Endogenous endophthalmitis: new insights from a 12-year cohort study.

OBJECTIVES: To describe the clinical and microbiological characteristics of patients with endogenous endophthalmitis (EE), determine factors associated with outcome and propose a management plan for EE. METHODS: Retrospective case series in two tertiary referral centres from 2010 to 2022. RESULTS: Sixty-four eyes of 53 patients were included. Bilateral involvement occurred for 11/53 patients (21%). Ocular symptoms were the only first manifestation of the disease in 36/53 (68%) of cases; signs of sepsis were evident in 17/53 (32%). Imaging tests detected at least one extraocular focus of infection in 34/53 patients (64%), with contrast-enhanced thoraco-abdominopelvic computed tomography showing relevant findings in 28/50 (56%) of cases. EE was microbiologically confirmed in 43/53 patients (81%); the organisms involved were: Gram-positive bacteria (19/53, 36%), Gram-negative bacteria (13/53, 25%) and Candida sp. (11/53, 21%). Klebsiella pneumoniae was the most common bacteria (10/32, 31%). Blood cultures were positive in 28/53 patients (53%) and eye samples in 11/41 eyes (27%). All patients were treated with systemic antimicrobial therapy, 39/64 eyes (61%) received anti-infective intravitreal injection(s) and 17/64 eyes (27%) underwent vitrectomy. Four patients (8%) died due to uncontrolled systemic infection. Final visual acuity (VA) was < 20/400 in 28/57 eyes (49%) and ocular structural loss (bulbar phthisis or enucleation/evisceration) was reported in 18/64 eyes (28%). In multivariate analysis, initial VA was the only parameter associated with visual and/or structural loss of the eye (OR = 24.44 (4.33-228.09) and 5.44 (1.33-26.18) respectively). CONCLUSIONS: EE remains a severe infection with a poor ocular outcome. We propose a standard protocol to improve diagnosis and medical management.

Listeria monocytogenes-associated spontaneous bacterial peritonitis in France: a nationwide observational study of 208 cases.

BACKGROUND: Listeriosis is a foodborne infection caused by Listeria monocytogenes. Three main forms of listeriosis are well characterised, but little is known about L monocytogenes-associated spontaneous bacterial peritonitis. We used data from the French national surveillance of listeriosis to perform a nationwide retrospective study. METHODS: All patients with L monocytogenes isolated by culture from a peritoneal fluid sample in France between April 1, 1993, and Dec 31, 2022, were included. Individuals for whom bacterial peritonitis was not confirmed and those who also had another type of invasive listeriosis were excluded. A standardised checklist was used to collect demographic, clinical, and biological data as well as antibiotic treatment and follow-up data. The primary outcome was to determine the characteristics of L monocytogenes-associated spontaneous bacterial peritonitis. We did descriptive analyses and assessed risk factors for 1-month mortality using an exploratory multivariable Cox model analysis. FINDINGS: Among the 8768 L monocytogenes cases reported, 208 (2%) were patients with L monocytogenes-associated spontaneous bacterial peritonitis. Mean age was 65 years (SD 13), 50 (24%) of 208 patients were female, and 158 (76%) were male (no data on race or ethnicity were available). 200 (98%) of 205 patients with L monocytogenes-associated spontaneous bacterial peritonitis with available data had immunosuppressive comorbidities, including cirrhosis (148 [74%] of 201 with available data), ongoing alcoholism (58 [62%] of 94), and ongoing neoplasia (60 [31%] of 195). Causes of ascites included cirrhosis (146 [70%] of 208), ongoing neoplasia (26 [13%]), end-stage heart failure (13 [6%]), and peritoneal dialysis (11 [5%]). Among those with available data, presentation was pauci-symptomatic and non-specific; only 67 (50%) of 135 patients presented with fever, 49 (37%) of 132 with abdominal pain, and 27 (21%) of 129 with diarrhoea. 61 (29%) of 208 patients were dead at 1 month, 92 (44%) were dead at 3 months, and 109 (52%) were dead at 6 months after diagnosis. Ongoing neoplasia (hazard ratio 2·42 [95% CI 1·05-5·56]; p=0·039), septic shock (8·03 [2·66-24·02]; p=0·0021), and high blood leukocyte count (1·05 [1·00-1·09]; p=0·045) were independently associated with 1-month mortality. INTERPRETATION: Despite the non-specific and mild presentation of L monocytogenes-associated spontaneous bacterial peritonitis, the outcome is poor and similar to that of neurolisteriosis, and so identification of L monocytogenes in ascitic fluid samples requires urgent parenteral amoxicillin-based treatment to avoid a fatal outcome. FUNDING: Institut Pasteur, Inserm, and French Public Health Agency. TRANSLATION: For the French translation of the abstract see Supplementary Materials section.

