Improved Classification Of Patients Research Articles

An exploratory data-driven approach to classify subgroups of patients with temporomandibular disorders based on pain mechanisms

Temporomandibular disorders (TMDs) are a common musculoskeletal condition, presenting treatment challenges due to their non-specific nature. Categorizing patients with TMDs into clusters based on neurobiological pain mechanisms could provide a promising approach to facilitate targeted treatments. This observational study (1) used a network analysis (NA) to explore the complexity of TMDs by investigating relationships among biopsychosocial variables, and (2) validated potential TMD subgroups based on mechanism-specific pain categories. One hundred and two patients with TMD were included. Biopsychosocial variables covered: general health, psychosocial features, TMD pain, and TMD characteristics. A NA evaluated the associations between variables and determined the role of each feature within the network. Hierarchical clustering was used to identify TMD subgroups. The NA revealed significant correlations primarily within the same feature domains, indicating a strong interplay between symptoms and psychological factors. Cluster analysis identified two subgroups driven by nociceptive and nociplastic pain mechanisms; the nociplastic group exhibited higher levels of anxiety, depression, pain catastrophization, central sensitization, pain duration, and more pain locations, along with poorer sleep quality, quality of life, and health status. In contrast, the nociceptive group exhibited restricted maximal mouth opening (MMO), heightened pain during TMJ palpation and mouth opening, and a greater positive response to manual therapy. Across all features, psychological factors, pain locations, and MMO primarily contributed to the separation of subgroups. By adopting a data-driven approach, these results support the significant role of considering the neurobiological basis of pain to improve patient classification. This knowledge may facilitate clinical reasoning and personalized treatments. PerspectiveThis study used a network analysis to explore the complex biopsychosocial interactions present in people with TMDs, identifying important variables such as the Central Sensitization Inventory and pain-free maximal mouth opening. The findings distinguish potential nociceptive and nociplastic pain subgroups, offering important insights for targeted therapeutic strategies.

Optimizing Patient Stratification in Healthcare: A Comparative Analysis of Clustering Algorithms for EHR Data

Advanced data analytics are increasingly being employed in healthcare research to improve patient classification and personalize medicinal therapies. In this paper, we focus on the critical problem of clustering electronic health record (EHR) data to enable appropriate patient categorization. In the era of personalized medicine, optimizing patient classification is critical to healthcare analytics. This research presents a comparative assessment of different clustering algorithms for Electronic Health Record (EHR) data, with the goal of improving the efficacy and productivity of patient clustering methods. Our study focuses on Fuzzy Technique for Order of Preference by Similarity to Ideal Solution (Fuzzy TOPSIS) as a Multi-Criteria Decision-Making (MCDM) strategy, includes an in-depth assessment of eight clustering algorithms: K-Means, DBSCAN, Hierarchical Clustering, Mean Shift, Affinity Propagation, Spectral Clustering, Gaussian Mixture Models (GMM), as well as Self-Organizing Maps. The evaluation factors used for evaluation in this research are Cluster Quality Metrics, Scalability, Robustness to Noise, Cluster Shape and Density, Interpretability, Cluster Number, Dimensionality, and Consistency and Stability. These criteria and alternatives were chosen after conducting a thorough assessment of the literature and consulting with domain experts. All participated specialists actively engaged in the decision-making process, bringing unique insights into the best clustering algorithms for healthcare data. The results of this study illustrate each algorithm’s strengths and weaknesses in the setting of patient stratification, providing insight into their performance across multiple dimensions. The fuzzy TOPSIS MCDM strategy is a reliable instrument for synthesizing expert opinions and methodically evaluating the found clustering alternatives. This study advances healthcare analytics by giving practitioners and researchers with informative perspectives on the selection of clustering algorithms designed to address the unique problems of patient stratification utilizing EHR data.

Prognostic Value of a Multi-mRNA Signature for 1-Year All-Cause Death in Hospitalized Patients With Heart Failure With a Preserved Ejection Fraction.

