Quinolone Resistance in Enterotoxigenic Escherichia coli Causing Diarrhea in Travelers to India in Comparison with Other Geographical Areas

Abstract

Enterotoxigenic Escherichia coli isolates were identified as a cause of traveler's diarrhea in 82 of 520 (16%) patients and tested for resistance to seven antimicrobial agents. Thirty patients (36%) needed antimicrobial therapy: 17 (56%) for persistence of symptoms and 13 (44%) for severity of symptoms. Ampicillin, tetracycline, and trimethoprim-sulfamethoxazole resistance was high. Chloramphenicol showed moderate activity, and amoxicillin plus clavulanic acid, nalidixic acid, and ciprofloxacin showed very good activity. Five nalidixic acid-resistant strains were isolated, four from patients visiting India.

Enterotoxigenic Escherichia coli (ETEC) is the major cause of traveler's diarrhea in people from industrialized countries visiting less-developed countries (6, 7) and is an important cause of dehydrating diarrhea in infants and children in less-developed countries (13). Traveler's diarrhea caused by ETEC strains is usually a mild, self-limited disease, but for severe traveler's diarrhea, early treatment with loperamide and an antibiotic such as trimethoprim-sulfamethoxazole, doxycycline, or a fluoroquinolone has been recommended (4).

Many previous studies of antimicrobial susceptibility of ETEC involved a small number of isolates from a single geographic location. We, therefore, performed antimicrobial susceptibility testing of ETEC isolates causing traveler's diarrhea originating from diverse geographical locations. Investigation of the mechanisms of acquisition of quinolone resistance in the nalidixic acid-resistant ETEC strains was also performed.

During the period from 1994 to 1997, stool specimens from 520 adult patients with traveler's diarrhea were analyzed. The patients were recruited from the traveler's clinic of the Tropical Medicine Department of the Hospital Clinic. All patients had diarrhea on arrival in Spain or within 2 days after their return. The stool specimens were cultured for E. coli and other bacterial enteropathogens by conventional methods (11). Single-colony subcultures of all different lactose-fermenting colonial morphotypes growing on MacConkey agar were identified by conventional criteria. The E. coli isolates were tested by PCR to detect the heat-stable (ST) and heat-labile (LT) toxin genes (15). ETEC strains were isolated from 82 patients (16%) with traveler's diarrhea. The distribution of these strains according to the type of enterotoxin synthesized was as follows: 58 strains (71%) produced the ST, 11 strains (13%) produced LT, and 13 strains (16%) produced both toxins (LT/ST). ETEC strains were isolated from stool samples of patients traveling to different tropical and subtropical areas, except for Central and South Africa. The range of prevalence was from 7.5% to 31%, with West Africa and the Indian Subcontinent being the two geographic areas where ETEC strains were more prevalent, at 31 and 22%, respectively. Thirty patients (36%) needed antimicrobial therapy: 17 patients (56%) because of persistence of symptoms and 13 (44%) because of severity of symptoms. In 14 treated patients, the cause of diarrhea was ST-producing ETEC (ETEC-ST), representing 24% of the total number of strains of ETEC-ST, whereas in 7 (63%) and 9 (69%) patients, the cause was LT- and LT/ST-producing ETEC, respectively (P < 0.02). All patients were empirically treated with ciprofloxacin, and in all, the duration of diarrhea was shortened and the accompanying symptoms such as abdominal discomfort, flatulence, nausea, and vomiting were relieved. It is noteworthy that patients with ETEC-LT and ETEC-LT/ST strains required antimicrobial therapy more frequently than patients with ETEC-ST strains. However, this finding needs to be confirmed in future studies.

The MICs for the clinical isolates studied were determined by E-test according to standard practice. E. coli ATCC 25922 was used as a reference strain for quality control. The breakpoints considered to define a resistant strain were those recommended in reference 12. The MICs of antimicrobial agents for ETEC-ST, ETEC-LT, and ETEC-ST/LT strains are shown in Table 1. Randomized, controlled studies have demonstrated that a 3- to 5-day course of an antibiotic can reduce the duration of an acute diarrheal episode in travelers. Most studies done during the 1980s with tetracycline (8) or with trimethoprim alone (1) or in combination with sulfamethoxazole (1, 2) showed the effectiveness of these antimicrobial agents in the treatment of acute diarrhea caused by ETEC, shortening the duration of excretion of microorganisms, resulting in earlier termination of illness. However, the emergence and dissemination of resistance to cotrimoxazole in fecal E. coli were observed during oral administration of this agent (9). Resistance to cotrimoxazole and tetracycline is now present in >50% of ETEC strains isolated from tropical and subtropical countries (Table 1) (10). This fact is due to the extensive use of these antimicrobial agents in these countries (10) and precludes their empirical use for the treatment of traveler's diarrhea. Among the other antimicrobial agents tested in this study, ampicillin was moderately effective. Chloramphenicol presented good activity against ETEC, but quinolones such as nalidixic acid or ciprofloxacin and amoxicillin plus clavulanic acid showed the best activity against these microorganisms. In the last decade, fluoroquinolones have been shown to be highly effective in reducing the duration of traveler's diarrhea (3, 5, 14, 18). However, 5 (6%) of the 82 ETEC strains investigated in our study were resistant to nalidixic acid, and one of them was also resistant to ciprofloxacin (MIC of 32 μg/ml).

