Resistance to ciprofioxacin in non-typhoidal salmonellas from humans in England and Wales—the current situation

Abstract

Objective: To report on the prevalence of isolates with ciprofloxacin resistance in non-typhoidal salmonellas from humans in England and Wales in 1997.

Methods: All non-typhoidal salmonellas referred to the Laboratory of Enteric Pathogens in 1997 were screened for resistance to ciprofloxacin at 0.125 and 1.0 mg/L and nalidixic acid at 16 mg/L, and results were compared to those for 1994. Full minimal inhibitory concentrations (MICs) of these antimicrobials were also determined for a selection of isolates resistant to ciprofloxacin at 0.125 mg/L but sensitive at 1.0 mg/L, and for all isolates resistant at 1.0 mg/L.

Results: Since 1994 there have been increases in the occurrence of resistance to ciprofloxacin (MICs: 0.25–1.0 mg/L) in Salmonella enterica serotypes Enteritidis, Typhimurium, Virchow and Hadar. Of particular importance have been increases in the occurrence of resistance in multiresistant S. Typhimurium DT 104, and also in S. virchow, a serotype with a propensity for causing extraintestinal infections in humans. High-level resistance (MIC≥2.0 mg/L) was uncommon and was identified in only a few strains, all from patients with a history of recent foreign travel.

Conclusions: There is a strong temporal association between increases in the occurrence of ciprofloxacin resistance in Salmonella serotypes Typhimurium, Virchow and Hadar from humans in England and Wales and with the licensing for use in food animals in the UK of the related fluoroquinolone antibiotic enrofloxacin; in contrast, for S. enteritidis ciprofloxacin resistance was most common in a phage type associated with foreign travel. It is hoped that recent recommendations for the use of fluoroquinolone antimicrobials in food animals in the UK will result in a reduction in the occurrence of resistance to ciprofloxacin in zoonotic salmonellas causing infections in humans.

INTRODUCTION

The Laboratory of Enteric Pathogens (LEP) is the national reference center for salmonellas from humans for England and Wales and receives strains from public health laboratories, National Health Service laboratories and commercial laboratories. All isolates are identified by a range of phenotypic methods, including serotyping and, where appropriate, phage typing, and are screened for resistance to a panel of antimicrobial drugs. As well as being used to supplement serotyping and phage typing for epidemiologic investigations, the antibiogram data provide a framework for observing changes in the incidence of resistance and for investigating factors contributing to change.

Two antimicrobials which are included in the panel are the quinolone antibiotic nalidixic acid and its fluorinated derivative ciprofloxacin. Nalidixic acid is rarely used for the treatment of salmonellosis but ciprofloxacin is a recognized first-line drug for the treatment of invasive salmonellosis in adults [1]. In 1996 we reported on the increasing occurrence of resistance to ciprofloxacin in salmonellas from humans in England and Wales in the 5-year period 1991–94; of particular note were significant increases in the occurrence of resistance in Salmonella enterica serotypes Virchow, Hadar and Newport [2]. In 1997 we reported on the occurrence of resistance in non-typhoidal salmonellas from humans in England and Wales in 1996 and compared the findings with those for 1994; again, an increase in the occurrence of isolates with resistance to ciprofloxacin was noted, particularly in S. Typhimurium, S. Hadar and S. Virchow [3].

In both these studies the levels at which resistance to ciprofloxacin was tested for were 0.125 and at 1.0mg/L. Although the majority of ciprofloxacin-resistant salmonellas were resistant at 0.125 mg/L but not at 1.0 mg/L, in a few isolates the levels of resistance were greater than 1.0 mg/L. The recommended breakpoint for the testing of Enterobacteriaceae for resistance to ciprofloxacin is 1.0 mg/L [4]. However, although resistance below this level is not regarded as clinically significant, there have been several reports of treatment failures in cases of invasive salmonella infections in which the ciprofloxacin MIC has been below the recognized serum level of the antibiotic following treatment at the recommended dosage [5–8]. It has also been reported that borderline susceptibility to ciprofloxacin in vitro may predict treatment failure and the development of full resistance [9], and because of this the determination of the MIC of ciprofloxacin for isolates that show reduced susceptibility on disk testing has been recommended. Because of these recommendations and because of the possibility of treatment failures at levels below 1.0 mg/L, we have continued to screen for resistance to ciprofloxacin at both 0.125 and 1.0 mg/L.