Hypervirulent Klebsiella pneumoniae-related bacteremia in intensive care unit: A retrospective cohort study.

INTRODUCTION: Hypervirulent Klebsiella pneumoniae (hvKP) bloodstream infections (BSI) have rarely been reported in critically ill patients. METHODS: We conducted a retrospective study of KP-BSI between January 2016 and December 2020 in an adult medical intensive care unit (ICU) of our tertiary care hospital. Hypervirulent phenotype was defined by the detection of both rmpA and iutA. RESULTS: Seventy patients diagnosed with K. pneumonia BSI were included, of whom 9 (13 %) had hvKP infection. Pneumonia accounted for 56 % of hvKP-BSI and for 28 % of those with cKP. Fifty-six percent of patients with hvKP-BSI were homeless, versus 2 % of those with cKP-BSI (p < 0.001). The 30-day mortality rate reached 44 % for hvKP-BSI and 34 % for cKP-BSI (p = 0.7) and did not appear related to the hypervirulent phenotype in multivariable analysis. DISCUSSION: We here evidenced a new clinical entity of hvKP-BSI associated with pulmonary infection in homeless patients, which exhibits high mortality.

Increased Risk of Invasive Aspergillosis in Immunocompromised Patients With Persistent SARS-CoV-2 Viral Shedding >8 Weeks, Retrospective Case-control Study.

Background: Immunocompromised patients now represent the population most at risk for severe coronavirus disease 2019. Persistent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) viral shedding was reported in these patients ranging from several weeks up to 9 months. We conducted a bicentric retrospective case-control study to identify risk and prognostic factors associated with persistent viral shedding in immunocompromised patients. Material and Methods: Symptomatic immunocompromised adults with persistent SARS-CoV-2 viral shedding >8 weeks were retrospectively included between 1 March 2020 and 24 April 2022 at 2 university hospitals in Paris, France, and matched with a control group consisting of symptomatic immunocompromised patients without persistent viral shedding. Results: Twenty-nine immunocompromised patients with persistent viral shedding were compared with 40 controls. In multivariate analysis, fever and lymphocytopenia (<0.5 G/L) were associated with an increased risk of persistent viral shedding (odds ratio [OR]: 3.3; 95% confidence interval [CI], 1.01-11.09) P = .048 and OR: 4.3; 95% CI, 1.2-14.7; P = .019, respectively). Unvaccinated patients had a 6-fold increased risk of persistent viral shedding (OR, 6.6; 95% CI, 1.7-25.1; P = .006). Patients with persistent viral shedding were at risk of hospitalization (OR: 4.8; 95 CI, 1.5-15.6; P = .008), invasive aspergillosis (OR: 10.17; 95 CI, 1.15-89.8; P = .037) and death (log-rank test <0.01). Conclusions: Vaccine coverage was protective against SARS-CoV-2 persistent viral shedding in immunocompromised patients. This new group of immunocompromised patients with SARS-CoV-2 persistent viral shedding is at risk of developing invasive aspergillosis and death and should therefore be systematically screened for this fungal infection for as long as the viral shedding persists.

Can Chatbot Artificial Intelligence Replace Infectious Diseases Physicians in the Management of Bloodstream Infections? A Prospective Cohort Study.

