Additional Prognostic Information Research Articles

Differences Between Cystatin C- and Creatinine-Based Estimated Glomerular Filtration Rate and Association with Mortality and Cardiovascular Events: Results from Three Cohorts of Adults with Diabetes.

To explore the association between the differences between cystatin C- and creatinine-based estimated glomerular filtration rate (eGFRdiff) with the risk of mortality and cardiovascular (CV) events in individuals with diabetes. Three prospective cohorts analyzed data of adults with diabetes from the Incident, Development, and Prognosis of Diabetic Kidney Disease (INDEED) study (2016-2017 to 2020) in China, the National Health, Nutrition Examination Survey (NHANES, 1999-2004 to 2019) in the United States, and UK Biobank (UKB, 2006-2010 to 2022). Baseline eGFRdiff was calculated using both absolute difference between cystatin C- and creatinine-based calculations (eGFRabdiff), and the ratio between them (eGFRrediff). Cox proportional hazards regression models were used to investigate the association between eGFRdiff and outcomes including all-cause mortality and incident CV events. A total of 8,129 individuals from the INDEED (aged 60.7±10.0 years), 1,634 from the NHANES (aged 62.5±14.4 years), and 29,358 from the UKB (aged 59.4±7.3 years;) were included. At baseline, 43.6%, 32.4% and 42.1% of participants in the INDEED, NHANES and UKB had an eGFRabdiff value ≥15 ml/min/1.73 m2. During a median follow-up of 3.8 years for the INDEED, 15.2 years for the NHANES, and 13.5 years for the UKB, a total of 430, 936 and 6143 deaths and a total of 481, 183 and 5583 CV events occurred, respectively. Each 1-standard deviation higher baseline eGFRabdiff was independently associated with a lower risk of all-cause mortality and CV events, with hazard ratios (HRs) of 0.77 and 0.82 in the INDEED, 0.70 and 0.68 in the NHANES, and 0.66 and 0.78 in the UKB. Similar results were observed for eGFRrediff. eGFRdiff represents a marker of adverse events for diabetes among general population. Monitoring both eGFRcys and eGFRcr yields additional prognostic information and has clinical utility in identifying high-risk individuals for mortality and CV events.

Clinicopathological Characteristic of Radioactive Iodine-Refractory Differentiated Thyroid Carcinoma at Dr. Hasan Sadikin Hospital 2016-2021

Background Initial therapy for differentiated thyroid carcinoma (DTC) is thyroidectomy with or without cervical lymph node dissection. Furthermore, radioactive iodine (RAI) is given to patients by considering risk stratification and other patient factors. Although most cases of DTC have a good prognosis after standard therapeutic approaches, the risks of local recurrence and distant metastases can be as high as 20% and 10%. Among these patients, two-thirds showed RAI-refractory. This is concerning because 10-year survival rate is less than 10%. This study aimed to analyze the clinicopathological characteristics of RAI-refractory DTC. Methods This is a case-control study. Data was collected from the Department of Nuclear Medicine and Molecular Theranostics and Department of Anatomical Pathology, Faculty of Medicine, Universitas Padjadjaran/Dr. Hasan Sadikin Hospital, Bandung period 1 January 2016 - 31 December 2021. Results Clinicopathological factors associated with RAI-refractory DTC are age, sex, aggressive histologic subtype, LVI (lymphovascular invasion), m-ETE (microscopic extrathyroid extension), TNM (tumor, nodal, metastasis) stage, and ENE (extranodal extension), with p-value <0.05. Meanwhile, there was no significant difference in the histologic type between RAI-refractory and non-RAI-refractory groups. Conclusion In the pathology report, it is necessary to include prognostically relevant tumor histopathological characteristics. In addition to histologic type, histologic subtype, and tumor size, other features such as presence and extent of capsular invasion, LVI, microscopic and macroscopic ETE, ENE, and number and size of metastatic lymph nodes, have been shown to provide additional prognostic information and are required in standard pathology reports for DTC.

Influence of the Pre-shock State on the Prognosis of Medical Patients with Sepsis: A Retrospective Cohort Study

ObjectiveTo investigate the effects of the pre-shock state on the mortality of patients with sepsis. MethodsWe enrolled patients with sepsis admitted to the medical intensive care unit of a tertiary care university hospital. These patients were then classified into three groups: sepsis, pre-shock state, and septic shock. The primary outcome was the 28-day mortality rate. The secondary outcomes were the 90-day, 180-day, and 1-year mortality rates. ResultsA total of 303 patients (groups: sepsis 135 [44.6%]), pre-shock state (93 [30.7%]), and septic shock (75 [24.8%]) completed the 1-year follow-up. The mortality rates at 28 days, 90 days, and 180 days and 1 year were significantly higher in the pre-shock state group than those of the sepsis group, but significantly lower than those in the septic shock group, especially among older patients. When compared with the pre-shock state group, the sepsis group had significantly lower mortality risks at 28 days, 90 days, and 180 days and 1 year, whereas the sepsis shock group had higher mortality risks at these time points. ConclusionThe mortality rates of patients in the pre-shock state were notably different from those of patients with sepsis or septic shock. The introduction of a modified sepsis severity classification, which includes sepsis, pre-shock state, and septic shock, could offer valuable additional prognostic information.

Cholesterol Crystal Dissolution Rate of Serum Predicts Outcomes in Patients With Aortic Stenosis Undergoing Transcatheter Aortic Valve Replacement.

