Significant Publications on Infectious Diseases Pharmacotherapy in 2022

Article Selection Process

Members of HIDN were requested to nominate the most influential peer-reviewed ID pharmacotherapy (print and electronic) articles, excluding their own research, published between January 1 and December 31, 2022. Upon conclusion of nominations, the compiled list of publications was submitted to members of the Society of Infectious Diseases Pharmacists (SIDP) via electronic survey to vote for selecting 10 general ID and 1 HIV/AIDS articles that were felt to contribute significantly to ID pharmacotherapy. SIDP members were also afforded the opportunity to vote for articles not included in the initial list of nominations, with each selection counting as a single vote. Members were given the opportunity to abstain from voting in either category of article. The final ranking of publications was determined by total votes per survey response.

For the 16th iteration of the HIDN top 10 articles, sixty-eight general ID and three HIV/AIDS publications were nominated for submission to SIDP members. One-hundred fifty-three SIDP members in the United States participated in the survey with 128 and 30 votes among the general ID and HIV/AIDS articles, respectively. The rankings of the top 20 ID pharmacotherapy and 4 HIV/AIDS publications are listed in Tables 1 and 2, respectively. The top 10 general ID and top 1 HIV/AIDS papers are summarized here in alphabetical order of the first author.^16-39

Chuang YC et al. The combination of daptomycin with β-lactam antibiotics is more effective than daptomycin alone for vancomycin-resistant Enterococcus faecium bloodstream infection ²⁵

OPEN IN VIEWER

Table 1.

Results of SIDP Member ranking of significant publications on infectious diseases pharmacotherapy in 2022.

Rank	Reference	Article	Individuals ranking articles in top 10 (n = 128), n (%)
1	16	Jarvis JN et al. Single-dose liposomal amphotericin B treatment for cryptococcal meningitis	41 (32.0)
2	17	Howard-Anderson J et al. A desirability of outcome ranking analysis of randomized clinical trial comparing seven vs Fourteen days of antibiotics for uncomplicated Gram-negative bloodstream Infection	28 (21.9)
3	18	Moehring RW et al. Evaluation of an opt-out protocol for antibiotic de-escalation in patients with suspected sepsis: A multicenter, randomized, controlled trial	22 (17.2)
4	19	Thaden JT et al. Association of follow-up blood cultures with mortality in patients with Gram-negative bloodstream infections: A systematic review and meta-analysis	21 (16.4)
5	20	Tamma PD et al. Clinical impact of ceftriaxone resistance in Escherichia coli bloodstream infections: A multicenter prospective cohort study	20 (15.6)
6	21	Kaye KS et al. Colistin monotherapy vs combination therapy for carbapenem-resistant organisms	18 (14.1)
7	22	Veve MP et al. Prevalence and predictors of Pseudomonas aeruginosa among hospitalized patients with diabetic foot infections	18 (14.1)
8	23	Feuerstadt P et al. SER-109, an oral microbiome therapy for recurrent Clostridioides difficile Infection	18 (14.1)
9	24	Venugopalan V et al. Use of therapeutic drug monitoring to characterize cefepime‐related neurotoxicity	17 (13.3)
10	25	Chuang YC et al. The combination of daptomycin with β-lactam antibiotics is more effective than daptomycin alone for vancomycin-resistant Enterococcus faecium bloodstream infection	17 (13.3)
11	26	Choi SH et al. Efficacy of Doxycycline for mild-to-moderate community-acquired pneumonia in adults: A systematic review and meta-analysis of randomized controlled trials	17 (13.3)
12	27	Yoshimura J et al. Effect of Gram stain-guided initial antibiotic therapy on clinical response in patients with Ventilator-associated pneumonia	17 (13.3)
13	28	Alegria W et al. Early antibiotic discontinuation or de-escalation in high-risk patients with AML with Febrile Neutropenia and prolonged Neutropenia	16 (12.5)
14	29	Kaye KS et al. Effect of Cefepime/Enmetazobactam vs Piperacillin/Tazobactam on clinical cure and Microbiological Eradication in patients with Complicated urinary Tract Infection or acute pyelonephritis: A randomized clinical trial	15 (11.7)
15	30	Keller SC et al. A framework for implementing antibiotic stewardship in ambulatory care: Lessons learned from the Agency for health care research and Quality safety program for improving antibiotic use	15 (11.7)
16	31	Heffernan AJ et al. Pharmacodynamics of ceftriaxone for the treatment of methicillin-susceptible Staphylococcus aureus: is it a viable treatment option?	15 (11.7)
17	32	Fosse PE et al. Comparing outcomes among outpatients treated for pyelonephritis with oral cephalosporins vs first-line agents	13 (10.2)
18	33	Kooda K et al. Impact of Pharmacist-Led antimicrobial stewardship on appropriate antibiotic prescribing in the Emergency Department: A systematic review and meta-analysis	13 (10.2)
19	34	Mohamed A et al. Ceftriaxone vs cefazolin for the treatment of methicillin-susceptible S aureus bacteraemia	11 (8.6)
20	35	Buckley MS et al. Vancomycin with concomitant piperacillin/tazobactam vs cefepime or meropenem associated acute kidney injury in the critically ill: A multicenter propensity score-matched study	10 (7.8)

OPEN IN VIEWER

Table 2.

