ORAL PRESENTATION 963, THURSDAY OCTOBER 12, 2023. EXTERNAL VALIDATION OF THE 2023 DUKE - INTERNATIONAL SOCIETY FOR CARDIOVASCULAR INFECTIOUS DISEASES (ISCVID) DIAGNOSTIC CRITERIA FOR INFECTIVE ENDOCARDITIS (IE). PRESENTED BY T VAN DER VAART, MD Summary

The Duke Criteria for diagnosis of infective endocarditis (IE) were published in 1994 and revised in 2000. In August of 2023, an international expert panel (International Society for Cardiovascular Infectious Diseases [ISCVID]) released a revised set of 2023 Duke criteria.1 These criteria reflected newer diagnostic techniques such as cardiac positron emission tomography (PET) computed tomography (CT), new microbiologic data, and surgical assessment as major criteria.

The present authors performed an external validation of the 2023 Duke-ISCVID criteria through retrospective analysis of a prospective database of patients at Amsterdam University Medical Center in the Netherlands, referred to the ID service for presumed IE. They compared the sensitivity and specificity of the 2023 Duke-ISCVID criteria to the 2000 Duke criteria and to the 2015 European Society of Cardiology (ESC) criteria.2 They noted no “gold standard” for IE diagnosis, instead referenced an adjudication panel of 12 experts on IE to determine whether each patient's diagnosis was consistent with IE. Each case was adjudicated by 2 experts who had to be in agreement (reviewed by a third expert if there was disagreement) to determine whether cases were consistent with IE. The authors took “definite endocarditis” expert opinion as positive, while combining mixed and negative opinions as negative. The study controlled for interrater variability by Cohen's κ.

The group studied 595 patients with suspected IE over a 5-year period. Of these, 8 (1%) had intravenous drug use, 152 (26%) had native valve disease, 217 (36% had prosthetic heart valves), and 103 (17%) had cardiac implantable electronic devices. For imaging, 95% of patients had a transthoracic echocardiogram, 75% had a transesophageal echocardiogram, 56% had a cardiac PET and computed tomography, and 23% had cardiac CT. Three hundred ninety-nine were deemed to have IE by adjudication. The Cohen's κ was 0.72 (88% agreement between adjudicators). The diagnostic accuracy of the various criteria was compared (all at 95% CI), with 2000 modified Duke Criteria having a 74.9% sensitivity (P value <0.001 vs Duke ISCVID), 94.9% specificity (P value 0.16 vs Duke ISCVID). The 2015 ESC criteria had 80.0% sensitivity (P value <0.001 vs Duke ISCVID), 93.9% specificity (P value 1 vs Duke ISCVID). The 2023 Duke ISCVID criteria had the greatest sensitivity, 84.2%, and a specificity of 93.9%. The sensitivity improvement between the 2000 Modified Duke criteria and 2023 Duke ISCVID criteria was most significantly due to novel imaging modalities being added as a major criterion. The study was strengthened by use of expert adjudicator panel, frequent use of PET/CT, and population, which matched clinical practice. It was limited by being a single-center study and low intravenous drug use prevalence. The conclusion was that the Duke ISCVID criteria had better sensitivity and equal specificity when compared to the 2000 Modified Duke criteria and the 2015 ESC criteria.

Comments by Dr. Pomakov: This external validation study of the 2023 Duke ISCVID criteria for IE is an essential exercise in assessing the clinical utility of the novel diagnostic criteria. As with all new guidelines and recommendations, a certain amount of time will pass before the clinical impact is truly felt and adopted across the board by specialists used to the prior recommendations. The improvement in sensitivity of the new criteria over the 2000 Modified Duke criteria in this study appears to be mostly driven by the utilization of cardiac CT as an adjuvant or surrogate for echocardiogram. While this may incur additional healthcare costs, it can conceivably find utility in the clinical cases where patient either cannot undergo transesophageal echocardiogram or have an anatomical preclusion to obtaining a diagnostic quality transthoracic echocardiogram. While this external validation may be well extrapolated in the Netherlands, it would be interesting to perform a cross center validation where the general population has more correlation with the health inequities frequently encountered in the US healthcare system.

