central action and rapid absorption. 7,8 It has an oral bioavailability of approximately 40% 9 compared with 14% for oral sumatriptan. 10 Together, these characteristics offer the potential for additional clinical benefits for migraine sufferers. Sufferers seek rapid and effective pain relief, but they also seek a treatment that provides consistent efficacy, low headache recurrence, and minimal adverse events. 11,12 The central action and increased bioavailability of zolmitriptan may provide improved and more consistent clinical efficacy than sumatriptan, a rationale that underpinned the current study. 7,13

The results of double-blind, placebo-controlled clinical trials of zolmitriptan show that when used to treat a single attack, 2-hour headache response rates for zolmitriptan, 2.5 mg, range from 62% to 65%, 14,15 rising to 85% in an open-label, long-term trial. 16

Although the clinical effects of 5-HT agonists have been compared with placebo, there exists a continued need for comparative trials. To compare the relative response rates and assist clinicians in determining the most appropriate medication for their patients with migraine, zolmitriptan and sumatriptan were compared directly. The current study evaluates the headache response rate at 1, 2, and 4 hours in patients using either zolmitriptan or sumatriptan to treat up to six moderate/severe migraine attacks during a 6-month period. In addition, the study evaluated meaningful migraine relief 1, 2, and 4 hours after the first dose and pain relief over 24 hours in patients using zolmitriptan compared with sumatriptan.

Methods

This was a randomized, double-blind, parallel design study comparing zolmitriptan, 2.5 mg and 5 mg, with sumatriptan, 25 mg and 50 mg. It was conducted on an outpatient basis with patients recruited from primary care offices, neurology offices, and research clinics in the United States. All patients provided written informed consent, and the study protocol was approved by the Institutional Review Board associated with the individual study centers.

Patients enrolled into the study had established diagnoses of migraine according to the criteria of the IHS 17 with a history of migraine attacks for at least 1 year. Women were asked to use a reliable method of contraception.

Patients were excluded from the study if they had evidence or history of ischemic heart disease, arrhythmia, or accessory conduction pathway disorders (eg, Wolff-Parkinson-White syndrome); hypertension (systolic BP > 160 mm Hg or diastolic BP > 95 mm Hg); or any condition that may have put them at increased risk on exposure to study medication or that may have interfered with efficacy or safety assessments. Patients with a history of basilar, ophthalmoplegic, or hemiplegic migraine; with nonmigraine headache for 10 or more days per month in the previous 6 months; or who were using monoamine oxidase inhibitors, methysergide, methylergonovine, fenfluramine, or dexfenfluramine were also excluded. Other exclusion criteria included drug or alcohol abuse, clinically abnormal laboratory results at screening, lactation, unacceptable adverse events (in the opinion of the investigator) following previous use of any 5-HT 1B/1D receptor agonist, simultaneous participation in another clinical trial, or treatment with another investigational drug within 30 days before screening.

Once entered into the trial, patients used either zolmitriptan, 2.5 mg or 5 mg, or sumatriptan, 25 mg or 50 mg, to treat their first migraine. A second dose identical to the first was available to treat any recurrent migraine occurring 4 to 24 hours after the initial dose. Escape medications for the treatment of persistent headache (ie, a migraine that did not respond to study medication by 2 hours after treatment) were allowed no sooner than 2 hours after the last dose of study medication; nonsteroidal anti-inflammatory agents, analgesics, or sedatives were permitted, but acute antimigraine treatments such as sumatriptan, ergotamine, dihydroergotamine, and isometheptene were not permitted as escape medications.

Following treatment of the first migraine headache with randomized medication, patients returned to the center with a completed diary card and any unused medication. They were then given additional blinded medication identical to the treatment used for their first attack to treat the next migraine. This cycle was repeated for the treatment of six migraine attacks or until the study ended.

