Resistant pathogens as causes of traveller’s diarrhea globally and impact(s) on treatment failure and recommendations

Abstract

Background: Diarrhea is a frequent clinical syndrome affecting international travellers. Bacterial etiologic agents have a long history of emergent antimicrobial resistance against commonly used antibiotics. Current approaches applying first-line antimicrobial therapy are being challenged by increasingly resistant organisms. This review summarizes recent epidemiological and clinical evidence of antibiotic resistance among enteropathogens causing traveller’s diarrhea and the subsequent impact on current treatment recommendations.

Methods: The PubMed database was systemically searched for articles related to antibiotic susceptibility and diarrheal pathogens.

Results: Antibiotic resistance related to travellers’ diarrhea has increased in recent years. Most notably, fluoroquinolone resistance has expanded from the Campylobacter-associated cases well documented in Southeast Asia in the 1990s to widespread occurrence, as well as increases among other common bacterial enteropathogens including, enterotoxigenic and enteroaggregative Escherichia coli, Shigella and non-typhoidal Salmonella. Multidrug resistance among enteropathogenic Enterobacteriacae and Campylobacter species create further challenges with the selection of empiric therapy. Treatment failures requiring early use of alternative agents, as well as delayed recovery comparable to placebo rates emphasize the impact of antimicrobial resistance on effective treatment.

Conclusions: Although there are limitations in the available data, the increasing antibiotic resistance and adverse impact on clinical outcome require continued surveillance and reconsideration of practice guidelines.

Evolution of Antimicrobial Use in the Treatment of Traveller’s Diarrhea (TD)

Diarrhea is one of the most common clinical syndromes impacting travellers, of which 80–90% of attacks are the result of bacterial pathogens (e.g., enterotoxigenic and enteroaggregative Escherichia coli [ETEC and EAEC, respectively], Campylobacter spp., Shigella spp. and non-typhoidal Salmonella spp.).¹ There are also well-recognized regional differences in the distribution of pathogens,² so travel destination often influences treatment recommendations. Specifically, Campylobacter spp. is commonly isolated from patients in Southeast Asia (particularly Thailand),^3–5 while ETEC is frequently the causative agent for TD among travellers in the Latin America, Africa, south Asia and the Middle East.^6–9

Although TD is a self-limited illness commonly resolving within 5 days when left untreated (quicker when antibiotics are prescribed), it frequently results in short-term morbidity adversely impacting travel plans, as well as potential for long-term enteric and extra-intestinal complications.¹⁰ Antibiotic therapy has been proved to be effective in treating patients with TD by both significantly reducing associated symptoms and shortening the illness duration.^2,11 As antibiotic therapy remains the most effective treatment, it is common for clinicians to prescribe antibiotics to international travellers for self-treatment if they experience diarrheal symptoms while abroad. Over the past two decades, enteropathogens have shown increasing resistance to first-line antibiotics, particularly fluoroquinolones, in endemic regions.^1,4,12–17 Decreased susceptibility of preferred antibiotics is an evolving challenge for clinicians requiring interval reconsideration of clinical practice guidelines. This review considers recent information on bacterial enteropathogen antimicrobial resistance against current first-line antibiotics, impact on clinical outcomes, and the effect on modification of clinical practice guidelines.

Factors Impacting Antimicrobial Effectiveness in TD Treatment

The effectiveness of an antibiotic to rapidly lead to clinical resolution with or without microbiological eradication is based on several factors. Randomized clinical trials provide point estimates of treatment efficacy under controlled circumstances, frequently highlighting an important subset of subjects with suboptimal (such as delayed time to resolution approximating placebo recovery rates) or outright treatment failure early in the antibiotic course or late recurrence.¹⁸ Antibiotic selection and the relationship to enteropathogen antibiotic susceptibility alone are not the only factors impacting clinical response.

