Patients, Materials, and Methods

PatientsThe study was conducted at the Dhaka Hospital of the Center for Health and Population Research at the International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR,B) during November 1999–June 2001. The patients were selected from the outpatient department and were admitted into the research ward for 72 h

The only patients admitted into the study were male adults, 15–60 years old, who fulfilled the following criteria: (1) presented with the abrupt occurrence of ⩾3 unformed stools (soft or watery) within the preceding 24 h in association with ⩾1 symptom (e.g., nausea, vomiting, abdominal pain/cramps, excess gas/flatulence, urgency, and incontinence); (2) had >3 unformed stools in the preceding 24 h; (3) had been otherwise healthy; (4) had been able to take oral medication and ORS; and (5) had no history of treatment with antimicrobial or antidiarrheal agents within the preceding 7 days. Informed consent was obtained from all patients. Only male adults were studied because stool could easily be collected separately from urine and because there is little difference between the sexes with regard to therapeutic response in cholera. The Ethical Review Committee of the ICDDR,B approved the study protocol

A total of 135 patients who met the admission criteria were initially selected, and a computer-generated table of random numbers was used to assign the patients to 1 of 2 treatment groups. The sample size estimation was based on an assumption of at least a 30% reduction in the volume of stool in the patients given ORS with l-histidine at a 5% significance level with 80% statistical power. Of the 135 patients enrolled, 126 successfully completed the 5-day treatment protocol; 9 were withdrawn from the study because of failure to isolate V. cholerae from their stool or the development of concurrent infections

Pretreatment observations (screening)A complete history of the current illness was obtained, and a physical examination was performed. The patients were placed on a cholera cot, to facilitate the separate collection of stool and urine. During the initial screening, before assignment to a treatment group, patients were rehydrated with either ORS or intravenous (iv) fluid and were observed for 6 h, during which time their baseline clinical data (e.g., body weight; number, volume, and consistency of stool; urine volume; intake of ORS and iv fluid; status of hydration; and any other gastrointestinal-related complaints) were recorded

Fresh stool samples were obtained for darkfield microscopic examination, to detect the presence of motile V. cholerae. Only patients who had V. cholerae in their stool samples and who produced ⩾4 mL/kg of body weight/h during this 6-h observation period were admitted into the study

Treatment protocolThe patients were randomly assigned to 1 of 2 treatment groups: those given CeraLyte-90 (CERA Products) supplemented with l-histidine (2.5 g/L of ORS) (Ajinomoto Chemical) and those given CeraLyte-90 without l-histidine. The assignment of patients to treatment groups was blinded from the clinical observers by a process in which a random number was written on a slip of paper, the slip was concealed in an opaque envelope, and the envelope was opened only by the pharmacist on duty at the time that a patient was allocated to a treatment group. The 2 ORS preparations looked identical and were indistinguishable by taste or color. l-histidine is a tasteless, odorless, whitish crystalline powder that can be easily dissolved in water. CeraLyte-90 is a rice-based oral electrolyte solution that was supplied as 50-g sachets. It contains rice syrup solids, trisodium citrate, sodium chloride, and potassium chloride; when reconstituted to a volume of 1.0 L, it contains 165 calories, is composed of 40g of carbohydrate, 90 mmol/L sodium, and 20 mmol/L potassium, and has an osmolality of 255 mOsm

ORS was provided at the beds to the patients, who drank the solution freely under the observation of the nursing staff. A physician evaluated the patients twice daily and specifically assessed the clinical severity of their illnesses. Each patient’s weight of stool, vomit, and urine and his physical characteristics were recorded by specially trained staff. Records of amounts of fluid intake (of ORS, water, and iv fluid), and output (of stool, urine, and vomit), assessed at 8-h intervals, were maintained daily until the end of the study. Because cholera is an acute disease of short duration, collecting stool and reporting the frequency and weight of stool output at 8-h intervals has been found to be a useful method of comparing data and detecting short-term therapeutic effects. Study personnel also assessed adverse effects and checked stool consistency every 8 h. The clinical staff also evaluated gastrointestinal symptoms and assessed daily fluid requirements to correct dehydration and maintain homeostasis by matching the intake of ORS volume-for-volume with loss of stool. As a standard antimicrobial treatment for cholera, oral ciprofloxacin was given to all patients at a dose of 500 mg every 12 h for 3 days, beginning at the time of admission into the study

Laboratory investigationsStool specimens were taken at the time of admission into the study, for isolation of V. cholerae, Shigella and Salmonella species by use of standard microbiological methods [18]. Venous blood (3–5 mL) was obtained for the determination of levels of serum electrolytes, hematocrit, and complete blood counts at the time of admission into the study and at the time of discharge from the hospital. Liver function tests and renal function tests (to measure levels of serum bilirubin, glutamic oxalacetic transaminase, glutamic pyruvic transaminase, and creatinine) were performed at the time of admission into the study and at the time of discharge from the hospital, to measure general metabolic condition

