Doxycycline post-exposure prophylaxis could theoretically select for resistance to various antimicrobials in 19 pathobionts: an in silico analysis

Objectives
Doxycycline post exposure prophylaxis (PEP) has been shown to reduce the incidence of bacterial STIs. However, if there is genetic linkage between resistance to tetracycline and other antimicrobials, then it could also select for resistance to these other antimicrobials. We therefore undertook to evaluate if there is an association between the minimum inhibitory concentrations (MICs) of tetracycline and other antimicrobials in 19 clinically important bacterial species.

Methods
Mixed-effects linear regression was used to assess if minocycline MICs were associated with the MICs of eight other antimicrobials (ceftriaxone, ampicillin, oxacillin, vancomycin, erythromycin, levofloxacin, amikacin, and trimethoprim-sulfamethoxazole) in 19 bacterial species in the Antimicrobial Testing Leadership and Surveillance (ATLAS) database.

Results
With the notable exception of vancomycin, where no association was found, strong positive associations were typically found between the MICs of minocycline and each of the eight antimicrobials in each of the species assessed. For example, the minocycline MICs of all the Gram-positive species were positively associated with ampicillin, ceftriaxone, oxacillin and erythromycin MICs (all P-values < 0.001). The only exceptions were ampicillin for Streptococcus pyogenes and ceftriaxone for S. dysgalactiae, where no significant associations were found. Similarly in the Gram-negative species, the minocycline MICs of all the species except Haemophilus influenzae and Stenotrophomonas maltophilia were positively associated with the MICs of ceftriaxone, ampicillin, levofloxacin and amikacin (all P-values < 0.001).

Conclusions
There is a theoretical risk that doxycycline PEP could select for resistance not only to tetracyclines but to a range of other antimicrobials in each of the 19 pathobionts assessed.