Reconsidering azithromycin disc diffusion interpretive criteria for Salmonellae in view of azithromycin MIC creep among typhoidal and nontyphoidal salmonella.

PURPOSE: Enteric fever continues to be an important public health challenge for the developing world. With the emergence of fluoroquinolone resistance in Salmonellae spp. azithromycin is increasingly being used for oral treatment of enteric fever. We investigated the antibiotic susceptibility pattern of azithromycin in Salmonellae spp. isolates from a tertiary care hospital to detect emerging resistance.
METHODS: The study assessed the reliability of disc diffusion as a screening test to detect azithromycin resistance by comparing it with the minimum inhibitory concentrations (MICs) of the drug in 100 Salmonellae spp. strains. The strains of Salmonellae spp. showing resistance to azithromycin were further investigated for resistance markers - mphA, mphB, and mef B genes.
RESULTS: This study was conducted on 100 Salmonella enterica strains recovered from blood culture samples between 2013 and 2017. Among these isolates, 18 showed resistance to azithromycin by disc diffusion methodology with zones of inhibition <13 mm. MIC of 6 of these isolates were ≥32 mg/L. The mean MIC of azithromycin increased from 5 mg/L in 2013 to 24 mg/L in 2017. Azithromycin consumption as defined daily doses per 1000 patient days also showed an increase over the past 4 years.
CONCLUSION: Azithromycin disc diffusion diameter interpretations as recommended by Clinical and Laboratory Standards Institute can mislabel a few sensitive strains as resistant. Azithromycin resistance is emerging in typhoidal and nontyphoidal Salmonella. MphA gene is associated with high MICs in nontyphoidal Salmonella spp.

Introduction

Enteric fever remains an important public health challenge for the developing world. Geographically, south central and southeast Asia have the highest incidence of typhoid fever with an estimated 100 cases per 100,000 person-year.[1] Risk factors commonly associated with a high incidence of typhoid fever include poor sanitation, limited access to clean drinking water and low socioeconomic status.[2] Changing susceptibility patterns of typhoidal Salmonella spp. has also added to the challenges faced by the treating physician. While multidrug-resistant Salmonella spp. has become a thing of the past in the Indian subcontinent, fluoroquinolone-resistant Salmonellae spp. have emerged as a formidable problem.[34] Consequently, intravenous ceftriaxone and oral azithromycin are increasingly being used in the empirical treatment of typhoid fever.[5] Although ceftriaxone continues to be effective against Salmonella spp., it is parenteral administration, cost and longer time for defervescence are factors that limit its use as an ideal antibiotic for enteric fever.[4]

Azithromycin has been shown to be equally effective, or in some cases, a superior treatment alternative to chloramphenicol or fluoroquinolones by several randomized control trials.[67] It has the ability to achieve intracellular concentrations which are 50–100 times higher than the serum level of the antibiotic.[7]

Guidelines for testing azithromycin susceptibility for Salmonella Typhi were released by Clinical and Laboratory Standards Institute (CLSI) in 2015.[8] EUCAST does not prescribe any clinical breakpoints for azithromycin testing but mentions that azithromycin has been used in the treatment of S. Typhi infections with minimum inhibitory concentration (MIC) ≤16 mg/L.[9] The emergence of azithromycin resistance in Salmonellae spp. has been documented by several studies.[10]

We investigated the antibiotic susceptibility pattern of azithromycin in Salmonella spp. isolates from a tertiary care hospital to detect emerging resistance and change in MICs over a 5-year period. The study also assessed the reliability of disc diffusion as a screening test to detect azithromycin resistance by comparing it with the MICs of the drug.

Materials and Methods

Bacterial strains

The study was conducted in a 1200 bed tertiary care hospital in Southern India between January 2013 and December 2017. A total of 100 Salmonella enterica strains recovered from blood samples were included in the study. Isolates were identified by standard biochemical reactions and VITEK 2 compact system (bioMérieux). The confirmation of identification was done by agglutination with specific antisera (Denka Seiken).