Results of SIDP Member ranking of significant publications on HIV/AIDS pharmacotherapy in 2022.

Rank	Reference	Article	Individuals ranking articles in top 10 (n = 30), n (%)
1	36	Delany-Moretlwe S et al. Cabotegravir for the prevention of HIV-1 in women: Results from HPTN 084, a phase 3, randomised clinical trial	16 (53.3)
2	37	Chounta V et al. Indirect comparison of 48-week efficacy and safety of long-acting cabotegravir and rilpivirine maintenance every 8 weeks with daily oral standard of care antiretroviral therapy in participants with virologically suppressed HIV-1-infection	11(36.7)
3	38	O’Halloran JA et al. Integrase strand transfer inhibitors are associated with incident diabetes mellitus in people with human immunodeficiency virus	1 (3.3)
4	39	Henderson HI et al. Predicting risk of multidrug-resistant Enterobacterales infections among people with HIV	1 (3.3)

Vancomycin-resistant Enterococcus (VRE) was designated a “serious” threat in the 2019 Antibiotic Resistance Threats Report from the US Centers for Disease Control and Prevention (CDC). Nearly 30% of hospital-associated enterococcal infections are vancomycin resistant, with VRE infections occurring more commonly in patients with weakened immune systems. ⁴⁰ The bactericidal activity of daptomycin makes it an ideal choice for treating VRE infections, and data suggests the addition of select β-lactams may lead to enhanced killing against some enterococcal strains. ⁴¹ However, questions remain about the exact role of β-lactam synergy for VRE infections particularly in isolates with elevated daptomycin minimum inhibitory concentration (MIC).

Chuang et al. conducted a multicenter, prospective observational study evaluating daptomycin alone (DAP) vs daptomycin plus a β-lactam (DAP + BL) for vancomycin-resistant E. faecium bloodstream infections (BSI) occurring between 2010 and 2021. Hospitalized adult patients with VRE BSI were eligible for inclusion if they were initiated on ≥8 mg/kg of daptomycin within five days of index culture collection and treated for ≥72 hours. Additionally, patients were included in the DAP + BL group if any β-lactam was initiated within 24 hours of starting DAP and maintained for ≥24 hours. Exclusion criteria included pneumonia, concurrent fosfomycin therapy, and non-E. faecium species. The primary outcome was clinical success, which was a composite of 14-day mortality, microbiologic failure, and change from DAP to linezolid.

A total of 430 eligible patients were treated with DAP (n = 45) or DAP + BL (n = 385). Baseline and clinical characteristics, including demographics, comorbidities, source of infection, and daptomycin MIC distributions, were comparable between the DAP and DAP + BL groups except for DAP duration (mean 2.7 vs 2.2 days, respectively, P = .02) and doses (mean 9.6 vs 10 mg/kg, respectively, P = .03). Carbapenems accounted for the majority of concurrent β-lactam use followed by cephalosporins and penicillins. Clinical success was significantly more common in patients who received DAP + BL compared to DAP alone (63.4 vs 42.2%, P = .006). The multivariable regression analysis demonstrated DAP + BL to be an independent predictor for clinical success (aOR 3.19, 95% CI 1.61-6.33, P = .001). A univariable logistic regression analysis comparing all clinical successes and failures showed no significant difference with respect to daptomycin dose or MIC (≤1, 2, or ≥4 mg/L). However, when evaluating clinical successes with DAP vs DAP + BL, the marginal effect analysis found a significant increase in clinical success for DAP + BL arm when daptomycin MICs were ≥2 mg/L (26.3%; 95% CI 9.3%-43.3%, P = .002). Creatine kinase (CK) elevations (>250 U/L) were observed in 8.8% of all patients, with seven patients within the DAP + BL arm (0 vs 7, P = .99) experiencing levels exceeding 2000 U/L. Adverse events, including CK elevations and new-onset thrombocytopenia, were comparable for the two treatment groups.

Limitations of this study include the observational study design, the relatively small sample size for the DAP monotherapy group, and composite nature of the primary outcome. This study adds to the growing body of literature which appears to support the use of daptomycin plus β-lactams for VRE BSI, most notably for isolates with daptomycin MICs that were ≥2 mg/L.

Delany-Moretlwe S et al. Cabotegravir for the Prevention of HIV-1 in Women: Results from HPTN 084, a Phase 3, Randomised Clinical Trial ³⁶

Two oral antiretroviral combination therapies are commonly utilized for pre-exposure prophylaxis (PrEP), which include tenofovir disoproxil fumarate and emtricitabine (TDF-FTC) for at-risk HIV-uninfected individuals via sexual acquisition and persons who inject drugs, and tenofovir alafenamide and emtricitabine (TAF-FTC) in the assigned male at birth, at-risk via sex population.^42,43 Both therapies have been proven to be effective but failures are common due to non-adherence. Despite the World Health Organization (WHO) recommending tenofovir-based regimens and over 70 countries adopting PrEP, young women in sub-Saharan African countries account for a disproportionately high number of new HIV infections as they face significant barriers like non-adherence, stigma, judgment, and the fear of violence from their community.^44,45