Aortic stenosis has pathophysiological similarities with atherosclerosis, including the deposition of cholesterol-containing lipoproteins. The resulting cholesterol crystals activate the NLRP3 (NOD-like receptor protein 3) inflammasome, leading to inflammation and cardiovascular diseases. We aimed to investigate the cholesterol crystal dissolution rate (CCDR) of serum in patients with aortic stenosis and to assess the prognostic value of this biomarker. The study included 348 patients with aortic stenosis undergoing transcatheter aortic valve replacement. The CCDR was measured using flow cytometry to enumerate cholesterol crystals that were added to a serum solution, at baseline and after 2 hours of incubation. Based on the median CCDR, the cohort was stratified into high and low cholesterol crystal dissolvers. The incidence of the primary end point, a composite of 1-year all-cause mortality and major vascular complication, was significantly lower in the high CCDR group (7.3 per 100 person-years) compared with the low CCDR group (17.0 per 100 person-years, P=0.01). This was mainly driven by a lower 1-year mortality rate in patients with a high CCDR (7.3 versus 15.1 per 100 person-years, P=0.04). Unplanned endovascular interventions were significantly less frequent in high cholesterol crystal dissolvers (12.8 versus 22.6 per 100 person-years, P=0.04). Although low-density lipoprotein cholesterol levels were comparable in both groups (101.8±37.3 mg/dL versus 97.9±37.6 mg/dL, P=0.35), only patients with a low CCDR showed a benefit from statin treatment. In multivariate analysis, low CCDR (hazard ratio, 2.21 [95% CI, 0.99-4.92], P=0.04) was significantly associated with 1-year mortality. The CCDR is a novel biomarker associated with outcome in patients with aortic stenosis undergoing transcatheter aortic valve replacement. It may provide new insights into patients' anti-inflammatory capacity and additional prognostic information beyond classic risk assessment.

Cardiovascular magnetic resonance imaging-derived intraventricular pressure gradients in ST-segment elevation myocardial infarction: a long-term follow-up study

Abstract Aims Recently, novel post-processing tools have become available that measure intraventricular pressure gradients (IVPGs) on routinely obtained long-axis cine cardiac magnetic resonance (CMR) images. IVPGs provide a comprehensive overview of both systolic and diastolic left ventricular (LV) functions. Whether IVPGs are associated with clinical outcome after ST-elevation myocardial infarction (STEMI) is currently unknown. Here, we investigated the association between CMR-derived LV-IVPGs and major adverse cardiovascular events (MACE) in a large reperfused STEMI cohort with long-term outcome. Methods and results In this prospectively enrolled multi-centre cohort study, 307 patients underwent CMR within 14 days after the first STEMI. LV-IVPGs (from apex-to-base) were estimated on the long-axis cine images. During a median follow-up of 9.7 (5.9–12.5) years, MACE (i.e. composite of cardiovascular death and de novo heart failure hospitalisation) occurred in 49 patients (16.0%). These patients had larger infarcts, more often microvascular injury, and impaired LV-IVPGs. In univariable Cox regression, overall LV-IVPG was significantly associated with MACE and remained significantly associated after adjustment for common clinical risk factors (hazard ratio (HR) 0.873, 95% confidence interval (CI) 0.794–0.961, P = 0.005) and myocardial injury parameters (HR 0.906, 95% CI 0.825–0.995, P = 0.038). However, adjusted for LV ejection fraction and LV global longitudinal strain (GLS), overall LV-IVPG does not provide additional prognostic information (HR 0.959, 95% CI 0.866–1.063, P = 0.426). Conclusion Early after STEMI, CMR-derived LV-IVPGs are univariably associated with MACE and this association remains significant after adjustment for common clinical risk factors and measures of infarct severity. However, LV-IVPGs do not add prognostic value to LV ejection fraction and LV GLS.

Estimated pulse wave velocity added additional prognostic information in general population: Evidence from National Health and Nutrition Examination Survey (NHANES) 1999–2018

BackgroundAs an indicator of arterial stiffness, there is controversy over whether estimated pulse wave velocity (ePWV) add additional prognostic information other than cardiovascular risk factors or traditional risk estimation model in general population. MethodsData from National Health and Nutrition Examination Survey in 1999–2018 was analyzed. Cardiovascular risk factors were collected and Framingham Risk Score (FRS) was calculated. Using all-cause and cardiovascular mortality as outcomes, Cox and restricted cubic spline (RCS) analysis was performed. Receiver operator characteristic (ROC) curves, Harrell's C-statistic and net reclassification index (NRI) analysis were used to assess whether ePWV adds additional predictive value. ResultsThe association between ePWV and outcomes was independent of cardiovascular risk factors (HR = 1.23 [95%CI 1.23–1.50] per m/s for all-cause mortality, and 1.52 [1.30–1.78] for cardiovascular mortality) and FRS (1.22 [1.12–1.32] for all-cause mortality, and 1.32 [1.10–1.59] for cardiovascular mortality). Except for ePWV and all-cause mortality adjusted by FRS, a liner association was found between ePWV and outcomes. For predictive value, the area under ROC and C-index of the model added with ePWV was higher than the one with FRS or risk factors alone (P < 0.01). The elevated ePWV upgraded 1338456 subjects from high-intermediate to high FRS category, and NRI was 3.61 % and 2.62 % for all-cause and cardiovascular deaths, respectively (all P < 0.001). ConclusionsIn general population, the present study demonstrated the association between ePWV and all-cause, cardiovascular mortality is independent of cardiovascular risk factors and traditional risk estimated model. ePWV also added additional information to them in predicting clinical outcomes.