Heart failure with preserved ejection fraction is a major global public health problem, while effective risk stratification tools are still lacking. We sought to construct a multi-mRNA signature to predict 1-year all-cause death. We selected 30 patients with heart failure with preserved ejection fraction who died during 1-year follow-up and 30 who survived in the discovery set. One hundred seventy-one and 120 patients with heart failure with preserved ejection fraction were randomly selected as a test set and a validation set, respectively. We performed mRNA microarrays in all patients. We constructed a 5-mRNA signature for predicting 1-year all-cause death. The scores of the 5-mRNA signature were significantly associated with the 1-year risk of all-cause death in both the test set (hazard ratio, 2.72 [95% CI, 1.98-3.74]; P<0.001) and the validation set (hazard ratio, 3.95 [95% CI, 2.40-6.48]; P<0.001). Compared with a reference model, which included sex, ASCEND-HF (Acute Study of Clinical Effectiveness of Nesiritide in Decompensated Heart Failure) score, history of HF and NT-proBNP (N-terminal pro-B-type natriuretic peptide), the 5-mRNA signature had a better discrimination capability, with an increased area under the curve from 0.696 to 0.813 in the test set and from 0.712 to 0.848 in the validation set. A composite model integrating the 5-mRNA risk score and variables in the reference model demonstrated an excellent discrimination capability, with an area under the curve of 0.861 (95% CI, 0.784-0.939) in the test set and an area under the curve of 0.859 (95% CI, 0.755-0.963) in the validation set. The net reclassification improvement and integrated discrimination improvement indicated that the composite model significantly improved patient classification compared with the reference model in both sets (P<0.001). The 5-mRNA signature is a promising predictive tool for 1-year all-cause death and shows improved prognostic power over the established risk scores and NT-proBNP in patients with heart failure with preserved ejection fraction.

Abstract 6843: Comprehensive multi-omics approaches for identification of genetic alterations in head and neck squamous cell carcinoma

Abstract Background: Head and neck squamous cell carcinoma (HNSCC) is one of the most difficult and challenging malignancies, with a high level of heterogeneity and a wide range of treatment responses, regardless of the clinical stage. Oral squamous cell carcinoma (OSCC) is the most common head and neck cancer and accounts for over 90% of cancers that develop in the mucosal epithelium of the oral cavity. Additional carcinogenic risk factors, including tobacco, alcohol, and HPV infection are also associated with the pathogenesis of oral carcinogenesis. Unfortunately, today the failure of standard treatment modalities such as surgery, radiotherapy, and chemotherapy underscore the need for comprehensive genomic profiling to provide a better picture of the genetic alterations and gain an improved understanding of the complex signaling networks involved in the development of potential treatment resistance as well as the identification of novel biomarkers that discriminate between smokers and non-smokers. Methods: A retrospective analysis was performed on biopsied Formalin-Fixed Paraffin-Embedded (FFPE) samples from OSCC patients with and without a history of smoking and treated +/- chemotherapy. Patient samples representing matched normal tissue to serve as controls were compared to tumor squamous cell carcinoma samples and profiled using a single assay panel that includes 517 genes by both DNA and RNA NGS, microsatellite instability (MSI), and tumor mutational burden (TMB). Further analysis was also performed using the NanoString nCounter® PanCancer Immune Profiling Panel (PCIP), 770-comprehensive gene expression panel Results: The Neo ComprehensiveTM Solid Tumor pan cancer assay enabled simultaneously analysis of both DNA and RNA in one integrated workflow with an accompanying DragenTM bioinformatics analysis platform and provided comprehensive genomic profiling of sequence and structural variants, as well as genomic signatures such as TMB and MSI in OSCC patients. Utility of the NanoString PCIP highlighted immune-relevant gene expression changes within the different sample cohorts. The detailed molecular information highlighted by our study emphasizes the need for improved patient classification based on smoking status in the management of OSCC and also in establishing potential therapeutic options based on the many altered cell signaling pathways that we identified. By combining a multi-omics approach with expression data for immune-related genes and associated overall survival data, we provide an effective strategy and workflow for the identification of prognostic biomarkers in OSCC. Citation Format: Kirsteen Maclean, Qinqin Zha, Lakshmi Chandramohan, Anna Juncker-Jensen. Comprehensive multi-omics approaches for identification of genetic alterations in head and neck squamous cell carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6843.

