Background: Hyaluronan receptor LYVE-1 is expressed by liver sinusoidal endothelial cells (LSEC), lymphatic endothelial cells and specialized macrophages. Besides binding to hyaluronan, LYVE-1 can mediate adhesion of leukocytes and cancer cells to endothelial cells. Here, we assessed the impact of LYVE-1 on physiological liver functions and metastasis. Methods: Mice with deficiency of Lyve-1 (Lyve-1-KO) were analyzed using histology, immunofluorescence, microarray analysis, plasma proteomics and flow cytometry. Liver metastasis was studied by intrasplenic/intravenous injection of melanoma (B16F10 luc2, WT31) or colorectal carcinoma (MC38). Results: Hepatic architecture, liver size, endothelial differentiation and angiocrine functions were unaltered in Lyve-1-KO. Hyaluronan plasma levels were significantly increased in Lyve-1-KO. Besides, plasma proteomics revealed increased carbonic anhydrase-2 and decreased FXIIIA. Furthermore, gene expression analysis of LSEC indicated regulation of immunological pathways. Therefore, liver metastasis of highly and weakly immunogenic tumors, i.e. melanoma and colorectal carcinoma (CRC), was analyzed. Hepatic metastasis of B16F10 luc2 and WT31 melanoma cells, but not MC38 CRC cells, was significantly reduced in Lyve-1-KO mice. In vivo retention assays with B16F10 luc2 cells were unaltered between Lyve-1-KO and control mice. However, in tumor-free Lyve-1-KO livers numbers of hepatic CD4⁺, CD8⁺ and regulatory T cells were increased. In addition, iron deposition was found in F4/80⁺ liver macrophages known to exert pro-inflammatory effects. Conclusion: Lyve-1 deficiency controlled hepatic metastasis in a tumor cell-specific manner leading to reduced growth of hepatic metastases of melanoma, but not CRC. Anti-tumorigenic effects are likely due to enhancement of the premetastatic hepatic immune microenvironment influencing early liver metastasis formation.

In patients with cardiogenic shock (CS) due to myocardial infarction, elevated lactate levels are known to be negative predictors. Studies regarding the prognostic impact in patients with CS complicated by out-of-hospital cardiac arrest (OHCA) are limited. Two hundred and sixty-three consecutive patients with CS were included. The prognostic value of lactate on days 1, 2, 3, 4 and 8 was tested stratified by OHCA and non-OHCA. Statistical analyses included the univariable t-test, Spearman’s correlation, C-statistics, Kaplan–Meier analyses, as well as multivariable mixed analysis of variance (ANOVA) and Cox proportional regression analyses. The primary endpoint of all-cause mortality occurred in 49.4% of the non-OHCA group and in 63.4% of the OHCA group. Multivariable regression models showed an association of lactate values with 30-day all-cause mortality in the non-OHCA (p = 0.024) and OHCA groups (p = 0.001). In Kaplan–Meier analyses, patients with lactate levels ≥ 4 mmol/L (log-rank p = 0.001) showed the highest risk for 30-day all-cause mortality in the non-OHCA as well as in the OHCA group. However, in C-statistics lactate on days 1 and 8 had a better discrimination for 30-day all-cause mortality in the OHCA group compared to the non-OHCA group. In conclusion, patients presenting with CS lactate levels showed a good prognostic performance, with and without OHCA. Especially, lactate levels on days 1 and 8 were more accurate in the discrimination for all-cause mortality in CS-patients with OHCA.

This study examines the prognostic impact of C-reactive protein (CRP) and white blood cell (WBC) counts in patients with cardiogenic shock (CS). Data regarding the prognostic impact of inflammatory biomarkers in CS are scarce. All consecutive patients with CS from 2019 to 2021 admitted to a cardiac intensive care unit (ICU) were included at one institution. Laboratory measurements were retrieved from the day of admission (i.e., day 1), as well as days 2, 3, 4, and 8. The primary endpoint was 30-day all-cause mortality. Statistical analyses included univariate t-tests, Spearman’s correlations, C-statistics, Kaplan–Meier, and Cox regression analyses. From a total of 240 consecutive patients admitted with CS, 55% died within 30 days. CRP levels on days 3 to 8 were associated with reliable discrimination for 30-day all-cause mortality (area under the curve (AUC): 0.623–0.754), whereas CRP on day 1 was not (AUC = 0.514). In line, CRP > 100 mg/L on day 3 (56% vs. 37%; log-rank p = 0.023; HR = 1.702; 95% CI 1.060–2.735; p = 0.028) and especially a CRP increase of at least 200% from days 1 to day 3 (51% vs. 35%; log-rank p = 0.040; HR = 1.720; 95% CI 1.006–2.943; p = 0.048) were associated with an increased risk of all-cause mortality. Furthermore, WBC on day 1 discriminated 30-day all-cause mortality (AUC = 0.605; p = 0.005) with an increased risk of all-cause mortality in patients admitted with WBC > 10 × 10⁶/mL (59% vs. 40%; log-rank p = 0.036; HR = 1.643; 95% CI 1.010–2.671; p = 0.045). In conclusion, WBC count on admission as well as CRP levels during the course of ICU treatment were associated with 30-day all-cause mortality. Specifically, an increase of CRP levels by at least 200% from day 1 to day 3 during the course of ICU treatment was associated with an increased risk of 30-day all-cause mortality. The present study is one of the first to describe the prognostic value of inflammatory biomarkers in consecutive all-comer CS patients treated at a cardiac ICU.