We now provide an update on the occurrence of resistance to ciprofloxacin in salmonellas isolated from humans in England and Wales in 1997, with particular reference to multiresistant S. Typhimurium definitive phage type (DT) 104 (=DT 104), epidemic in humans and food animals in the UK since 1990 [10–12], and the zoonotic serotypes Enteritidis, Virchow and Hadar.

MATERIALS AND METHODS

All salmonellas received in the LEP in 1997 were identified by serotyping when appropriate [13], and the epidemiologically important serotypes S. Enteritidis, S. Typhimurium, S. Virchow and S. Hadar were further subdivided by phage typing [14–19]. All isolates were screened for resistance to ampicillin, chloramphenicol, gentamicin, kanamycin, streptomycin, sulfonamides, tetracyclines, trimethoprim, furazolidone, nalidixic acid and ciprofloxacin, using an agar dilution breakpoint method in Isosensitest agar [20]. The final plate concentrations of the respective antibiotics were (mg/L): ampicillin, 8 and 128; chloramphenicol, 8; gentamicin, 4; kanamycin, 16; streptomycin, 16 and 128; sulfonamides, 64; tetracyclines, 8 and 128; trimethoprim, 2; nalidixic acid, 16; ciprofloxacin, 0.125 and 1.0. Full minimal inhibitory concentrations (MICs) of nalidixic acid and ciprofloxacin, based on growth of cultures on doubling dilutions of the respective antibiotics in Isosensitest agar, were also determined for five randomly chosen isolates of S. Enteritidis, 18 of S. Typhimurium (including 15 of multiresistant S. Typhimurium DT 104), four of S. Virchow and four of S. Hadar resistant to ciprofloxacin at 0.125 mg/L but sensitive at 1.0 mg/L, and for all isolates resistant at 1.0 mg/L.

RESULTS

Overall occurrence of resistance to ciprofloxacin

In 1997 the four most common serotypes of S. enterica from humans in England and Wales were Enteritidis, Typhimurium, Virchow and Hadar, comprising 89% of isolates identified (Table 1). For S. Enteritidis, 1.3% of the isolates were resistant to ciprofloxacin at 0.125 mg/L, which is a three-fold increase since 1994. For S. Typhimurium, the occurrence of ciprofloxacin-resistant isolates increased in 1997 from 1% to 13%, for S. Virchow from 5% to 14% and for S. Hadar from 40% to 50%. For the remaining 230+ serotypes, the overall occurrence of ciprofloxacin resistance has increased from 1% to 3% since 1994, but because of the low numbers involved, the significance of the results for individual serotypes is difficult to assess and will not be discussed further in this paper.

Table 1. Resistance to ciprofloxacin in Salmonella serotypes Enteritidis, Typhimurium, Virchow and Hadar isolated from humans in England and Wales in 1994 and 1997

Serotype	Year	Total received	Percentage ciprofloxacin resistant^a
^aMIC: ≥0.25 mg/L. ^bProvisional figures. All isolates resistant to ciprofloxacin with MICs ≥0.25 mg/L were also resistant to nalidixic acid with MICs ≥256 mg/L.
S. Enteritidis	1994	17701	0.4
	1997	22723^b	1.3
S. Typhimurium	1994	5603	1
	1997	4690^b	10
S. Virchow	1994	2797	5
	1997	704^b	14
S. Hadar	1994	753	40
	1997	692^b	50
Others	1994	4293	1
	1997	3315^b	3

Only three isolates were identified which were resistant to ciprofloxacin at 1.0 mg/L. These isolates, which were all of serotype Typhimurium, did not react with the S. Typhimurium typing phages and were therefore designated as untypeable. All three isolates were from patients with a history of recent return from countries outside the UK.