Biomarkers in Ovarian Cancer: Towards Personalized Medicine.

Ovarian cancer is one of the deadliest cancers in women. The lack of specific symptoms, especially at the initial stages of disease development, together with the malignancy heterogeneity, lower the life expectancy of patients. Aiming to improve survival rates, diagnostic and prognostic biomarkers are increasingly employed in clinics, providing gynecologists and oncologists with new tools to guide their treatment decisions. Despite the vast number of investigations, there is still an urgent need to discover more ovarian cancer subtype-specific markers which could further improve patient classification. To this end, high-throughput screening technologies, like mass spectrometry, are applied to deepen the tumoral cellular landscape and describe the malignant phenotypes. As for disease treatment, new targeted therapies, such as those based on PARP inhibitors, have shown great efficacy in destroying the tumoral cells. Likewise, drug-nanocarrier systems targeting the tumoral cells have exhibited promising results. In this narrative review, we summarize the latest achievements in the pursuit of biomarkers for ovarian cancer and recent anti-tumoral therapies.

Prehospital mSOFA Score for Quick Prediction of Life-Saving Interventions and Mortality in Trauma Patients: A Prospective, Multicenter, Ambulance-based, Cohort Study.

Background: Prehospital emergency medical services (EMS) are the main gateway for trauma patients. Recent advances in point-of-care testing and the development of early warning scores have allowed EMS to improve patient classification. We aimed to identify patients presenting with major trauma involving life-saving interventions (LSI) using the modified Sequential Organ Failure Assessment (mSOFA) score in the prehospital scenario, and to compare these results with those of other trauma scores. Methods: This was a prospective, ambulance-based, multicenter, training-validation study in trauma patients who were treated in a prehospital setting and subsequently transported to a hospital. The study involved six Advanced Life Support units, 38 Basic Life Support units, and four hospitals. The primary outcome was LSI performed at the scene or en route and intensive care unit (ICU) admission and all-cause two-day in-hospital mortality. We collected epidemiological variables, creatinine, lactate, base excess, international normalized ratio, and vital signs. Discriminative power (area under the receiver operating characteristic curve [AUC]), calibration (observed vs predicted outcome agreement), and decision-curve analysis (DCA, clinical utility) were used to assess the reliability of the mSOFA in comparison to other scores. Results: Between January 1, 2020-April 30, 2022, a total of 763 patients were selected. The mSOFA score's AUC was 0.927 (95% confidence interval [CI] 0.898-0.957) for LSI, 0.845 (95% CI 0.808-0.882) for ICU admission, and 0.979 (95% CI 0.966-0.991) for two-day mortality. Conclusion: The mSOFA score outperformed the other scores, allowing a quick identification of high-risk patients. The routine implementation in EMS of mSOFA could provide critical support in the decision-making process in time-dependent trauma injuries.

Peering through the PSMA PET Lens: The Role of the European Association of Urology Biochemical Recurrence Risk Groups after Radical Prostatectomy.

(1) Background: The European Association of Urology (EAU) biochemical recurrence (BCR) risk grouping relies on data from historical cohorts that used conventional imaging techniques. In the era of PSMA PET/CT, we compared the patterns of positivity in the two risk groups and provided insight into positivity predictive factors. (2) Methods: Data from 1185 patients who underwent 68Ga-PSMA-11PET/CT for BCR was analyzed, out of which 435 patients treated initially treated by radical prostatectomy were included in the final analysis. (3) Results: A significantly higher rate of positivity in the BCR high-risk group was observed (59% vs. 36%, p < 0.001). BCR low-risk group demonstrated more local (26% vs. 6%, p < 0.001) and oligometastatic (100% vs. 81%, p < 0.001) recurrences. The BCR risk group and PSA level at the time of PSMA PET/CT were independent predictive factors of positivity. (4) Conclusions: This study confirms that the EAU BCR risk groups have different rates of PSMA PET/CT positivity. Even with a lower rate in the BCR low-risk group, oligometastatic disease was 100% in those with distant metastases. Given the presence of discordant positivity and risk classification, integrating PSMA PET/CT positivity predictors into risk calculators for BCR might improve patient classification for subsequent treatment options. Future prospective studies are still needed to validate the above findings and assumptions.