Azithromycin susceptibility testing

Antibiotic susceptibility testing was performed by Kirby Bauer disc diffusion method using azithromycin discs (15 μg). Azithromycin discs from two different manufacturers were used for disc diffusion (HiMedia and BD BBL sensi disc). Testing was done in triplicate for each isolate and disc. Disc diffusion diameters were recorded as mean of three readings. MIC of azithromycin was determined using Etest (bioMérieux). CLSI (2015–2017) guidelines were used to interpret azithromycin susceptibility-sensitive ≥13 mm and ≤16 mg/L and resistant ≤12 mm and ≥32 mg/L. Since CLSI mentions azithromycin breakpoints for S. Typhi only, these were used for interpreting Salmonella Paratyphi A and Group B Salmonella spp. susceptibility also. MIC50 and MIC90 of the isolates were determined. Escherichia coli ATCC 25922 and Staphylococcus aureus ATCC 25923 were used as the control strain for all susceptibility testing.

Errors in susceptibility were defined as follows:

Very major error: False susceptible result by disc diffusion compared to the MIC value

Major error: False resistant result produced by disc diffusion compared to MIC value

Minor errors: A difference of >2 mm in disc diffusion diameters while using two different discs.

Synergy testing

The evaluation of 10 strains of Salmonella Typhi for in vitro synergy between azithromycin and ceftriaxone was performed. Mueller-Hinton agar plates were inoculated with the suspension of the study strains grown to an optical density of 0.5 McFarland units. For each isolate, MIC of azithromycin and ceftriaxone was determined individually and in combination (AB Biodisk Etest Customer Information Sheet EAS023). For combination testing, E strip A (azithromycin) was put on the inoculated MHA plate and left for 1 h at room temperature. After an hour, this strip was removed after marking its outline, was washed with alcohol and stored as MIC reading scale. E strip B (ceftriaxone) was placed on the imprint of E strip A immediately and incubated at 35°C for 18 h. MIC scales were used to read the combination MIC gradients.

Fractional inhibitory concentration index (FIC index) calculations were made according to the formula:

FIC index = MICAB/MICA + MICBA/MICB

Synergy was defined as MIC of combination ≥2 dilutions lower than MIC of the most active drug alone or FIC index ≤0.5.

Polymerase chain reaction

Mechanism responsible for azithromycin resistance was studied by amplifying mphA, mphB, and mefB gene using previously published primers.[11] The thermal method of DNA extraction from Salmonellae spp., described by Gibson and McKee, was used with minor modifications.[12] Briefly, isolates were grown on brain heart infusion agar plates for 18–24 h. 2–3 colonies were scooped using an inoculation loop and suspended in 100 μl of sterile distilled water. The suspension was heated in a dry bath at 95°C for 15 min. After centrifugation the 1 μl of supernatant solution was used as DNA template for polymerase chain reaction (PCR).

PCR reactions were done in 25 μl volumes using EmeraldAmp® MAX PCR Master Mix, 100 ng of each primer and the extracted whole DNA as the template. The PCR was carried out in Primus 25 thermal cycler (Peqlab) starting with an initial denaturation at 95°C for 3 min. This was followed by 30 cycles of denaturation at 95°C for 30 s, annealing at 60°C for 30 s, and elongation at 72°C for 30 s. A final elongation step was run at 72°C for 10 min. The PCR products were subjected to 2% agarose gel electrophoresis stained using ethidium bromide solution and visualized under trans-UV illumination.

Results

This study was conducted on 100 Salmonella enterica strains which included S. Typhi (n =46), S. Paratyphi A (n = 12), and Group B Salmonella spp. (n = 42) recovered from blood culture samples between 2013 and 2017.