The primary efficacy end point was headache response 2 hours after initial treatment. Headache response was defined as a reduction in headache intensity from severe or moderate at baseline to mild or none at 1, 2, or 4 hours after treatment. The intensity of each migraine and the accompanying symptoms were recorded immediately before and 1, 2, and 4 hours after the initial dose. Recurrence was defined as a headache that initially responded to the first dose of trial medication by 4 hours and then recurred with moderate or severe intensity 4 to 24 hours after the first dose. The intensity of the headache recurrence was measured immediately before and 2 hours after treatment with the second dose.

The secondary end point of pain relief over 24 hours was assessed to provide a complete picture of migraine management following the initial dose. The categories for response were the following: full (patient had a headache response at 2 hours and maintained that response for 24 hours without having a recurrent headache or taking escape medication); partial (patient had a response at 2 hours and then had a recurrence or took escape medication); and none (patients with attacks of moderate-to-severe intensity who had no response to treatment at 2 hours). Pain relief over 24 hours provides a combined assessment of the proportion of patients with a full response and the proportion of patients with a full or partial response. Sustained pain relief over 24 hours was a full response, as defined above.

Meaningful migraine relief, which encompasses all migraine symptoms, and overall treatment satisfaction (ie, excellent, good, fair, poor) were subjective assessments recorded by patients on diary cards. Consistency of response was defined as the percentage of patients responding in 80% or more treated attacks. Patients who reported that they achieved a headache response or meaningful migraine relief were included in the analysis only if no escape medication had been taken prior to the assessment.

Statistical analysis

To be included in the statistical analysis, patients had to treat at least two headaches. Headache response rates, meaningful migraine relief, and sustained pain relief over 24 hours were analyzed using the generalized estimating equations (GEE) method for ordinal data with the logistic mean link in SAS. Pain relief over 24 hours was analyzed using GEE for ordinal data with the cumulative logistic mean link in SAS. Global rating of treatment satisfaction was analyzed the same way using SAS logistic regression. The effects of treatment, region (combination of centers), and individual baseline severity were included in all the models except for global rating of treatment satisfaction where individual baseline severity was not included because it was an overall evaluation over multiple attacks (up to six). Consistency of headache response was analyzed using the chi-square test. Estimated treatment odds ratios (ie, the ratio of the odds for the two treatment groups being compared), P values, and 95% confidence intervals were obtained. Odds ratios were tested at the 5% level of significance. Statistical significance was obtained when P < .05.

The following secondary parameters were not subject to formal statistical analysis because they were not based on the original randomized sample. However, descriptive statistics were provided. These parameters were the following: data relating to the incidence of headache recurrence, time to recurrence, use of additional medication to treat recurrence, improvement in associated baseline headache symptoms, and incidence of adverse events.

The intent-to-treat population (ie, patients treating at least two headaches) was the primary focus of efficacy.

Results

A total of 1445 patients with migraine were enrolled from 61 centers. Of these patients, 1212 treated at least two migraine attacks (intent-to-treat population) and 1043 completed the study (ie, they treated two to six migraine headaches or remained in the trial for 6 months). The intent-to-treat population treated a total of 6187 migraine attacks. Demographic characteristics were similar among the four treatment groups (Table 1).

Overall, a higher percentage of patients who received either dose of zolmitriptan had a headache response at 2 hours than those who received either dose of sumatriptan. In the zolmitriptan treatment groups, 67.1% of those receiving the 2.5-mg dose and 64.8% of those receiving the 5-mg dose reported a 2-hour headache response, compared with 59.6% of those receiving sumatriptan, 25 mg, and 63.8% of those receiving sumatriptan, 50 mg.

The differences in response rates between patients using zolmitriptan, 2.5 mg or 5 mg, and those using sumatriptan, 25 mg, were statistically significant at 2 hours ( P , .001, both zolmitriptan doses). Furthermore, patients taking zolmitriptan, 2.5 mg, had a significantly higher response rate at 2 hours when compared with those taking sumatriptan, 50 mg ( P 5 .017) (Table 2).

