TRANSACTIONSOF THEROYAL S~CIETYOFTROPICAL MEDICINEAND HYGIENE(~~~~) 89,103-106 103

Comparative trial of five antimicrobial compounds in the treatment of cholera in adults

Wasif Ali Khan*, Monira Begum, Mohammed Abdus Salam, Pradip Kumar Bardhan, Mohammed Rafiqul Islam and Dilip Mahalanabis Clinical Sciences Division, International Centre for Diarrhoeal Disease Research, Bangladesh, GPO Box 128, Dhaka 1000, Bangladesh

Abstract To compare the efficacy of ciprofloxacin, erythromycin, nalidixic acid and pivmecillinam in the treatment of tetracycline-resistant strains of Vibrio cholerae 01 in adults, a randomized, open, clinical trial was con- ducted. A tetracycline group was used for comparison. Seventy-five adult men infected with V. cholerae 01 were randomly assigned to receive either 400 mg pivmecillinam or 500 mg of one of each of the other drugs. Ciprofloxacin was given every 12 h and the others every 6 h for 3 d. The mean total stool volume per kg was 155 mL for the ciprofloxacin group, 212 mL for the erythromycin and pivmecillinam groups, 246 mL for nalidixic acid, and 293 mL for tetracycline. The difference between ciprofloxacin and tetracychne was signi- ficant (P=OXM5). After 72 h, diarrhoea had stopped in 14 patients (93%) in the ciprofloxacin group and 12 (80%) in the erythromycin group, compared to 5 (42%) of those receiving tetracycline (P=OXlO6 and 0.049, respectively). Bacteriological clearance was 100% at 24 h in patients treated with ciprofloxacin compared to 20% and 8.3% (P<OXtOl for both comparisons) in the erythromycin and tetracycline groups. Ciprofloxacin in conjunction with appropriate fluid therapy was the most effective treatment for cholera in adults; erythro- mycin was the next best.

Keywords: cholera, chemotherapy, antimicrobials, Bangladesh

Introduction Cholera is a severely dehydrating disease caused by

Vibrio cholerae 01 (see HOLMGREN, 1981). Correction of fluid and electrolyte deficits through appropriate in- travenous or oral rehydration therapy remains the most imnortant component of case management of cholera pa- tients; appropriate antimicrobial agents have been found effective in reducing stool volume, duration of diarrhoea, and duration of vibrio excretion (WALLACE et al., 1968; DE et al.. 19761. This reduces intravenous fluid reauire- ment, which can be very large, enabling hydration\0 be more easily maintained with oral rehydration therapy once initial hydration has been completed and the dura- tion of the hospital stay to be shortened. In controlled clinical trials of the treatment of cholera, tetracycline, chloramphenicol, erythromycin, trimethoprim-sulpha- methoxazole and furazolidone have been found to reduce the volume and duration of diarrhoea (LINDENBAUM et al., 1967; GHARAGOZLOO et al., 1970; CASH et al., 1973; BURANS et al., 1989; RABBANI et al., 1989). Tetracycline has been the drug of choice in the treatment of cholera patients in Bangladesh for more than 2 decades. Recentlv in Bangladesh,-however, in a span of less than one

i! ear

(Sentember 1991 to Tune 1992) the orevalence o V. thoierae strains resistant to tetracycline and trimethop- rim-sulphamethoxazole increased from 2% to 90%, and from 18% to 90%., respectively. This rapid and rather dramatic increase m the emergence of tetracycline-resis- tant strains prompted us to look for effective alternative antimicrobial compounds for use in the treatment of V. cholerae 01 infections.

Research, Bangladesh (ICDDR,B) from August 1992 to October 1992. Patients eligible for the studv were adult males, aged 18 to 65 years, who had a history of acute watery diarrhoea of not more than 24 h duration and moderate or severe dehydration according to the stand- ard criteria on arrival at the treatment centre (BENNISH, 1994). Patients who had received antimicrobial therapy for this episode of diarrhoea, or those with concomitant infections, were excluded from the study. Patients as- signed to the tetracycline group who were infected with tetracycline-sensitive strains of V. cholerae 01 were also excluded.