Fowler VG, Durack DT, Selton-Suty C, Athan E, et al. The 2023 Duke-International Society for Cardiovascular Infectious Diseases Criteria for Infective Endocarditis: Updating the Modified Duke Criteria. Clin Infect Dis 2023;77(4):518126 Habib G, Lancellotti P, Antunes MJ, Grazia Bongiorni M, et al. 2015 ESC Guidelines for the management of infective endocarditis: The Task Force for the Management of Infective Endocarditis of the European Society of Cardiology. Europ Hear Journ 21 Nov 2015; 36(44): 3075–3128 ORAL PRESENTATION. SESSION 1952, OCTOBER 13, 2023. SECOND TIME'S THE CHARM? LOW YIELD OF 3 AFB SMEARS AMONG INPATIENTS EVALUATED FOR PULMONARY TUBERCULOSIS. PRESENTED BY C. DUGDALE, MD Summary

Investigators conducted a retrospective study at Massachusetts General Hospital to estimate the sensitivity of 1, 2, or 3 acid fast bacilli smears plus or minus nucleic acid amplification test in a low prevalence setting. The Cepheid Xpert MTB/RIF assay employed. Data was analyzed from 7445 inpatients hospitalized between July 2016 and December 2022, of which 52 patients (0.7%) had a positive mycobacterial culture for Mycobacterium tuberculosis on one or more respiratory specimens. The first acid fast bacteria (AFB) smear was positive in 22/52 cases (42%), and a second AFB smear was positive in 5/24 cases (21%). There were no cases in which a third AFB smear was positive after 2 initial negative smears. There were 38 patients (73%) with culture-positive tuberculosis (TB) detected by either AFB smear or Xpert. There were 1523 instances in which a third specimen was obtained in a patient with or without tuberculosis, and there was 1 case in which that sample yielded a diagnosis of tuberculosis by culture. The sensitivity to detect a positive AFB smear or Xpert among patients with tuberculosis was equivalent at 52% between 2 and 3 AFB smears. Finally, the sensitivity to detect tuberculosis prior to a positive culture was highest at 67% with the addition of an Xpert to 2 or 3 AFB smears.

Comments by Dr. Pillai: The prevalence of M. tuberculosis in the US, along with ongoing concern for healthcare-associated transmission necessitates prompt identification of patients with suspected tuberculosis and placement in airborne isolation rooms. In addition, timely de-escalation of precautions when tuberculosis workup is negative is also critical to preserve hospital capacity and reduce length of stay. Current Centers for Disease Control and Prevention guidelines recommend collection of 3 respiratory specimens 8–24 hours apart for AFB smear and mycobacterial culture along with at least one nucleic acid amplification test. This study showed that in a low incidence environment, tuberculosis screening with 2 AFB smears is equivalent in diagnostic yield to 3 AFB smears. The addition of Xpert to AFB smear-based screening increased the sensitivity for detection of tuberculosis. While this study is promising to potentially reduce duration of patient isolation and burden on microbiology labs, it is important to note that a quarter of patients with tuberculosis diagnosed by mycobacterial culture were not detected by AFB smear or polymerase chain reaction. Therefore, it is critical to factor clinical suspicion and awareness of tuberculosis prevalence and screening practices into the decision to discontinue isolation.