More patients responded at 2 hours to zolmitriptan, 5 mg than to sumatriptan, 50 mg, although this failed to reach statistical significance (odds ratio [OR] 5 1.16, P 5 .064).

Patients taking zolmitriptan, 2.5 mg, had a significantly higher response rate at 4 hours (83.3%) compared with sumatriptan, 25 mg (75.8%; P , .001) and 50 mg (80.8%; P 5 .039) (Table 2). At 1 hour post dose, patients taking zolmitriptan, 2.5 mg, had a headache response rate of 35% compared with 32.9% for those taking sumatriptan, 25 mg ( P 5 .061), and 34.7% for sumatriptan, 50 mg ( P 5 .461). The differences in headache response rates were significant at 1 and 4 hours between zolmitriptan, 5 mg, and sumatriptan, 25 mg (37.4% versus 32.9% and 83.6% versus 75.8%, respectively; both P , .001), and between zolmitriptan, 5 mg, and sumatriptan, 50 mg, at 1 and 4 hours (37.4% versus 34.7%, P 5 .002 and 83.6% versus 80.8%, P 5 .012, respectively) (Table 2).

Overall, more patients in the zolmitriptan treatment groups reported meaningful migraine relief, a subjective assessment recorded by patients on diary cards, at 1, 2, and 4 hours (43.4%, 72.2%, and 82.3% for zolmitriptan, 2.5 mg, respectively; 45.5%, 72.2%, and 83.1% for zolmitriptan, 5 mg, respectively) than patients in the sumatriptan treatment groups (39.2%, 66.2%, and 74.5%, respectively, for sumatriptan, 25 mg; 41.7%, 67.9%, and 80.1%, respectively, for sumatriptan, 50 mg). Patients receiving zolmitriptan, 2.5 mg, were significantly more likely to achieve meaningful migraine relief at 1, 2, and 4 hours after treatment compared with those taking sumatriptan, 25 mg ( P 5 .014 at 1 hour, P , .001 at 2 and 4 hours), and at 2 hours after treatment compared with those taking sumatriptan, 50 mg ( P 5 .009) (Table 2). At 1 and 4 hours post dose, patients taking zolmitriptan, 2.5 mg, were at least as likely to have meaningful migraine relief as those taking sumatriptan, 50 mg (OR 5 1.07, P 5 .325 at 1 hour; OR 5 1.16, P 5 .129 at 4 hours). Patients taking zolmitriptan, 5 mg, were significantly more likely to achieve meaningful migraine relief at 1, 2, and 4 hours compared with sumatriptan, 25 mg and 50 mg ( P , .001 at all time points for zolmitriptan, 5 mg, versus sumatriptan, 25 mg; P 5 .017 at 1 hour, P 5 .005 at 2 hours, and P 5 .026 at 4 hours for zolmitriptan, 5 mg, versus sumatriptan, 50 mg) (Table 2). A summary of all statistical comparisons is given in Table 3.

Both the 2.5-mg and 5-mg doses of zolmitriptan provide significantly better pain relief over 24 hours for up to six attacks treated (67.1% and 64.8%, respectively, with a full or partial response) compared with sumatriptan, 25 mg (59.4%; OR 5 1.47 and 1.54, respectively; P , .001), or sumatriptan, 50 mg (63.8%; OR 5 1.17, P 5 .021; OR 5 1.22, P 5.005 respectively). Pain relief over 24 hours is a composite measure that takes into consideration 2-hour headache response, subsequent headache recurrence, and use of escape medications for headache over 24 hours.