All tetracycline-resistant V. cholerae isolates in Dhaka to date remain susceptible to erythromycin, which is rec- ommended by the World Health Organization for treat- ing tetracycline-resistant V. cholerae 01 (WHO, 1983a). All V. cholerae 01 isolates tested in Dhaka also remain susceptible to nalidixic acid, pivmecillinam and the newer quinolones. In a recently conducted clinical study, norfloxacin was found to be more effective than trimeth- oprim-sulphamethoxazole in the treatment of cholera in adults (BHATTACHARYA et al., 1990).

We report here the results of a comparative, ran- domized study of 5 different antimicrobial agents in the treatment of tetracycline-resistant V. cholerae infections.

Patients and Methods This study was conducted at the Dhaka treatment

centre of the International Centre for Diarrhoeal Disease

*Author for correspondence and offprint requests.

Patients who fulfilled the initial selection criteria, and who provided informed consent, were weighed, then re- hydrated with an intravenous polyelectrol te solution containing 133 mmoYL of sodium, 13 mm0 6 L of potas- sium, 98 mmoYL of chloride and 48 mmol/L of bicarbo- nate (WHO, 1983a). A fresh stool sample was sent for dark-field microscopy for detection of V. chokrae.

Once the patients were fully hydrated (usually within 3 h of admission), thev were observed for 4 h to deter- mine the severity ‘of their diarrhoea. Patients with stool volumes of 20 mL/kg or more during the 4 h observation period and whose stool contained V. cholerae were en- rolled in the study. Patients were randomly assigned to receive one of 5 treatment regimens, recorded in sealed envelopes. A block randomized method, with a block size of 10, was used. Treatment regimens were (i) cipro- floxacin 500 ma everv 12 h. (ii) ervthromvcin 500 me every 6 h, (iii) &lid& acid 5dO mg every 6&h, (iv) tetra: cycline 500 mg every 6 h, or (v) pivmecillinam 400 mg every 6 h. The drugs were all in tablet form, and were given for 3 d.

On entry into the study a fresh stool specimen was sent for inoculation on to tellurite taurocholat gelatin agar, both immediately and after 6 h incubation in alkaline peptone water, for detection of V. cholerae bv standard procedures (W-HO, 1983b); drug susceptibility tests were done bv the disk diffusion method (Kirbv-Bauer) for ciprofloxacin,. erythromycin, tetracycline, nalidixic’acid and pivmecillmam. The disk potencies for ciprofloxacin, erythromycin, tetracycline, nalidixic acid and pivmecilli- nam were 5 pg, 15 pg, 30 pg, 30 pg and 25 1.18 respec- tively, and the zone duuneters used to determine suscep- tibility were 319 mm, >18 mm, 219 mm, >19 mm and >21 mm respectively. For V. cholerae resistant to tetra- cycline the zone diameter was > 14 mm. All patients then received their first dose of antimicrobial therapy accord- ing to their sealed envelope. Packed cell volume was measured for each patient at the beginning of the initial

104

hydration and after the end of the 4 h observation period. Stool cultures were also made, on each of 3 subsequent days after commencement of antimicrobial therapy, in al- kaline peptone water for detection of V. choZew 01. Physical examinations were made at entry, at enrohnent into the study, and then daily until discharge of the pa- tients. All findings were recorded on predesigned case re- port forms. Fluid intake and output charts were main- tained for each 8 h period from the time of enrolrnent for 72 h. Vital signs were recorded every 8 h during the period in hospital. Duration of diarrhoea was defined as the interval between the administration of the first dose of the study drug and the end of the last 8 h period dur- ing which a watery stool was passed. If watery stools con- tinued after 72 h of therapy, treatment was considered to have failed. Such patients were treated with oral rehydra- tion solutions (ORS) and, in addition, if purging was 300 mL or more during the first 8 h of study day 4, patients were treated with erythromycin 500 mg every 6 h for 3 d.