ORAL PRESENTATION SESSION 31. OCTOBER 12, 2023. EFFECT OF ORAL FOSFOMYCIN COMPARED TO THE INTRAVENOUS BETA-LACTAM/BETA-LACTAMASE INHIBITOR OR CARBAPENEM IN STEP-DOWN TREATMENT OF PATIENTS WITH COMPLICATED URINARY TRACT INFECTION DUE TO ESBL-PRODUCING ENTEROBACTERIACEAE: A MULTICENTER, RANDOMIZED, PROSPECTIVE, OPEN-LABEL, NON-INFERIORITY TRIAL. PRESENTED BY J. SEO, MD, PHD. Summary

Investigators enrolled 199 participants at 4 tertiary medical hospitals nationwide in Korea with a complicated urinary tract and a confirmed fosfomycin sensitive extended spectrum beta lactamase (ESBL) producing enterobacteriaceae isolate to receive either oral fosfomycin or intravenous (IV) therapy with a carbapenem or beta-lactam/beta-lactamase inhibitor after identification of the isolate. A complicated urinary tract infection (UTI) was defined as objective or subjective fever with urinary tract symptoms, pyuria or bacteriuria, and an additional complicating factor. All patients received an empiric IV carbapenem or beta-lactam/beta-lactamase inhibitor prior to isolate identification and received 10 days total of therapy. The primary endpoint was clinical resolution of urinary tract infection related symptoms within 4 days after end of treatment and fosfomycin found to be noninferior to IV therapy with 4 failures in the fosfomycin group and 3 in the comparison. Secondary end points included microbiologic cure,urinary tract infection related re-admission, Clostridioides difficile colitis, intensive care unit admission, and mortality. There were no significant differences in the secondary endpoints.

Comments by Dr. Stephen: With rising rates of ESBL infections, physicians are frequently faced with the task of treating ESBL infections. As many of these patients improve before a complete course of IV therapy is administered, there is a push to finish treatment with oral options for stewardship, side effects, and hospital space reasons. While fosfomycin is well studied for simple UTIs, prospective data for complicated UTIs is limited. This study will hopefully add substantial weight toward more standardized use for this indication. The long-term fight against resistant organisms is only helped by having a wider variety of drugs available for treatment. Having another generally well-tolerated oral step-down therapy with data supporting its use is another important tool for treating these resistant bacteria.

SESSION 77, OCTOBER 12, 2023. “GOING BACKWARD? A CRITICAL EXAMINATION AT PET AND PROPHYLAXIS FOR PREVENTION OF CMV DISEASE IN SOT RECIPIENTS” PRESENTED BY A. LIMAYE MD. Summary

Among transplanted patients, cytomegalovirus (CMV) seropositive donor/seronegative recipient (D+/R−) status has been associated with worse outcomes, independent of CMV disease. Prophylactic therapy for CMV has been an acceptable standard approach in this patient population. However, with the increasing proportion of CMV D+/R− patients, the option of “test and treat” is currently being explored as a way to reduce drug exposure, costs, risk of drug toxicity, and resistance. This approach has been less well studied and has no standardized implementation approach.

Comparison of Antiviral Prevention Strategies in Liver Transplant was a randomized controlled trial aimed at comparing preemptive treatment (PET) versus prophylactic treatment approaches in high-risk, liver transplant patients. The PET approach study arm received weekly monitoring for “DNAemia,” and any degree of detection by polymerase chain reaction was treated as viremia. After 12 months, the study found no significant differences in mortality, graft failure, acute rejection, or incidence of opportunistic infection. In the PET approach arm, use of antiviral therapy was reduced by 40%. Among high-risk liver transplant patients, the use of PET therapy instead of prophylactic therapy was also associated with a survival benefit in the post-hoc analysis.

Comments by Dr. Russo: Our approach to managing solid organ transplant patients with high-risk CMV features remains divided. The Comparison of Antiviral Prevention Strategies in Liver Transplant study showed 1 potential implementation strategy for PET, which demonstrated favorable outcomes. In doing so, it also generates questions regarding alternate approaches to implementation, such as biweekly rather than weekly monitoring. If outcomes were still comparable, this monitoring approach might make implementation more feasible. Additionally, this study focuses on CMV D+/R− status patients but does not assess the PET model in moderate-risk serostatus groups. Additional studies incorporating moderate-risk serostatus groups to assess PET versus standardized prophylaxis would broaden applicability and potentially further validate the above results. Lastly, although the rates of indirect adverse effects associated with high-risk CMV serostatus (such as graft failure, acute rejection, and incidence of opportunistic infection) remained unchanged with PET, the mechanism behind these indirect effects remains unclear. One proposed hypothesis is that the indirect effects are related to multifunctional CD8 T-cells. However, validation studies are needed to further assess this proposed mechanism.

