Systematic Evaluation of Clinical, Nutritional, and Fecal Microbial Factors for Their Association With Colorectal Polyps.

INTRODUCTION: The identification of risk factors for precursor lesions of colorectal cancer (CRC) holds great promise in the context of prevention. With this study, we aimed to identify patient characteristics associated with colorectal polyps (CPs) and polyp features of potential malignant progression. Furthermore, a potential association with gut microbiota in this context was investigated. METHODS: In this single-center study, a total of 162 patients with CPs and 91 control patients were included. Multiple variables including information on lifestyle, diet, serum parameters, and gut microbiota, analyzed by 16S-rRNA gene amplicon sequencing and functional imputations (Picrust2), were related to different aspects of CPs. RESULTS: We observed that elevated serum alkaline phosphatase (AP) levels were significantly associated with the presence of high-grade dysplastic polyps. This association was further seen for patients with CRC. Thereby, AP correlated with other parameters of liver function. We did not observe significant changes in the gut microbiota between patients with CP and their respective controls. However, a trend toward a lower alpha-diversity was seen in patients with CRC. Interestingly, AP was identified as a possible clinical effect modifier of stool sample beta diversity. DISCUSSION: We show for the first time an increased AP in premalignant CP. Furthermore, AP showed a significant influence on the microbial composition of the intestine. Relatively elevated liver enzymes, especially AP, may contribute to the detection of precancerous dysplastic or neoplastic changes in colorectal lesions. The association between elevated AP, premalignant CP, and the microbiome merits further study.

A deep learning model enables accurate prediction and quantification of pulmonary edema from chest X-rays.

BACKGROUND: A quantitative assessment of pulmonary edema is important because the clinical severity can range from mild impairment to life threatening. A quantitative surrogate measure, although invasive, for pulmonary edema is the extravascular lung water index (EVLWI) extracted from the transpulmonary thermodilution (TPTD). Severity of edema from chest X-rays, to date is based on the subjective classification of radiologists. In this work, we use machine learning to quantitatively predict the severity of pulmonary edema from chest radiography. METHODS: We retrospectively included 471 X-rays from 431 patients who underwent chest radiography and TPTD measurement within 24 h at our intensive care unit. The EVLWI extracted from the TPTD was used as a quantitative measure for pulmonary edema. We used a deep learning approach and binned the data into two, three, four and five classes increasing the resolution of the EVLWI prediction from the X-rays. RESULTS: The accuracy, area under the receiver operating characteristic curve (AUROC) and Mathews correlation coefficient (MCC) in the binary classification models (EVLWI < 15, ≥ 15) were 0.93 (accuracy), 0.98 (AUROC) and 0.86(MCC). In the three multiclass models, the accuracy ranged between 0.90 and 0.95, the AUROC between 0.97 and 0.99 and the MCC between 0.86 and 0.92. CONCLUSION: Deep learning can quantify pulmonary edema as measured by EVLWI with high accuracy.

Adjunctive Hydrocortisone Improves Hemodynamics in Critically Ill Patients with Septic Shock: An Observational Study Using Transpulmonary Thermodilution.