Among these isolates, 18 showed resistance to azithromycin by disc diffusion methodology with zones of inhibition <13 mm. Of these 18 isolates, 10 were S. Typhi, three were S. Paratyphi A and five were Group B Salmonella spp. MICs of six of these isolates were ≥32 mg/L [Table 1].

Table 1

Minimum inhibitory concentration values of Salmonella spp. showing azithromycin resistance by disc diffusion

Identification	Disc diffusiona (mm)	Disc diffusionb (mm)	MIC (mg/Ll)
S. Typhi	12	12	4
S. Typhi	12	11	12
S. Typhi	9	8	16
S. Typhi	12	12	24
S. Typhi	6	6	32
S. Typhi	6	6	32
S. Typhi	6	6	32
S. Typhi	6	6	96
S. Typhi	12	11	16
S. Typhi	6	6	32
S. Paratyphi A	6	6	8
S. Paratyphi A	6	8	8
S. Paratyphi A	10	11	24
S. Paratyphi B	10	11	12
Salmonella spp.	10	12	12
Salmonella spp.	10	12	12
Salmonella spp.	12	12	12
Salmonella spp.	8	10	64

Disc diffusiona: HiMedia; Disc diffusionb: BD BBL Sensi disc. S. Typhi = Salmonella Typhi, S. Paratyphi = Salmonella Paratyphi, MIC = Minimum inhibitory concentration

MIC 50 and MIC 90 of azithromycin for Salmonella spp. was found to be 4 mg/L and 12 mg/L, respectively. Errors in susceptibility were evaluated when disc diffusion was compared with MIC values. Major error was seen in 12 isolates which were labeled as resistant by disc diffusion and showed MICs in the susceptible zone [Table 2].

Table 2

Error classification of Salmonella spp. showing discordant results by disc diffusion and minimum inhibitory concentration

Identification	Very major error	Major error	Minor error
S. Typhi	0	5	0
S. Paratyphi A	0	3	0
Salmonella Group B	0	1	0
Salmonella species	0	3	0
Total	0	12	0

S. Typhi= Salmonella Typhi, S. Paratyphi = Salmonella Paratyphi

Synergy testing was inconclusive. MIC of azithromycin showed reduction when combined with ceftriaxone. However, the FIC indices did not show values supporting synergy [Supplemental Table 1].

Supplemental Table 1

Synergy testing of Ceftriaxone and Azithromycin

MICa	MICb	MICab	MICba	FIC Index
24	0.125	0.25	0.25	2.01
32	0.125	0.25	0.25	2.007
8	0.125	0.25	0.25	2.03
16	0.047	0.125	0.125	2.607
12	0.25	0.25	0.25	1.02
4	0.125	0.125	0.125	1.02
8	0.047	0.25	0.25	5.33
32	0.125	0.25	0.25	2.007
8	0.125	0.25	0.25	2.03
32	0.125	0.25	0.25	2.007

MICa: MIC of azithromycin; MICb: MIC of ceftriaxone; MICab: MIC of the combination when ceftriaxone overlays azithromycin strip; MICba: MIC of the combination when azithromycin overlays ceftriaxone strip; FIC Index=MICab/MICa+MICba/MICb

Isolates which showed resistance by disc diffusion and MIC were investigated for mphA gene. A single isolate of Group B Salmonella spp. was positive for mphA gene. All the isolates were negative for mphB and mef B gene.

Mean MICs of azithromycin of the Salmonella spp. isolates from 2013 to 2017 were calculated and the mean MIC increased from 5 mg/L in 2013 to 24 mg/L in 2017. Azithromycin consumption as defined daily doses per 1000 patient days also showed an increase over the past 4 years [Figure 1]. Mean MIC of ceftriaxone was also calculated and the ceftriaxone consumption was compared from 2013 to 2017. Mean MIC of ceftriaxone over the 4 years were around 0.13 mg/L. Ceftriaxone consumption also remained stable around 0.74 DDD/1000 patient days.

