Antibiotic resistance and molecular characterization of non-invasive clinical Haemophilus influenzae isolates in Germany 2019 and 2020.

Abstract

Background: Haemophilus influenzae (Hi) is known as a cause of invasive and non-invasive diseases. Especially ear, nose and throat (ENT) infections are common reasons for antibiotic prescriptions in outpatient settings in Germany. Therefore, antibiotic resistance surveillance is important to provide the basis of recommendations for the empirical usage of antibiotic agents.

Objectives: To provide data on susceptibility rates of oral antibiotics for non-invasive clinical Hi isolates in Germany and to investigate molecular resistance patterns of β-lactams, ciprofloxacin, doxycycline and trimethoprim/sulfamethoxazole.

Methods: Isolates were collected from a sentinel network of diagnostic laboratories in a prospective multicentre prevalence study. Antibiotic susceptibility testing was done with a commercial broth microdilution kit. MICs were interpreted according to EUCAST guidelines. Resistance gene sequencing and WGS were performed to analyze molecular antibiotic resistance patterns and genetic relationships between the isolates.

Results: In total, 215 Hi isolates were collected from 23 laboratories across Germany. The highest resistance rates were found for amoxicillin (n = 30; 14%), cefuroxime (n = 40; 18.6%) and trimethoprim/sulfamethoxazole (co-trimoxazole) (n = 34; 15.8%). Resistance to amoxicillin was mainly due to bla_TEM-1 (n = 29; 96.7%). PBP3 alterations were found in 39 of 40 cefuroxime-resistant isolates (97.5%). Two of the cefuroxime-resistant isolates harboured PBP3 mutation patterns that have not yet been associated with cefuroxime resistance; in one of them, a known lpoA mutation was found. One isolate showed no mutations in PBP3 or lpoA. All co-trimoxazole-resistant isolates (15.8%) showed known mutations in folA and its promoter region. Additionally, point mutations in folP were identified in a subset of these isolates. The most frequent sequence types (STs) were ST57 (n = 10) and ST103 (n = 10). Genetic cluster analysis identified six clusters, but no epidemiological link could be confirmed.

Conclusion: Resistance to oral antibiotics in non-invasive clinical Hi isolates in Germany was generally low. Amoxicillin is estimated to cover 86% of infections involving non-invasive Hi and, therefore, is still effective for the first-line empirical treatment for ENT infections in Germany. Further surveillance of antimicrobial susceptibility in non-invasive Hi isolates is important to ensure the data basis for guidelines of antibiotic usage.