BACKGROUND: The development of chatbot artificial intelligence (AI) has raised major questions about their use in healthcare. We assessed the quality and safety of the management suggested by Chat Generative Pre-training Transformer 4 (ChatGPT-4) in real-life practice for patients with positive blood cultures. METHODS: Over a 4-week period in a tertiary care hospital, data from consecutive infectious diseases (ID) consultations for a first positive blood culture were prospectively provided to ChatGPT-4. Data were requested to propose a comprehensive management plan (suspected/confirmed diagnosis, workup, antibiotic therapy, source control, follow-up). We compared the management plan suggested by ChatGPT-4 with the plan suggested by ID consultants based on literature and guidelines. Comparisons were performed by 2 ID physicians not involved in patient management. RESULTS: Forty-four cases with a first episode of positive blood culture were included. ChatGPT-4 provided detailed and well-written responses in all cases. AI's diagnoses were identical to those of the consultant in 26 (59%) cases. Suggested diagnostic workups were satisfactory (ie, no missing important diagnostic tests) in 35 (80%) cases; empirical antimicrobial therapies were adequate in 28 (64%) cases and harmful in 1 (2%). Source control plans were inadequate in 4 (9%) cases. Definitive antibiotic therapies were optimal in 16 (36%) patients and harmful in 2 (5%). Overall, management plans were considered optimal in only 1 patient, as satisfactory in 17 (39%), and as harmful in 7 (16%). CONCLUSIONS: The use of ChatGPT-4 without consultant input remains hazardous when seeking expert medical advice in 2023, especially for severe IDs.

Defining standard and high dosages for ß-lactam agents administered by intermittent, prolonged or continuous infusion: a PK/PD simulation study.

BACKGROUND: The new definitions of antimicrobial susceptibility categories proposed by EUCAST in 2020 require the definition of standard and high dosages of antibiotic. For injectable ß-lactams, standard and high dosages have been proposed for short-infusion regimens only. OBJECTIVES: To evaluate dosages for ß-lactams administered by prolonged infusion (PI) and continuous infusion (CI). METHODS: Monte Carlo simulations were performed for seven injectable ß-lactams: aztreonam, cefepime, cefotaxime, cefoxitin, ceftazidime, piperacillin and temocillin. Various dosage regimens based on short infusion, PI or CI were simulated in virtual patients. Pharmacokinetic (PK) profiles and PTAs were obtained based on reference population PK models, as well as PK/pharmacodynamic targets and MIC breakpoints proposed by EUCAST. Alternative dosage regimens associated with PTA values similar to those of recommended dosages up to the breakpoints were considered acceptable. RESULTS: Adequate PTAs were confirmed for most EUCAST short-infusion dosage regimens. A total of 9 standard and 14 high dosages based on PI (3 to 4 h) or CI were identified as alternatives. For cefepime and aztreonam, only PI and CI regimens could achieve acceptable PTAs for infections caused by Pseudomonas spp.: 2 g q8h as PI of 4 h or 6 g/24 h CI for cefepime; 2 g q6h as PI of 3 h or 6 g/24 h CI for aztreonam. CONCLUSIONS: These alternative standard and high dosage regimens are expected to provide antibiotic exposure compatible with new EUCAST definitions of susceptibility categories and associated MIC breakpoints. However, further clinical evaluation is necessary.

Analysis of 15 cases from a monocentric cohort of 307 liver abscesses.

BACKGROUND: Out of the context of haematological patients, Candida sp. is rarely retrieved from pyogenic liver abscesses (PLA). OBJECTIVES: Our objective was to assess the risk factors for occurrence, and clinical, microbiological characteristics, management and outcome of Candida pyogenic liver abscesses (C-PLA). PATIENTS/METHODS: We retrospectively analysed C-PLA cases and compared them to pyogenic liver abscesses exclusively due to bacteria (B-PLA) included in our monocentric database on liver abscesses. Unfavourable course was defined as the occurrence of a primary treatment failure (PTF), recurrence after an initial cure, or death within 3 months after diagnosis. RESULTS: Between 2010 and 2018, 15 C-PLA and 292 B-PLA were included. All C-PLA had a biliary origin and were polymicrobial. All patients with C-PLA had at least one comorbidity at risk for Candida infection and 7 (53.3%) presented with sepsis requiring an admission in intensive care unit. Median duration of antifungal treatment was 42 days [24-55]. In multivariate analysis, compared with B-PLA, a medical history of malignancy (OR 4.16; 95%CI 1.15-18.72) or liver abscess (OR 7.39; 95%CI 2.10-26.62), and sepsis with severity criteria (OR 3.52; 95%CI 1.07-11.90) were independently associated with the occurrence of C-PLA. In multivariate analysis, C-PLA was associated with a higher risk of recurrence (HR 3.08; 95%CI 1.38-11.22). CONCLUSION: Candida liver abscesses in non-neutropenic is a rare and severe disease. The high rate of recurrence should lead to discuss a more intensive treatment.