Data regarding the prognostic value of cardiac biomarkers in patients suffering from sepsis or septic shock is scarce. Studies investigating the prognostic role of cardiac biomarkers in patients with sepsis and septic shock were commonly published prior to the sepsis-3 criteria and were often not restricted to septic patients only, too. This study investigated the diagnostic and prognostic value of the aminoterminal pro-B-type Natriuretic Peptide (NT-pro BNP) and cardiac troponin I (cTNI) in patients with sepsis and septic shock. Consecutive patients with sepsis and septic shock were included from 2019 to 2021. Blood samples were retrieved from the day of disease onset (i.e., day 1), day 2 and 3. Firstly, the diagnostic value of the NT-pro BNP and cTNI to diagnose sepsis or septic shock was tested. Secondly, the prognostic value of the NT-pro BNP and cTNI was examined with regard to the 30-day all-cause mortality. The statistical analyses included univariable t-tests, Spearman’s correlations, C-statistics, Kaplan–Meier analyses and Cox proportional regression analyses. A total of 162 patients were included prospectively, of which 57% had a sepsis and 43% a septic shock. The overall rate of all-cause mortality at 30 days was 53%. With an area under the curve (AUC) of 0.658 on day 1 and 0.885 on day 3, cTNI expressed a better diagnostic value than NT-pro BNP, especially on day 3 (ΔAUCd3 = 0.404; p = 0.022). Furthermore, cTNI displayed a moderate but slightly better prognostic value than NT-pro BNP on all examined days (AUC for cTNI, d1 = 0.635; 95% CI 0.541–0.729; p = 0.007 vs. AUC for NT-pro BNP, d1 = 0.582; 95% CI 0.477–0.687; p = 0.132). In conclusion, cTNI was a reliable diagnostic parameter for the diagnosis of sepsis and septic shock, as well as a reliable prognostic tool with regard to 30-day all-cause mortality in patients suffering from sepsis and septic shock.

Background: In recent years, remarkable progress has been made in deep learning technology and successful use cases have been introduced in the medical domain. However, not many studies have considered high-performance computing to fully appreciate the capability of deep learning technology. Objective: This paper aims to design a solution to accelerate an automated Gram stain image interpretation by means of a deep learning framework without additional hardware resources. Methods: We will apply and evaluate 3 methodologies, namely fine-tuning, an integer arithmetic–only framework, and hyperparameter tuning. Results: The choice of pretrained models and the ideal setting for layer tuning and hyperparameter tuning will be determined. These results will provide an empirical yet reproducible guideline for those who consider a rapid deep learning solution for Gram stain image interpretation. The results are planned to be announced in the first quarter of 2021. Conclusions: Making a balanced decision between modeling performance and computational performance is the key for a successful deep learning solution. Otherwise, highly accurate but slow deep learning solutions can add value to routine care. International Registered Report Identifier (IRRID): DERR1-10.2196/16843

Numerous immunoassays for detecting antibodies directed against SARS-CoV-2 have been rapidly developed and released. Validations of these have been performed with a limited number of samples. The lack of standardisation might lead to significantly different results. This study compared ten automated assays from six vendors in terms of sensitivity, specificity and reproducibility. This study compared ten fully automated immunoassays from the following vendors: Diasorin, Epitope Diagnostics, Euroimmun, Roche, YHLO, and Snibe. The retrospective part of the study included patients with a laboratory-confirmed COVID-19 infection, and controls comprised patients with a suspected infection, in whom the disease was excluded. Furthermore, biobanked sera were taken as negative controls (n = 97). The retrospective part involved four groups: (1) laboratory-confirmed COVID-19 infection (n = 183); (1B) suspected COVID-19 infection (n = 167) without a qRT-PCR result but positive serological results from at least two different assays, and suspected COVID-19 infection due to a positive serological result from the Roche assay (n = 295); (2) biobanked sera obtained from patients before the emergence of SARS-CoV-2 (n = 97) as negative controls; and (2A) probably COVID-19-negative sera with negative serological results from at least two different assays (n = 152). Overall diagnostic sensitivities were: Euroimmun (IgA) 87%; Epitope Diagnostics (IgG) 83%; YHLO (IgG) 77%; Roche (IgM/IgG) 77%; Euroimmun (IgG) 75%; Diasorin (IgG) 53%; Epitope Diagnostics (IgM) 52%; Snibe (IgG) 47%; YHLO (IgM) 35%; and Snibe (IgM) 26%. Diagnostic specificities were: YHLO (IgG) 100%; Roche, 100%; Snibe (IgM/IgG) 100%; Diasorin (IgG) 97%; Euroimmun (IgG) 94%; YHLO (IgM) 94%; Euroimmun (IgA) 83%. Assays from different vendors substantially varied in terms of their performance. These findings might facilitate selection of appropriate serological assays.