Data were entered into a personal computer and ana- lysed using Stat Pat Gold@, version 3.2 (Walonick Asso- ciates, Minneapolis, Minnesota, USA). The significance of differences between means of the 5 groups were tested with ANOVA and Student’s t test. When the data were skewed, the t test was performed after log transforma- tion. The ~2 and Fisher’s exact tests, when appropriate, were used to test the significance of differences between categorical variables. The Kaplan-Meier method and the log-rank test were used to test significance of differences between the treatment groups.

Results A total of 75 patients was initially enrolled into the

study; 3 patients (infected with tetracycline-sensitive strains and assigned to the tetracycline group) withdrew from the study on day 1 (2 patients) or were excluded from the analysis (one patient).

Thus, 72 patients were eligible for analysis; of these, 15 patients each received ciprofloxacin, erythromycin,

nalidixic acid or pivmecillinam and 12 patients received tetracycline. Eighty-three per cent of the patients were severely dehydrated at the time of arrival and all were in- fected with V. &.&rue 01, biotype El Tor and serotype Ogawa. Of the 72 patients, 54 (75%) were infected with tetracycline-resistant strains of V. cholerue 0 1.

Admission characteristics were comparable among the 5 treatment groups (Table l), as was severity of dehydra- tion (determined clinically and estimated by packed cell volume), on arrival at the treatment centre.

There were similar high rates of purging in all 5 treat- ment ately fl

roups (Table 2) and all patients remained adequ- ydrated during the 4 h observation period, as

determined clinically and by packed cell volume. Total stool volumes during the first 24 h period were also com- parable. The total stool volume during the entire period of diarrhoea was least in the ciprofloxacin group, fol- lowed by the pivmecillinam and erythromycin groups. Stool volume was high in the nalidixic acid and tetracy- cline groups. Mean stool output was significantly lower in the ciprofloxacin group than in the tetracycline group. Total ORS and other fluid (plain water and milk) intakes also differed significantly between the ciprofloxacin and tetracycline groups (Table 2).

In the ciprofloxacin group 14 patients (93%) were clinically cured by 72 h after the start of therapy; in the erythromycin, tetracycline, nalidixic acid and pivmecilli- nam groups the proportions were 80%, 42% 60% and 60% respectively. Cure rate at 72 h was significantly higher in the patients receiving ciprofloxacin and ery- thromycin than in those given tetracycline. Of the 23 pa- tients whose diarrhoea continued after 72 h, the single patient in the ciprofloxacin group, all 3 of those in the erythromycin group, one of those receiving nalidixic acid and 4 given tetracycline produced ~300 mL of watery, stool during first 8 h of study day 4 and the diarrhoea ceased by the end of day 4 without adding erythromycin to the treatment regimen; 5 patients in the nalidixic acid group, 6 in the pivmecillinam group and 3 in the tetracy-

Table 1. Admission characteristics of patients with Vibrio choferae 01

Treatment groupa Ciprofloxacin Erythromycin Nalidixic Acid Pivmecillinam Tetracycline

Number Age (years) Body weight (kg) Duration of illness (h) No. with severe dehydration Packed cell volume (%)

32:ll 29:12 34Yl5 32:12 3&o 42+5 42kll 40+5 44&7 42+5 16k5 13+5 13f6 15+6 15+6

‘;yy’ l&y4 W~y4 wuw 1; gw

‘Values are means&standard deviations (except for number and those with severe dehydration).