Immunoexpression of p53 and Ki-67 in Urinary Bladder Carcinoma and Their Association with Tumor Grade and Muscle-invasiveness

Background: Urinary bladder carcinoma is one of the threatening causes of human morbidity and mortality. Various molecular studies and prognostic biomarkers are helpful to assess the outcome of urinary bladder carcinoma. High expression of p53 and Ki-67 represents the aggressiveness of the disease and indicates poor prognosis. Methods: This cross-sectional study was conducted at the Department of Pathology, Sylhet MAG Osmani Medical College from September 2018 to July 2020 to observe the immunoexpression of p53 and Ki-67 in urinary bladder carcinoma and their association with tumor grade and muscle-invasiveness. Total 57 transurethral resection (TURBT) samples of radiologically and clinically suspected bladder tumor cases were diagnosed histopathologically included in the study. Results: Age of the study patients ranged from 35 to 80 years with a mean of 60.60 ± 11.14 and male to female ratio was 4.7:1. There were 31 (54.4%) high grade and 26 (45.6%) low grade urinary bladder carcinomas on histopathological examination. Thirty (52.6%) cases were histopathologically muscle invasive bladder carcinoma (MIBC) and 27 (47.4%) cases were non-muscle invasive bladder carcinoma (NMIBC). Positive p53 expression was observed in 24 (77.4%) cases of high grade carcinoma while only in 6 (23.1%) cases of low grade urinary bladder carcinoma. Positive p53 expression was seen in 22 (73.3%) cases of MIBC and 8 (29.6%) cases of NMIBC. Ki-67 expression was found positive in 27 (87.1%) cases of high grade carcinoma while only in 7 (26.9%) cases of low grade urinary bladder carcinoma. Positive Ki-67 expression was seen in 25 (83.3%) cases of MIBC and 9 (33.3%) cases of NMIBC. Statistically significant association (P&lt;0.05) was observed in histopathological grades and muscle invasiveness of UBC with p53 and Ki-67 expression. Conclusion: p53 and Ki-67 immunomarker may be helpful to provide additional prognostic information in urinary bladder carcinoma.

Clinical utility of DNA methylation profiling for choroid plexus tumors.

Choroid plexus tumors (CPTs) are rare, potentially aggressive CNS tumors with defined histologic criteria for grading. In recent years, several patients within our practice have demonstrated discordance between the histologic diagnosis and clinical behavior. DNA methylation profiling has emerged as a potential diagnostic adjunct for aiding the clinical approach. We reviewed the clinical and pathologic data of all CPTs diagnosed at Boston Children's Hospital from 1995 to 2023. All cases with available material (38/48) underwent DNA methylation profiling at NIH/NCI, and the classifier results were correlated with the WHO histologic grade and patient outcomes. Survival information was analyzed using Kaplan-Meier curves. There was good correlation (11/12, 92%) between methylation class and WHO histologic grade for choroid plexus carcinomas (CPC); one histologic CPC grouped with choroid plexus papilloma (CPP) group pediatric (P). Five CPPs grouped with methylation class CPC (5/17, 29%). In the group of atypical CPPs (n = 9), there were two that grouped with methylation class CPC. Survival analysis showed utility of methylation classes in the prediction of biologic behavior. Results indicated that methylation profiling may serve as a valuable tool in the clinical decision-making process for patients with CPTs, providing additional prognostic information compared to WHO histologic grade alone. The value of methylation array analysis is particularly important given the lack of consensus on treatment regimens for CPTs.

Confidence intervals of survival predictions with neural networks trained on molecular data

In medicine, an important objective is predicting patients’ survival based on their molecular and clinical characteristics. In this context, neural networks have recently been used for their ability to capture complex interactions in the data. Measuring the uncertainty associated with survival estimates obtained by neural networks is essential to enhance predictions’ reliability. We compared four methods adapted to multilayer perceptrons (MLPs) for building confidence intervals at the patient level. The methods were based either on bootstrap with Boot (Efron, 1979), ensembling with DeepEns (Lakshminarayanan et al., 2016), or Monte-Carlo Dropout with MCDrop and BMask (Gal and Ghahramani, 2016; Mancini et al., 2020). A comparison was made through MLP-based survival models: CoxCC and CoxTime (Kvamme et al., 2019) in a continuous time framework, DeepHit (Lee et al., 2018) and PLANN (Biganzoli et al., 1998) in a discrete time framework. We applied the methods to a simulation study, enabling us to estimate a coverage rate of the estimated confidence intervals. We also applied them to real-world datasets, and predicted the survival probability for patients with breast cancer and patients with lung cancer.In the simulation study, CoxCC and CoxTime obtained the mean C-indices numerically closest to those from the Oracle model (mean C-index of 0.723 for CoxCC, 0.726 for CoxTime, versus 0.743 for the Oracle model). Regarding the confidence intervals of survival probabilities, Boot with CoxCC obtained a coverage rate of 96.5%, the closest to the nominal value of 95%. MCDrop was slightly anticonservative and obtained a coverage rate of 89.8% with CoxTime. This method may represent a reasonable compromise in terms of coverage with regards of computational time. In the breast cancer cohort, MLPs had difficulty capturing additional prognostic information from the molecular data. In contrast, in the lung cancer cohort, the models led to substantially stronger discrimination values when adding molecular data to the clinical variables. In conclusion, we were able to represent uncertainty in the survival estimates at particular time points at the patient level using MLPs in the form of 95% confidence intervals. We recommend using CoxTime with either Boot or, for a less intensive computation time, MCDrop.

Combined utility of Ki-67 index and tumor grade to stratify patients with pancreatic ductal adenocarcinoma who underwent upfront surgery

ObjectiveTo investigate the prognostic prediction of a new indicator, combined by tumor grade and Ki-67, in patients with resected pancreatic ductal adenocarcinoma (PDAC).MethodsData were retrospectively collected from consecutive patients who underwent primary resection of pancreas from December 2012 to December 2017. Tumor grade and Ki-67 were reviewed from routine pathological reports. G-Ki67 was classified as three categories as I (G1/2 and Ki-67 < 40%), II (G1/2 and Ki-67 ≥ 40%), and III(G3/4 and all Ki-67).ResultsCox regression analyses revealed that tumor stage (II vs. I: hazard ratio (HR), 3.781; 95% confidence index (CI), 2.844–5.025; P < 0.001; III vs. I: HR, 7.476; 95% CI, 5.481–10.20; P < 0.001) and G-Ki67 (II vs. I: HR, 1.299; 95% CI, 1.038–1.624; P = 0.022; III vs. I: HR, 1.942; 95% CI, 1.477–2.554; P < 0.001) were independent prognostic factors in the developing cohort. The result was rectified in the validation cohort. In subgroups analysis, G-Ki67 (II vs. I: HR, 1.866 ; 95% CI, 1.045–3.334; P = 0.035; III vs. I: HR, 2.333 ; 95% CI, 1.156–4.705; P = 0.018) also had a high differentiation for survival prediction.ConclusionOur findings indicate that three-categories of G-Ki67 in resectable PDAC according to the routine pathological descriptions provided additional prognostic information complementary to the TNM staging system.