Advances in antibacterial treatment of adults with high-risk febrile neutropenia.

BACKGROUND: High-risk febrile neutropenia (HR-FN) is a life-threatening complication in patients with haematological malignancies or receiving myelosuppressive chemotherapy. Since the last international guidelines were published over 10 years ago, there have been major advances in the understanding and management of HR-FN, including on antibiotic pharmacokinetics and discontinuation/de-escalation strategies. OBJECTIVES: Summarizing major advances in the field of antibacterial therapy in patients with HR-FN: empirical therapy, pharmacokinetics of antibiotics and antibiotic stewardship. SOURCES: Narrative review based on literature review from PubMed. We focused on studies published between 2010 and 2023 about the pharmacokinetics of antimicrobials, management of antimicrobial administration, and discontinuation/de-escalation strategies. We did not address antimicrobial prophylaxis, viral or fungal infections. CONTENT: Several high-quality publications have highlighted important modifications of antibiotic pharmacokinetics in HR-FN, with standard dosages exposing patients to underdosing. These recent clinical and population pharmacokinetics studies help improve management protocols with optimized initial dosing and infusion rules for ß-lactams, vancomycin, daptomycin and amikacin; they highlight the potential benefits of therapeutic drug monitoring. A growing body of evidence also shows that antibiotic discontinuation/de-escalation strategies are beneficial for bacterial ecology and patients' outcome. We further discuss methods and limitations for implementation of such protocols in haematology. IMPLICATIONS: We highlight recent information about the management of antibacterial therapy in HR-FN that might be considered in updated guidelines for HR-FN management.

Surgical site infection after pelvic bone and soft tissue sarcoma resection: Risk factors, microbiology, and impact of extended postoperative antibiotic prophylaxis.

BACKGROUND: Pelvic bone and/or soft tissue sarcoma removal surgeries are associated with a high rate of surgical site infection (SSI). The recommended antibiotic prophylaxis (ABP) duration is 24-48 h. We aimed to assess the impact of extended ABP (5 days) on the SSI rate and describe the microbiology of SSI in bone and/or soft tissue pelvic sarcomas. METHODS: We retrospectively included all consecutive patients who underwent pelvic bone and/or soft tissue sarcoma removal surgery between January 2010 and June 2020. RESULTS: We analyzed 146 patients with pelvic bone (45, 31%) or soft tissue (101, 69%). Sixty patients (41%) developed SSI. SSI occurred in 13/28 (46.4%) in the extended ABP group versus 47/118 (39.8%) in the standard group (p = 0.53). In multivariable analysis, risk factors for SSI were surgery duration (OR: 1.94 [1.41-2.92] per h), stay in postoperative ICU for more than 2 days (12.0 [2.8-61.3]), and shred or autologous skin flap (39.3 [5.8-409.5]). Extended ABP was not associated with SSI. SSI were mainly polymicrobial with Enterobacterales (57.4%) and Enterococcus (45%). CONCLUSIONS AND DISCUSSION: Pelvic bone and/or soft tissue sarcoma removal surgery is highly prone to postoperative infection. Extending the ABP to 5 days does not reduce the level of SSI.

Antimicrobial for 7 or 14 Days for Febrile Urinary Tract Infection in Men: A Multicenter Noninferiority Double-Blind, Placebo-Controlled, Randomized Clinical Trial.