Delany-Moretlwe et al. conducted a randomized, double-blind, double-dummy, superiority trial from November 2017 to 2020 in seven sub-Saharan African countries to evaluate the efficacy and safety of long-acting, intramuscular (IM) cabotegravir compared with TDF-FTC. Sexually-active individuals assigned female sex at birth aged 18 to 45 years at risk for HIV infection, per the modified Vaginal and Oral Interventions to Control the Epidemic (VOICE) scoring tool, and agreeable to contraception were included.^46,47 Patients were excluded if they were pregnant or breastfeeding during randomization. The primary efficacy and safety endpoint were incidence of HIV infections and grade 2 or greater adverse events on therapy, respectively.

Among 3178 participants randomized who completed follow-up visits (median [IQR], age 25 years [22-30]), 1592 received cabotegravir starting with an oral lead-in period followed by IM injected every 4-8 weeks, and 1586 received TDF-FDC daily. Both treatment arms were comparable at baseline. In the intention-to-treat (ITT) population, a total of 40 incident HIV infections were identified with 4 in the cabotegravir arm and 40 in the comparator (HR .12, 95% CI .05-.31, P < .0001). In a post-hoc analysis excluding 1 patient in the cabotegravir group who had an HIV infection at study enrollment, only 3 HIV infections were identified in the cabotegravir group compared with 36 in the comparator (HR .09, 95% CI .04-.27, P < .0001) demonstrating a 91% risk reduction for HIV infection in favor of cabotegravir. An absolute risk difference between the two groups was -1.6% (95% CI -1.0 to -2.3%). All incident infections were found in patients with poor or non-adherence confirmed by plasma drug concentrations. No major resistance mutations were detected in the cabotegravir arm. Though they were common (92.2%), no differences were seen in grade 2 or greater adverse events. Most adverse events in the treatment arm included injection site reactions (38.0% vs 10.8%) and non-clinically relevant laboratory value abnormalities. Fifteen patients (.9%) receiving cabotegravir required discontinuation of therapy due to hepatic-related events, which was similar in the comparator (1.1%). Weight gain was higher in the cabotegravir arm (.4 kg, 95% CI .27-.51, P < .0001). Thus, cabotegravir demonstrated superiority over TDF-FTC in HIV prevention with improved convenience and discretion without increased risk for adverse events.

Limitations of this study included unblinding in patients who became pregnant or were breastfeeding, inadequate resources for HIV viral load and resistance mutation testing in resource-poor areas, and medication access. This study adds evidence for use of an alternative PrEP agent in a high-risk population where a great need has been identified.

Feuerstadt P et al. SER-109, an Oral Microbiome Therapy for Recurrent Clostridioides difficile Infection ²³

Clostridioides difficile infections remains an urgent threat to human health causing significant morbidity, mortality, and increased health care costs.^48,49 Current CDI treatment options are unable to eradicate C. difficile spores or repair the dysbiosis in the gastrointestinal microbiome that lead to infections. ⁵⁰ Fecal microbiota transplantation (FMT) aims to address this latter issue, and studies have shown beneficial yet variable effects on clinical response through recovery of bowel flora. ⁵¹ Additionally, FMT products are not approved by the Food and Drug Administration (FDA) and risk the transmission of occult pathogens, including multi-drug resistant bacteria and viruses.^52,53 Live biotherapeutic products (LBPs) are a new class of drugs that have been formulated to restore beneficial flora and reduce CDI recurrence while mitigating the risks of FMT. ⁵⁴

ECOSPOR III was a randomized, double-blind, placebo-controlled trial in the US and Canada that compared four capsules daily of SER-109, an LBP of live purified Firmicutes spores formulated from donor stool after extensive screening and processing, for 3 days to placebo in patients with recurrent CDI. Adult patients with ≥3 CDI episodes within 12 months identified with a positive C. difficile toxin enzyme immunoassay during a diarrheal episode and resolution of symptoms on standard-of-care antibiotics were included. Patients were stratified according to age and CDI treatment received and followed for 8 weeks after randomization. The primary efficacy endpoint was superiority of SER-109 in reducing CDI recurrence up to 8 weeks following administration. Exploratory analyses of stool microbiome-related changes were also conducted.

A total of 182 patients were randomized in the ITT population to receive SER-109 (n = 89) or placebo (n = 93). Patients were mostly white, female outpatients with a mean age of 65.5 ± 16 years. Approximately three-fourths of patients received treatment with vancomycin, and 39% of patients had ≥4 CDI episodes in the preceding 12 months. SER-109 was found to be superior to placebo in reducing recurrences (12% vs 40%, relative risk (RR) .32, 95% CI .18-.58, P < .001, NNT 3.6). This benefit remained in both the age-stratified and the CDI treatment-stratified group analyses. Additionally, SER-109 recipients experienced a higher sustained clinical response without CDI recurrence during the follow-up period compared with placebo (88% vs 60%, respectively). No serious adverse events attributable to SER-109 were observed, and mild-moderate gastrointestinal disorders were seen commonly (88%). Engraftment of SER-109, defined as the number of dose species detected not present at baseline, occurred by week 1, endured through week 8, and remained higher than placebo throughout the follow-up period. Secondary bile acids increased and proinflammatory Enterobacterales decreased in the SER-109 group.