Abstract B017: Clinical utility of next generation sequencing in advanced colorectal cancer: The earlier the better

Abstract Background. Since next generation sequencing (NGS) was approved in solid tumor in Korea, it has been usually performed when patients were progressed on standard chemotherapy in advanced colorectal cancer (CRC). Recently, new targeted agents have been developed rapidly, so the NGS is recommended as early as possible. In this study, we analyzed NGS results and investigated clinical benefit in advanced CRC. Patient and method. Between October 2018 to December 2022, NGS data in 85 advanced CRC patients in St. Vincent’s Hospital, Suwon, Korea were retrospectively analyzed. Results. Seven patients had metachronous, and others had synchronous metastasis. The median age was 58 (34-84), and 59 (69.4%) were male. Fourteen (16.5%) were right-side, 40 (47.1%) were left-side colon, and 30 (35.3%) were rectum. The median time from diagnosis of advanced CRC to NGS was 13 (0-81) months, and median time to death after NGS was 8 (0.3-43.0) months. NGS platform were ‘Oncomine comprehensive assay Plus’ in 71 (83.5%), and the others were various platforms in K-master project. NGS analyses revealed new therapeutic target in 17 (20%): high tumor mutational burden (n = 2), KRAS G12C (n = 4), KRAS (n = 1) or BRAF (n = 2) mutation not routinely screened at initial, BRAF class II mutation (n = 2), Her2/3 mutation (n = 2), and NTRK fusion (n = 1). Among those, two patients with KRAS G12C mutation enrolled in clinical trial of KRAS G12 inhibitor, and one patient with NTRK fusion received NTRK inhibitor (3.5% of total). Among RAS wild type population (n=30), we identified hyperselected patients (n = 18, [60%], absence of either KRAS, NRAS, PTEN, and extracellular domain EGFR mutations, HER2 and MET amplifications, and ALK, RET or NTRK1 fusions by PARADIGM trial), which showed longer survival in cetuximab-treated population (p&amp;lt;0.001). Specific mutation including KRAS, NRAS, BRAF, TP53, or APC did not impact on prognosis. High TMB (≥10Mut/Mb) were associated with longer survival (p = 0.020). For patients with oligometastasis, hyperselected patients (p = 0.062) showed good prognosis. Conclusion. The NGS results provided additional predictive or prognostic information in advanced CRC patients. Application of NGS as early as possible might help to make a clinical decision. Citation Format: Ho Jung An, Hyunho Kim, Hyung Soon Park, Hyeon-Min Cho, Bong-Hyeon Kye, Ri Na Yoo, Ji-Yeon Moon, Sung Hwan Kim, Jonghoon Lee, Hyo Chun Lee. Clinical utility of next generation sequencing in advanced colorectal cancer: The earlier the better [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr B017.

Value of CRP/albumin ratio as a prognostic marker of acute pancreatitis: a retrospective study.

the C-reactive protein (CRP) to albumin ratio is an inflammatory marker that has shown promise in the prognosis of critically ill patients. This study is aimed to assess the value of CRP/albumin ratio to predict severity in acute pancreatitis. a retrospective study was performed using a prospectively collected database of patients diagnosed with AP admitted to the Department of Gastroenterology between March 2014 and December 2021. among 722 patients included in the study, 78.67 % had mild, 15.65 % had moderately severe, and 5.67 % had severe acute pancreatitis. The CRP/albumin ratio was significantly associated with severe AP (OR 1.02; 95 % CI: 1.01-1.03; p < 0.001), and each ten-unit increase in the ratio was associated with a 20 % increased likelihood of severe acute pancreatitis. The area under the ROC curve (AUC) value of the CRP/albumin ratio in severe acute pancreatitis was 0.68 (95 % CI: 0.58-0.77), which was higher than that of the Ranson criteria (0.62). The optimal cut-off value for predicting severe acute pancreatitis was 7.51, with a sensitivity of 63.4 % and specificity of 65.6 %. despite its low sensitivity and specificity, the CRP/albumin ratio could be used as a complementary marker to the current scoring systems for the initial assessment of acute pancreatitis prognosis. It is easily obtainable and can provide additional prognostic information to clinicians.

The prognostic significance of tumor-immune microenvironment in ascites of patients with high-grade serous carcinoma.

High-grade serous carcinoma (HGSC) is often associated with ascites at presentation. Our objective was to quantify immune cells (ICs) in ascites prior to any treatment was given and evaluate their impact on progression-free survival (PFS) and overall survival (OS). Forty-seven patients with primary HGSC and ascites were included. Flow-cytometric analysis was performed to detect percentages of CD3+ T cells (CD4+, CD8+, Tregs, and NKT cells), B cells, NK cells (CD56brightCD16- and CD56dimCD16+ subsets), macrophages and dendritic cells (DCs). Furthermore, CD103 expression was analyzed on T cells and their subsets, while PD-1 and PD-L1 expression on all ICs. Cut-off of low and high percentages of ICs was determined by the median of variables, and correlation with PFS and OS was calculated. CD3+ cells were the predominant ICs (median 51%), while the presence of other ICs was much lower (median ≤10%). CD103+ expression was mostly present on CD8+, and not CD4+ cells. PD-1 was mainly expressed on CD3+ T cells (median 20%), lower expression was observed on other ICs (median ≤10%). PD-L1 expression was not detected. High percentages of CD103+CD3+ T cells, PD-1+ Tregs, CD56brightCD16- NK cells, and DCs correlated with prolonged PFS and OS, while high percentages of CD8+ cells, macrophages, and PD-1+CD56brightCD16- NK cells, along with low percentages of CD4+ cells, correlated with better OS only. DCs were the only independent prognostic marker among all ICs. Our results highlight the potential of ascites tumor-immune microenvironment to provide additional prognostic information for HGSC patients. However, a larger patient cohort and longer follow-up are needed to confirm our findings.

