Citations

Abstract Background Gram stain (GS) is essential for preliminary bacterial identification. The lack of a standardized quality control method for GS is a significant drawback. By utilizing a deep convolutional neural network (CNN) design, we present a proof-of-concept for a novel solution to GS quality assurance. Using this approach, current manual GS, and potentially automated GS methods, may possess a standardized method of quality control. Methods Twenty-nine GS from positive blood cultures encompassing the spectrum of GS quality (acceptable, over/under decolorization, smear thickness, stain artifact) were selected by two expert microbiologists. A total of 197 images were obtained using a Cognex D-900 camera and Olympus BX-43 microscope at 1000X magnification. Images were classified using deep learning software (Cognex, ViDi Suite 4.1). Eighty-one images were classified as suboptimal and 116 were classified as optimal by expert review. To address background variation due to manual GS, we implemented a screening classification tool to separate images into 5 distinct classes (pale, rose, orange, indigo, variegated). The training and test sets consisted of 100 and 97 images respectively. Using a chaining technique, a second classification tool was added after the screening classification tool. The second tool further divided the 197 images into two dataset classes: optimal and suboptimal. The second tool used 105 images for training and 92 images for the test set. Model performance was compared to expert microbiology review, with GS verified by organism identification using MALDI-TOF. Results Cognex ViDi Suite 4.1 could discriminate between optimal and suboptimal GS images. The chained classification tool was able to be trained with 100% recall and 100% accuracy, with a resulting F-score of 1.00. Figure 1 VisionPro GUI with optimal image (MRSA) Conclusion Deep learning enabled image analysis is an effective method for sorting GS which are acceptable for further interpretation and suboptimal smears which require rework. Additional study is warranted to develop deep learning image analysis tools for the interpretation of GS on multiple specimen types and incorporate into automated GS systems. Disclosures Jane Hata, PhD , D(ABMM), Roche Molecular Systems: Grant/Research Support Kyle Magee, MLS(ASCP), SeluxDx: Employee

Abstract Background Cytomegalovirus (CMV) infects millions in the US and is a source of comorbidity among immunocompromised patients. Despite its prevalence, there is limited data on clinical characteristics or outcomes among HIV-negative, non-transplant (NHNT) patients in population-based cohorts. We aim to evaluate a multicenter US-based network to obtain clinical characteristics and outcomes of NHNT patients with CMV disease and explore potential prognostic clinical characteristics implicated in increased morbidity and mortality. Methods This is a US-based multicenter, population-based, retrospective cohort study. We queried TrinetX, a global research network, to identify patients with a CMV disease by ICD-10-CM codes and serum and PCR > 5000 IU/mL. We captured comorbidities diagnosed within 90 days of CMV infection diagnosis, symptoms noted within 30 days, and outcomes of interest at 30 days and 1 year. Primary outcome was all-cause mortality. Secondary outcomes were hospitalization and CMV disease-related complications. We performed a propensity score-matched analysis comparing clinical characteristics among patients who survived versus non-survivors at 90 days. Results We found 1123 patients with evidence of CMV disease. The mean patient age was 53 years, 48% men and 50% white. Aplastic anemia and neoplasms were the most common comorbidities (Table 1). Thromboembolism was a common complication (15%). Half of the patients had dyspnea, fever, abdominal pain, and nausea. CMV disease was associated with 14%, 22%, and 26% mortality at 30 days, 90 days, and one year, respectively. 35% of patients required hospitalization within 30 days. After propensity score matching, we had 278 patients in each cohort. Dyspnea and weakness were linked to increased 90-day mortality (Figure 1). Purpura, encephalopathy, and sepsis were more common in patients who died. High CMV serum viral load and ferritin were associated with increased mortality. Conclusion NHNT patients with CMV disease often have underlying comorbidity, commonly aplastic anemia, and neoplasms. CMV disease was associated with thrombosis and a high risk of hospitalization and mortality. Vascular disease, sepsis, high viral load, and cell lysis markers were prognostic factors for mortality in this cohort. Disclosures Raymund R. Razonable, MD, Allovir: Endpoint Adjudication Committee|American Society of Transplantation: Board Member|Gilead: Grant/Research Support|Novartis: DSMB|Regeneron: Grant/Research Support|Roche: Grant/Research Support