Limitations of this study included minimal minority representation and lack of a pre-CDI-antibiotic stool specimen, which inhibited a full assessment of SER-109’s effect on the microbiome. Overall, this study showed SER-109 was superior in reducing CDI recurrence vs placebo and is a promising start for this new drug class in decreasing CDI burden in the US.

Howard-Anderson J et al. A Desirability of Outcome Ranking Analysis of Randomized Clinical Trial Comparing Seven vs Fourteen Days of Antibiotics for Uncomplicated Gram-Negative Bloodstream Infection ¹⁷

Randomized controlled trials (RCTs) have demonstrated that 7-day antibiotic courses for Gram-negative bacteremia (GNB) are noninferior to the conventional 14-day duration; however, the question between which treatment duration is associated with a safer and more desirable clinical outcome has yet to be elucidated.^55-57 Desirability of Outcome Ranking (DOOR), in addition to the clinical outcome of the intervention, takes into consideration treatment-related harms and benefits at the individual patient level beyond the discrete time on therapy.^58,59

Howard-Anderson et al. applied a post-hoc DOOR analysis on the study conducted by Yahav and colleagues, ⁵⁶ which was a multicenter, randomized, open-label, noninferiority trial comparing 90-day all-cause mortality, clinical failure, and readmission or prolonged hospitalization of a 7 vs 14 day antibiotic course for uncomplicated GNB. In the present study, DOOR outcomes adapted treatment failure, infectious complications relating to the index infection, and adverse events unrelated to infection in addition to the clinical outcome. The primary outcome of this current DOOR analysis was to validate the desirable outcome using the ITT population. Thus, patients were assigned DOOR rankings between from 1 to 5 where 1 represents the most desirable outcome, which is survival without any adverse events, and 5 is the least desirable, or mortality. The ranks in between 1 and 5 were assigned based on the number of undesirable events, up to 3, patients experienced. Finally, if the DOOR probabilities remained at 50%, this would indicate that there is no difference among the groups.

A total of 604 patients randomized to either the 7-day (n = 306) or 14-day (n = 298) courses were included in this analysis. Among all patients, 248 (41%) patients received rankings of 1 while 68 (11%) died thus receiving a rank of 5. DOOR distribution probabilities were similar in both 7 and 14-day durations among the ranks (51.1%, 95% CI, 46.7%–55.4%). The primary outcome to comparing DOOR ranks demonstrated no difference between the 2 treatment durations with mean difference of 6 days of antibiotic exposure, which may translate to overall improved antimicrobial stewardship curbing the rise of antibiotic resistance. ⁶⁰

Limitations of this study include the retrospective, post-hoc fitting of the variable from the original study into the DOOR analysis, the possibility of censored desirability outcomes limited by the 90-day follow-up for infection recurrence and long-term sequelae, and limited generalizability to uncomplicated GNB largely secondary to a urinary source (∼68%). Based on this analysis, it appears that patients with uncomplicated GNB may be treated with a 7-day course of antibiotics without compromising efficacy or safety and minimizing future collateral damage with antimicrobial resistance. This study demonstrated the potential advantages of DOOR analysis, which may prove to be beneficial in randomized, clinical trials.

Jarvis JN et al. Single-Dose Liposomal Amphotericin B Treatment for Cryptococcal Meningitis ¹⁶

Infections and deaths due to Cryptococcus remain high among patients with HIV/AIDS, with meningitis cases exceeding 150 000 worldwide each year. ⁶¹ The burden of disease is particularly high in sub-Saharan Africa accounting for the majority of cases and 75% of global deaths. ⁶² Standard of care for the treatment of cryptococcal meningitis has historically been associated with substantial toxicities, while alternative regimens are associated with poor efficacy.^63,64

Jarvis et al. conducted a phase III randomized, controlled, noninferiority trial to assess the safety and efficacy of an experimental induction regimen for cryptococcal meningitis across five African countries. ⁶⁵ Cerebral spinal fluid (CSF) samples with a positive India ink stain and/or cryptococcal antigen test were used to diagnose cryptococcal meningitis. HIV-positive patients ≥18 years of age with a first episode of cryptococcal meningitis were eligible for inclusion. Select exclusion criteria included pre-screening receipt of >2 doses of amphotericin B or fluconazole (≥800 mg), pregnancy, breastfeeding, and baseline laboratory abnormalities for alanine aminotransferase level, polymorphonuclear leukocyte count, and/or platelet count. Patients were randomly assigned in a 1:1 ratio to receive the experimental regimen (one dose of liposomal amphotericin B 10 mg/kg with 14 days of flucytosine 100 mg/kg/day and fluconazole 1200 mg/day) or the WHO-recommended regimen (amphotericin B deoxycholate 1 mg/kg/day with flucytosine 100 mg/kg/day for 7 days, followed by fluconazole 1200 mg/day for 7 more days). Upon completion of the 14-day induction phase, participants from both groups received fluconazole 800 mg/day for 8 weeks followed by 200 mg/day. The primary efficacy endpoint was all-cause mortality 10 weeks after randomization.