Feature-based clustering of the left ventricular strain curve for cardiovascular risk stratification in the general population

ObjectiveIdentifying individuals with subclinical cardiovascular (CV) disease could improve monitoring and risk stratification. While peak left ventricular (LV) systolic strain has emerged as a strong prognostic factor, few studies have analyzed the whole temporal profiles of the deformation curves during the complete cardiac cycle. Therefore, in this longitudinal study, we applied an unsupervised machine learning approach based on time-series-derived features from the LV strain curve to identify distinct strain phenogroups that might be related to the risk of adverse cardiovascular events in the general population.MethodWe prospectively studied 1,185 community-dwelling individuals (mean age, 53.2 years; 51.3% women), in whom we acquired clinical and echocardiographic data including LV strain traces at baseline and collected adverse events on average 9.1 years later. A Gaussian Mixture Model (GMM) was applied to features derived from LV strain curves, including the slopes during systole, early and late diastole, peak strain, and the duration and height of diastasis. We evaluated the performance of the model using the clinical characteristics of the participants and the incidence of adverse events in the training dataset. To ascertain the validity of the trained model, we used an additional community-based cohort (n = 545) as external validation cohort.ResultsThe most appropriate number of clusters to separate the LV strain curves was four. In clusters 1 and 2, we observed differences in age and heart rate distributions, but they had similarly low prevalence of CV risk factors. Cluster 4 had the worst combination of CV risk factors, and a higher prevalence of LV hypertrophy and diastolic dysfunction than in other clusters. In cluster 3, the reported values were in between those of strain clusters 2 and 4. Adjusting for traditional covariables, we observed that clusters 3 and 4 had a significantly higher risk for CV (28% and 20%, P ≤ 0.038) and cardiac (57% and 43%, P ≤ 0.024) adverse events. Using SHAP values we observed that the features that incorporate temporal information, such as the slope during systole and early diastole, had a higher impact on the model's decision than peak LV systolic strain.ConclusionEmploying a GMM on features derived from the raw LV strain curves, we extracted clinically significant phenogroups which could provide additive prognostic information over the peak LV strain.

Leukemic Stem Cells (LSCs) As Peri-Transplant MRD Assessment in AML Patients Undergoing Allogeneic Stem Cell Transplantation in CR

Background: Several studies have shown that the leukemic stem cells (LSCs: CD34 +CD38 -CD45RA +/Combi-6 +) play a crucial role in the development of relapses in AML patients (pts). LSCs are less sensitive to conventional therapy; moreover, their phenotypical characteristics may allow immune evasion promoting relapses after allo-SCT. Although ELN guidelines recommend including LSCs into MRD assessment, this approach has not yet been studied extensively in the allogeneic setting. Here we suggested that peri-transplant LSCs detection may provide additional prognostic information to conventional (c)MRD in AML pts allografted in CR. Methods: 90 pts (male, n=46; median, 60.5 years, 19-76) with AML in CR and available pre-transplant cMRD data (multicolor flow cytometry, “different from normal” approach, according to ELN guidelines), who received allografts (matched, n=73; mismatched, n=17) during 2019-2022 years at the Department of Stem Cell Transplantation at University Medical Centre Hamburg were included. LSCs were assessed according to Zeijlemaker et al. 2019 before and after transplant at days +30 and +100. Both methods (cMRD and LSCs detection) were investigated solely and combined for their prognostic impact. Results: Majority of pts had de novo AML (76%), adverse ELN risk (48%), received matched grafts (related, n=73, 81%; unrelated, n=17, 19%) and MAC regimen (n=55, 61%). ATG was used for GvHD prophylaxis in 70 pts (78%). There were 47 cMRD neg (52%) and 43 cMRD pos (48%) pts. LSCs were detected in 39 (56%) pts with a median proportion of 0.01% (0.005-0.37%) of WBC. There was no correlation between LSCs proportion and cMRD status. The cMRD pos pts were more frequently LSCs pos (67% vs 33%), and cMRD neg were more frequently LSCs neg (33% vs 67%, p=0.002). Before allo-SCT, 34 pts (38%) were cMRD neg/LSCs neg, 26 pts (29%) experienced cMRD pos/LSCs pos, 17 pts (19%) had cMRD neg/LSCs pos and 13 pts (14%) had cMRD pos/LSCs neg. The conformity between both assays was 67%. During a median follow up of 10 months (1-29) there were 17 mortalities, 13 relapses and 11 NRM events. The 1-year OS and LFS were significantly better in LSCs neg than in LSCs pos pts: 91% (77-97%) vs 73% (55-86%, p=0.012) and 88% (72-96%) vs 52% (35-68%, p=0.001), respectively. This was due to a higher relapse rate in LSCs pos group: 33% (18-52%) vs 4% (1-20%, p=0.002). The median time to relapse was 180 days (40-395). The difference in NRM was not significant. The area under the ROC curve for relapses was 0.74 (0.61-0.88, p=0.006). Regarding the combined used of LSCs/cMRD, the best 1-year OS (p=0.038) and LFS (p=0.002) were observed for the double negative group: 94% (77-99%) and 88% (65-97%); followed by cMRD pos/LSCs neg: 86% (59-96%) and 86% (61-97%); cMRD neg/LSCs pos: 77% (43-94%) and 71% (39-90%); and the double positive group: 63% (43-79%) and 42% (25-61%)., respectively ( Figure 1). This was due to in the differing relapse rate that was highest in cMRD neg/LSCs pos group: 47% (17-79%), followed by double positive: 29% (14-50%), double negative: 7% (1-32%), and cMRD pos/LSCs neg group: 0% (p=0.02). Post-transplant cMRD/LSCs data on days +30 and +100 were available in 70 consecutive pts. The post-transplant conversion rates (positive &amp;gt; negative) for LSCs, cMRD and cMRD/LSCs were 49%, 61%, and 63%, respectively. The majority of pts converted at day +30. We observed a higher conversion rate to LSC neg status after MAC vs RIC (12/16, 75% vs 5/11, 45%, p=0.12). Of nine pts who were MRD pos at day +30, six were LSC neg and three were LSC pos. One MRD pos/LSC neg pts died due to NRM while remaining five pts survived without relapse. All three MRD pos/LSC pos pts developed relapses with two mortalities. Conclusion: Peri-transplant detection of LSCs has a strong predictive value for early post-transplant relapses in AML pts. This approach can improve pre-transplant risk assessment for cMRD pos patients by defining a population with better survival and low relapses (cMRD pos/LSCs neg) and one with worse survival and more relapses (cMRD pos/LSCs pos). The post-transplant clearance of residual disease is associated with better survival regardless of the flow strategy being used (cMRD, LSCs, or both). Pts with persisting cMRD early after allo-SCT may be subdivided into different risk groups according to their LSCs status. Figure 1. Leukemia-free survival according to pre-transplant cMRD/LSCs status in 90 AML patients.