Abstract Background Empiric antibiotic selection in outpatient care often relies on inpatient antibiograms which may not accurately reflect ambulatory susceptibility rates. To address this, Novant Health (NH) and Labcorp analyzed urine culture results from North Carolina (NC) to create an ambulatory antibiogram and assess temporal and regional differences. Methods Cross-sectional study of urine culture data collected 1/1/20-12/31/21 from ambulatory sites in NC. Testing and antibiogram preparation were performed at Labcorp (Burlington, NC) based on CLSI guidelines (M100 S28). Pediatric and adult data were compiled separately. NH and state-wide data were analyzed using the chi-square or Fisher exact test with Bonferroni correction for multiple comparisons. Results In 2021, Escherichia coli was the most common isolate for NH (n=18937 adult, n=624 pediatric) and state-wide (n=87,226 adult; n=2696 pediatric). E. coli was ≥ 80% susceptible to all tested antibiotics except ampicillin; Klebsiella pneumoniae was ≥ 80% susceptible to all except nitrofurantoin; and Proteus mirabilis/penneri was ≥ 80% susceptible to all antibiotics tested. Susceptibility rates were similar between NH and state-wide data except for lower susceptibility to ceftriaxone (94% vs 95%; p< .001). Compared to 2020, state-wide adult E. coli isolates from 2021 had increased susceptibility for amoxicillin/clavulanic acid (86% vs 85%), ampicillin (60% vs 59%), cephalosporins (84-93% vs 82-92%), and quinolones (87% vs 86%) (p< .001 all). NH isolates did not show significant changes. Susceptibility rates by region were similar, except for lower susceptibility to cefuroxime among adult E. coli isolates in the Triad region relative to state-wide (82% vs 84%; p< .001). No significant temporal or regional differences in susceptibility were observed for K. pneumoniae, Proteus spp., P. aeruginosa, and among pediatric samples. Conclusion Susceptibility data were largely similar in NH vs. state-wide data, and between regions. Small but statistically significant temporal changes were found in state-wide data. Overall, our results support use of a state-wide antibiogram to inform treatment choices across all regions and within specific healthcare systems. Disclosures Dusica Curanovic, PhD, Labcorp: employed by labcorp during study|Labcorp: Stocks/Bonds Chris Garcia, MD, Labcorp: Prior employee|Labcorp: Stocks/Bonds Catherine Wright, PhD, Labcorp: Employee|Labcorp: Stocks/Bonds Rita Stainback, MS, MT (ASCP) SM, DLM, Labcorp: Employee|Labcorp: Stocks/Bonds Melissa Martin, BS, Labcorp: Employee Samia Naccache, PhD, D(ABMM), Labcorp: Employee|Labcorp: Stocks/Bonds Kristen Smith, PhD, bioMerieux: Employee|Labcorp: Stocks/Bonds Howard Engler, PhD, D(ABMM), Labcorp: Employee|Labcorp: Stocks/Bonds

Abstract Background The worldwide prevalence of cytomegalovirus (CMV) infection is estimated to be 83-100%.1,2 In pregnancy, transmission to the fetus can occur through the placenta with 0.4-2.3% of all newborns affected.3 The phenotype of congenital CMV (cCMV) has a spectrum of severity; 10-20% of exposed neonates show severe clinical manifestations.4 CMV has a wide range of tropism and can infect epithelial, myeloid, and endothelial cells. However, the cellular mechanisms of cCMV infection, and how they influence disease severity in exposed fetuses during gestation are not fully outlined,5 Here, we used the NanoString™ Digital Spatial Profiler (DSP) to characterize the cell-specific proteomic profile in placental tissue from infants with cCMV. Methods Placental tissue from 5 CMV-affected and 4 CMV-unaffected pregnancies were retrieved from the Mayo Clinic tissue repository. Immune cells were identified by CD45 positivity and cytotrophoblasts (CTBs) identified by CK7. The cell-specific proteome of these tissues was identified using the NanoString™ immune cell and cell death panels. Data was analyzed by NanoString™ GeoMx technology. Results Unsupervised clustering illustrated a distinct proteome in the