A total of 814 patients were randomized to the intervention (n = 407) and conventional treatment arms (n = 407) in the ITT population between January 2018 and February 2021. Baseline characteristics were similar between groups with the majority being male patients (60.3%) weighing a median of 53 kg (IQR, 47-60). Mortality at 10 weeks was similar between the experimental group (24.8%, 95% CI 20.7 to 29.3) and WHO group (28.7%, 95% CI 24.4 to 33.4) (absolute difference -3.9%; upper bound 95% CI 1.2, P < .001 for noninferiority). Fungal clearance from CSF was similar between groups (difference .017 log₁₀ colony-forming units/mL/day; 95% CI −.001 to .036), while grade 3 or 4 adverse events were significantly less common in patients treated with the experimental regimen (50 vs 62.3%, P < .001). Notably, patients treated with the experimental regimen experienced lower relative increases in serum creatinine from baseline to day 7 (20.2 vs 49.7%, P < .001).

Limitations to this study include the applicability to treatment of cryptococcal meningitis in patients without HIV given the potential differences in pathogenesis and host-immune response. Despite this limitation, the promising data in HIV-positive patients prompted the WHO to add the experimental regimen from this study into the “Guidelines for diagnosing, preventing and managing cryptococcal disease among adults, adolescents and children living with HIV” with a strong recommendation for induction treatment of cryptococcal meningitis. ⁶⁶

Kaye KS et al. Colistin Monotherapy vs Combination Therapy for Carbapenem-Resistant Organisms ²¹

Multi-drug resistant Gram-negative organisms, including extensively drug-resistant (XDR) Acinetobacter baumannii, XDR Pseudomonas aeruginosa, and carbapenem-resistant Enterobacterales (CRE) cause severe infections across the globe and are considered urgent and serious threats by the CDC. ⁴⁰ While treatment options for CRE have increased in the past decade, there are still gaps in the market for XDR A. baumannii and XDR P. aeruginosa leading to continued use of salvage regimens, which often include combination therapy with colistin and/or carbapenems. Colistin, however, is associated with significant toxicities, including nephrotoxicity and neurotoxicity, and efficacy concerns. Additionally, colistin in combination with carbapenem or rifampicin may result in higher rates of adverse effects and resistance.^67-71

The OVERCOME study was a randomized, double-blind, placebo-controlled trial that aimed to assess combination therapy of colistin and carbapenems vs colistin alone. The study was performed at sites in seven countries between October 2012 and August 2020. Adult patients with colistin-susceptible (MIC ≤2 mg/L) XDR A. baumannii, XDR P. aeruginosa, or CRE pneumonia and/or BSI were eligible for inclusion. Patients were excluded if they received ≥72 hours of polymyxin therapy within 96 hours of enrollment or if they were not expected to live past 24 hours of enrollment. Patients received a colistin loading dose of 5 mg/kg followed by 1.67 mg/kg every eight hours plus a carbapenem (imipenem 500 mg every six hours or meropenem 1000 mg every eight hours either infused over 30 minutes) or colistin plus placebo for 7-14 days. Patients who received at least one dose of trial medication and had pathogens with colistin susceptibility confirmed by broth microdilution were included in the modified intention-to-treat population (mITT), which was used for the primary outcome of all-cause 28-day mortality.

A total of 423 patients were included in the mITT population and randomized to the combination (n = 210) or monotherapy (n = 213) arms. Patients were mostly white males with a mean age of 68 ± 16.0 years admitted to the intensive care unit (69%) with pneumonia (70%). The most common comorbidities included hypertension (62%), diabetes (39%), and chronic lung disease (39%). The most prevalent pathogen was A. baumannii (78%). No difference was seen between combination therapy and monotherapy in 28-day mortality (37 vs 43%, respectively, P = .17) or time to mortality. This non-significant difference remained true in subgroup analyses of infection type, including pneumonia or BSI, and pathogen isolated. For clinical failure, there was no difference between the combination and monotherapy groups (58 vs 65%, 95% CI -3.1 to 16.6). Of the 334 patients included in the microbiologic analysis, microbiologic cure was achieved in 60% and 65% (difference 4.8%, 95% CI -5.6 to 15.2) of combination and monotherapy patients, respectively. Acute kidney injury rates were similar between groups (combination 49% vs monotherapy 52%, P = .55).

Limitations of this study include difficulties in accurately diagnosing pneumonia in critically ill patients. Similarly, mortality can be confounded by several factors that are often present in critically ill patients with XDR Gram-negative infections. Lastly, the carbapenem dosing used was lower than current recommendations thus not optimized to meet pharmacokinetic parameters.^72,73 Overall, colistin combination therapy did not demonstrate a mortality difference when compared with monotherapy, which highlights the need for continued investment into new treatment options for XDR pathogens.