Definitions of clinical variables“Liquid stool” was defined as feces that took the shape of the container and could be easily transferred without sticking to the surface of the container. “Soft stool” was defined as feces that retained its shape, did not stick to the container, and had no mucus. “Duration of diarrhea” was defined as the number of hours that elapsed between initiation of ORS treatment and passage of the last liquid stool that was not followed by another abnormal stool within 12 h. The World Health Organization standard criteria for “some” and “severe” dehydration were used: “some” dehydration was defined as 5%–10% loss of body weight (BW) accompanied by poor skin turgor, thirst, irritability, restlessness, and sunken eyes; “severe” dehydration was defined as >10% loss of BW accompanied by poor skin turgor, thirst, irritability, restlessness, sunken eyes, lethargy, and unconsciousness

Evaluation of adverse effectsDuring the assessments performed every 8 h, patients were asked to report any adverse effects they experienced and to describe their general condition. Any reported unusual symptoms were immediately brought to the attention of the investigators. A list of general clinical symptoms was carefully checked and reviewed every day throughout the treatment period

Statistical analysisStatistical analysis of data was performed using SPSS software (version 11.0; SPSS), to compare the differences between the 2 treatment groups with regard to their baseline characteristics and their outcome variables. Appropriate statistical tests were used to determine the probability level of significant differences between the groups; Student’s t test, χ² test, and Wilcoxon&amp;rank sum test (nonparametric) were used in most cases. P<.05 was considered to be significant

Results

Clinical characteristicsTable 1 shows the clinical characteristics of the study patients at the time of admission to the hospital. Most of these adult male patients were ∼30 years old and presented with watery stool, dehydration, and vomiting of short duration (10–12 h). Some patients presented with shock, as indicated by deviations in their vital signs (pulse rate, respiration rate, and blood pressure) from reference values. As was expected, fever was present in only a small proportion of patients

After admission to the hospital, 62 patients received CeraLyte-90 ORS with l-histidine, and 64 patients received CeraLyte-90 ORS without l-histidine. After the pretreatment observation period of 4–6 h, in which patients received iv and/or oral rehydration, the hydration status improved in all patients, as was indicated by changes in vital signs, gain in body weight, and establishment of urine flow. As their hydration status improved, patients started producing stool in large amounts (∼10–12 mL/kg of body weight/h), and the frequency of stool output during the 6-h observation period was also high (∼20 bowel movements). As was consistent with the results of the initial darkfield examination of the stool samples, all patients had V. cholerae isolated from their stools. V. cholerae El Tor O1 was the predominant isolate (in 68%–77% of patients), and V. cholerae O139 was isolated from other patients

At the time of admission to the hospital, patients in the 2 treatment groups were comparable with regard to their clinical characteristics. During the pretreatment observation period, the number of stools and the weight of stool output were comparable in both groups. There were no statistically significant differences between the 2 treatment groups in the distribution of any of the clinical characteristics listed in table 1. Patients infected with either V. cholerae El Tor O1 or V. cholerae O139 did not differ in their pretreatment characteristics, number of stools, and output of stool after initiation of ORS treatment

Reduction of stoolThe frequency of stool output is shown in figure 1. During the pretreatment observation period, the frequency of stool output was comparable in patients in both groups. However, during the 72-h treatment period, patients given ORS with l-histidine had a significant reduction in the frequency of stool output, compared with that in patients given ORS without l-histidine. Administration of ORS with l-histidine significantly (P<.05) reduced the frequency of stool output during 32–48 h and 56–64 h of ORS treatment, compared with that in the control group ([all data are means±SD] 32–48 h, 11.5±6.9 mL/kg vs. 18.8±16.0 mL/kg; 40–48 h, 6.7 ± 4.4 mL/kg vs. 11.5±9.7 mL/kg; and 56–64 h, 6.3±5.8 mL/kg vs. 7.8±4.1 mL/kg)

The decline in stool output started from the first 8-h period after initiation of ORS treatment, but the differences were significant (P<.05) only at 32 h after initiation of ORS treatment. However, during the first 32 h of ORS treatment, there was a persistent reduction in stool output in patients given ORS with l-histidine, compared with that in patients given ORS without l-histidine, but these differences were not statistically significant