SESSION 106. OCTOBER 13, 2023. LINEZOLID DOSING: CAN LESS BE MORE? PRESENTED BY P COJUTTI, MD, PHD, Y TSUJI, PHD. MODERATORS: JA JUSTO, PHARMD, Y DOI, MD, PHD

Linezolid is a very useful antibiotic, but a high incidence of thromboctyopenia often limits its duration of use. Although linezolid is often prescribed as a fixed dose of 600 mg by mouth twice daily, there is a growing body of data from Europe and Japan that custom dosing based on renal function, drug-drug interactions, and actual linezolid levels may achieve good outcomes while lessening side effects. Researchers in Europe and Japan who have access to plasma linezolid levels have shown that there is great variability in linezolid concentrations seen with fixed dosing. Supratherapeutic levels of linezolid were seen more frequently in patients with creatinine clearance <40 ml/min. Thrombocytopenia occurred more frequently in patients with creatinine clearance<60 ml/min.1,2 Bin et al recommend therapeutic drug monitoring (TDM) for linezolid in critically ill patients, children, renal insufficiency, cirrhosis, advanced age, obesity, and in patients taking medications known to interact with linezolid, such as proton pump inhibitors and rifampin. Their target trough concentration is 2–8 mg/L.3

Cojutti et al looked at a proactive TDM program for linezolid. In their study, 61 patients received linezolid. Of 29 patients who had transiently elevated levels of linezolid, whose doses were adjusted to the recommended range, only 3 developed thrombocytopenia. Of 4 patients who had persistently high levels of linezolid despite dose adjustments, 3 developed significant thrombocytopenia. Median levels of linezolid did indeed correlate with thrombocytopenia.4

Tsuji et al published data that linezolid induced thrombocytopenia is highly likely when the platelet count decreases by at least 2.3% within the first 96 hours; in the absence of such a drop, thrombocytopenia is likely to happen when the linezolid concentration is greater than or equal to >13.5 mg/L, while some authors felt a lower level was still associated with thrombocytopenia.5

Comments by Dr. Louie: There is now ample evidence that linezolid dosing should be reduced in selected patients, particularly patients with renal insufficiency. Therapeutic drug monitoring reduces toxicity and more often allows for treatment >28 days. However, linezolid levels are not widely available in the United States. We should advocate for use of TDM in selected patients. In the absence of TDM, we will have to consider empiric dose reductions in patients with risk factors for accumulating linezolid, and perhaps in those patients who have a diminishing platelet count.

REFERENCES 1. Pea F, Cojutti P, Baraldo M. A 10 year experience of therapeutic drug monitoring of linezolid in a hospital-wide population of patients receiving conventional dosing: is there enough evidence for suggesting TDM in the majority of patients? Basic Clin Pharmacol Toxicol. 2017;121(4):303–308. 2. Abdul-Aziz M, Alffenaar JC, Bassetti M, et al. Antimicrobial therapeutic drug monitoring in critically ill adult patients: a position paper. Intensive Care Med. 2020;46(6):1127–1153. 3. Bin L, Hu Y, Xu P, et al. Expert consensus statement on therapeutic drug monitoring and individualization of linezolid. Front Public Health. 2022;10:967311. 4. Cojutti PG, Merelli M, Bassetti M, et al. Proactive therapeutic drug monitoring may be helpful in managing long-term treatment with linezolid safety: findings from a monocentric, prospective, open-label, interventional study. J Antimicrob Chemother. 2019;74(12):3588–3595. 5. Tsuji Y, Holford N, Kasai H, et al. Population pharmacokinetics and pharmacodynamics of linezolid-induced thrombocytopenia in hospitalized patients. Br J Clin Pharmacol. 2017;83:1758–1772.