Combining Gene Mutation with Transcriptomic Data Improves Outcome Prediction in Myelodysplastic Syndromes

Background and Aim. Myelodysplastic syndromes (MDS) are myeloid neoplasms characterized by peripheral blood cytopenias and risk of progression to acute myeloid leukemia (AML). Disease management is challenged by heterogeneity in clinical courses and survival probability. Recently, the genomic screening integration (by Molecular International Prognostic Scoring System, IPSS-M) into patient's assessment has resulted into a significant improvement in predicting clinical outcomes compared to the conventional prognostic score (Revised IPSS, IPSS-R). Many of the consequences of genetic and cytogenetic alterations will affect gene expression by means of transcriptional and epigenetic instability and altered microenviromental signaling. The aim of this project conducted by GenoMed4All and Synthema EU consortia is to link genomic information with transcriptomic data for possibly improving the prediction of clinical outcomes in MDS patients. Patients and Methods.Clinical, cytogenetic, genomic (somatic mutations screening of 31 target genes) and transcriptomic (bulk RNA-seq of CD34 + bone marrow cells) data were collected at diagnosis in 389 MDS patients. Transcriptomic and genomic profiles were processed and the former were normalized before Principal Component Analysis (PCA) dimensionality reduction to mine the interdependency of expression-wide perturbation and recurrent genomic alterations. The prognostic impacts of genetic, cytogenetic, transcriptomic, clinical and demographic features were assessed with a penalized Cox's proportional hazards model [Gerstung M et al, Nat Commun. 2015. 6, 5901] considering the Overall Survival (OS) as primary end point. A 5-fold cross-validating (CV) scheme was exploited to control bias in risk estimation. Model accuracy was assessed using Harrell's concordance index (C-index). An independent validation of the results on 202 patients was planned. Results.We first processed each data layer assessing data robustness, removed not informative variables and scaled quantitative ones. We considered recurrent genomic and cytogenetic lesions (present in ≥5 patients), platelets, hemoglobin and bone marrow blasts (%), age and sex as covariates. To explore the main patterns of expression changes, PCA was performed to reduce multidimensional correlated expression features (20 PCs was selected, explaining 42% of the total transcriptomic variability). To evaluate the prognostic power of each data layer we grouped all available features into five groups: gene mutations (n=15), cytogenetic alterations (n=7), expression data (n=20), blood counts (n=3) and demographic variables (n=2). Within a 5-fold CV we combined these variables in our integrative model to calculate MDS patients risk. The obtained predictive accuracy (C-index) for OS was 0.83, underlying that transcriptomic data significantly improved the current standard prognostic scoring systems. Accordingly, in our patient population, the C-index of the conventional IPSS-R score and the new IPSS-M were 0.68 and 0.76, respectively. A similar improvement by adding transcriptomic data was observed in prediction of the risk of AML evolution. Moreover, by analyzing the contribution of each feature category to the OS probability ( Figure 1), in term of explained variance, the relative impact of transcriptomic is 40%, with the remaining prognostic information distributed among genomic features (somatic gene mutations and cytogenetics lesions, 24%), demographics (20%) and clinical features (15%). An independent validation of these results on 202 patients is currently ongoing. Figure 2 shows an example of personalized survival prediction using patients from the study population. In two subjects with same clinical phenotype and mutations leading to a similar IPSS-M prognosis, the integrative model captures additional prognostic information and efficiently predicts clinical outcome. Given the complexity of our model, specific technological support is needed to combine data at individual patient level and to translate it into a personalized outcome prediction. To this aim, we created a prototype web portal based on our dataset for user-defined genomic/transcriptomic and clinical features. Conclusion. In predicting survival of MDS patients, genomic, transcriptomic and diagnostic clinical variables all have utility, with a significant contribution from the transcriptome.