Additional considerations include host factors, such as immunocompromised states, which may decrease efficacy, as well as add risk for persistent and recurrent illness; adjunctive therapy with antimotility agents (e.g., loperamide), which is well documented to decrease duration of diarrheal episodes;¹⁹ patient adherence to prescribed regimen; frequent co-infection challenging interpretation of attribution with resistant pathogens; and potentially non-antimicrobial effects of the therapeutic agents.²⁰ The clinical syndrome encompassing TD includes the most common acute watery diarrhea (AWD) presentation, in addition to dysentery or febrile AWD and even viral gastroenteritis cases that manifest with diarrhea along with the common predominance of vomiting.¹ Overall efficacy estimates for first-line therapies are reported by pooling these various clinical presentations into common TD endpoints, which tends to lose discrimination of important differences in time to recovery. Furthermore, the time of clinical presentation from illness onset is known to impact time to resolution in these self-limited illnesses.²¹ Trials attempt to control for this by restricting duration of illness at time of enrolment; however, this still may fail to consider relative differences given that these illnesses can be quite short lived. Lastly, antimicrobial resistance and the effect on clinical outcomes are extremely important, but should be placed in context with the above considerations.

Current Status of Major TD Enteropathogen Antimicrobial Resistance Epidemiology

The TD bacterial enteropathogen antimicrobial resistance data derive from a number of sources, which have strengths and limitations. Direct assessment of clinical microbiology from a TD treatment trial is most likely to report the relationship of clinical outcomes relative to susceptibility. Other sources of information stem from epidemiological surveys of travellers with TD, either at the travel destination or upon return to their home country, as well as, surveys on non-travellers (e.g., children from developing worlds in their country of residence), which are often extrapolated to inform regional infection threats. Irrespective of the information source, a significant challenge in translating this information to clinical practice guidance is interpretation of the susceptibility criteria.²² Interpretative criteria relies on clinical breakpoint development generated by standardized in vitro susceptibility testing, pharmacological parameters, and, importantly, clinical outcome studies, which are often lacking. Further complicating interpretation is the extremely high faecal antibiotic levels relative to a pathogen’s minimum inhibitory concentration (MIC) for both systemic and non-absorbable antibiotics as follows: azithromycin with a 11–14 h half-life has ∼50% active drug excreted in stool with high tissue levels,^23,24 ciprofloxacin has 100- to 1000-fold faecal levels above serum levels with 19% oral dose excreted as metabolites in faeces, as well as high tissue levels,²⁵ and rifaximin is <0.4% absorbed with 4000–8000 µg/g faecal levels at the 400 mg two times daily dose.²⁶

The following summarizes recent reports focusing on traveller populations highlighting the continued evolution of antimicrobial resistance against first-line agents. The summary first considers the TD-causing Enterobacteriaceae species followed by Campylobacter jejuni/coli.

TD-Causing Enterobacteriaceae

Antibiotic resistance among enteric bacterial pathogens has been steadily increasing over the last two decades.^4,16,27–32 A recent review of EAEC epidemiology and clinical manifestations summarized the increasing rates of multidrug resistance including loss of fluoroquinolone activity as well as extended-spectrum beta-lactamase (ESBL) production.³³ Moreover, a comprehensive study assessed ETEC and EAEC isolate changing antibiotic susceptibility trends over a decade among adult travellers to Mexico, India, and Guatemala with TD.²⁷ The older agents, such as ampicillin, trimethoprim-sulfamethoxazole, and doxycycline, retained high levels of resistance. Fluoroquinolones and azithromycin MICs were significantly higher than when assessed 10 years prior. Regional variation in ETEC fluoroquinolone resistance was also demonstrated with higher rates in India compared with Central America. The EAEC isolates from Central America showed increased resistance across most antibiotics tested. In addition to the observed increases in fluoroquinolone resistance, 4- to 10-fold increases in MIC₉₀ values for ceftriaxone and azithromycin were also seen for both ETEC and EAEC. The non-absorbable antibiotic rifaximin had no significant changes in ETEC or EAEC MICs.

In recent years, a rise in clinical resistance, often multidrug-resistance, to first-line antibiotics (e.g., fluoroquinolones and azithromycin) and alternative agents (e.g., third-generation cephalosporins) has been reported in Shigella spp. and Salmonella spp. diarrheal patients.^34–44 The past 10–15 years have seen higher rates of fluoroquinolone resistance in surveys among industrialized countries with a large proportion associated with foreign travel.^36,38,39 Multidrug resistance is also being reported, including decreased azithromycin susceptibility, ESBL production, coupled with loss of fluoroquinolone activity.³⁷ The strong association with quinolone resistance and foreign travel has also been reported with non-typhoidal Salmonella species through the US antimicrobial resistance monitoring systems.⁴³ Furthermore, resistance to ciprofloxacin was reported in 20.9% of Salmonella enterica isolates collected in the UK and associated with treatment failure.⁴² As seen in Shigella spp.,⁴⁰ non-typhoidal Salmonella also have increasing evidence in MDR strains of losing available alternative agents, such as cefotaxime.⁴⁴