Minimal inhibitory concentrations of ciprofloxacin and nalidixic acid

For the 31 strains resistant to ciprofloxacin at 0.125 mg/L but sensitive at 1.0 mg/L which were subjected to full MIC determination, the MICs of ciprofloxacin ranged from 0.5 to 1.0 mg/L, and of nalidixic acid from 256 to 512 mg/L (Table 2). The corresponding MICs for the three isolates of S. Typhimurium with high-level resistance to ciprofloxacin ranged from 32 to 64 mg/L for ciprofloxacin, and from 512 to 1024 mg/L for nalidixic acid.

Table 2. MICs of ciprofloxacin and nalidixic acid for salmonellas with resistance to ciprofloxacin at 0.125 mg/L

		MICs (mg/L)
Serotype/Phage type	No. tested	Ciprofloxacin	Nalidixic acid
In parentheses are numbers of isolates with resistance at the specified MICs.
S. Enteritidis	5	0.5 (3)	256 (3)
		1.0 (2)	512 (2)
S. Typhimurium
DT 104	15	0.5–1.0 (15)	512 (15)
Others	6	0.5–1.0 (3)	256–512 (3)
		36 (2)	512 (1), 1024 (1)
		32 (1)	1024 (1)
S. Virchow	4	0.5–1.0 (4)	512 (4)
S. Hadar	4	0.5–1.0 (4)	512 (4)

Distribution of ciprofloxacin-resistant strains within phage type

The distribution of phage types of S. Enteritidis, S. Typhimurium, S. Virchow and S. Hadar with isolates resistant to ciprofloxacin is presented in Table 3.

Table 3. Distribution of resistance to ciprofloxacin within phage type in Salmonella serotypes Enteritidis, Typhimurium, Virchow and Hadar, 1997

		Phage types with ciprofloxacin resistance
Serotype	No. of phage types identified	PT	No. of isolates^a	Percentage ciprofloxacin resistant
^aProvisional figures. PT, phage type.
S. Enteritidis	45	1	928	19
		6a	1137	14
		4	15210	0.4
		6	1669	0.4
		Others	3779	0.6
		(10 PTs)
S. Typhimurium	90	104	2956	14
		104b	240	5
		U302	206	4
		193	286	3
		Others	1002	0.3
		(4 PTs)
S. Virchow	25	45	31	68
		31	38	29
		26	222	21
		8	191	8
		Others	222	3
		(5 PTs)
S. Hadar	23	21	17	94
		2	268	80
		9	83	10
		Others	324	28
		(13 PTs)

S. Enteritidis

For S. Enteritidis, resistance to ciprofloxacin was observed in 14 of 45 phage types (PTs) identified in 1997. Resistance was most common in PT1, and in this phage type 19% of isolates were ciprofloxacin resistant. In contrast, for PT4, less than 0.5% of isolates were resistant to ciprofloxacin. For both PT1 and PT4, ciprofloxacin resistance was commonly present as a single resistance and was rarely identified in combination with other resistances.

S. Typhimurium

For S. Typhimurium, ciprofloxacin resistance was observed in nine of 90 phage types but was most common in multiresistant DT 104 (Table 3). In this phage type, 14% of isolates were resistant to this antimicrobial. This should be compared to 1994, when only 0.6% of isolates of DT 104 were resistant [11]. Resistance to ciprofloxacin was also common in the related multiresistant phage types DT 104b and U302 but was rare in other phage types.