Moehring RW et al. Evaluation of an Opt-Out Protocol for Antibiotic De-Escalation in Patients with Suspected Sepsis: A Multicenter, Randomized, Controlled Trial ¹⁸

Prompt administration of antibiotic therapy is pivotal in the management of patients with septic shock. ⁷⁴ Part of the Centers for Medicare & Medicaid’s (CMS) SEP-1 core measure includes timely antibiotic administration but does little to emphasize the importance of de-escalation and discontinuation of antibiotics, which may result in overutilization of broad spectrum antibiotic therapy.^75-78

Moerhing et al. conducted a multicenter, patient-level, randomized controlled trial from September 2018 to May 2020. Non-intensive care adult patients with suspected sepsis were evaluated by a 5-step protocol: eligibility screen, safety check, randomization, opt-out procedure, and guided de-escalation discussion. Suspected sepsis was defined as having blood cultures that demonstrated no growth 48 to 96 hours after collection for a patient with an active order for broad-spectrum antibiotics. Charts were reviewed to apply a 23-item safety checklist to further identify patients for randomization to receive usual care (no intervention) or the opt-out discussion (intervention). The opt-out protocol provided scripted discussion points with ordering providers to ascertain the appropriateness of antibiotic de-escalation. The primary outcome was post-enrollment days of antibacterial therapy (DOT). Select secondary outcomes included 30-day safety events: CDI, readmission, transfer to intensive care, death, and length of stay. Secondary antibiotic use outcomes included total DOT and DOT by agent rank. Antibiotics were ranked in groupings of 1 through 4 according to the spectrum of activity with ranking 3 (extended spectrum) and 4 (protected) having the broadest spectrum.

Of 9606 patients screened for eligibility, 767 (8%) were randomized. Many patients were excluded for having not met the safety check commonly due to receipt of antibiotics prior to blood culture collection (35%), positive culture (26%), or new oxygen requirement (21%). Baseline characteristics were similar between the two groups. Post-enrollment antibiotic use was lower in the intervention group compared to the control group (301 vs 324 patients, respectively; OR .68, 95% CI .47 to .98, P = .04). In patients with antibiotics continued, there was no difference in DOT (mean 10.4 vs 9.9, 95% CI .88-1.26) or length of therapy (mean 8.3 vs 7.5, 95% CI .95-1.33) between the groups. Fewer patients in the intervention group received rank 3 antibiotics (138 vs 167, OR .73, 95% CI .55-.98); however, no difference was noted between rank 4 antibiotics (13 vs 16, OR .81, 95% CI .38-1.70). The sum of major safety events was numerically lower in the intervention group compared to the control group. No difference was observed for DOOR/response adjusted for duration of antibiotic risk (RADAR) outcomes (52%, 95% CI 48%–56%).

Limitations of this study include the overall feasibility of implementing such a screening tool in real-world practice given the time commitment of study investigators to complete each safety check. Additionally, this tool only identified low risk patients across various practice settings, which only accounted for 8% of those screened and is thus not widely applicable. Lastly, antibiotic selection was not assessed. The authors concluded the opt-out antibiotic stewardship intervention resulted in lower post-enrollment antibiotic use; however, no difference in DOT was seen in those patients who continued antibiotic therapy.

Tamma PD et al. Clinical Impact of Ceftriaxone Resistance in Escherichia coli Bloodstream Infections: A Multicenter Prospective Cohort Study ²⁰

Cases of ceftriaxone-resistant (CRO-R) Escherichia coli BSIs are rising in the United States (US). ⁷⁹ Infections caused by CRO-R organisms are often difficult to treat as there are limited antibiotic options available. ⁸⁰ Several retrospective studies have demonstrated that CRO-R BSIs are associated with significant morbidity and mortality as compared to ceftriaxone-susceptible (CRO-S) E. coli BSIs.^81-85 However, it is unknown if worse clinical outcomes persist after adjusting prospectively for confounders. ⁸⁶

Tamma et al. conducted a prospective, multicenter study across 14 acute care hospitals in the US from November 2020 to April 2021 investigating outcomes of adults with monomicrobial CRO-R and CRO-S E. coli BSIs. Patients with E. coli isolates that were intermediate to ceftriaxone (MIC 2 μg/mL), infection with E. coli non-susceptible to ≥1 carbapenem, isolates unavailable for confirmatory testing, and polymicrobial BSIs were excluded from enrollment. If a patient with a CRO-R E. coli BSI was enrolled, a sequential patient at the same participating site with a CRO-S E. coli BSI was sequentially enrolled. The clinical evaluation was DOOR based on an ordinal scale of disposition at day 30 after collection of the index blood culture in patients with CRO-R E. coli BSI vs CRO-S E. coli BSI.

