We retrospectively reviewed medical records from 2 tertiary care centers and 1 university hospital in northeast France. Our main purpose was to assess how Gordonia bacteremia cases had been managed. Our secondary purpose was to identify risk factors associated with infection. We requested information from the laboratory database of each center to collect all blood cultures testing positive for Gordonia spp. during January 1, 2010–May 31, 2021. We included patients with blood cultures positive for Gordonia sampled while hospitalized in one of the centers during the study period. We sought informed consent by mailing information about the study to patients or the parents of pediatric patients. We excluded persons opposed to participation and adults under legal guardianship or conservatorship. From medical files, we collected data on underlying diseases and immunodeficiency; presence and type of long-lasting intravascular device use, if any; clinical characteristics; antimicrobial susceptibility; treatments; and outcomes. We considered outcomes favorable when complete clinical cure was achieved, as determined by normalization of blood tests and repeated negative blood cultures. We sought relapses that occurred through May 31, 2022, or death of the patient.

We included 10 patients (median 49 years of age), among whom 7 had immunodeficiency and 6 had immunosuppression factors, most commonly severe malnutrition or long-term corticosteroid therapy (Tables 1, 2). Fever was the most common clinical sign; no secondary infection site was reported among case-patients. Healthcare providers considered Gordonia a contaminant in medical records from 5 cases; 4 patients had no central venous catheter, 3 were immunocompetent, and 1 had controlled HIV infection. We identified subspecies G. sputi, G. bronchialis, and G. aichiensis. Despite 16S rRNA gene sequencing, among 3 case-patients bacterium identification was limited to the genus level (i.e., Gordonia spp.). Nine patients were empirically prescribed antimicrobial therapy, mostly with third-generation cephalosporins and β-lactam/β-lactamase inhibitor combinations, for a median duration of 15 days. All Gordonia spp. were susceptible to treatment in all cases. In the 3 cases in which a long-term catheter had been removed, apyrexia without relapse was obtained within a few days of device removal; median treatment period was shorter, 13 days (range 0–15 days), for those patients who had their catheters removed than for patents who had not had their catheters removed (15 days, range 7–32 days). One patient was cured, as assessed by clinical improvement and negative repeated blood sample, by device removal alone without antimicrobial treatment. Of 2 patients who experienced relapses, 1 recovered promptly after delayed catheter removal, and the other exhibited febrile relapses with no other obvious cause of fever (e.g., venous thrombosis), clinical signs of catheter infection, and repeated blood cultures positive for Gordonia.

In this study, based on clinical information from an extensive series of Gordonia bacteremia infections among patients from several hospitals, we analyzed risk factors regardless of patient immune status. We found the main risk factors were immunosuppression and indwelling catheters of any type, consistent with findings from previous reviews (4,5,7,9). The preferred treatment, with third-generation cephalosporins or β-lactam/β-lactamase inhibitor combinations for 1–2 weeks, was of shorter duration than in most previous cases and did not match the proposed treatment described earlier (4,7). However, the antimicrobial drugs used were effective against all Gordonia species. Among 8 case-patients with favorable outcomes, only 2 received an aminoglycoside, suggesting it might be reserved for use in more severe cases. The strains’ susceptibility to many antimicrobials, except for lesser susceptibility to sulfamethoxazole in some other studies, agrees with data published elsewhere (4,7,10,11). Effects after removing the central venous catheter were inconsistent but removal generally resulted in favorable outcomes. Positive effects from catheter removal has also been reported in other cases and reviews (7,9,12).

Our retrospective analysis of case management reflects difficulties related to late identification and uncertainty about the truly infectious nature of Gordonia spp. bacteria. Differences exist in the literature about whether a single positive blood culture should be considered contamination or actual infection. Some subspecies, such as G. terrae or G. sputi, seem more likely to be responsible for bacteremias; other species, such as G. bronchialis, are rarely represented (4,5). In this study, G. terrae was present in 4 patients, but G. bronchialis was found in at least 2 cases and was strongly considered a probable cause of true infection.

In reported cases in which Gordonia was considered a contaminant, empiric antimicrobial therapy was stopped after the few days required to achieve negative blood culture results, after which clinical signs regressed and repeated blood samples were negative in most cases (9,13). Whatever the significance of this microbiological result, whether specific Gordonia spp. are agents of infection or contaminants, short-term treatment may suffice in the absence of a central catheter or if subsequent blood samples are consistently negative, as described elsewhere (13).

One limitation of our retrospective study was that data were collected over >11 years, during which time variations in risk level and management strategies may have altered the experience of the small number (n = 10) of included case-patients. In addition, including cases in which Gordonia spp. were considered potential contaminates rather than causes of infection may have positively biased treatment outcomes. No cultures were taken from catheters to confirm their role in bacteremia, and antimicrobial susceptibility was not tested in all cases. Thus, some resistant strains might exist that have not yet been identified.

In the absence of consensus recommendations, standardizing management of bacteremias caused by Gordonia spp. should be pursued. In cases with positive blood cultures, Gordonia should be considered infective in immunosuppressed case-patients, especially when a central venous catheter is present. β-lactam monotherapy can be empirically prescribed with aminoglycoside added in severe cases. After any central venous catheter has been removed and clinical signs improve, discontinuing antimicrobial drugs after 1–2 weeks should be considered. However, further studies are needed to clarify treatment duration and make more precise recommendations about managing a central catheter, especially when it is not removable. Studies on the effectiveness of antimicrobial lock therapy would help, possibly combined with lengthened systemic antimicrobial therapy. Given the sometimes serious nature of infection, especially among immunocompromised patients, developing treatment strategies and increasing awareness among clinicians of Gordonia-caused bacteremias would be of public health benefit.