Introduction: Septic shock is associated with high mortality and hemodynamic impairment. The use of corticoids is a common therapeutic tool in critically ill patients. However, data on the mechanisms and prognostic ability of hemodynamic improvement by adjunctive steroids are rare. This study primarily aimed to evaluate short-term effects of hydrocortisone therapy on catecholamine requirement and hemodynamics derived from transpulmonary thermodilution (TPTD) in 30 critically ill patients with septic shock and a 28 days mortality rate of 50%. Methods: Hydrocortisone was administered with an intravenous bolus of 200 mg, followed by a continuous infusion of 200 mg per 24 h. Hemodynamic assessment was performed immediately before as well as 2, 8, 16, and 24 h after the initiation of corticoids. For primary endpoint analysis, we evaluated the impact of hydrocortisone on vasopressor dependency index (VDI) and cardiac power index (CPI). Results: Adjunctive hydrocortisone induced significant decreases of VDI from 0.41 (0.29-0.49) mmHg-1 at baseline to 0.35 (0.25-0.46) after 2 h (P < .001), 0.24 (0.12-0.35) after 8 h (P < .001), 0.18 (0.09-0.24) after 16 h (P < .001) and 0.11 (0.06-0.20) mmHg-1 after 24 h (P < .001). In parallel, we found an improvement in CPI from 0.63 (0.50-0.83) W/m2 at baseline to 0.68 (0.54-0.85) after 2 h (P = .208), 0.71 (0.60-0.90) after 8 h (P = .033), 0.82 (0.6-0.98) after 16 h (P = .004) and 0.90 (0.67-1.07) W/m2 after 24 h (P < .001). Our analyses revealed a significant reduction in noradrenaline requirement in parallel with a moderate increase in mean arterial pressure, systemic vascular resistance index, and cardiac index. As a secondary endpoint, our results showed a significant decrease in lung water parameters. Moreover, changes in CPI (ΔCPI) and VDI (ΔVDI) after 24 h of hydrocortisone therapy revealed accurate prognostic ability to predict 28 days mortality (AUC = 0.802 vs 0.769). Conclusion: Adjunctive hydrocortisone leads to a rapid decrease in catecholamine requirement and a substantial circulatory improvement in critically ill patients with septic shock.

Accurate prediction of histological grading of intraductal papillary mucinous neoplasia using deep learning.

BACKGROUND: Risk stratification and recommendation for surgery for intraductal papillary mucinous neoplasm (IPMN) are currently based on consensus guidelines. Risk stratification from presurgery histology is only potentially decisive owing to the low sensitivity of fine-needle aspiration. In this study, we developed and validated a deep learning-based method to distinguish between IPMN with low grade dysplasia and IPMN with high grade dysplasia/invasive carcinoma using endoscopic ultrasound (EUS) images. METHODS: For model training, we acquired a total of 3355 EUS images from 43 patients who underwent pancreatectomy from March 2015 to August 2021. All patients had histologically proven IPMN. We used transfer learning to fine-tune a convolutional neural network and to classify "low grade IPMN" from "high grade IPMN/invasive carcinoma." Our test set consisted of 1823 images from 27 patients, recruiting 11 patients retrospectively, 7 patients prospectively, and 9 patients externally. We compared our results with the prediction based on international consensus guidelines. RESULTS: Our approach could classify low grade from high grade/invasive carcinoma in the test set with an accuracy of 99.6 % (95 %CI 99.5 %-99.9 %). Our deep learning model achieved superior accuracy in prediction of the histological outcome compared with any individual guideline, which have accuracies between 51.8 % (95 %CI 31.9 %-71.3 %) and 70.4 % (95 %CI 49.8-86.2). CONCLUSION: This pilot study demonstrated that deep learning in IPMN-EUS images can predict the histological outcome with high accuracy.

B-Lines Scores Derived From Lung Ultrasound Provide Accurate Prediction of Extravascular Lung Water Index: An Observational Study in Critically Ill Patients.

INTRODUCTION: Visualization of B-lines via lung ultrasound provides a non-invasive estimation of pulmonary hydration. Extravascular lung water index (EVLWI) and pulmonary vascular permeability index (PVPI) assessed by transpulmonary thermodilution (TPTD) represent the most validated parameters of lung water and alveolocapillary permeability, but measurement is invasive and expensive. This study aimed to compare the correlations of B-lines scores from extensive 28-sector and simplified 4-sector chest scan with EVLWI and PVPI derived from TPTD in the setting of intensive care unit (primary endpoint). METHODS: We performed scoring of 28-sector and 4-sector B-Lines in 50 critically ill patients. TPTD was carried out with the PiCCO-2-device (Pulsion Medical Systems SE, Maquet Getinge Group). Median time exposure for ultrasound procedure was 12 minutes for 28-sector and 4 minutes for 4-sector scan. RESULTS: Primarily, we found close correlations of 28-sector as well as 4-sector B-Lines scores with EVLWI (R2 = 0.895 vs. R2 = 0.880) and PVPI (R2 = 0.760 vs. R2 = 0.742). Both B-lines scores showed high accuracy to identify patients with specific levels of EVLWI and PVPI. The extensive 28-sector B-lines score revealed a moderate advantage compared to simplified 4-sector scan in detecting a normal EVLWI ≤ 7 (28-sector scan: sensitivity = 81.8%, specificity = 94.9%, AUC = 0.939 versus 4-sector scan: sensitivity = 81.8%, specificity = 82.1%, AUC = 0.902). Both protocols were approximately equivalent in prediction of lung edema with EVLWI ≥ 10 (28-sector scan: sensitivity = 88.9%, specificity = 95.7%, AUC = 0.977 versus 4-sector scan: sensitivity = 81.5%, specificity = 91.3%, AUC = 0.958) or severe pulmonary edema with EVLWI ≥ 15 (28-sector scan: sensitivity = 91.7%, specificity = 97.4%, AUC = 0.995 versus 4-sector scan: sensitivity = 91.7%, specificity = 92.1%, AUC = 0.978). As secondary endpoints, our evaluations resulted in significant associations of 28-sector as well as simplified 4-sector B-Lines score with parameters of respiratory function. CONCLUSION: Both B-line protocols provide accurate non-invasive evaluation of lung water in critically ill patients. The 28-sector scan offers a marginal advantage in prediction of pulmonary edema, but needs substantially more time than 4-sector scan.