Table 2. Clinical course of patients with Vibrio cbolerae 01

Ciprofloxacin

Number 15 Stool output (mL/kg/h)b lo&5 Packed cell volume (%) 42+6 Total stool output (mL/kg)d 134+76 Total watery stool output GmWkg) 155+ 106 Total ORS intake (mL/kg) Total intake of other fluids (mL/kg)g :20;g5 No. with diarrhoea s72 hh 14 (93%)

Treatment groupa Erythromycin Nalidixic Acid Pivmecillinam Tetracycline

12:5 llY7 9:s 12:5 43f3 43+4 44+5 43*4

137+ 106 162+63 137f71 174-t 106 212+213 246f 129 212+151 293+216 285+216 346f 162 296& 187 356t 187 140+ 100 185f115 151*94 179+83 12 (80%) 9 (60%) 9 (60%) 5 (42%)

aValues are means&standard deviations (except for total number and those with diarrhoea <72 h). bDuring first 4 h. ‘After 4 h. din first 24 h. ‘Significance of differences: ciprofloxacin vs. tetracycline, P=O.O45; erythromycin vs. tetracycline, P=O*136. ‘Significance of differences: ciprofloxacin vs. tetracycline, P=O.O36. gSignificance of differences: ciprofloxacin vs. tetracycline, P=OzO12. hOverall significance of different treatments, P=O*O38 (x2 test); significance for ciprofloxacin vs. tetracycline, P=O.O06; for erythromycin vs. tetracycline, P=O.O49.

1

0.6

0.6

0.4

0.2 0 __--- ye7

_- _--- ---- I. - - - - - - --

01

0 6 16 24 32 40 46 56 84 72

Houn Fig. 1. Duration of diarrhoea in patients with Vibrio choleme 01, accord- ing to treatment groups.

cline group received supplementary erythromycin. All these patients were discharged by day 5 when diarrhoea ceased and were asked to take the remaining doses of ery- thromycin at home. The duration of diarrhoea in patients according to treatment groups is shown in Fig. 1. The differences between the groups receiving ciprofloxacin and tetracycline, and between those receiving erythro- mycin and tetracycline, were significant (log-rank test, P=O-004 and 0.043 respectively).

V. chokrae could not be isolated from the stool culture of any of the 15 patients receiving ciprofloxacin after 24 h of therapy. In comparison, V. cholerae could still be iso- lated from 80% of the erythromycin group, 92% of the tetracycline group, and 100% of the nalidixic acid and pivmecillinam groups (Fig. 2) after 24 h of therapy. Bac-

Admission Day 1 Day 2

Study days Day 3

m Ciprofloxacin r w Erythromycin Nalidlxic lcid

m Tetracycline 0 Pivmecillinam

Fig. 2. Isolation rates of Vibrio cholerae 01 from stool or rectal swab samples by treatment group and study day. On study days l-3 the dif- ferences between ciprofloxacin and tetracycline, nalidixic acid and pivme- cillinam were significant (P<O401). On study days 1 and 2 the differences between ciprofloxacin and erythromycin were significant (P<O.OOl and <0.05, respectively). On study days 2 and 3, the dif- ferences between erythromycin and tetracycline, nalidixic acid and piv- mecillinam were significant (PCO.01).

teriological cure was achieved significantly more often with ciprofloxacin therapy than with the other drugs after 24 h of therapy (PtO.OO1). At 72 h V. cholerae could not be isolated from any of the ciprofloxacin and erythromycin treated patients but was still present in 50% of the tetracycline group (P=O.O03). Although V. chokrae were susceptible in vitro to nalidixic acid and piv- mecillinam, bacteriological eradication was achieved in

105

only 27% and 47% of the patients in these 2 groups 72 h after initiation of therapy.

Discussion Patients treated with ciprofloxacin 500 mg every 12 h

for 3 d had significantly lower mean total stool volume and the number of patients whose diarrhoea continued beyond 72 h was lower than that of patients treated with tetracycline. In addition, vibrios were eradicated from the stools of patients treated with ciprofloxacin sooner than in the 4 other treatment groups.

