Phenotypic Collateral-Susceptibility and Cross-Resistance Signatures among Urinary Escherichia coli Isolates

Collateral effects occur when resistance to one antimicrobial is associated with altered susceptibility to another, creating either collateral susceptibility or cross-resistance. This relationship can provide phenotypic information beyond a conventional antibiogram. Objective to characterize collateral-susceptibility and cross-resistance signatures among urinary Escherichia coli using culture-based antimicrobial susceptibility testing. Eighty non-duplicate E. coli isolate records were evaluated with a nine-agent disk-diffusion panel. Broth microdilution minimum inhibitory concentrations (MICs) were analyzed for ciprofloxacin, ceftriaxone, gentamicin, and nitrofurantoin. Multidrug resistance (MDR), multiple-antibiotic-resistance (MAR) index, pairwise log2 MIC shifts, Mann–Whitney tests, Benjamini–Hochberg false-discovery-rate correction, Wilson confidence intervals, and Spearman correlations were applied. Resistance was highest to ampicillin (76.3%), trimethoprim–sulfamethoxazole (63.8%), ciprofloxacin (61.3%), and ceftriaxone (56.3%), whereas nitrofurantoin (7.5%) and meropenem (3.8%) retained high activity. Fifty-six isolates (70.0%) met the MDR definition. Ciprofloxacin-resistant isolates had a median nitrofurantoin MIC of 8 µg/mL compared with 16 µg/mL among ciprofloxacin-susceptible isolates (median shift −1 log2 dilution; FDR q<0.001). Ceftriaxone resistance showed the same one-dilution downward shift in nitrofurantoin MIC (q<0.001). In contrast, ciprofloxacin and ceftriaxone displayed strong bidirectional cross-resistance signatures. MAR index correlated positively with ciprofloxacin, ceftriaxone, and gentamicin MICs but inversely with nitrofurantoin MIC. A phenotypic analysis combining categorical susceptibility with quantitative MIC distributions can reveal resistance-associated susceptibility shifts that are not apparent from routine categorical reporting alone. Such associations should be interpreted as candidate collateral effects unless confirmed by longitudinal, isogenic, or mechanistic studies.