BACKGROUND: The optimal duration of antimicrobial therapy for urinary tract infections (UTIs) in men remains controversial. METHODS: To compare 7 days to 14 days of total antibiotic treatment for febrile UTIs in men, this multicenter randomized, double-blind. placebo-controlled noninferiority trial enrolled 282 men from 27 centers in France. Men were eligible if they had a febrile UTI and urine culture showing a single uropathogen. Participants were treated with ofloxacin or a third-generation cephalosporin at day 1, then randomized at day 3-4 to either continue ofloxacin for 14 days total treatment, or for 7 days followed by placebo until day 14. The primary endpoint was treatment success, defined as a negative urine culture and the absence of fever and of subsequent antibiotic treatment between the end of treatment and 6 weeks after day 1. Secondary endpoints included recurrent UTI within weeks 6 and 12 after day 1, rectal carriage of antimicrobial-resistant Enterobacterales, and drug-related events. RESULTS: Two hundred forty participants were randomly assigned to receive antibiotic therapy for 7 days (115 participants) or 14 days (125 participants). In the intention-to-treat analysis, treatment success occurred in 64 participants (55.7%) in the 7-day group and in 97 participants (77.6%) in the 14-day group (risk difference, -21.9 [95% confidence interval, -33.3 to -10.1]), demonstrating inferiority. Adverse events during antibiotic therapy were reported in 4 participants in the 7-day arm and 7 in the 14-day arm. Rectal carriage of resistant Enterobacterales did not differ between both groups. CONCLUSIONS: A treatment with ofloxacin for 7 days was inferior to 14 days for febrile UTI in men and should therefore not be recommended. CLINICAL TRIALS REGISTRATION: NCT02424461; Eudra-CT: 2013-001647-32.

Hypervirulent Klebsiella Pneumoniae, an Emerging Cause of Endogenous Endophthalmitis in A French Center: A Comparative Cohort Study.

INTRODUCTION: Klebsiella pneumoniae (KP) is the most common cause of endogenous endophthalmitis (EE) in Asia, but data in Europe are scarce. We describe eight cases of KP EE compared to a cohort of EE in a French center. METHODS: EE cases were retrospectively studied between January 2014 and January 2021. KP EE cases were analyzed to assess clinical, microbiological features, and outcome. RESULTS: Among the 33 EE cases identified, the first causative agent (24%, n = 8) was KP, mainly (7/8) with hypervirulent phenotype (hvKP). All but one of these cases occurred from December 2019 to January 2021. Contrary to non-KP patients, KP patients had multiple extraocular infective foci (p = .006), all presented with liver abscesses (p < .001), 50% had cerebral involvement (p = .13). Visual outcome was poor in both groups. CONCLUSION: KP is an emerging cause of EE in a French center, consistently associated with liver abscesses, frequent cerebral involvement, and predominance of hvKP strains.

Large retrospective study analysing predictive factors of primary treatment failure, recurrence and death in pyogenic liver abscesses.

PURPOSE: Pyogenic liver abscess (PLA) is a severe disease, which unfavourable evolution remains frequent. Our objective was to assess predictive factors of unfavourable outcome in patients with PLA. METHODS: We conducted a retrospective study in a French tertiary care centre. All patients admitted for PLA between 2010 and 2018 were included. Unfavourable course was defined as the occurrence of a primary treatment failure (PTF), recurrence of PLA after an initial cure, or death within 3 months after diagnosis. Hazard ratios (95% CI) were calculated with multivariable Cox proportional hazard models. RESULTS: 302 patients were included among which 91 (30.1%) patients had an unfavourable outcome because of PTF, recurrence or death in 55 (18.2%), 28 (9.2%) and 32 (10.6%) patients, respectively. Hepatic metastases (HR 2.08; 95% CI 1.04-4.15), a nosocomial infection (2.25; 1.14-4.42), portal thrombosis (2.12; 1.14-3.93), and the isolation of Enterococcus spp. (2.18; 1.22- 3.90) were independently associated with PTF. Ischemic cholangitis (6.30; 2.70-14.70) and the isolation of Streptococcus spp. (3.72; 1.36-10.16) were associated with the risk of recurrence. Charlson comorbidity index (HR 1.30 per one point; 95% CI 1.15-1.46; p < 0.001), portal thrombosis (3.53; 1.65-7.56) and the presence of multi-drug-resistant organisms (3.81; 1.73-8.40) were associated with mortality within 3 months following PLA diagnosis. PLA drainage was the only factor associated with a lower mortality (0.14; 0.06-0.34). CONCLUSION: Identification of specific risk factors may help to improve the management of PLA and to elaborate targeted recommendations according to patient's and disease's characteristics.