The Utility of Nasal Challenges to Phenotype Asthma Patients.

Asthma is a heterogeneous disease in terms of both phenotype and response to therapy. Therefore, there is a great need for clinically applicable tools allowing for improved patient classification, and selection for specific management approaches. Some interventions are highly helpful in selected patients (e.g., allergen immunotherapy or aspirin desensitization), but they are costly and/or difficult to implement. Currently available biomarkers measurable in peripheral blood or exhaled air display many limitations for asthma phenotyping and cannot identify properly the specific triggers of the disease (e.g., aeroallergens or NSAID). The united airway concept illustrates the relevant epidemiological and pathophysiological links between the upper and lower airways. This concept has been largely applied to patient management and treatment, but its diagnostic implications have been less often explored. Of note, a recent document by the European Academy of Allergy and Clinical Immunology proposes the use of nasal allergen challenge to confirm the diagnosis of allergic asthma. Similarly, the nasal challenge with lysine acetylsalicylate (L-ASA) can be used to identify aspirin-sensitive asthma patients. In this review, we will summarize the main features of allergic asthma and aspirin-exacerbated respiratory disease and will discuss the methodology of nasal allergen and L-ASA challenges with a focus on their capacity to phenotype the inflammatory disease affecting both the upper and lower airways.

Prognostic Value of Multiple Circulating Biomarkers for 2-Year Death in Acute Heart Failure With Preserved Ejection Fraction.

Background: Heart failure with preserved ejection fraction (HFpEF) is increasingly recognized as a major global public health burden and lacks effective risk stratification. We aimed to assess a multi-biomarker model in improving risk prediction in HFpEF.Methods: We analyzed 18 biomarkers from the main pathophysiological domains of HF in 380 patients hospitalized for HFpEF from a prospective cohort. The association between these biomarkers and 2-year risk of all-cause death was assessed by Cox proportional hazards model. Support vector machine (SVM), a supervised machine learning method, was used to develop a prediction model of 2-year all-cause and cardiovascular death using a combination of 18 biomarkers and clinical indicators. The improvement of this model was evaluated by c-statistics, net reclassification improvement (NRI), and integrated discrimination improvement (IDI).Results: The median age of patients was 71-years, and 50.5% were female. Multiple biomarkers independently predicted the 2-year risk of death in Cox regression model, including N-terminal pro B-type brain-type natriuretic peptide (NT-proBNP), high-sensitivity cardiac troponin T (hs-TnT), growth differentiation factor-15 (GDF-15), tumor necrosis factor-α (TNFα), endoglin, and 3 biomarkers of extracellular matrix turnover [tissue inhibitor of metalloproteinases (TIMP)-1, matrix metalloproteinase (MMP)-2, and MMP-9) (FDR < 0.05). The SVM model effectively predicted the 2-year risk of all-cause death in patients with acute HFpEF in training set (AUC 0.834, 95% CI: 0.771–0.895) and validation set (AUC 0.798, 95% CI: 0.719–0.877). The NRI and IDI indicated that the SVM model significantly improved patient classification compared to the reference model in both sets (p < 0.05).Conclusions: Multiple circulating biomarkers coupled with an appropriate machine-learning method could effectively predict the risk of long-term mortality in patients with acute HFpEF. It is a promising strategy for improving risk stratification in HFpEF.

Abstract 9731: Improved Risk Stratification with Vendor Independent Coronary Calcium Scores: Data from the Multi-Ethnic Study of Atherosclerosis (mesa)