immune cell compartment of CMV-positive compared to CMV-negative placental tissue (Figure 1). Linear Mixed Model (LMM) analysis revealed that immune cell related proteins were significantly overexpressed in CMV+ tissue as were four proteins associated with cell death: BIM, BCLXL, PARP, neurofibromin (Figure 2). Clustering did not demonstrate a distinct proteome in the CTB compartment (Figure 3), but LMM showed that two immune markers (CD44 and HLA-DR) and a cell death protein (BCL6) were overexpressed by CTBs in CMV+ tissue (Figure 4). The proteome of immune cells, and to a lesser extent in CTBs in CMV-infected placentae differs from CMV-uninfected placentae with overexpression of proteins associated with immune activation and cell death. With new screening initiatives set to improve identification of cCMV, our results provide a starting point to identify specific biomarkers and immune pathways associated with disease severity. These biomarkers and pathways may serve as novel antiviral targets in the treatment and management of cCMV. The proteome of immune cells, and to a lesser extent in CTBs in CMV-infected placentae differs from CMV-uninfected placentae with overexpression of proteins associated with immune activation and cell death. With new screening initiatives set to improve identification of cCMV, our results provide a starting point to identify specific biomarkers and immune pathways associated with disease severity. These biomarkers and pathways may serve as novel antiviral targets in the treatment and management of cCMV. Fig 3 - The proteome of immune cells, and to a lesser extent in CTBs in CMV-infected placentae differs from CMV-uninfected placentae with overexpression of proteins associated with immune activation and cell death. With new screening initiatives set to improve identification of cCMV, our results provide a starting point to identify specific biomarkers and immune pathways associated with disease severity. These biomarkers and pathways may serve as novel antiviral targets in the treatment and management of cCMV. Fig 4 - Volcano plot illustrating the differential proteomic expression in CK7+ tissue compartment of CMV-positive placental tissue compared to CMV-negative placental tissue. Conclusion The proteome of immune cells, and to a lesser extent in CTBs in CMV-infected placentae differs from CMV-uninfected placentae with overexpression of proteins associated with immune activation and cell death. With new screening initiatives set to improve identification of cCMV, our results provide a starting point to identify specific biomarkers and immune pathways associated with disease severity. These biomarkers and pathways may serve as novel antiviral targets in the treatment and management of cCMV. Disclosures Mark R. Schleiss, MD, Moderna Vaccines: Grant Support to my Division but no personal honoraria

Abstract Background Deep convolutional neural networks (DNN) have been successfully applied to the detection and segmentation of objects in gigapixel digital slides. In this study, trained laboratory technologists utilized a DNN-assisted digital workflow (Techcyte, Inc.; Orem, UT) to evaluate clinical stool specimens for the presence and identification of protozoan parasites. Methods Slides were prepared from 142 stool specimens by spreading a thin layer of concentrated stool onto a slide, staining with modified-trichrome, and coverslipping with permanent mounting medium. Slides were then scanned using a NanoZoomer 360 digital slide scanner (Hamamatsu Photonics K.K., Japan) at 40x apparent magnification and prospectively analyzed by the Techcyte DNN. Objects of interest were identified and classified by the DNN (protozoa, leukocytes, erythrocytes) and presented to a laboratory technologist for interpretation. (See example image below.) DNN-assisted results were compared to those from conventional microscopic examination. To determine analytical sensitivity, slides were prepared from serial, 2-fold dilutions of 3 unique stool specimens (6 slides at each dilution/specimen) and analyzed by the DNN-assisted method and conventional microscopy. Example of Techcyte DNN analysis review screen Objects of interest most likely to be significant are presented to the technologist for review in an intuitive web-based format. Parasites are confirmed as present or absent by a trained technologist after reviewing these images. Results Digital analysis was highly