Campylobacter jejuni/coli

Although an increasing level of fluoroquinolone resistance has been reported with many enteric pathogens, there is less information connecting resistance to clinical treatment failure. Presently, the association between emergent resistance and adverse clinical outcomes is best understood with Campylobacter spp. (Table 1). The emergence of fluoroquinolone resistance among C. jejuni/coli started earlier (i.e., 1990s)^45–47 than the increasing trends over the past decade observed among diarrheagenic E. coli, Shigella, and Salmonella strains. Thailand has been a region with particularly rapid emergence with fluoroquinolone resistance among Campylobacter spp. increasing from zero in 1990 to 84% in 1995.⁴⁵ In one analysis, travellers with confirmed bacterial enteritis within one month after travel to the tropics were examined. Campylobacter jejuni was the etiologic agent in 41% of cases, of which 53% of isolates were resistant to norfloxacin. Clinical failure occurred in 33% of patients who were prescribed fluoroquinolines.²⁹ Surveillance of Campylobacter spp. isolates over a 10-year period (2001–10) in Peru found ciprofloxacin resistance of C. jejuni and Campylobacter coli in Lima significantly increased from 73.1 to 89.8% and 48.1 to 87.4%, respectively. Resistance of C. jejuni to azithromycin and erythromycin also significantly increased from 2.2 to 14.9% and 3.2 to 14.9%, respectively, in Iquitos, Peru.³¹

Table 1.