Extracorporeal multiorgan support including CO2-removal with the ADVanced Organ Support (ADVOS) system for COVID-19: A case report.

A substantial part of COVID-19-patients suffers from multi-organ failure (MOF). We report on an 80-year old patient with pulmonary, renal, circulatory, and hepatic failure. We decided against the use of extracorporeal membrane oxygenation (ECMO) due to old age and a SOFA-score of 13. However, the patient was continuously treated with the extracorporeal multi-organ- "ADVanced Organ Support" (ADVOS) device (ADVITOS GmbH, Munich, Germany). During eight 24h-treatment-sessions blood flow (100-300 mL/min), dialysate flow (160-320 mL/min) and dialysate pH (7.6-9.0) were adapted to optimize arterial PaCO2 and pH. Effective CO2 removal and correction of acidosis could be demonstrated by mean arterial- versus post-dialyzer values of pCO2 (68.7 ± 13.8 vs. 26.9 ± 11.6 mmHg; p < 0.001). The CO2-elimination rate was 48 ± 23mL/min. The initial vasopressor requirement could be reduced in parallel to pH-normalization. Interruptions of ADVOS-treatment repeatedly resulted in reversible deteriorations of paCO2 and pH. After 95 h of continuous extracorporeal decarboxylating therapy the patient had markedly improved circulatory parameters compared to baseline. In the context of secondary pulmonary infection and progressive liver failure, the patient had a sudden cardiac arrest. In accordance with the presumed patient will, we decided against mechanical resuscitation. Irrespective of the outcome we conclude that extracorporeal CO2 removal and multiorgan-support were feasible in this COVID-19-patient. Combined and less invasive approaches such as ADVOS might be considered in old-age-COVID-19 patients with MOF.

Comparison of global end-diastolic volume index derived from jugular and femoral indicator injection: a prospective observational study in patients equipped with both a PiCCO-2 and an EV-1000-device.

Transpulmonary thermodilution (TPTD)-derived global end-diastolic volume index (GEDVI) is a static marker of preload which better predicted volume responsiveness compared to filling pressures in several studies. GEDVI can be generated with at least two devices: PiCCO and EV-1000. Several studies showed that uncorrected indicator injection into a femoral central venous catheter (CVC) results in a significant overestimation of GEDVI by the PiCCO-device. Therefore, the most recent PiCCO-algorithm corrects for femoral indicator injection. However, there are no systematic data on the impact of femoral indicator injection for the EV-1000 device. Furthermore, the correction algorithm of the PiCCO is poorly validated. Therefore, we prospectively analyzed 14 datasets from 10 patients with TPTD-monitoring undergoing central venous catheter (CVC)- and arterial line exchange. PiCCO was replaced by EV-1000, femoral CVCs were replaced by jugular/subclavian CVCs and vice-versa. For PiCCO, jugular and femoral indicator injection derived GEDVI was comparable when the correct information about femoral catheter site was given (p = 0.251). By contrast, GEDVI derived from femoral indicator injection using the EV-1000 was obviously not corrected and was substantially higher than jugular GEDVI measured by the EV-1000 (846 ± 250 vs. 712 ± 227 ml/m2; p = 0.001). Furthermore, measurements of GEDVI were not comparable between PiCCO and EV-1000 even in case of jugular indicator injection (p = 0.003). This is most probably due to different indexations of the raw value GEDV. EV-1000 could not be recommended to measure GEDVI in case of a femoral CVC. Furthermore, different indexations used by EV-1000 and PiCCO should be considered even in case of a jugular CVC when comparing GEDVI derived from PiCCO and EV-1000.