reproducible, showing 100% slide level agreement (positive/negative) within and between runs using positive and negative slides. There was 98.7% and 98.3% positive and negative agreement, respectively, between conventional and DNN-assisted results. Of the 142 slides conventional microscopy detected parasites in 82, while the DNN-assisted method detected 83. The DNN-assisted method also detected more slides with leukocytes (29 vs 24) and erythrocytes (12 vs 8). All discrepant results were confirmed by repeat conventional microscopy. Analytical sensitivity of the DNN-assisted method was equivalent or superior to conventional microscopy review, detecting down to 1:64 dilutions in 6/6 slides per specimen (conventional = 1:32 to 1:64), extending for some slides to detection at dilutions of 1:2048. Conclusion The DNN-assisted method facilitates rapid screening of negative specimens, while providing sensitive and accurate detection and classification of protozoa. Disclosures All Authors: No reported disclosures

Abstract Background There is scarce data surrounding the epidemiology of invasive fungal infections (IFI) after pancreas transplantation (PT). Our centers’ PT protocols recommend universal fluconazole prophylaxis for at least 1 month, which is extended to 1 year for recipients who live in Coccidioides endemic areas. The purpose of our study was to evaluate the epidemiology of IFIs after PT while on our antifungal prophylaxis protocol. Methods A retrospective cohort study was performed among adult patients who underwent PT at Mayo Clinic sites in Arizona, Florida, or Minnesota between January 1, 2010 and December 31, 2020. IFI was defined based on EORTC/MSGERC criteria. The primary aim was to describe epidemiology and timing of IFI after PT, particularly in the context of the universal antifungal prophylaxis protocol. Results Among 482 pancreas transplant recipients, 35 developed IFIs (7.3%). All patients received antifungal prophylaxis per protocol for a median duration of 33 days (Figure 1). Median duration from time of transplant to IFI was 16 (IQR 10.5-28.2) days for Candida IFIs versus 529 (IQR 192.5-1195.0) days for non-Candida IFI. 20 IFIs (57.1%) were secondary to Candida species with 45% of these cases caused by either C. glabrata or C. krusei (Figure 2). 90% were on antifungal prophylaxis (85% fluconazole) at time of diagnosis of invasive candidiasis. All the Candida IFIs were intra-abdominal infections. Amongst the 15 non-Candida IFIs (42.9%), 9 were secondary to endemic mycoses and 4 were secondary to invasive molds. Conclusion Our study shows a high incidence of IFIs secondary to Candida, with few patients diagnosed with non-Candida IFIs (Figure 3). Non-Candida IFIs, mostly secondary to endemic mycoses, occurred more than a year after transplant. Patients are at highest risk for Candida IFIs within the first 30 days after PT, likely secondary to organ space surgical site infections, despite receipt of antifungal prophylaxis. Breakthrough invasive candidiasis occurred in 90% of cases. Though our study lacked a comparator group without prophylaxis, these data question the effectiveness of fluconazole prophylaxis after PT. Disclosures All Authors: No reported disclosures

Abstract Background HIV prevention continues to be a care gap in developing countries, like Dominican Republic. The lack of sufficient programs for prescription and information of general populations are barriers for better control of HIV infection in this country. The knowledge of HIV prevention strategies such as Pre-Exposure Prophylaxis (PrEP) or Post-Exposure Prophylaxis (PEP) among medical professionals and medical students is unknown. The goal of ending HIV Pandemic by 2030 is to rely on prevention strategies. Our study aims to determine the level of knowledge of medical students in a pre-preclinical cycle and in a clinical cycle regarding HIV/AIDS, PrEP and PEP Methods We survey medical students in the Dominican Republic to measure knowledge on HIV, PrEP/PEP. We compared mean knowledge for each topic based on a score out of 10. We analyzed mean scores by using T-Tests and ANOVA with Post Hoc or Tukey Analysis, assuming significance when p< 0.05. Results A total of 503 students participated in the survey. ANOVA results showed that there was a statistically significant difference in mean knowledge scores for HIV, PEP, and PrEP (F (2, 1506) = [196.6], p < 0.001). Post hoc analysis revealed that medical