Introduction: Coronary calcium score (CCS) plays a role in cardiovascular disease (CVD) risk calculators, but also reclassifies patients when considering statin therapy. However, CCS is still calculated following the Agatston methodology as developed on electron beam tomography (EBT) in the early 1990s. EBT has been replaced by multidetector computed tomography (MDCT). It is known that CCS differs substantially between MDCT-systems from different manufacturers, which can lead to risk reclassifications for 6.5% of examined individuals Aims: Our first aim was to identify CT system specific conversion factors to calculate a vendor-neutral Agatston score (rCCS). Second, we aimed to validate rCCS risk stratification for participants of the Multi-Ethnic Study of Atherosclerosis (MESA). Methods: As a first step, we calculated conversion factors between six state-of the-art CT systems and the reference EBT system used in the MESA study. For this purpose, we scanned two calcium-containing inserts, which were placed in a small and large anthropomorphic thorax phantom. Linear regression was used to find a model which converts vendor-specific CCS to vendor-independent rCCS. As a second step, we recalculated the non-zero Agatston scores of 550 MESA participants. Following the Youden method, we calculated an optimal cut-off point for rCCS. Results: The linear model for each CT system allowed for prediction of Agatston scores as measured on the EBT system (p=0.001). rCCS for GE LightSpeed, used in MESA, was calculated as rCCS=0.049+1.97xMESA Agatston score and rCCS=1.903+1.37xMESA Agatston score for small and large size, respectively. The median MESA Agatston score was 100.5 (26.3-379.5), and the median rCCS was 135.8 (35.6-508.6). Participants were assigned to a high Agatston category with Agatston scores exceeding 100 (50.2%) and 187 (43.1%) for CCS and rCCS, respectively. The hazard ratio for CVD of individuals with high Agatston scores increased from 1.88 (95%CI: 1.31-2.68, p=0.001) with MESA scores to 2.15 (95%CI: 1.52-3.05, p=0.001) with rCCS. rCCS reclassified 39 participants into the lower risk group, without statin therapy. Conclusions: In conclusion, based on these preliminary results, rCCS may improve patient classification for statin therapy.

T-Cell Specificity Influences Disease Heterogeneity in Multiple Sclerosis.

Background and ObjectivesEncouraged by the enormous progress that the identification of specific autoantigens added to the understanding of neurologic autoimmune diseases, we undertook here an in-depth study of T-cell specificities in the autoimmune disease multiple sclerosis (MS), for which the spectrum of responsible autoantigens is not fully defined yet. The identification of target antigens in MS is crucial for therapeutic strategies aimed to induce antigen-specific tolerance. In addition, knowledge of relevant T-cell targets can improve our understanding of disease heterogeneity, a hallmark of MS that complicates clinical management.MethodsThe proliferative response and interferon gamma (IFN-γ) release of CSF-infiltrating CD4+ T cells from patients with MS against several autoantigens was used to identify patients with different intrathecal T-cell specificities. Fresh CSF-infiltrating and paired circulating lymphocytes in these patients were characterized in depth by ex vivo immunophenotyping and transcriptome analysis of relevant T-cell subsets. Further examination of these patients included CSF markers of inflammation and neurodegeneration and a detailed characterization with respect to demographic, clinical, and MRI features.ResultsBy testing CSF-infiltrating CD4+ T cells from 105 patients with MS against seven long-known myelin and five recently described GDP-l-fucose synthase peptides, we identified GDP-l-fucose synthase and myelin oligodendrocyte glycoprotein (35-55) responder patients. Immunophenotyping of CSF and paired blood samples in these patients revealed a significant expansion of an effector memory (CCR7− CD45RA−) CD27− Th1 CD4+ cell subset in GDP-l-fucose synthase responders. Subsequent transcriptome analysis of this subset demonstrated expression of Th1 and cytotoxicity-associated genes. Patients with different intrathecal T-cell specificities also differ regarding inflammation- and neurodegeneration-associated biomarkers, imaging findings, expression of HLA class II alleles, and seasonal distribution of the time of the lumbar puncture.DiscussionOur observations reveal an association between autoantigen reactivity and features of disease heterogeneity that strongly supports an important role of T-cell specificity in MS pathogenesis. These data have the potential to improve patient classification in clinical practice and to guide the development of antigen-specific tolerization strategies.