Antibiotic resistance of Campylobacter jejuni/coli and clinical implications

Region	Location	Antibiotica		Sourceb	Date of isolate collection	Clinical implicationsc	Reference
Asia
	Asia	Fluoroquinolones		Travellers	1994–2000; 2001–2006	Increased from 36% resistant to 71% resistant (statistically significant)	[49]
	Asia (excluding China)	Fluoroquinolones/ Nalidixic acid		Travellers	2005–2011	78% resistant	[50]
		Macrolides		Travellers	2005–2011	6% resistant	[50]
	East Asia	Fluoroquinolones		Travellers	1994–2000; 2001–2006	Increased from 48% resistant to 61% resistant (significance not stated)	[49]
	Southeast Asia (Thailand)	Fluoroquinolones		Travellers	1990	All pretreatment isolates susceptible to ciprofloxacin (MIC = 0.2 μg/ml) 2 patients (4%) had clinical enteritis relapse associated with development of ciprofloxacin resistance (MIC >16 μg/ml after treatment)	[46]
				Travellers	1993	50% of isolates were resistant to ciprofloxacin (MIC ≥4 μg/ml) 48% recovered within 48 h 10% clinical failure; 29% bacteriologic failure	[47]
				Domestic	1981–1995	Significant increase from zero in 1990 to 84% resistant in 1995	[45]
				Travellers	1998	96% resistant; MICs ranged from 4 to 64 μg/ml, but no association with treatment failure Sub-optimal treatment response observed in 17% of cases	[5]
				Travellers	2000–2001	93% resistant ciprofloxacin; MIC50 16 μg/ml; MIC90 >64 μg/ml 50% resistant to levofloxacin; MIC50 6 μg/ml; MIC90 >64 μg/ml 38% cure rate with levofloxacin within 48 h (71% within 72 h); 13% clinical failure	[4]
				Travellers	2000–2006	64% resistant	[29]
		Macrolides		Travellers	1993	All isolates were susceptible (MIC ≤2 μg/ml) 62% recovered within 48 h No treatment or bacteriologic failure	[47]
				Domestic	1981–1995	7–15% resistant in 1994/1995 All isolates resistant to azithromycin were also resistant to ciprofloxacin	[45]
				Travellers	1998	No resistance	[5]
				Travellers	2000–2001	No resistance; 96% cure rate for single dose within 72 h; 96% microbiological cure	[4]
				Travellers	2000–2006	No resistance	[29]
	China	Fluoroquinolones/ Nalixidic acid		Travellers	2005–2011	100% resistant	[50]
		Macrolides		Travellers	2005–2011	11% resistant	[50]
	Indian subcontinent	Fluoroquinolones		Travellers	1994–2000;2001–2006	Increased from 29% resistant to 79% resistant (statistically significant)	[49]
Europe and North America
	Eastern Europe	Fluoroquinolones/ Nalidixic acid		Travellers	2005–2011	79% resistant	[50]
		Macrolides		Travellers	2005–2011	5% resistant	[50]
	Belgium	Fluoroquinolones		Domestic	1994–2000; 2001–2006	Increased from 9% resistant to 43% resistant (statistically significant)	[49]
	Western Europe	Fluoroquinolones/ Nalidixic acid		Travellers	2005–2011	50% resistant	[50]
		Macrolides		Travellers	2005–2011	1% resistant	[50]
	North America (excluding United States)	Fluoroquinolones/ Nalidixic acid		Travellers	2005–2011	12% resistant	[50]
		Macrolides		Travellers	2005–2011	Zero resistance	[50]
Africa
	Africa	Fluoroquinolones		Travellers	1994–2000; 2001–2006	Increased from 14% resistant to 32% resistant (statistically significant)	[49]
		Fluoroquinolones/ Nalidixic acid		Travellers	2005–2011	37% resistant	[50]
		Macrolides		Travellers	2005–2001	1% resistant	[50]
	Sub-Saharan Africa	Fluoroquinolones		Travellers	2000–2006	31% resistant	[29]
		Macrolides		Travellers	2000–2006	No resistance	[29]
	West Africa	Fluoroquinolones		Travellers	1994–2000; 2001–2006	Increased from 7% resistant to 29% resistant (statistically significant)	[49]
Central/Latin America and Tropics
	Mexico		Fluoroquinolones/ Nalidixic acid	Travellers	2005–2011	62% resistant	[50]
			Macrolides	Travellers	2005–2011	4% resistant	[50]
	Central America (excludes Mexico)		Fluoroquinolones/ Nalidixic acid	Travellers	2005–2011	51% resistant	[50]
			Macrolides	Travellers	2005–2011	3% resistant	[50]
	South America		Fluoroquinolones/ Nalidixic acid	Travellers	2005–2011	83% resistant	[50]
			Macrolides	Travellers	2005–2011	4% resistant	[50]
	Peru		Fluoroquinolones	Domestic	2001–2006; 2006–2010	Significantly increased in Lima from 73% resistance to 90% for C. jejuni Significant increase of 48% resistance to 87% for C. coli	[31]
			Macrolides	Domestic	2001–2006; 2006–2010	Significantly increased in Iquitos from 2% azithromycin resistance to 15% for C. jejuni Significant increase of 3% erythromycin resistance to 15% for C. jejuni; 15% of isolates resistant to both azithromycin/erythromycin and ciprofloxacin in Iquitos in 2006–2010	[31]
	Caribbean and Central/ South America		Fluoroquinolones	Travellers	1994–2000;2001–2006	Increased from 28% resistant to 61% resistant (statistically significant)	[49]
	Combined tropics		Fluoroquinolones	Travellers	2000–2006	33% clinical failure	[29]
				Travellers	2002–2003	MIC >0.25 μg/ml in 42% of isolates	[53]
			Rifaximin	Travellers	2002–2003	Pre-treatment: MIC ranged 0.06–0.5 μg/ml; Post-treatment: MIC ranged 0.06–0.125 μg/ml	[53]

^aAntibiotics used in the studies are classified as fluoroquinolones, macrolides, and rifaximin. Macrolides includes azithromycin and erythromycin. Fluoroquinolones include norfloxacin, levofloxacin, and ciprofloxacin.

^bSource indicates whether isolates were collected from travellers or residents of the location (domestic).

^cBolded text in clinical implications column indicates reported clinical outcomes.