students had a higher mean score for HIV knowledge (M= 8.9) than mean scores for PrEP (M= 7.6, p< 0.01) and PEP (M= 7.1, p< 0.01). Similarly, the difference in knowledge scores for PrEP was significantly higher than for PEP (p< 0.01). We further divided participants into Preclinical Years and Clinical years and found that participants within the Clinical Years of Med School had a significantly increased mean PEP knowledge (m= 7.5) than Preclinical Years participants (m= 6.7) (t (460) = 5.8, p< 0.001). Similar results were not appreciated with PrEP, where Clinical year participants had a non-significant increased mean knowledge (m=7.8) compared to their counterparts (m= 7.5) (t (486) = 1.4, p= 0.79). Conclusion Our findings might suggest that medical students acquire more knowledge about PEP as they progress into clinical training, but that PrEP knowledge may be emphasized equally in both pre-clinical and clinical training. Further studies should focus on identifying points for improvement and implementing strategies to tackle barriers. One of several limitations of this study is a response bias. Disclosures All Authors: No reported disclosures

Abstract Background The BioFire® FilmArray® Gastrointestinal Pathogen Panel (GIPP) tests for 22 potential pathogens in stool, including Clostridioides difficile. C. difficile nucleic acid is detected via nested polymerase chain reaction (PCR) targeting toxin A (TcdA) and/or toxin B (TcdB) genes. While PCR-based diagnostic methods are highly sensitive, they lack specificity to differentiate colonization from C. difficile infection (CDI). Mayo Clinic Arizona (MCA) uses enzyme immunoassay (EIA) testing for toxin A/B and glutamate dehydrogenase antigens with reflex to PCR for discordance as the primary algorithm for diagnosing CDI; however, GIPP PCR testing circumvents this algorithm. When used in patients with low pretest probability for CDI, clinical false positives can lead to inappropriate diagnosis and treatment of colonization. GIPA Results in the Electronic Medical Record Pre-Intervention MCA Standalone C. difficile Testing Algorithm Abbreviations – EIA, enzyme immunoassay; GDH, glutamate dehydrogenase; NAAT, nucleic acid amplification test Methods We convened a multidisciplinary quality improvement team to reduce the detection and treatment of C. difficile colonization secondary to GIPP test results. Surveys were sent to providers and nurses as part of a formal gap analysis and included questions around ordering patterns, knowledge of appropriate tests, and associated costs. Baseline data was compiled from GIPP orders in MCA from 1/1/2022-9/30/2022. Results There were 1978 GIPP orders for 1694 patients during the study period (∼7 GIPPs/day). Most orders (65%) originated from ambulatory settings. The most common positive target was C. difficile (355/723, 49%), which resulted in treatment for suspected CDI in 94.9% (337/355) of cases. Of the 49 providers who responded to the survey, 86% ordered the GIPP at least monthly. Testing for C. difficile was the 2nd most common reason for ordering the GIPP. Most providers underestimated the cost of the panel and were unfamiliar with the differences between the GIPP and EIA testing algorithm. Nurses sought clarification on stool samples suitable for GIPP testing. Following education and modifications to ordering in the electronic health record, the decision was made to remove the C. difficile target from the GIPP. Provider Understanding of C. difficile Testing Methods GIPA Results in the Electronic Medical Record Post-Intervention Conclusion Multiplex PCR panels are comprehensive but require careful implementation to facilitate diagnostic stewardship. Facilities with robust standalone C. difficile testing algorithms may consider removing the C. difficile target from the GIPP. Disclosures All Authors: No reported disclosures

