Contribution of Antibiotic Susceptibility Testing and CH Typing Compared to Next-Generation Sequencing for the Diagnosis of Recurrent Urinary Tract Infections Due to Genetically Identical Escherichia coli Isolates: a Prospective Cohort Study of Cystitis in Women.

Recurrent cystitis is a common disease in women, mainly due to uropathogenic Escherichia coli (UPEC). For decades, typing methods now considered obsolete suggested that relapse by the same clone is dominant over reinfection, most UPEC strains being otherwise fully susceptible to antibiotics. We aimed to update these data. Thanks to a prospective study over 17 months, we recruited 323 women with cystitis. Of these, 251 of them had sporadic infection and 72 had recurrence, with 2 to 9 episodes per patient for a total of 131 UPEC isolates and 145 UPEC pairs at patient level. Phylogroups B2 (52.4%) and D (14.1%) were overall dominant, as expected due to their particular urovirulence. CH typing identified 119 distinct profiles with no CH type particularly associated with recurrence. Relapse was attested by CH typing for only 30.6% (22 out of 72), with very diverse situations ranging from all episodes due to the same clone to alternating reinfections and relapses. Next-generation sequencing confirmed the clonality for all but two of the 145 UPEC pairs. Antibiotic resistance was common for recurrent cystitis isolates (only 25 [17.2%] out of 145 UPEC pairs were fully susceptible), allowing us to predict UPEC clonality. Indeed, antibiotic susceptibility profile matched CH typing for 104 (71.7%) pairs. Finally, we demonstrated a large genetic diversity among UPEC isolates responsible for cystitis in women, even in cases of recurrence for which reinfection appeared dominant over relapse. Recurrent cystitis appears to be a heterogeneous disease requiring tailored treatment and prevention. More than half of women will experience cystitis during their lifetime. Among these women, 25% will experience a second episode within the following 6 months. It is epidemiologically important to discriminate relapses from reinfections. Relapse identification relies on long and laborious methods and might influence treatment. Therefore, the designation of time- and cost-effective strategies for this goal is of particular interest. Our work suggests using CH typing and antibiotic susceptibility profiles to type Escherichia coli, the main uropathogen.