Multiparametric Flow Cytometry Based Quantification of the Marrow Hematopoietic Stem and Progenitor Compartment in Post Allogeneic Stem Cell Transplant Recipients with Poor Graft Function to Evaluate Hematopoietic Stem Cell Quality and Predict Clinical Outcomes: A Pilot Study

Poor graft function (PGF) is a poorly defined, commonly occurring (estimated 5 to 27%) 1 and difficult to treat complication of allogeneic stem cell transplantation (Allo-HSCT). Thrombopoietin agonists such as eltrombopag (EPAG) or CD34-selected cell infusions are treatments offered for persistent cytopenias. However, with limited knowledge of pathophysiology, there is a lack of reliable prognostic biomarkers. We aimed to investigate if assessment of the BM hematopoietic stem and progenitor cell (HSPC) landscape by multiparametric flow cytometry (MP-FCM) could help evaluate stem cell quality and predict clinical outcomes in PGF patients. We screened the UHN biobank repository to identify banked BM samples of patients who underwent diagnostic BM aspirate for PGF post allo-HSCT. Patients who had progressive cytopenia in one or more lineages after achieving engraftment while having full donor chimerism were included. Relapse as a cause of cytopenia was excluded. Controls in 1:1 ratio were Allo-SCT recipients with stable counts with banked day 60 BM aspirate. The blinded BM samples were analyzed with a 14-parameter FCM adapted from panels previously utilized to assess HSPC and mature lineages in cord blood xenograft studies (2). We selected 12 each of PGF and controls. The mean age of all recipients was 50 years and donors was 30.5 years. 23/24 cases had AlloSCT for hematological malignancies with 15 having AML or MDS. There were 14 unrelated (9 matched, 5 mismatched) and 10 related (5 matched sibling, 5 haploidentical) donors. Mean cell dose of infused graft was 6.66(± 1.98)x 10E6/kg CD34 cells. The significant difference between PGF and controls were recipient age (median 57y vs 28y), allo-SCT number (3 vs 0 with 2 nd transplant), intensity of conditioning regimen (myeloablative in 16% PGF vs 83% controls), and incidence of ≥ grade 3 acute graft vs host disease (GVHD) (n=3 vs 0). Most common causes of PGF were GVHD (n=6), infections (n=4) and drug effects (n=5). 11 PGF patients were treated with EPAG with 5 patients (A3, B2, B3, B5, C4) showing complete response, 3 (A6, B6, C5) showing partial response, and 3 (A2, A4, C9) showing no response. Patient C1 showed complete recovery without the use of EPAG, while C9, who initially did not respond to EPAG, showed a slow spontaneous recovery almost a year later. Amongst non-responders A2 and A4 received CD34 boost; with A4 showing a good response to the second boost. Analyzing the HSPC population by MP-FCM, we found that the PGF vs control groups show significant differences in the percentage of CD34+CD19+ B-Lymphoid precursors (B-LP) and CD19+ B lymphocytes. The mean ± SD percentage of B Lymphocytes and B-LPs in the PGF cohort were 6.76% ± 0.096% and 0.39% ± 0.005% respectively, while in the control group they were 20.09% ± 0.136% and 1.17% ± 0.012% respectively. When the 50 th centile value of the B-LPs (0.375%) or CD19+ B cells (10.89%) was investigated as a cut off to predict the clinical syndrome, this strategy achieved a 75% success rate with 3 PGF (blue bars) or control (black bars) samples misclassified (Fig1). Next, we investigated if these HSPC fractions could help predict the clinical outcome of PGF cases. The cases A2 and A4 who demonstrated no count recovery and needed CD34 boost lie in the lower quartile of both B Lymphocytes as well as B Lymphoid precursors content (Fig 1 red arrows). Conversely the three PGF cases in the upper 50 centile of these cell populations (patient B3, C4 and C9) all showed either a spontaneous or EPAG induced recovery of counts. We further investigated if the proportion of immunophenotypic CD34+CD38-CD45RA-CD90+CD49f+ long-term HSC (LT-HSC) could provide additional prognostic information (Fig2). Again, patients A2 and A4 had LT-HSC in the lower 50 th centile. Patient C1 who showed a spontaneous recovery of counts without the need of EPAG had the highest proportion of LT-HSCs despite lying in the second quartile of the cohort for both B cells as well as B-LPs. In conclusion, MP-FCM based analysis of the HSPC landscape in PGF marrow could be a robust prognostic biomarker in Allo-SCT patients.

PATH-25. EXPLORING THE PROGNOSTIC IMPLICATIONS OF MOLECULAR ALTERATIONS IN ASTROCYTOMAS USING COMPREHENSIVE GENOMIC PROFILING (CGP)

Abstract BACKGROUND Astrocytomas display molecular heterogeneity with prognostic implications for genomic alterations, including IDH, EGFR, TERT promoter and CDKN2A/B. Increasing access to comprehensive genomic profiling (CGP) provides an opportunity to characterize the genomic landscape of these tumors and their clinical implications. METHODS Patients with astrocytomas undergoing CGP through the Australian Molecular Screening and Therapeutics (MoST) precision oncology program between March 2017 and January 2022 were identified; the frequencies of genomic alterations were assessed. Associations between overall survival (OS) and alterations in EGFR, TERT promoter and/or CDKN2A/B were estimated using Kaplan-Meier methods. RESULTS 214 patients with astrocytomas were included. Median age was 49 years; majority were ECOG performance status 0-1 at screening (N=200/214, 93%) and histologically grade 4 (N=184/214, 86%). The median OS (mOS) for the cohort was 26.2mo (95%CI 24.0-31.5). For histologically grade 2/3 IDH-mutant tumors, mOS was 31.3mo (N=1) and 122.6mo (N=11, 95%CI 110-NA) for those with or without alterations in EGFR, TERT promoter and/or CDKN2A/B, respectively. For histologically grade 2/3 IDH-wildtype tumors, mOS was 19.9mo (N=7, 95%CI 15.8-NA) and 118.5mo (N=10, 95%CI 77.3-NA), respectively. For histologically grade 4 IDH-mutant tumors, mOS was 81.3mo (N=4, 95%CI 32.7-NA) and 82.8mo (N=15, 95%CI 71.0-NA), respectively. For histologically grade 4 IDH-wildtype tumors, mOS was 19.6mo (N=120 [56% TERT promoter, 42% EGFR, 14% CDKN2A/B], 95%CI 18.0-24.8) compared to 32.1mo (N=45, 95%CI 22.0-58.6; logrank P=0.0015), respectively. In grade 4 IDH-wildtype tumors additional co-occurring alterations involved PTEN (39%), TP53 (33%), NF1 (21%), RB1 (13%), CDK4 (13%) and PIK3CA (5%) genes. For the overall cohort, potentially actionable variants were seen in BRAF (5%), KIT (5%), MSH2 (2%), FGFR1 (1%) and FGFR3 (4%). Tumor mutational burden was ≥5 mutations/ megabase in 14%, without significant OS impact. CONCLUSION CGP provides additional prognostic information that complements histological grading. This demonstrates an emerging role in understanding the molecular heterogeneity of astrocytomas.