Abstract Background The treatment of invasive fungal infections (IFI) is often complicated by drug-drug interactions, toxicities, and host comorbidities. Isavuconazole (ISA) has a favorable side effect profile and has FDA approval for invasive aspergillosis and mucormycosis, but there are fewer data on its efficacy for other microbes and non-pulmonary infections. We report our experience of ISA in IFI treatment, with a particular focus on non-pulmonary sites and other microbial etiologies. Methods We identified patients who received ISA as part of IFI treatment from 1/1/2015–4/1/2020 at Mayo Clinic. We categorized them by organ involvement and by pathogens, and identified patterns of ISA use and clinical outcomes. Results A total of 131 patients were identified. Similar to the existing literature, the majority were pulmonary infections (60%), with 12% invasive sinus infection, 12% central nervous system (CNS) infection, and 7% fungemia. Aspergillus was the most common pathogen (39%), yeast comprised 19%, dimorphic fungi 15%, Mucormycosis 12%, phaeohyphomycosis 5%, and other molds 3%. ISA was used as secondary/salvage therapy in 89% of cases. The response rate varied by organ involvement and pathogen. ISA had complete/partial resolution or achieved stability in 57% of fungemia cases, 63% of invasive sinus infections, 64% of pulmonary infections, and 73% of CNS infections. Aspergillus species had a 67% response rate, with 72% response in yeast infections, 65% rate in dimorphic fungi, 53% for mucormycosis, and 86% in dematiaceous molds. Conclusion Although limited by a small sample size, the subgroup review based on organ site and microbial etiology did seem to show roughly similar results to approved uses. The CNS infections treated with ISA were particularly of interest, given the limited options available with reliable CNS penetration. The current available literature has little to say on the efficacy of ISA in these off-label uses, with the majority being similar case series. This report highlights the importance of prospective trials comparing ISA to other antifungals in a greater variety of indications to better identify when this better-tolerated agent can be safely used. Disclosures Courtney E. Harris, MD, Dynamed: Advisor/Consultant

Abstract Background Antimicrobial resistance (AMR) is a threat to healthcare, and infections with antimicrobial resistant organisms carry significant morbidity and mortality. Owing to high rates of AMR, the Centers for Disease Control and Prevention listed Acinetobacter baumannii as a top priority pathogen for research and development of new antimicrobial agents. Antimicrobial susceptibility testing is challenging for this organism, and susceptibility profiles are variable, leading to need of a potential guidance document with assistance in selecting effective empiric therapeutic regimens. Methods We retrospectively reviewed the microbiology and antimicrobial susceptibility testing (AST) results for Acinetobacter spp. from blood culture isolates submitted to Mayo Clinic Laboratories from December 2012 to 2022. AST was performed using agar dilution and interpretations following Clinical & Laboratory Standards Institute (CLSI) guidelines. Isolates were identified by MALDI mass spectrometry and/or 16S sequencing, as available (Figure 1).Figure 1:Isolate identification Isolate identification description. Isolates were analyzed based on this classification used in out laboratory. Results 804 blood culture isolates were included. A. baumannii/calcoaceticus complex (27%) was most frequently isolated, followed by A. lwoffii (22%), and showed the highest resistance across all antimicrobials (73% susceptible or below) (Figure 2). A. baumannii/calcoaceticus complex showed rising resistance rates over time from 2012 to 2022 for meropenem, minocycline and ampicillin-sulbactam (Figure 3). Cefiderocol non-susceptible rate was 35% for A. baumannii/calcoaceticus complex.Figure 2:Antimicrobial susceptibility profiles for Acinetobacter spp. Description of susceptibility profiles for Acinetobacter spp. isolates, based on the classification used by our laboratory. Antimicrobials are presented at the top and isolates on the left. Number of isolates analyzed are presented in the table with the percentage of the isolates susceptible to each antibiotic class in parenthesis.Figure 3:Acinetobacter baumannii/calcoaceticus complex isolates and resistance, trend Trend of isolates per year and resistance percentage to meropenem, colistin, minocycline and amp-sulb. Number of isolates per year at the base of each bar; percentage of resistance for antimicrobials expressed on lines. Amp-sulb = ampicillin-sulbactam. Conclusion A. baumannii/calcoaceticus complex showed high levels of resistance ( >75%) across multiple antimicrobials and increasing rates of resistance over time from 2012 to 2022. A high number of cefiderocol non-susceptible organisms is also concerning (Figure 2). Disclosures Audrey N. Schuetz, MD, Merck: Advisor/Consultant