Zolmitriptan, 2.5 mg and 5 mg, provided significantly better sustained pain relief over 24 hours for up to six attacks treated (40.7% and 42.5%, respectively) compared with sumatriptan, 25 mg (33.1%; OR51.46 and 1.67, respectively; both P,.001). Sustained pain relief for zolmitriptan, 2.5 mg, was numerically greater when compared with sumatriptan, 50 mg, although failing to reach statistical significance (40.7% versus 38.1%; OR51.15, P5.071). Zolmitriptan, 5 mg, provided significantly better sustained relief over 24 hours for up to six attacks treated, compared with sumatriptan, 50 mg (42.5% versus 38.1%; OR51.31, P,.001).

Both doses of zolmitriptan had greater apparent consistency of response and, therefore, potentially greater reliability over multiple attacks than either dose of sumatriptan. The percentage of patients who responded at 2 hours post dose in 80% to 100% of attacks (ie, consistency of response) was higher for zolmitriptan, 2.5 mg and 5 mg, compared with sumatriptan, 25 mg and 50 mg. The consistency of response was significantly higher for zolmitriptan, 2.5 mg (47.1%), and zolmitriptan, 5 mg (44.3%), compared with sumatriptan, 25 mg (33%; P,.001 and P5.004, respectively). The proportion of patients who responded to 80% of attacks or more at 2 hours was numerically higher for zolmitriptan, 2.5 mg (47.1%), and zolmitriptan, 5 mg (44.3%), than for sumatriptan, 50 mg (39.2%), although failing to reach statistical significance (P5.051, P5.206, respectively) (Figure 1).

More patients rated their treatment as “good” or “excellent” than as “fair” or “poor” across all treatment groups. Patients taking zolmitriptan, 2.5 mg or 5 mg, were more likely to give their treatment an overall higher rating when compared with patients taking sumatriptan, 25 mg (OR51.64, P5.002; OR51.68, P5.001, respectively). More patients taking zolmitriptan, 2.5 mg or 5 mg, rated their treatment higher than those taking sumatriptan, 50 mg, although the difference was not statistically significant (OR51.19, P5.265; OR 51.22, P5.199, respectively).

The odds ratios for all the efficacy comparisons were greater than 1 in favor of both doses of zolmitriptan, showing zolmitriptan to be at least numerically superior to sumatriptan, 25 mg and 50 mg. The majority of the end points also reached statistical significance.

Headache recurrence, which can only occur in those patients who have had an initial headache response, was also assessed. Overall recurrence rates for attacks 2 through 6, combined (39.3% and 37.9%), were slightly lower and, in patients with recurrence, initial headache relief lasted slightly longer (13 to 15 hours) for patients taking zolmitriptan, 2.5 mg and 5 mg, than for patients taking sumatriptan, 25 mg or 50 mg (recurrence rates 45.6% and 43.9%, respectively; headache relief lasted 10 to 13.5 hours).

Roughly three quarters of the patients who experienced recurrence did not take additional medication for recurrence, even though patients were instructed that this was permitted. Patients in the zolmitriptan treatment groups were less likely to take additional medication for recurrence (15% and 20% for zolmitriptan, 2.5 mg and 5 mg, respectively) than patients in the sumatriptan, 25 mg, treatment group (30.2%). Additional medication for recurrent headache was taken by 21.4% of patients in the sumatriptan, 50 mg, treatment group.

In all treatment groups, nausea, photophobia, and phonophobia were substantially decreased to similar levels 2 hours after dosing compared with predose assessments. The percentage of patients across the four treatment groups with nausea before treatment ranged from 48% to 54%, decreasing to 24% to 26% 2 hours post dose. No apparent differences were observed across the 4 treatment groups. The percentage of patients with photophobia was reduced substantially from 75% to 80.4% across the treatment groups before treatment to 39.3% to 43.8% 2 hours post dose, as was the case for patients with phonophobia (range from 61.4% to 68.5% before treatment reducing to 29% to 33.7% 2 hours post dose). Sumatriptan, 25 mg, was associated with a slightly higher presence of photophobia and phonophobia post dose.