Figure 2 shows the cumulative mean weight of stool during 72 h after initiation of ORS treatment. The patients given ORS with l-histidine had a consistent and increasing trend of a reduction in the mean weight of stool at each 8-h interval throughout the treatment period, compared with that in the patients given ORS without l-histidine, although the differences between the 2 groups were not statistically significant. During 72 h after initiation of ORS treatment, the total mean weight of stool was lower in patients given ORS with l-histidine than in patients given ORS without l-histidine ([all data are means ± SD] 8.6±3.6 kg vs. 10.5±6.5 kg [P<.05]). During 0–24 h after initiation of ORS treatment, the volume of stool in patients given ORS with l-histidine was 12%–20% less than that in patients given ORS without l-histidine, and during 24–48 h after initiation of ORS treatment, the volume of stool in patients given ORS with l-histidine was 30%–35% less than that in patients given ORS without l-histidine. An overall reduction of 25% in the mean volume of stool was observed in patients given ORS with l-histidine, compared with that in patients given ORS without l-histidine

Reduction of unscheduled iv fluidFigure 3 shows the requirement of unscheduled iv fluid during the treatment periods. There was a significant (P<.01) reduction in the mean volume of unscheduled iv fluid required to maintain hydration in patients given ORS with l-histidine, compared with that required in patients given ORS without l-histidine. Patients given ORS with l-histidine required ∼50% less iv fluid during 0–24 h and 24–48 h after initiation of ORS treatment, compared with the amount of fluid required in patients given ORS without l-histidine; however, no patients required iv fluid during 48–72 h after initiation of ORS treatment because of decreasing amounts and frequency of diarrhea ([all data are means ± SD] 0–24 h, 82.5±44.4 mL of fluid/kg of body weight vs. 158.6 ± 72.2 mL of fluid/kg of body weight [P<.01]; and 24–48 h, 41.6±40.4 mL of fluid/kg of body weight vs. 52.5±22.1 mL of fluid/kg of body weight [P<.05])

Reduction in duration of diarrheaThe duration of diarrhea was significantly shorter in patients given ORS with l-histidine than in patients given ORS without l-histidine ([all data are means±SE] 42.7±1.7 h vs. 47.0±1.8 h [P<.05]). Of the patients given ORS with l-histidine, 90% stopped having watery stools within 48 h of initiation of ORS treatment, whereas only 78% of the patients given ORS without l-histidine did so

ORS intakeThe total amount of ORS consumed by patients given ORS with l-histidine was consistently smaller than that consumed by patients given ORS without l-histidine at each 8-h interval. However, these differences were statistically significant only during 48–72 h after initiation of ORS treatment ([all data are means±SD] 0–24 h, 214±67 mL ORS/kg of body weight vs. 219±74 mL ORS/kg of body weight; 24–48 h, 108±55 mL ORS/kg of body weight vs. 126±78 mL ORS/kg of body weight; and 48–72 h, 41±25 mL ORS/kg of body weight vs. 64±48 mL ORS/kg of body weight [P = .05]) (figure 4)

Safety evaluation of ORS withl-histidineBefore initiation of ORS treatment, almost all patients had normal serum levels of liver enzymes (alanine aminotransferase [ALT], <40 U/L; aspartate aminotransferase [AST], <37 U/L; alkaline phosphatase, <90 U/L; and bilirubin, <17 μmol/L). However, after 72 h of ORS treatment, ALT and AST levels were significantly increased in both groups (table 2). Of the patients given ORS with l-histidine, ∼10% had values for ALT and AST that were twice the reference values. Similar increases were also observed in patients given ORS without l-histidine. No significant differences in these values were observed between the groups. Also, no significant differences in bilirubin and alkaline phosphatase concentrations were observed before or after initiation of ORS treatment in either group. No clinical signs or symptoms related to dysfunction of the hepatobiliary system were observed. Regular screening of peripheral blood counts also revealed an increase in the percentage of eosinophils (∼7%–8%), relative to these values at the time of admission to the hospital, after initiation of ORS treatment in both groups

Discussion

In the present study, we observed that l-histidine given with a rice-based ORS significantly reduced the frequency and weight of stool output, duration of illness, and requirement of iv fluid in a group of adult men with severe cholera in Bangladesh. A reduction of >20% in the volume of stool was observed primarily during 32–48 h after initiation of ORS treatment. These observations need to be interpreted in the light of 3 important characteristics of cholera

First, cholera has a short incubation period (a few hours to a few days) and a short duration (2–3 days), and most fluid loss occurs during the first 48 h of the disease, making it the most critical time period for treatment [19]. In patients given ORS with l-histidine, a declining trend in stool output was seen at 8 h after initiation of ORS treatment, and this trend persisted throughout the treatment period. We observed a statistically significant reduction during 32–48 h after initiation of ORS treatment, which is within the critical time period of the disease. Because a large volume of fluid loss occurs in a relatively short period of time (usually 10–30 L in 3 days), even a modest reduction in this loss could be highly useful, especially when supplies for treatment are limited during epidemics in rural areas