Influence of Sustained Low-Efficiency Dialysis Treatment on Isavuconazole Plasma Levels in Critically Ill Patients.

Isavuconazole plasma concentrations were measured before and after sustained low-efficiency dialysis (SLED) treatment in 22 critically ill adult patients with probable invasive aspergillosis and underlying hematological malignancies. Isavuconazole levels were significantly lower after SLED treatment (5.73 versus 3.36 µg/ml; P < 0.001). However, even after SLED treatment, isavuconazole concentrations exceeded the in vivo MICs for several relevant Aspergillus species.

Body surface and body core temperatures and their associations to haemodynamics: The BOSTON-I-study: Validation of a thermodilution catheter (PiCCO) to measure body core temperature and comparison of body surface temperatures to thermodilutionderived Cardiac Index.

Assessment of peripheral perfusion and comparison of surface and body core temperature (BST; BCT) are diagnostic cornerstones of critical care. Infrared non-contact thermometers facilitate the accurate measurement of BST. Additionally, a corrected measurement of BST on the forehead provides an estimate of BCT (BCT_Forehead). In clinical routine BCT is measured by ear thermometers (BCT_Ear). The PiCCO-device (PiCCO: Pulse contour analysis) provides thermodilution-derived Cardiac Index (CI_TD) using an arterial catheter with a thermistor tip in the distal aorta. Therefore, the PiCCO-catheter might be used for continuous BCT-measurement (BCT_PiCCO) in addition to intermittent CI-measurement. To the best of our knowledge, BCT_PiCCO has not been validated compared to standard techniques of BCT-measurement including measurement of urinary bladder temperature (BCT_Bladder). Therefore, we compared BCT_PiCCO to BCT_Ear and BCT_Bladder in 52 patients equipped with the PiCCO-device (Pulsion; Germany). Furthermore, this setting allowed to compare different BSTs and their differences to BCT with CI_TD. BCT_PiCCO, BCT_Ear (ThermoScan; Braun), BCT_Bladder (UROSID; ASID BONZ), BCT_Forehead and BSTs (Thermofocus; Tecnimed) were measured four times within 24h. BSTs were determined on the great toe, finger pad and forearm. Immediately afterwards TPTD was performed to obtain CI_TD. 32 (62%) male, 20 (38%) female patients; APACHE-II 23.8 ±8.3. Bland-Altman-analysis demonstrated low bias and percentage error (PE) values for the comparisons of BCT_PiCCO vs. BCT_Bladder (bias 0.05 ±0.27° Celsius; PE = 1.4%), BCT_PiCCO vs. BCT_Ear (bias 0.08 ±0.38° Celsius; PE = 2.0%) and BCT_Ear vs. BCT_Bladder (bias 0.04 ±0.42° Celsius; PE = 2.2). While BCT_PiCCO, BCT_Ear and BCT_Bladder can be considered interchangeable, Bland-Altman-analyses of BCT_Forehead vs. BCT_PiCCO (bias =-0.63 ±0.75° Celsius; PE = 3.9%) Celsisus, BCT_Ear (bias = -0.58 ±0.68° Celsius; PE = 3.6%) and BCT_Bladder (bias = -0.55 ±0.74° Celsius; PE = 3.9%) demonstrate a substantial underestimation of BCT by BCT_Forehead. BSTs and differences between BCT and BST (DCST) significantly correlated with CI_TD with r-values between 0.230 and 0.307 and p-values between 0.002 and p < 0.001. The strongest association with CI_TD was found for BST_forearm (r = 0.307; p < 0.001). In a multivariate analysis regarding CI_TD and including biometric data, BSTs and and their differences to core-temperatures (DCST), only higher temperatures on the forearm and the great toe, young age, low height and male gender were independently associated with CI_TD. The estimate of CI based on this model (CI_estimated) correlated with CI_TD (r = 0.594; p < 0.001). CI_estimated provided large ROC-areas under the curve (AUC) regarding the critical thresholds of CI_TD ≤ 2.5 L/min/m2 (AUC = 0.862) and CI_TD ≥ 5.0 L/min/m2 (AUC = 0.782). 1.) BCT_PiCCO, BCT_Ear and BCT_Bladder are interchangeable. 2.) BCT_Forehead significantly underestimates BCT by about 0.5° Celsius. 3.) All measured BSTs and DCSTs were significantly associated with CI_TD. 4.) CI_estimated is promising, in particular for the prediction of critical thresholds of CI.