A total of 300 patients with positive E. coli CRO-R (n = 150) and CRO-S (n = 150) blood cultures were included. Baseline characteristics between the two arms were similar in age and sex. Patients with CRO-R E coli BSI were more likely to be admitted from nursing homes, have higher Pitt bacteremia scores, and greater comorbid conditions, including severe immunocompromise, as compared to the CRO-S E coli group; although, differences were not statistically significant. The most common sources of infection were urinary (62%) and intra-abdominal (10%). Adequate source control was achieved similarly in CRO-R and CRO-S E. coli cohorts (78 vs 79%, respectively). The median time to appropriate therapy was 12 hours (IQR 1-35 hours) in the CRO-R group and 1 hour in the CRO-S group (IQR 0-6 hours, P < .001). CRO-R BSI had a 58% probability (95% CI 52%-63%) for worse clinical outcomes as compared to CRO-S BSI in the unadjusted DOOR analysis. Similar trends were seen in the inverse probability weighting adjusted DOOR analysis, with CRO-R BSI having a 54% probability (95% CI 47%-61%) for worse outcomes as compared to CRO-S. Patients in the CRO-R E coli BSI group were more likely to have prolonged hospital stays, remain in the hospital by day 30, and be newly transferred to long-term care facilities.

Limitations of this study include inability to assess the effect of individual antibiotics used for either infection type, potential for prescribing bias, and being underpowered to detect a mortality difference. Although attempts were made to control for external variables, confounders may be present. In conclusion, this prospective study demonstrated that patients with CRO-R E. coli BSIs are more likely to have worse clinical outcomes than patients with CRO-S E. coli BSIs.

Thaden JT et al. Association of Follow-up Blood Cultures with Mortality in Patients with Gram-Negative Bloodstream Infections: A Systematic Review and Meta-analysis ¹⁹

Routinely foregoing follow-up blood cultures (FUBC) due to GNB has recently become a more common occurrence. Previous literature has evaluated FUBC for GNB and found the practice to be associated with prolonged hospitalizations, increased blood culture contamination rates, and increased duration of antimicrobial therapy with no additional mortality difference vs patients who did not have follow up cultures obtained.^87-89 However, recent literature describes an association with decreased mortality in patients with follow up blood cultures performed for GNB.^90,91 This systematic review and meta-analysis aimed to ultimately answer two questions: Key Question 1 (KQ1) was defined as: “for patients with GNB, is obtaining FUBCs after the initial positive blood culture associated with decreased patient mortality?” and Key question 2 (KQ2) was defined as: “for patients with GNB and FUBCs, are positive FUBCs associated with increased mortality relative to negative FUBCs?”

A total of 3495 studies were initially screened of which 3451 were excluded after title and abstract screening. Out of the 44 remaining studies assessed, 15 studies were included in the final analysis. The studies resulted in a total of 8007 patients for KQ1 and 3243 patients for KQ2. Most of the studies were cohort studies; none were RCTs. There was variation in how FUBCs were defined, with the most common definition being those drawn between 24 hours and 7 days after the index culture. The studies addressing the primary analysis of survival bias took measures to ensure accuracy, such as excluding patients who died ≤72 hours of the index blood culture and matching patients with and without FUBCs. The studies evaluating both KQ1 (n = 5) and KQ2 (n = 4) were deemed to have low risk of bias.

The primary analysis combining the results of 5 studies (n = 4378) addressing KQ1 showed that obtaining FUBCs was associated with decreased mortality (HR .56, 95% CI .45-.71) with little heterogeneity (I ² = 0). Utilizing 2 studies meeting criteria for KQ2, an association between positive FUBCs and increased mortality was found in patients with GNB (OR 2.01, 95% CI 1.07-3.76, P = .03). A Cox proportional regression model demonstrated a similar association with positive FUBCs and mortality (HR 1.56, 95% CI 1.13-2.16, P = .007).

The study is not without limitations, including the absence of randomized clinical trials, making it necessary to rely on primarily observational data. The primary analysis used observational data that attempted to address selection and survival bias through methods such as matching, statistical adjustment, and exclusion criteria, but these methods may still be subject to bias. A meta-analysis of individual patient data was not conducted due to the significant differences in the covariates used to describe comorbidities, illness, and differing FUBCs drawing protocols in the included studies. There is a shortage of studies that examine the relationship between FUBCs and mortality in subgroups of patients, which warrant additional studies to determine which patients with GNB would most benefit from FUBCs.

Venugopalan V et al. Use of Therapeutic Drug Monitoring to Characterize Cefepime‐Related Neurotoxicity ²⁴

Cefepime-related neurotoxicity (CRN) is commonly documented as an adverse effect in critically ill patients. ⁹² Neurotoxicity risk has been shown to correlate with reduced renal function due to a decrease in cefepime clearance as well as increased central nervous system penetration due to blood-brain barrier dysfunction.^92-94 The symptoms usually present as depressed consciousness, encephalopathy, aphasia, myoclonus, and seizures. ⁹⁵ The exact mechanism has not been well elucidated but is theorized to be due to a concentration-dependent inhibition of GABA-A receptors, which, in turn, leads to central excitotoxicity.^95,96 The use of therapeutic drug monitoring (TDM) has been limited to antimicrobials with relatively known narrow therapeutic windows such as vancomycin and aminoglycosides. Recent data with TDM for β-lactams have shown improved drug exposure and clinical outcomes although there is a lack of evidence when identifying efficacy and toxicity thresholds. ⁹⁷ This study aimed to further determine the incidence of CRN and the association between cefepime concentrations and CRN.

