As illustrated by the present case and previous reports, severe A urinae infections can occur, and the contributory factors deserve consideration. In our patient, the actual mechanism for bacteremia remains unclear. The initial concern for acute pyelonephritis was prompted by a computed tomography finding of bilateral perinephric fat stranding. This finding was questioned because it is common in older patients without infection, hence, is highly nonspecific. A correlation with urinary outflow obstruction may be an important clue in cases like this one.25,26
Furthermore, whether the urinary tract truly was the source of the patient’s bacteremia is clouded by the differing antimicrobial susceptibility patterns of the A urinae blood and urine isolates. The simplest explanation for this discordance may be that all the isolates shared a common initial origin but adapted to different environments in the host (perhaps over time) or laboratory, producing phenotypic variation. Alternatively, the infection could have been polyclonal from the onset, with sampling error leading to the differing detected susceptibility patterns, or the blood and urine isolates may have represented independent acquisition events, involving distinct A urinae strains. Unfortunately (from an academic perspective), given patient preferences and recommendations from the infectious disease consultant, no bone biopsy was done for histology and culture to confirm infection and to allow comparative strain identification if A urinae was isolated.
Optimal treatment for A urinae spondylodiscitis has yet to be established. β-lactams have shown good clinical efficacy despite being bacteriostatic in vitro.27 Early in vitro studies showed synergistic bactericidal synergistic activity with penicillin plus aminoglycoside combination therapies.27-30 Cases of endocarditis have been successfully treated mainly with the combination of a β-lactam plus aminoglycoside combination therapy.30,31 Previous cases of spondylodiscitis have been treated successfully with diverse antimicrobial agents, including clindamycin, β-lactams, cephalosporins, fluoroquinolones, and aminoglycosides.14
Our patient improved rapidly while receiving empiric therapy with vancomycin and ceftriaxone and tolerated a rapid transition to oral amoxicillin and levofloxacin. This is the shortest IV treatment course for A urinae spondylodiscitis reported to date. We suspect that such rapid IV-to-oral transitions will suffice in most stable patients with A urinae spondylodiscitis or other invasive A urinae infections in line with the results of the OVIVA and POET trials.32,33
Conclusions
We believe A urinae UTI in the absence of obvious predisposing factors should prompt evaluation for urinary outflow obstruction. Despite improved laboratory diagnostic techniques, spondylodiscitis related to A urinae remains a rare entity and thus definitive treatment recommendations are difficult to make. However, we suspect that in many cases it is reasonable to extrapolate from the results of the POET and OVIVA trials and rapidly transition therapy of A urinae spondylodiscitis from IV to oral antibiotics. We suspect a review of the US Department of Veterans Affairs population might uncover a higher incidence of A urinae infection than previously estimated due to the population demographics and the epidemiology of A urinae.