Isolation of ciprofloxacin-resistant Legionella pneumophila in a patient with severe pneumonia

L. pneumophila, Legionnaires' disease, susceptibility

Sir,

Legionella species are responsible for 1%–5% of cases of community-acquired pneumonia. Legionella pneumophila serogroup 1 (SG1) accounts for >90% of Legionnaires' disease (LD) in North America and Europe¹ and is the cause of significant mortality. The mortality rate among patients with L. pneumophila infections continues to be high, up to 26%.² The antimicrobial agents most commonly used for treatment of LD are fluoroquinolones (e.g. ciprofloxacin or levofloxacin) and macrolides.³ In recent studies, we established wild-type distributions and determined the epidemiological cut-off values (ECOFFs) in clinical L. pneumophila SG1 isolates for 10 antimicrobials commonly used for the treatment of Legionella infections.⁴

A patient sought care at his general practitioner after several days of falling and body pains. On examination, the patient appeared ill and was sent to the emergency department of a nearby hospital. The initial chest radiograph demonstrated an infiltrate of the left lower lung field. The patient was admitted to the intensive care unit. Blood cultures were taken and antibiotic treatment was started with cefazolin and gentamicin. Urine was examined for the presence of Legionella antigens and when this test was reported positive, treatment was switched to 400 mg of ciprofloxacin intravenously twice daily. After initial improvement the clinical condition of the patient deteriorated, leading to intubation and mechanical ventilation. A new chest radiograph revealed a diffuse interstitial pneumonia. Bronchoalveolar lavage (BAL) was performed 4 days after treatment with ciprofloxacin was started and the patient slowly recovered. Eventually, culture of the BAL grew L. pneumophila SG1 after 4 days of ciprofloxacin treatment. After 10 days, the patient could be transferred to the ward. Therapy was then switched to 500 mg of clarithromycin orally twice daily. The patient's further recovery was uncomplicated.

The L. pneumophila SG1 strain was sent, as part of a national Legionella outbreak detection programme, to the reference laboratory for Legionella in Haarlem, The Netherlands.⁵ Susceptibility testing for ciprofloxacin was performed with Etest and an MIC value of ciprofloxacin of 2 mg/L was found. This value is outside the wild-type distribution range ECOFF = 1 mg/L as previously described and therefore potentially resistant.⁴

For sequencing of gyrA and gyrB (DNA gyrase) and parC and parE (topoisomerase IV) genes, extraction of L. pneumophila DNA was performed. The DNA extraction was performed by use of Qiagen's BioRobot^® EZ1 (Hilden, Germany) according to the manufacturer's instructions. The sequencing reaction was performed twice by using primer systems previously described for the L. pneumophila SG1 strain Paris.⁶ A comparative analysis of the obtained sequences was done using the published L. pneumophila genomes and data from the literature describing mutations in the quinolone resistance-determining region (QRDR) of type II topoisomerase of L. pneumophila by using DNAStar (WI, USA) software and the NCBI database.^6,7 For control experiments, the wild-type strain MTZ OLDA and a spontaneous quinolone-resistant mutant of this strain were used. MTZ OLDA is an environmental isolate (L. pneumophila SG1), isolated from the water supply of a large building.⁸ A point mutation in the QRDR of the gyrA gene was identified and this mutation led to an amino acid exchange at position 83 (Escherichia coli numbering system). The result of this amino acid exchange is a change in ciprofloxacin susceptibility. Mutation at the same position (amino acid 83) has also been reported for other spontaneous quinolone-resistant mutants (Table 1).

Table 1.

Amino acid position 83 in the QRDR of the gyrA gene from L. pneumophila SG1 (E. coli numbering system)

Strain	Amino acid codon	Amino acid	MIC (mg/L)			Method used	Reference/accession number
			ciprofloxacin	levofloxacin	moxifloxacin
Philadelphia 1	ACA	Thr	0.012	ND	ND	microdilution	8
Paris	ACA	Thr	ND	ND	0.0625	microdilution	6
Lens	ACA	Thr	ND	ND	ND		CR628337a
Corby	ACA	Thr	0.014	ND	ND	microdilution	8
Alcoy	ACA	Thr	ND	ND	ND	—	CP001828a
HL06041035	ACA	Thr	ND	ND	ND	—	FQ958211a
Lorraine	ACA	Thr	ND	ND	ND	—	FQ958210a
Strain 1
wild-type	ACA	Thr	0.015	ND	ND	microdilution	7
in vitro mutant	AAA	Lys	0.25	ND	ND
Strain 2
wild-type	ACA	Thr	0.015	ND	ND	microdilution	7
in vitro mutant	ATA	Ile	0.25	ND	ND
LP4
wild-type	ACA	Thr	ND	ND	0.0625	microdilution	6
in vitro mutant	ATA	Ile	ND	ND	32
MTZ OLDA
wild-type	ACA	Thr	0.38	0.094	0.38	Etest	this study
in vitro mutantb	ATA	Ile	2	0.5	1.5
Wild-type isolate #123	GCA	Ala	2c	0.75d	0.75d	Etest	this study

ND, not determined.

^aNBCI accession number.

^bMutants selected by ciprofloxacin.

^cDetermined in both laboratories.

^dBruin et al.⁴

It is known that, in general, pathogens can become resistant during the course of a patient's therapy and also induction of resistance upon exposure to antibiotics has been described.^6,7 The origin of resistance in the clinical isolate is as yet unclear. There are two possibilities. The first is that the patient contracted an L. pneumophila SG1 strain with this point mutation from the environment. Alternatively, the mutation occurred during the course of the patient's ciprofloxacin therapy. The start of antimicrobial therapy with ciprofloxacin was 4 days before collecting BAL material for culturing of Legionella and it could well be that the mutation did occur during that period. However, the initial response of the patient to therapy was poor and it was only with sustaining therapy in intensive care that recovery occurred. Although Legionella infections often result in severe disease, the elevated MIC of ciprofloxacin for this isolate may have contributed here. Our finding emphasizes the need for culture of clinical samples for Legionella and susceptibility testing of Legionella isolates.

This is, to our knowledge, the first report of a ciprofloxacin-resistant L. pneumophila isolated from a clinical specimen. Clinicians should be aware that a lack of clinical response during treatment with quinolones could be due to resistance. There is a clear need to perform susceptibility testing for clinical L. pneumophila isolates.

Funding

This study was funded by the Regional Laboratory of Public Health, Haarlem, The Netherlands.

Transparency declarations

None to declare.

Acknowledgements

We thank Dr F. Heilman and Dr J. Blaauw for sending the isolate and the case information.