Comparison of cardiac function index derived from femoral and jugular indicator injection for transpulmonary thermodilution with the PiCCO-device: A prospective observational study.

INTRODUCTION: Cardiac function index (CFI) is a trans-pulmonary thermodilution (TPTD)-derived estimate of systolic function. CFI is defined as the ratio of cardiac output divided by global end-diastolic volume GEDV (CFI = CO/GEDV). Several studies demonstrated that the use of femoral venous access results in a marked overestimation of GEDV, while CFI is underestimated. One study suggested a correction formula for femoral venous access that markedly reduced the bias for GEDVI. Therefore, the last PiCCO-algorithm requires information about the CVC-position which suggests a correction of GEDV for femoral access. However, a recent study demonstrated inconsistencies of the last PiCCO algorithm using incorrected GEDV to calculate CFI despite obvious correction of GEDV. Nevertheless, this study was based on mathematical analyses of data displayed in a total of 15 patients equipped with only a femoral, but not with a jugular CVC. Therefore, this study compared CFI derived from the femoral indicator injection TPTD to data derived from jugular indicator injection in 28 patients with both a jugular and a femoral CVC. METHODS: 28 ICU-patients with PiCCO-monitoring were included. Each dataset consisted of three triplicate TPTDs using the jugular venous gold standard access and the femoral access with and without information about the femoral indicator injection to evaluate, if correction for femoral GEDV also pertains to CFI. (CFI_jug: jugular indicator injection; CFI_fem: femoral indicator injection; CFI_fem_cor: femoral indicator injection with correct information about CVC-position; CFI_fem_uncor: femoral indicator injection with uncorrect information about CVC-position; CFI_fem_uncor_form = CFI_fem_uncor * (GEDVI_fem_uncor/GEDVI_fem_cor)). RESULTS: CFI_fem_uncor was significantly lower than CFI_jug (4.28±1.70 vs. 5.21±1.91 min-1; p<0.001). Similarly, CFI_fem_cor was significantly lower than CFI_jug (4.24±1.62 vs. 5.21±1.91 min-1; p<0.001). This is explained by the finding that CFI_fem_uncor was not different to CFI_fem_cor (4.28±1.70 vs. 4.24±1.62 min-1; p = 0.611). This suggests that correction for femoral CVC does not pertain to CFI. Calculative correction of CFI_fem_uncor by multiplying CFI_fem_uncor by the ratio GEDVI_fem_uncor/GEDVI_jug resulted in CFI_fem_uncor_form which was slightly, but significantly different from the gold standard CFI_jug (5.51±2.00 vs. 5.21±1.91 min-1; p = 0.024). The agreement of measurements classified in the same category of CFI (decreased (<4.5), normal (4.5-6.5) and increased (>6.5 min-1)) was high for CFI_jug and CFI_fem_uncor_form (identical categories in 26 of 28 comparisons; p = 0.49). By contrast, the agreement with CFI_jug was significantly lower for CFI_fem_cor (14 out of 28; p<0.001) and CFI_fem_uncor (15 out of 28; p<0.001). CONCLUSIONS: While the last PiCCO algorithm obviously corrects GEDVI for femoral indicator injection, this correction is not applied to CFI. Therefore, femoral TPTD indicator injection results in substantially lower values for CFI compared to TPTD using a jugular CVC. Necessarily, uncorrected CFI-values derived from femoral TPTD are misleading and have to be corrected.