Since all patients in the tetracycline group were in- fected with tetracycline-resistant strains of V. cholerae, we can consider that these 12 patients did not receive an effective antimicrobial agent. The mean total stool out- put in the ciprofloxacin group was only 53% of the mean output in the tetracycline group, a reduction of 47%, which is similar to the reductions seen in an earlier study, comparing tetracycline 500 mg every 6 h for 72 h to placebo (LINDENBAUM et al., 1967). The mean stool output in the erythromycin, pivmecillinam, and nalidixic acid groups was 72%, 72% and 84% of that in the tetra- cycline group, respectively. These differences were not changed appreciably if we included in the analysis the one patient in the tetracycline group infected with tetra- cycline-sensitive V. cholerae 01. In the tetracycline group, all of whom were infected with V. d&rue 01 re- sistant to tetracycline, the mean stool output until diar- rhoea stopped was only 293 mL/kg compared to 499 mL/kg in patients treated with an inactive placebo in a previous trial (ISLAM, 1987). Although the percentage re- duction in stool output in the tetracycline group com- pared to that in the ciprofloxacin group was similar to that seen in the earlier placebo trial, the difference in mean stool output between patients with tetracycline-re- sistant V. cholerae 01 who were treated with tetracycline and those receiving the inactive placebo was marked.

Our primary objective was to look for alternative drugs to tetracycline for the treatment of cholera. Although nalidixic acid and pivmecillinam had some effect on total stool volume, compared to tetrac cline in the treatment of tetracycline-resistant strains o ry cholera the effect was not significant. In our view, these drugs should be ke in reserve for the treatment of shigellosis. The Word P

t

Health Organization recommends erythromycin for pa- tients with tetracycline-resistant strains of V. chokrae and our results support this. Ciprofloxacin was more effective than erythromycin in rapidly clearing V. cholerae 01 from faeces. Due to the increased frequency of isolation of tetracycline-resistant strains of V. chokrae 01 in dif- ferent parts of the world, erythromycin and ciprofloxacin may be effective alternatives for the treatment of multi- resistant strains. The very rapid bacteriological cure achieved with ciprofloxacin suggests that short-course therapy with this drug may be effective for the treatment of cholera. If so, it may become a suitable alternative for the treatment of a large number of cholera cases during epidemics.

Until short-course ciprofloxacin has been evaluated in larger clinical studies, erythromycin should remain the drug of choice in the treatment of tetracycline-resistant strains of V. cholerae 0 1, because it is less expensive than ciprofloxacin. Furthermore, ciprofloxacin should not be used indiscriminately at this stage, since it is an import- ant reserve drug for treatment of multi-resistant enteric and other pathogens.

Acknowledgements This research was supported by the International Centre for

Diarrhoeal Disease Research, Bangladesh (ICDDR,B). The ICDDR,B is supported by countries and agencies which share its concern for the health problems of developing countries. Current donors include: the aid agencies of the Governments of Australia, Bangladesh, Belgium, Canada, China, Japan, Saudi Arabia, Sweden, Switzerland, the United Kingdom and the United States; international or anizations including the Arab Gulf Fund, United Nations thildren’s Fund (UNICEF), the

106

United Nations Development Programme and the United Na- tions Pooulation Fund.

We thank Mrs Josephine Sack for editing the manuscript, Mr Humayun Kabir for assistance with data entry and analysis, and the staff of the Diarrhoeal Treatment Centre of the ICDDR,B for their excellent patient care.

References Bennish, M. L. (1994). Cholera: pathophysiology, clinical fea-

tures, and treatment. In: Vibrio cholerae and Cholera: Molecu- lar to-Global Perspectives, Wachsmuth, I. K., Blake, P. A. & Olsvik, 0. (editors). Washington, DC: American Society for Microbiology, pp. 229-255.

