Toward the development and evaluation of sewage surveillance of antibiotic resistance in Klebsiella pneumoniae

Abstract

Antibiotic resistance is one of the greatest threats to modern medicine. It severely limits treatment options and is associated with a higher risk of treatment failure, morbidity and mortality. Representative surveillance data are essential to address this challenge, but in many parts of the world such data are lacking, often due to resource limitations. Because sewage contains pooled bacteria from a large number of people, sewage analysis is emerging as a potentially resource-efficient approach for monitoring antibiotic resistance. While previous studies have shown promise for Escherichia coli, other important pathogens, including Klebsiella pneumoniae, can also be detected in sewage. K. pneumoniae is an opportunistic pathogen within the K. pneumoniae species complex (KpSC) and is associated with high mortality and resistance levels, including to last-line antibiotics. This thesis aimed to develop and evaluate a sewage-based system for surveillance of antibiotic resistance in K. pneumoniae at the population level. Four studies addressed key steps toward this goal. Study I demonstrated that the sampling strategy can significantly influence the bacterial diversity and representativeness of the sewage sample and the subsequently generated resistance data. Study II evaluated different protocols for isolating K. pneumoniae from sewage. This study identified a culture-based protocol that isolates K. pneumoniae from sewage samples with high efficiency and demonstrated its applicability for sewage-based surveillance studies. By applying the developed sewage monitoring system, Studies III and IV assessed how well sewage analyses could reflect resistance in K. pneumoniae from patients. Study III showed strong associations between resistance rates in clinical and hospital sewage isolates, whereas associations with resistance rates in municipal sewage isolates were weaker. It also indicated that the distribution of KpSC members differs by sewage types, with hospital sewage more closely resembling human samples. Study IV further supported the presence of associations between sewage and patient isolates by revealing genetic similarities and overlaps between ESBL-producing K. pneumoniae isolates from sewage and patients. Altogether, the findings of this thesis could open up promising opportunities for using sewage analyses to monitor antibiotic resistance in clinical K. pneumoniae and to track clinically important lineages at the population level. However, further validations are needed, particularly in settings with different resistance landscapes, before such a system can be implemented.