Antimicrobial stewardship in high-risk febrile neutropenia patients.

BACKGROUND: The 2011 4th European Conference on Infections in Leukemia (ECIL4) guidelines recommend antibiotics de-escalation/discontinuation in selected febrile neutropenia (FN) patients. We aimed to assess the impact of an antimicrobial stewardship (AMS) program based on these guidelines on antibiotics use and clinical outcomes in high-risk FN patients. METHODS: We conducted an observational study in the hematology department of Cochin University Hospital in Paris, France. An ECIL4-based antibiotics de-escalation and discontinuation strategy was implemented jointly by the hematologists and the AMS team. The pre-intervention (January-October 2018) and post-intervention (January-October 2019) periods were compared. We retrospectively collected clinical and microbiological data. We compiled antibiotics consumptions via hospital pharmacy data and standardized them by calculating defined daily doses per 1000 patient-days. We analyzed the two-monthly antibiotic consumption using an interrupted time series method and built a composite endpoint for clinical outcomes based on transfer to the intensive care unit (ICU) and/or hospital death. RESULTS: Overall, 273 hospital stays (164 patients) in the pre-intervention and 217 (148 patients) in the post-intervention periods were analyzed. Patients were mainly hospitalized for intensive chemotherapy for acute leukemia or autologous stem-cell transplant for myeloma. Patients were slightly younger in the pre-intervention compared to the post-intervention period (median age 60.4 vs 65.2 years, p = 0.049), but otherwise comparable. After implementation of the AMS program, glycopeptide and carbapenem use decreased by 85% (p = 0.03) and 72% (p = 0.04), respectively. After adjustment on confounders, the risk of transfer to the ICU/death decreased significantly after implementation of the AMS program (post-intervention period: odds-ratio = 0.29, 95% Confidence Interval: 0.15-0.53, p < 0.001). CONCLUSION: Implementation of a multidisciplinary AMS program for high-risk neutropenic patients was associated with lower carbapenem and glycopeptide use and improved clinical outcomes.

Oral levofloxacin: population pharmacokinetics model and pharmacodynamics study in bone and joint infections.

OBJECTIVES: This study aimed at characterizing the pharmacokinetics (PK) of oral levofloxacin in adult patients in order to optimize dosing scheme and explore the PK/pharmacodynamics (PD) of levofloxacin in bone and joint infections (BJIs). METHODS: From November 2015 to December 2019, all patients hospitalized in Cochin Hospital, treated with levofloxacin and who had at least one dosage for therapeutic drug monitoring were included. PK was described using non-linear mixed-effect modelling. In a subgroup of patients with BJIs, the association between PK, MIC for the isolated pathogen and clinical outcome was investigated. Monte Carlo simulations investigated dosing regimens to achieve the PK/PD target (AUC/MIC ratio >100). RESULTS: One hundred and two patients were included (199 measurements), including 32 treated for BJI. A one-compartment model with first-order absorption and elimination best described the data. Effects of estimated creatinine clearance (eCLCR) and age were significant on levofloxacin clearance. In BJI patients, no significant association was found between levofloxacin PK/microbiological parameters and either clinical outcome or adverse events. Based on our model, we proposed optimized oral levofloxacin dosing regimens according to renal function, to reach the PK/PD target AUC/MIC ratio >100 for three frequent causative pathogens (Staphylococcus aureus, Enterobacterales and Pseudomonas aeruginosa). CONCLUSIONS: Our results reinforce the need of determining the MIC and using therapeutic drug monitoring in complex infections caused by P. aeruginosa.