Safety results

All patients who were randomized to treatment and who treated at least one headache (ie, 1338 patients treating 6315 attacks) were included in the safety analysis. Of these patients, 683 (51%) reported at least one adverse event across the attacks treated (up to six). The rate was similar among all treatment groups and decreased with number of attacks treated (Table 4). Descriptive statistics are provided for adverse events (Table 4).

Few patients withdrew from the trial because of adverse events (1.8%, 3.6%, 2.7%, and 2.1% for zolmitriptan, 2.5 mg and 5 mg, and sumatriptan, 25 mg and 50 mg, respectively), suggesting that the benefits outweighed the adverse events. As expected for 5-HT1 agonists, the most frequently reported drug-related adverse events were dizziness, nausea, paresthesia, somnolence, and sensation of tightness of the throat, chest, or jaw (Table 4).

Ten patients (four in the zolmitriptan treatment groups and six in the sumatriptan treatment groups) reported 13 serious adverse events, none of which were considered by the investigator to be related to trial treatment. Twenty-one of the 74 adverse events leading to withdrawal were considered by the investigators to be of severe intensity; only 2 were considered serious and had no causal relationship to study treatment (moderate drug dependency [1 patient, sumatriptan, 25 mg] and severe eye hemorrhage [1 patient, zolmitriptan, 5 mg] in a patient who had a history of ophthalmic hemorrhage disorder since 1996).

Comments

The development of zolmitriptan grew from a desire to provide an antimigraine therapy with an improved clinical profile in comparison with previously available treatments.18 Zolmitriptan offers good oral bioavailability and dual central and peripheral action, properties that theoretically may provide more consistent and higher efficacy, respectively.18 The high efficacy rates and consistent response to zolmitriptan shown in this and prior studies14–16,19 demonstrate the potential of zolmitriptan to address the important needs of migraine sufferers, namely, rapid, consistent, and effective pain relief.11,12

Patients taking zolmitriptan in this trial were more likely to have a headache response at 2 hours than those taking sumatriptan. The average 2-hour headache responses in this trial were 67.1% and 64.8% for zolmitriptan, 2.5 mg and 5 mg, respectively, compared with 59.6% for sumatriptan, 25 mg, and 63.8% for sumatriptan, 50 mg. In a previous trial with similarly sized groups, the percentages of patients with headache responses were 65% and 67% for zolmitriptan, 2.5 mg and 5 mg, respectively.14

In this comparative study, zolmitriptan exhibited a higher headache response over multiple attacks compared with sumatriptan. Patients taking zolmitriptan for repeated attacks (up to six) responded more quickly with a significant difference between treatments seen by 1 hour (zolmitriptan, 5 mg, versus sumatriptan, 25 mg and 50 mg). Furthermore, zolmitriptan offers an additional benefit over sumatriptan in terms of pain relief over 24 hours. Considering that the median duration of untreated or unsuccessfully treated migraine headaches is 24 hours (mean 31 hours),20 the ability of zolmitriptan to provide greater pain relief over 24 hours than sumatriptan is of clinical interest. The consistency of response over multiple attacks demonstrated by zolmitriptan in this trial was similar to that seen in other studies.16,19

Zolmitriptan and sumatriptan were both well tolerated, and few patients withdrew from the study because of adverse events.

Conclusions

Overall, a 2.5-mg or 5-mg dose of zolmitriptan was at least as effective as sumatriptan, 25 mg or 50 mg, for all parameters studied. A 2.5-mg dose of zolmitriptan was significantly more effective than sumatriptan, 50 mg, in terms of headache response at 2 and 4 hours. In addition, patients taking zolmitriptan were significantly more likely to have pain relief over 24 hours than those taking sumatriptan.

The tolerability of zolmitriptan, 2.5 mg and 5 mg, was similar to that seen in previous trials and, in conjunction with the efficacy profile, confirms zolmitriptan, 2.5 mg, as the optimal dose. The results from this comparative study suggest that zolmitriptan may be more efficacious than sumatriptan for the acute oral treatment of migraine.

Other studies