Higher-order connections between stereotyped subsets: implications for improved patient classification in CLL

Chronic lymphocytic leukemia (CLL) is characterized by the existence of subsets of patients with (quasi)identical, stereotyped B-cell receptor (BcR) immunoglobulins. Patients in certain major stereotyped subsets often display remarkably consistent clinicobiological profiles, suggesting that the study of BcR immunoglobulin stereotypy in CLL has important implications for understanding disease pathophysiology and refining clinical decision-making. Nevertheless, several issues remain open, especially pertaining to the actual frequency of BcR immunoglobulin stereotypy and major subsets, as well as the existence of higher-order connections between individual subsets. To address these issues, we investigated clonotypic IGHV-IGHD-IGHJ gene rearrangements in a series of 29 856 patients with CLL, by far the largest series worldwide. We report that the stereotyped fraction of CLL peaks at 41% of the entire cohort and that all 19 previously identified major subsets retained their relative size and ranking, while 10 new ones emerged; overall, major stereotyped subsets had a cumulative frequency of 13.5%. Higher-level relationships were evident between subsets, particularly for major stereotyped subsets with unmutated IGHV genes (U-CLL), for which close relations with other subsets, termed “satellites,” were identified. Satellite subsets accounted for 3% of the entire cohort. These results confirm our previous notion that major subsets can be robustly identified and are consistent in relative size, hence representing distinct disease variants amenable to compartmentalized research with the potential of overcoming the pronounced heterogeneity of CLL. Furthermore, the existence of satellite subsets reveals a novel aspect of repertoire restriction with implications for refined molecular classification of CLL.

Evaluation of blood gene expression levels in facioscapulohumeral muscular dystrophy patients

Facioscapulohumeral muscular dystrophy (FSHD) is caused by the expression of DUX4 in skeletal muscles. A number of therapeutic approaches are being developed to antagonize the events preceding and following DUX4 expression that leads to muscular dystrophy. Currently, the possibility to evaluate treatment response in clinical trials is hampered by the lack of objective molecular biomarkers connecting the disease cause to clinical performance. In this study we employed RNA-seq to examine gene expression in PAXgene tubes obtained from two independent cohorts of FSHD patients. Analysis of gene expression profiles did not lead to the identification of genes or pathways differentially expressed in FSHD patients, or associated with disease severity. In particular, we did not find evidence that the DUX4 and PAX7 signatures were differentially expressed. On the other hand, we were able to improve patient classification by including single genes or groups of genes in classification models. The best classifier was ROPN1L, a gene known to be expressed in testis, coincidentally the typical location of DUX4 expression. These improvements in patient classification hold the potential to enrich the FSHD clinical trial toolbox.

Neuronal pentraxin receptor-1 is a new cerebrospinal fluid biomarker of Alzheimer’s disease progression

Background: Alzheimer’s disease (AD) is the most common type of dementia, with progressive onset of clinical symptoms. The main pathological hallmarks are brain deposits of extracellular amyloid beta plaques and intracellular neurofibrillary tangles (NFT). Cerebrospinal fluid reflects pathological changes in the brain; amyloid beta 1-42 is a marker of amyloid plaques, while total and phosphorylated tau are markers of NFT formation. Additional biomarkers associated with disease pathogenesis are needed, for better prognosis, more specific diagnosis, prediction of disease severity and progression and for improved patient classification in clinical trials. The aim of the present study was to evaluate brain-specific proteins as potential biomarkers of progression of AD. Methods: Overall, 30 candidate proteins were quantified in cerebrospinal fluid (CSF) samples from patients with mild cognitive impairment (MCI) and mild, moderate and severe AD dementia (n=101) using mass spectrometry-based selected reaction monitoring assays. ELISA was used for neuronal pentraxin receptor-1 (NPTXR) confirmation. Results: The best discrimination between MCI and more advanced AD stages (moderate and severe dementia) was observed for protein NPTXR (area under the curve, AUC=0.799). A statistically different abundance of this protein was observed between the two groups, with severe AD patients having progressively lower levels (p<0.05). ELISA confirmed lower levels in AD, in a separate cohort that included controls, MCI and AD patients. Conclusions: We conclude that NPTXR protein in CSF is a novel potential biomarker of AD progression and could have important utility in assessing treatment success in clinical trials.