This retrospective study was performed at a tertiary, academic center between 2016 and 2018. TDM of β-lactams was established prior to the study period and was performed at the discretion the treating team. ⁹⁸ Adult patients receiving cefepime during their hospital encounter during the study period were included with key exclusion criteria of undetectable cefepime troughs and inappropriately drawn serum concentrations.

A total of 10 513 patients who received cefepime were evaluated for inclusion; however, only a few met criteria (n = 503) and underwent cefepime TDM. CRN was reported in 4.4% of encounters. The findings most consistent with CRN included neurology consultations (n = 17) describing this phenomenon and electroencephalographic (EEG) findings suggestive of generalized background rhythm slowing, generalized periodic discharges, and generalized slowing. For the primary outcome, the investigators found increased cefepime trough concentrations in patients who experienced CRN compared to those who did not (61.8 ± 33.7 vs 30 ± 27.7 mg/L, P = .0002). A logistic regression analysis found renal impairment and elevated cefepime trough concentrations to be significant predictors of neurotoxicity.

A limitation to consider with this study includes making an association of cefepime and neurotoxicity rather than establishing causality with certainty due to the retrospective nature of the study. Additionally, when comparing patients who experienced CRN to those who did not, there was a statistically significant difference in renal function without the expected difference in total daily doses. This may suggest that inappropriate dosing in patients with renal dysfunction was one of the driving factors for cefepime-induced neurotoxicity although appropriateness of dosing was not assessed. Lastly, AUC estimations made in the present study were made from a single cefepime trough level, which may not accurately reflect the dynamic nature of renal function among the critically ill. Additional studies are warranted to evaluate the frequency of cefepime neurotoxicity in the setting of optimized dosing.

Veve MP et al. Prevalence and Predictors of Pseudomonas aeruginosa Among Hospitalized Patients with Diabetic Foot Infections ²²

Guidelines suggest that empiric therapy targeted at P. aeruginosa is generally unnecessary in diabetic foot infections unless the patient presents with certain risk factors. ⁹⁹ Nevertheless, these risk factors remain ill-defined and the frequency at which P. aeruginosa is isolated is not well categorized leading to overuse of broad spectrum antibiotics.^100,101

Veve et al. conducted a multicenter, retrospective cohort study of hospitalized patients across five acute care centers located throughout the US with diabetic foot infections (DFI) from January 2013 to December 2020. Eligible adult patients must have been hospitalized for ≥24 hours with documented signs/symptoms of infection, have had accessible positive DFI cultures, and received antibiotic therapy. Uninfected ulcers or those with bone involvement were excluded. The primary outcome was prevalence of DFI cultures positive for P. aeruginosa. Secondary objectives included evaluating DFI antibiotic prescribing with a focus on anti-P. aeruginosa therapy and describing adverse drug events attributable to antibiotic utilization.

Of the 292 patients included in the study, 27 (9%) had P. aeruginosa isolated from DFI culture. Patients with P. aeruginosa DFI at baseline had a higher median age-adjusted Charlson Comorbidity Index (6 [IQR 5-8] vs 5 [IQR 4-7], P = .04), increased median glycosylated hemoglobin A1c (8.5 [IQR 7.1 – 10.4] vs 7.3% [IQR 6.5 – 8.0], P = .004), were active smokers (37 vs 19%, P = .04), and had an immunocompromising condition (67 vs 30%, P < .001) compared to non-P. aeruginosa DFIs. Patients with P. aeruginosa were also more likely to have failed outpatient DFI antibiotic treatment within the previous 90 days (41 vs 22%, P = .011). There were 568 organisms isolated with 147 (54%) of them being polymicrobial. Gram-positive pathogens were most commonly isolated accounting for 61% of the pathogen distribution followed by Gram-negatives (30%) and anaerobes or fungi (7%). Methicillin-susceptible Staphylococcus aureus (MSSA) (35%), Streptococcus species (32%), methicillin-resistant S aureus (MRSA) (20%), coagulase-negative Staphylococcus species (14%), Enterococcus faecalis (11%), Klebsiella species (10%), and P. aeruginosa (9%) were among the most isolated microorganisms. Despite the low incidence of P. aeruginosa, two-hundred fifty-seven (88%) patients received empiric anti-P. aeruginosa treatment. Multivariate regression analysis found that immunosuppressed status (aOR 4.6, 95% CI 1.3 to 16.7, P = .019) and failed outpatient DFI antibiotics within the prior 90 days (aOR 4.8, 95% CI 1.9 to 11.9, P = .001) were associated with increased odds of having P. aeruginosa DFI. With regards to the safety endpoint, three patients (1%), all in the non-P. aeruginosa group, developed CDI.

A notable limitation of the study was the exclusion of DFI with bone involvement, which may have a higher incidence of P. aeruginosa given the severity and chronicity of infection, though suggests that mild DFIs likely are not caused by P. aeruginosa. Additionally, the study did not report antimicrobial susceptibilities, which may explain the broad use of empiric therapy with anti-P. aeruginosa agents. In conclusion, this study demonstrated that P. aeruginosa isolation from DFI is low, and it is reasonable to follow guideline-based recommendations avoiding empiric anti-P. aeruginosa agents unless patients have significant risk factors.