Bhattacharya, S. K., Bhattacharya, M. K., Dutta, D., De, S. P., Sikder, S. N., Maitra, A., Dutta, A. & Pal, S. C. (1990). Double-blind, randomized, controlled clinical trial of norflox- r3kp4; cholera. Antimicrobial Agents and Chemothera~, 34,

Burans, J.’ P., Podgore, J., Mansour, M. M., Farah, A. H., Abbas, S., Abu-Elyaxeed, R. & Woody, J. N. (1989). Com- parative trial of erythromycin and sulphatrimethoprim in the treatment of tetracycline-resistant Vibrio cholerae 0 1. Transac- tions of the Royal Society of Tropical Medicine and Hygiene, 83, 836-838.

Cash, R. A:, Northrup, R. S. & Rahman, A. S. M. (1973). Tri- methoprun and sulphamethoxaxole in clinical cholera: com- f;e;;3with tetracycline. Journal of Infectious Diseases, 128,

De, S., Chaudhuri, A., Dutta, P., Dutta, D., De, P. S. & Pal, S. C. (1976). Doxycycline in the treatment of cholera. Bulletin of the World Health Organization, 54,177-179.

Gharagozloo, R. A., Naficy, K., Mouin, M., Nasstrzadeh, M. N. & Yalda, R. (1970). Comparative trial of tetracycline, chloramphenicol, and trimethoprim-sulphamethoxazole in er- adication of Vibrio cholerae El Tor. British MedicalJournal, iv, 281-282.

Holmgren, J. (1981). Actions of cholera toxin and the preven- tion and treatment of cholera. Nature, 292., 413-417.

Islam, M. R. (1987). Single dose tetracychne in cholera. Gut, 28,1029-1032 --_ ----.

Linden Ibaum, J., Greenough, W. B., III & Islam, M. R. (1967). Antibiotic therapy of cholera. Bulletin of the World Health Or- ganization, 36,871-883.

Rabbani, G. H., Islam, M. R., Butler, T.. Shahrier. M. & Alam, K. (1989). Single dose.treatment of cholera with fu- raxolidone or tetracycline in a double-blind randomized trial. Antimicrobial Agents and Chemotherapy, 33,1447-1450.

Wallace, C. K., Anderson, P. N., Brown, T. C., Khanra, S. R., Lewis, G. W., Pierce, N. F., Sanyal, S. N., Segre, C. V. & Waldman, R. H. (1968). Optimal antibiotic therapy in cho- lera. Bulletin of the World Health Organization, 39,239-245.

WHO (1983a). A manual for the treatment of acute diarrhoea. Geneva: World Health Organization, mimeographed publica- tion no. WHOCDDSER83.2.

WHO (1983b). Manual for laboratory investieation of acute enteric infection. Geneva: World Health drgani&on, mimeographed publication no. CDD/83.3.

Received 4 May 1994; revised 30 June 1994; accepted for publication 3OJune 1994

1 Announcement 1

African Index Medicus (AIM) Project An international index to African health literature and information sources

In order to give access to information published in or related to Africa and to encourage local publishing, the Association for Health Information and Libraries in Africa (AHILA), with the technical support of the World Health Organization, has initiated a project to create an international index to African health litera- ture and information sources-the African Imlex Medicus.

The creation of the regional index is a collaborative, participatory process. Firstly, national databases are being built, using a common methodology, in African countries. From them, local information services and products will be provided for national health professionals. National production should ensure self-suffi- ciency and sustainability at country level and the tailoring of services according to local needs.

The various national databases are then merged into a regional data base to which are added bibliographic records relating to health in Africa from other international existing sources such as WHO’s WHOLIS, MEDLINE, POPLINE etc. to produce the Aj-ican Index Medicus in printed or electronic form, eventually CD-ROM. It is distributed to African countries and as part of an affiliated membership to AHILA for institutions outside the region.

At this stage, AHILA, with support from WHO, is looking for further sponsoring partners at bilateral level with African countries not yet participating in the Project. Sponsorship comprises equipment and training of staff and could be part of an information component of a health-related project in the country, which may also include use of communications and CD-ROM.

Further information can be obtained from Dr Deborah Avriel, World Health Organization, Library, 1211 Geneva 27, Switzerland; fax +41 22 791 0746.