Figure 1

Azithromycin consumption trends in the tertiary care hospital from 2013 to 2017 and mean minimum inhibitory concentration of Salmonellae spp. during that period

The clinical outcomes of the patients harboring six strains showing true resistance were analyzed. All the patients recovered with ceftriaxone treatment. Two patients received prolonged ceftriaxone therapy (14and 28 days). One patient gave a history of gall bladder polyps and repeated episodes of enteric fever [Table 3].

Table 3

Clinical details of patients with azithromycin resistant salmonella infection

Patient	Gender/age	Isolate from blood	Provisional diagnosis	Associated comorbidities	Clinical features	Ceftriaxone susceptibility	Ciprofloxacin susceptibility	Treatment	Outcome
1	Female/18 years	S. Typhi	Enteric fever	None	Low grade fever with mild chills, no rigor (1 weeks duration)No history of abdominal pain, loose stools	S	R	Ceftriaxone for 7 days, Cefixime for 7 days	Recovered
2	Male/57 years	S. Typhi	PUO	HypertensionCoronary Artery DiseaseMixed airway diseaseGall Bladder polyp	Low grade fever loss of appetite and weight lossTreated for enteric fever 1 week back	S	R	Ceftriaxone 28 daysAzithromycin for 7 days	Recovered
3	Male/55 years	S. Typhi	Enteric fever	None	Low grade fever loss of appetite and weight loss stools; 2-3 episodes of vomiting	S	R	Ceftriaxone for 14 days	Recovered
4	Female/17 years	S. Typhi	PUO	None	High grade fever for 10 days, nausea, vomiting loose stools	S	R	Ceftriaxone for 14 days and Azithromycin for 5 days	Recovered
5	Male/30 years	S. Typhi	Enteric fever	None	Fever for 7 days with headache, 1 episode of vomiting	S	R	Ceftriaxone for 14 days	Recovered
6	Female/70 years	Group B Salmonella	Retroperitoneal liposarcoma	Diabetes mellitusHypertensionSplenectomy	Neuropathic pain	S	S	-	-

S. Typhi: Salmonella Typhi, S. Paratyphi = Salmonella Paratyphi

Discussion

Enteric fever due to nalidixic acid resistant strains of typhoidal Salmonellae spp. requires ceftriaxone or azithromycin for its treatment.[513] While resistance to extended spectrum cephalosporins is uncommon, the need for parenteral therapy limits their use as a preferred first line treatment.[14] Azithromycin appears to be an attractive oral alternative for treatment of uncomplicated enteric fever. However, as cautioned by Misra and Prasad, irrational antibiotic therapy and easy over the counter availability of azithromycin could contribute to its emerging resistance and subsequent treatment failure in Salmonellae spp.[14]

Our study was conducted on 58 typhoidal Salmonella spp. and 42 nontyphoidal Salmonella spp. Among these isolates, 18% showed resistance to azithromycin if disc diffusion interpretive criteria were used to determine resistance. However, only 6% of the isolates showed true resistance as their MICs were ≥32 mg/L. An analysis of the discordant disc diffusion diameters showed that of the 12 isolates, eight (66.7%) had diameters between 10–12 mm. Of the 12 isolates showing discordant results between disc diffusion and MIC of azithromycin-five were S. Typhi, three Salmonella Paratyphi A and four Salmonella spp. As CLSI interpretive criteria were specific for S. Typhi, our study had extrapolated these criteria for S. Paratyphi A and Salmonellae spp. also. While this can be a major reason for the large number of discordant results, the presence of a considerable number of S. Typhi showing discordance is a matter of concern. Zone diameters of 10–12 mm can therefore be considered as grey areas for determining azithromycin resistance for Salmonella spp. Therefore, unlike other studies which found good correlation between disc diffusion diameters and MIC values, our study found disc diffusion diameters for determining azithromycin resistance as unreliable.[1415] The results of this study indicate that isolates with diameters between 10–12 mm should be reconfirmed with respective MIC values before arriving at any conclusion.