Second, all patients received an effective antibiotic, ciprofloxacin, which alone reduces the volume of stool by >50% in patients with cholera [19]. Similarly, rice-based ORS has also been shown to reduce the volume of stool in patients with cholera [20–22]. Therefore, the additional reduction (>20%) in the volume of stool in patients given ORS with l-histidine suggests that l-histidine is highly efficacious. Because withholding an effective antimicrobial in the treatment of cholera is considered to be unethical, l-histidine should be considered an effective adjunct treatment when administered with fluids and antibiotics

Administration of ORS with l-histidine also reduced the requirement for iv fluid by 50%–60%, and this leads to a net savings of 3–4 L of fluid/day/patient. This substantial savings in iv fluid and its associated cost of administration is clinically important, because iv fluid is given only to severely ill patients whose ORS treatment fails. During outbreaks of cholera, there will be many severe cases of cholera that put patients at risk of death, and administration of ORS with l-histidine could be lifesaving for them

The observed reduction in the need for ORS that was associated with administration of the l-histidine–supplemented preparation could be substantial, because a large number of patients with cholera are treated each year at any health center in an affected area. Administration of ORS with l-histidine also reduced the duration of diarrhea and vomiting significantly, resulting in a net reduction of 5 h of severe illness. Thus, patients could be released sooner than is typical (within 48 h of the time of admission to the hospital) and thereby offer a substantial savings in the cost of hospital care. For example, at the ICDDR,B, a net savings of $10.40/patient/day (35% of a hospital stay at $30/day) has been estimated, and this calculation indicates that there would be a substantial savings for an average of 100,000 patients with cholera treated per year

l-histidine is an essential amino acid that has been well studied for its nutritional and metabolic effects. Recent observations have indicated that l-histidine possesses potent antioxidant and antiinflammatory activities [12]. In cholera, secretion from the small-bowel epithelium is induced by the stimulation of ecosanoids, including prostaglandin E₂, by CT, and these ecosanoids can alter levels of water in the intestinal mucosa and can affect the transport of electrolytes [10, 11]. CT is also known to stimulate levels of adenylate cyclase, 3,5-cAMP, and PGE₂ in the intestines, and it also provides a role for 5HT and the enteric nervous system [12–15]

We have published elsewhere the results of a study on the molecular effects of l-histidine and imidazole on CT-induced PGE₂ activity and secretion of fluids [16]. In a murine model, l-histidine reduced the net secretory response mediated by the inhibitory effects of l-histidine adducts on intestinal adenylate cyclase. l-histidine and imidazole reacted with various prostaglandins, and this leads to the formation of new molecular species that can block cAMP formation by inhibiting adenylate cyclase. An l-histidine–mediated reduction in the secretion of fluids could also be explained by other mechanisms not yet described

Compared with other antisecretory agents (chlorpromazine, nicotinic acid, and berberine), whose effects have been reported elsewhere [5–8], l-histidine seems to be a more promising compound. A few other amino acids, including glycine and alanine, that have been used as sodium promoters in ORS have shown some initial promise and need to be studied further [23, 24]

The effects of l-histidine in noncholera diarrheas—such as those caused by E. coli, rotavirus, and Shigellae—are not known, and these need to be studied. Whether l-histidine would be similarly effective when given with glucose-based ORS, particularly for diarrhea in children, also needs to be established. Another challenging task would be to examine whether administration of a rice-based ORS with l-histidine could obviate the need for antimicrobial agents in the treatment of cholera

l-histidine is an amino acid that is safe for human consumption. Our adult patients who received l-histidine at a total dose of 20–30 g over 3 days did not report any adverse effects. Only a few complained of mild soreness of skin, which disappeared spontaneously. There were no clinical abnormalities with regard to liver and kidney function and hematologic characteristics, as indicated by laboratory findings. However, ∼15%–20% of patients in both groups had mild to moderate increases in their serum levels of liver enzymes, although these changes are unlikely to have been related to the administration of l-histidine. We have also evaluated the administration of l-histidine (180 mg/kg for 5 days) in young Bangladeshi children and found no safety concerns (G.H.R., unpublished data). No researchers have reported safety concerns other than the mild symptoms observed in a few subjects who received large doses of l-histidine [25–27]. Our patients had marginal increases in the number of circulating eosinophils during ORS treatment. The mechanism of eosinophilia during treatment of cholera is not known but has been observed in cholera in children (D.A.S., unpublished data)

In conclusion, l-histidine was found to be a clinically useful and safe antisecretory agent when given with a rice-based ORS as an adjunct treatment of severe cholera in adults. Further studies are needed to determine its effectiveness in children and in patients who have diarrhea caused by noncholera intestinal infections