TABLE 1.

MICs of seven antimicrobial agents for 82 ETEC clinical isolates

Antibiotic	Result for strainsa
	ETEC-ST (n = 58)			ETEC-ST/LT (n = 13)			ETEC-LT (n = 11)
	MIC50	MIC90	% Susceptible	MIC50	MIC90	% Susceptible	MIC50	MIC90	% Susceptible
Ampicillin	4	256	55	2	256	77	256	256	46
Amoxicillin-clavulanic acid	3	6	100	3	4	100	3	3	100
Chloramphenicol	4	256	83	4	256	77	4	256	73
Ciprofloxacin	0.012	0.023	98	0.012	0.016	100	0.012	0.016	100
Nalidixic acid	3	6	95	3	3	92	3	12	91
Tetracycline	256	256	42	4	256	54	256	256	37
Trimethoprim-sulfamethoxazole	32	32	47	0.047	32	85	32	32	27

^aMIC₅₀ and MIC₉₀, MICs at which 50 and 90% of the isolates tested are inhibited, respectively. % Susceptible, percentage of strains susceptible to the breakpoint concentration. 

The quinolone resistance in E. coli has mainly been associated with mutations in the gyrA and parC genes (16, 17). All nalidixic acid-resistant ETEC strains were analyzed for mutations in the gyrA and/or parC genes by PCR and DNA sequencing of the quinolone resistance-determining region as described in references 16 and 17 (Table 2). One ETEC-ST strain for which the MICs were 32 (nalidixic acid) and 0.032 (ciprofloxacin) μg/ml had a substitution in Ser-83 of the GyrA protein to Ala. Three strains (one each of ST, LT, and ST/LT) for which the MIC of nalidixic acid was 256 μg/ml and that of ciprofloxacin was between 0.19 and 0.25 μg/ml showed a substitution of Ser-83 by Leu. These results confirm that a substitution at Ser-83 is sufficient to confer a very high nalidixic acid resistance level. Finally, one strain for which the MIC of ciprofloxacin was 32 μg/ml and that of nalidixic acid was >256 μg/ml showed three mutations—two in the gyrA gene, which generated substitutions of Ser-83 to Leu and Asp-87 to Asn, and the third in the parC gene, producing a change in Ser-80 to Ile. In E. coli, a higher quinolone resistance level has been related to the presence of concomitant mutations in the gyrA and parC genes (16, 17).

TABLE 2.

Mutations in the gyrA and parC genes of quinolone-resistant clinical isolates of ETEC

Strain	MIC (μg/ml)		Amino acid change
	Ciprofloxacin	Nalidixic acid	GyrA		ParC
			Position 83	Position 87	Position 80	Position 84
Wild type			Ser	Asp	Ser	Glu
135-ST	0.032	32	Ala
265-ST	0.25	256	Leu
12-ST	32	>256	Leu	Asn	Ile	Gly
225-ST/LT	0.19	256	Leu
299-LT	0.19	256	Leu

Of the five nalidixic-acid resistant ETEC strains isolated, four (17%) were isolated from patients who had traveled to India. Nalidixic acid was introduced a few years ago as the first-line therapy for shigellosis in some areas of this country, and therefore ETEC strains resistant to nalidixic acid are now emerging.

In summary, ETEC strains, and in particular ETEC-LT strains, are a frequent cause of traveler's diarrhea requiring antimicrobial therapy. The resistance of these microorganisms to the classical agents such as cotrimoxazole, tetracycline, and ampicillin is very high due to the widespread use of these antimicrobial agents in tropical and subtropical countries. Fluoroquinolones are drugs of choice for most adults traveling to high-risk parts of the world. Although the level of resistance to nalidixic acid is generally low, in some geographical areas such as India, it is a matter for concern. Strains for which the MICs of nalidixic acid are high are likely to have at least one mutation of the gyrA gene. The induction of further mutations, e.g., by exposing such strains to fluoroquinolones, is likely to result in a significantly reduced susceptibility to fluoroquinolones. These agents should therefore be used with caution. Furthermore, it is important to continue the surveillance of quinolone resistance levels in these microorganisms.