Six isolates in this study showed true resistance to azithromycin. These isolates were studied for genes of azithromycin resistance - mphA, mphB, and mefB. The mphA gene was detected in a single isolate of non typhoidal Salmonella spp. Five S. Typhi isolates were negative for mphA gene while none of the isolates showed mphB and mefB genes. Azithromycin resistance in non typhoidal Salmonellae spp. has been reported from several centers.[1016] Elevated MICs of typhoidal Salmonella spp. have been reported occasionally.[17] The plasmid borne mphA gene has been reported as one of the reasons for azithromycin resistance.[18] Association of a chromosomal macrolide inactivation gene cluster mphA-mrx-mphr(A) has also been associated with azithromycin resistance in non typhoidal Salmonellae spp.[16] In the present study as well, a single high-level azithromycin-resistant isolate (MIC 64 mg/L) harbored the mphA gene, while other 5 (MIC: 16–32 mg/L) did not. These results indicate that the mphA gene may mediate a high level of resistance to azithromycin in Salmonella, as described previously in studies.[1016] In addition, azithromycin resistance could arise from other probable mechanisms such as mutations in the rlpV and rlpD genes.[19] Currently, few studies have investigated azithromycin resistance mechanisms in Salmonella, specifically, typhoidal salmonella.

Annual consumption of azithromycin in the hospital was tracked from 2014 to 2017. Azithromycin showed an increased consumption from 4.4DDD per 1000 patient days to 6.7 DDD per 1000 patient days. The mean MIC of Salmonella isolates also increased from 5 mg/L in 2013 to 24 mg/L in 2017 showing the MIC creep over 5 years. Therefore, the increasing utilization of this antibiotic probably had a considerable role to play in the development of resistance in these organisms. The easy over the counter availability of azithromycin and its widespread use for treating upper respiratory infections may have an important role to play in the emerging drug resistance among Salmonella species in our population. A comparison with the utilization of ceftriaxone in the hospital showed that the DDD s of ceftriaxone remained constant over the past 4 years. Therefore, our study reinforces that ceftriaxone can be used as an effective therapeutic option for culture proven enteric fever cases as well as empirical therapy for suspected cases of enteric fever.

The clinical outcome of six patients who harbored azithromycin resistant strains was analyzed. All patients were treated with injection ceftriaxone for 14 days. One patient showed gall bladder polyps and received an extended treatment with injection ceftriaxone for 28 days. All the patients growing typhoidal salmonella recovered with appropriate therapy. One patient received injection ceftriaxone and azithromycin in the initial course of therapy. Salmonella isolates with elevated azithromycin MICs have been isolated and reported from travellers from India and also from the Indian population. The clinical relevance of elevated MICs remains unknown. Clinically and microbiologically correlated breakpoints by Parry et al. indicates that azithromycin MIC of <16 mg/L likely predicts a favorable clinical outcome.[13] Azithromycin reaches 50–100 fold concentration within the macrophages and polymorphonuclear cells which is crucial in treating intracellular organisms of enteric fever; nevertheless, it achieves very low plasma levels. Hence, MIC might not be a good indicator of decreased susceptibility to azithromycin.[20]

Conclusion

This study has several limitations which include a small sample size, being a single center study, use of Estrip for MIC and use of interpretive criteria of S. Typhi for S. Paratyphi A and nontyphoidal salmonella. However, azithromycin disc diffusion diameter interpretations as recommended by CLSI can mislabel a few sensitive strains as resistant. Caution needs to be exercised when using disc diffusion breakpoints for S. Typhi to interpret nontyphoidal Salmonella breakpoints. MphA gene is seen in resistant nontyphoidal salmonella with high azithromycin MICs. It is not seen in isolates with lower MICs and in typhoidal Salmonella. While, further studies to understand the mechanism of resistance to azithromycin among Salmonella spp. is required, ceftriaxone continues to be a reliable drug for empirical treatment of enteric fever.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.