Among the confirmed cases of Campylobacter spp. infections affecting people in the United States (2005–11), 18% were associated with international travel. Data also suggest that a significant portion of fluoroquinolone resistance among Campylobacter spp. infections in the United States, as well as other countries,⁴⁸ may be the result of acquisition through international travel. Specifically, 60% of travel-related Campylobacter spp. infections were found to be fluoroquinolone-resistant, while it was only 13% of domestic cases. Macrolide resistance, which was significantly associated with an increased risk of hospitalization, was observed with 4% of travel-related Campylobacter spp. infections, which has been observed at similar, relatively stable, rates in other surveys.⁴⁹ In general, the majority of travel-related Campylobacter spp. infections resulted from travel to Asia, Mexico, or South America, and a high proportion were resistant to at least two fluoroquinolones (78, 62, and 83%, respectively).⁵⁰ Fluoroquinolone resistant C. jejuni/coli has now become a widespread problem among industrialized and developing countries.

Evidence of TD Enteropathogen Treatment Failure Related to Antimicrobial Resistance

Treatment response in a majority of randomized control trials utilizes the time from initial antibiotic dose to the last unformed stool (TLUS). This endpoint does correlate well with overall clinical resolution, but other important outcomes include clearance of associated symptoms along with functional recovery and return to regular activities.¹⁸ The eradication of pre-treatment isolates, typically days 5–9 after the last treatment day, has not correlated well with clinical outcome.⁵¹ Early studies comparing azithromycin and ciprofloxacin in patients with shigellosis demonstrated high faecal antibiotic levels for both drugs with no evidence of resistance; however, treatment failure rates among S. dysenteriae type 1 were as high as 29 and 17% in the azithromycin and ciprofloxacin groups, respectively.⁵² In comparison the non-dysenteriae shigellosis cases had overall treatment failure rates, irrespective of drug, of only 6%, suggesting impact of specific pathogen, relative severity of disease at presentation, or both.

In a randomized, double-blind treatment trial comparing azithromycin and levofloxacin regimens, military personnel presenting with acute diarrhea while serving in Thailand were enrolled. Campylobacter spp. were isolated from 64% of patients, of which C. jejuni was predominant (95%). The Campylobacter spp. isolates were all susceptible to azithromycin; however, 50% and 93% were resistant to levofloxacin and ciprofloxacin, respectively. The TLUS was significantly different among subjects with levofloxacin-resistant Campylobacter-associated diarrhea who were provided the levofloxacin regimen (76.4 h compared with 41.2 h in subjects with levofloxacin-susceptible isolates and 41–47 h in subjects receiving azithromycin; P = 0.008; Figure 1). Clinical cure (i.e., resolution of diarrhea and associated symptoms within 72 h) was 96% for patients receiving single dose of azithromycin and 70% for 3 days of levofloxacin.⁴

Figure 1.

Time to cure (after receipt of the first antibiotic dose), by treatment group, as stratified by diagnosis of Campylobacter infection. This figure has been reprinted from Tribble et al.⁴ by permission of Oxford University Press.

While rifaximin has been shown to be effective at treating TD, it should not be prescribed in cases of febrile invasive disease where there have been clearly documented cases of primary treatment failure.^53,54 Furthermore, rifaximin should specifically not be prescribed when Campylobacter is the etiologic agent due to its high level of resistance and clinical outcome evidence of treatment failure.⁵⁵ Rifaximin MICs for ETEC and EAEC have been shown to increase among cases with persistent infection; however, this has not correlated with a decline in treatment efficacy.⁵³

Impact of Resistant Pathogens on Treatment Recommendations

Current clinical practice patterns have been greatly impacted by the increasing levels of antibiotic resistance over the last two decades, as well as regional differences in the profile of resistant pathogens. Specifically, increased resistance of common enteropathogens to older first-line antibiotics, in addition to currently used absorbable systemic antibiotics, created a challenging situation for clinicians leading to past modification of practice guidelines. Although there are significant limitations in the available studies with outcome data, the findings show that the presence of a resistant phenotype often results in suboptimal clinical outcomes, including treatment failure. This is particularly evident with the case of the high level of fluoroquinolone-resistant Campylobacter spp. isolated in Southeast Asia. Further surveillance and assessment of outcomes are needed in order to examine trends in antibiotic resistance and modify treatment regimens to rapidly achieve clinical cure while balancing safety and emergent resistance concerns.

Funding

This work was supported by the Infectious Disease Clinical Research Program, a Department of Defense program executed through the Uniformed Services University of the Health Sciences, Department of Preventive Medicine and Biostatistics. This project has been funded by the National Institute of Allergy and Infectious Diseases, National Institute of Health [Inter-Agency Agreement Y1-AI-5072].