S. Virchow

Resistance to ciprofloxacin was observed in nine of 25 phage types, sometimes as a single resistance but more commonly together with other resistances. The phage types in which resistance was most common were PT45, PT31 and PT26, with 68%, 29% and 21% of isolates being ciprofloxacin resistant.

S. Hadar

For S. Hadar, ciprofloxacin resistance was observed in 16 of 23 phage types, most commonly in combination with at least two other resistances. The phage types in which resistance was most common were PT2, PT21 and PT9, with 80%, 94% and 10% of isolates being resistant. However, it is noteworthy that none of the isolates of PT5, the second most common S. Hadar phage type in 1997, were resistant to ciprofloxacin.

DISCUSSION

Since 1994, the occurrence of ciprofloxacin resistance has increased in the four most common salmonella serotypes from cases of human infection in England and Wales—Enteritidis, Typhimurium, Virchow and Hadar. For S. Enteritidis, resistance was most common in PT1, which is a phage type often associated with travel to countries in southern Europe. This should be contrasted with PT4, a phage type which is endemic in the UK and for which less than 0.5% of isolates were ciprofloxacin resistant.

For S. Typhimurium, the overall occurrence of ciprofloxacin resistance has increased from 1% to 10% since 1994. Resistance to this antimicrobial was most common in multiresistant DT 104, and in this phage type increased from 0.6% in 1994 to 13% in 1997. Data from the Veterinary Laboratory Agencies [12] have demonstrated that DT 104 with resistance to nalidixic acid is now common in turkeys, chickens and cattle. Data for resistance to ciprofloxacin or to the veterinary equivalent enrofloxacin were not provided, but recent reports from Piddock et al [21] and from Ridley and Threlfall [22] have demonstrated that food animal isolates of S. Typhimurium DT 104 with resistance to nalidixic acid are also resistant to ciprofloxacin at the levels described above. As S. Typhimurium DT 104 is a zoonotic pathogen, it is possible that food animals, and particularly poultry and cattle, are important sources of ciprofloxacin-resistant strains of this phage type in cases of human infection. S. Virchow and S. Hadar are also poultry-related serotypes [12] and it is also possible that poultry may have been an important source of ciprofloxacin resistant strains of these serotypes. The high incidence of resistance to ciprofloxacin in S. Virchow is of particular concern because of the propensity of this serotype for extraintestinal spread in humans [23].

One drug which has been extensively used in food animals in the UK since 1994 is the fluoroquinolone antibiotic enrofloxacin, which was approved for veterinary use in the UK in November 1993. Strains resistant to enrofloxacin show decreased susceptibility to ciprofloxacin, and it is possible that the use of this antimicrobial in food animals has contributed to the increases in the occurrence of ciprofloxacin resistance in zoonotic salmonellas from cases of human infection. In 1992 the Expert Group on Animal Feedingstuffs recommended that ‘not only should antibiotics giving cross-resistance to those in human medicine not be used as growth promoters but … their prophylactic use in animals be reconsidered’[24]. This recommendation, which was particularly relevant to fluoroquinolone antibiotics such as enrofloxacin, was not put into practice. However, a recent report by the House of Lords Select Committee on Science and Technology has recommended the rapid introduction by the veterinary profession of a code of practice on when fluoroquinolone antibiotics should be prescribed for food animals [25]. This has already been accepted by some pharmaceutical companies, and recently introduced recommendations for the use of enrofloxacin in livestock are encouraging. In light of a recent report from Denmark describing a lack of clinical response to fluoroquinolone antibiotics in patients infected with multiresistant S. Typhimurium DT 104 with additional resistance to quinolones in an outbreak associated with pork of Danish origin [8], it is hoped that these recommendations will rapidly be implemented.

Acknowledgments

An abridged version of this paper was presented to the WHO Meeting on the Use of Quinolones in Food Animals and Potential Impact on Human Health, WHO HQ, Geneva, Switzerland, 2–5 June 1998. We are grateful to Mr A. Graham and Mr J Skinner for assistance with these studies.