Serum cholinesterase level provides the additional long-time prognostic information over AHEAD risk score in patients with acute decompensated heart failure with preserved ejection fraction

Abstract Background and purpose In patients with heart failure with preserved ejection fraction (HFpEF), comorbidities strongly influence the prognosis. AHEAD (A: atrial fibrillation; H: hemoglobin; E: elderly; A: abnormal renal parameters; D: diabetes mellitus) score is a simple risk model based on several comorbidities. Its prognostic role has been reported in patients with HFpEF. Although malnutrition is one of the most important comorbidities, it is not included in AHEAD score. We previously reported the prognostic value of cholinesterase (CHE), which is a simple nutritional biomarker in daily practice, in patients with HFpEF. Here, we aimed to investigate whether serum CHE level provides the additional prognostic information over AHEAD risk score in patients with HFpEF. Methods We studied 888 patients admitted for acute decompensated heart failure (ADHF) in our multicenter-prospective HFpEF cohort study (PURSUIT-HFpEF registry). Laboratory data including CHE was obtained at discharge. We evaluated AHEAD risk score (range 0-5, atrial fibrillation, hemoglobin &amp;lt;13mg/dL for men and 12mg/dL for women, age &amp;gt;70years, creatinine &amp;gt;130μmol/L, and diabetes mellitus).The endpoint of this study was set as the composite of worsening heart failure readmission and all-cause mortality. Results During a follow up period of 1.6±1.3 years, 428 patients experienced primary composite endpoint. On multivariate Cox analysis, CHE (p&amp;lt;0.001) and AHEAD risk score (p = 0.003) were independently and significantly associated with primary composite endpoint. Kaplan-Meier analysis that patients with higher AHEAD risk score (defined as 4-5 point) and those with intermediate AHEAD score (defined as 2-3 point) had significantly greater risk of primary composite endpoint than those with low AHEAD score (defined as 0-1 point) (67% vs 54% vs 37%, log-rank p&amp;lt;0.001). Kaplan-Meier analysis in the subgroup with low AHEAD risk score revealed that primary composite event was significantly more frequently observed in patients with low CHE (&amp;lt;207 IU/L determined by the median of the whole cohort) (56% vs 28% p = 0.009). Moreover, in the subgroup with intermediate AHEAD risk score, the incidence of primary composite event was also significantly higher in patients with low CHE (56% vs 37% p&amp;lt;0.001). Finally, in the subgroup with high AHEAD risk score, the incidence of primary outcome was also significantly higher in patients with low CHE (69% vs 48% p&amp;lt;0.001). Receiver operating curve analysis showed that the addition of CHE significantly improved the risk prediction of AHEAD score (AUC: 0.641 vs 0.573, p&amp;lt;0.001). Conclusion Malnutrition, as evaluated by serum cholinesterase level, provided the additional long-time prognostic information beyond AHEAD risk score in patients with ADHF-HFpEF.

High-sensitivity troponin-T concentrations and long-term risk of death in patients with acute pulmonary embolism

Abstract Background Cardiac troponin elevation is associated with adverse short-term prognosis in patients with pulmonary embolism (PE). However, it is unknown whether high-sensitivity troponin concentrations predict long-term outcomes and whether serial measurements offer additional prognostic information in this setting. Purpose To determine short- and long-term prognostic implications of high-sensitivity troponin-T (hsTnT) concentrations and their changes from baseline, in patients with acute PE. Methods Retrospective cohort study based on Danish national registries. We identified all patients discharged from the hospital with a PE from 2013 through 2019 and merged them with all records of at least one hsTnT (99th percentile upper reference limit, 13.5 ng/l) measurement obtained during the same hospitalization. Absolute and relative risks for death from any cause at days 0-30 and 31-365 were calculated through multivariable logistic regression with average treatment effect modeling. Moreover, we tested whether hsTnT changes (in those who had a second measurement available ≥1 hour and ≤7 hours after the first measurement) were associated with additional prognostic information. Results At least one hsTnT measurement was available in 4611 individuals. Median age (25th-75th percentile) was 70.8 (58.6-79.5) years, and 2312 (50.1%) were women. One-hundred and forty-one (3.1%) patients had a prior venous thromboembolic event, 731 (15.9%) had known coronary artery disease, and 1001 (21.7%) had a history of cancer. Median baseline hsTnT was 29 (25th-75th percentile, 14-79) ng/l, and 1082 (76.5%) had an elevated concentration. Four-hundred and thirty-one (9.3%) individuals had died at 30 days, while an additional 614 (14.7% of 30-day survivors) died between days 31-365. hsTnT concentration displayed a significant, non-linear association with death (P&amp;lt;0.001). Figure 1 shows the unadjusted 30-day risk of death stratified for hsTnT quartiles, while Figure 2 shows the risk at days 31-365 (with 0 on the x-axis representing day 30 from the index event). Upon multivariable adjustment (including absence or presence of acute cor pulmonale), the risk of death at both timepoints remained significantly higher among individuals in each of quartiles 3 and 4 versus quartile 1 (P&amp;lt;0.05), with quartile 2 showing a trend (P=0.06 and P=0.07 for short- and long-term death, respectively). In 1247 subjects with two hsTnT measurements available, absolute and relative changes from baseline did not appear to carry additional prognostic information, nor did switching from normal to elevated hsTnT or vice versa, though the latter was only observed in very few persons. Conclusions In patients with acute PE, elevated hsTnT concentrations were associated with both short- and long-term mortality, but unlike in acute coronary syndromes, the utility of drawing a second sample appeared limited.Figure 1Figure 2