Validating a breast cancer score in Spanish women. The MCC-Spain study

A breast-risk score, published in 2016, was developed in white-American women using 92 genetic variants (GRS92), modifiable and non-modifiable risk factors. With the aim of validating the score in the Spanish population, 1,732 breast cancer cases and 1,910 controls were studied. The GRS92, modifiable and non-modifiable risk factor scores were estimated via logistic regression. SNPs without available genotyping were simulated as in the aforementioned 2016 study. The full model score was obtained by combining GRS92, modifiable and non-modifiable risk factor scores. Score performances were tested via the area under the ROC curve (AUROC), net reclassification index (NRI) and integrated discrimination improvement (IDI). Compared with non-modifiable and modifiable factor scores, GRS92 had higher discrimination power (AUROC: 0.6195, 0.5885 and 0.5214, respectively). Adding the non-modifiable factor score to GRS92 improved patient classification by 23.6% (NRI = 0.236), while the modifiable factor score only improved it by 7.2%. The full model AUROC reached 0.6244. A simulation study showed the ability of the full model for identifying women at high risk for breast cancer. In conclusion, a model combining genetic and risk factors can be used for stratifying women by their breast cancer risk, which can be applied to individualizing genetic counseling and screening recommendations.

Prognostic Value of Secretoneurin in Patients with Acute Respiratory Failure: Data from the FINNALI Study.

We examined whether secretoneurin (SN), a biomarker associated with cardiomyocyte Ca(2+) handling, provides prognostic information in patients with acute respiratory failure (ARF). We included 490 patients with ARF, defined as ventilatory support >6 h, with blood samples available on admission to the intensive care unit (ICU). SN concentrations were measured by RIA. A total of 209 patients (43%) were hospitalized with cardiovascular (CV)-related ARF, and 90-day mortality rates were comparable between CV- and non-CV-related ARF (n = 281): 31% vs 24%, P = 0.11. Admission SN concentrations were higher in nonsurvivors than in survivors in both CV-related (median 148 [quartile 1-3, 117-203] vs 108 [87-143] pmol/L, P < 0.001) and non-CV-related ARF (139 [115-184] vs 113 [91-139] pmol/L, P < 0.001). In patients with CV-related ARF, SN concentrations on ICU admission were associated with 90-day mortality [odds ratio (OR) 1.97 (95% CI, 1.04-3.73, P = 0.04)] after adjusting for established risk indices, including N-terminal-pro-B-type natriuretic peptide (NT-proBNP) concentrations. SN also improved patient classification in CV-related ARF as assessed by the net reclassification index: 0.32 (95% CI, 0.04-0.59), P = 0.03. The area under the curve (AUC) of SN to predict mortality in patients with CV-related ARF was 0.72 (95% CI, 0.65-0.79), and the AUC of NT-proBNP was 0.64 (0.56-0.73). In contrast, SN concentrations on ICU admission did not provide incremental prognostic value to established risk indices in patients with non-CV-related ARF, and the AUC was 0.67 (0.60-0.75). SN concentrations measured on ICU admission provided incremental prognostic information to established risk indices in patients with CV-related ARF, but not in patients with non-CV-related ARF.

Prognostic Value of Secretoneurin in Critically Ill Patients With Infections

Secretoneurin is produced in neuroendocrine cells, and the myocardium and circulating secretoneurin levels provide incremental prognostic information to established risk indices in cardiovascular disease. As myocardial dysfunction contributes to poor outcome in critically ill patients, we wanted to assess the prognostic value of secretoneurin in two cohorts of critically ill patients with infections. Two prospective, observational studies. Twenty-four and twenty-five ICUs in Finland. A total of 232 patients with severe sepsis (cohort #1) and 94 patients with infections and respiratory failure (cohort #2). None. We measured secretoneurin levels by radioimmunoassay in samples obtained early after ICU admission and compared secretoneurin with other risk indices. In patients with severe sepsis, admission secretoneurin levels (logarithmically transformed) were associated with hospital mortality (odds ratio, 3.17 [95% CI, 1.12-9.00]; p = 0.030) and shock during the hospitalization (odds ratio, 2.17 [1.06-4.46]; p = 0.034) in analyses that adjusted for other risk factors available on ICU admission. Adding secretoneurin levels to age, which was also associated with hospital mortality in the multivariate model, improved the risk prediction as assessed by the category-free net reclassification index: 0.35 (95% CI, 0.06-0.64) (p = 0.02). In contrast, N-terminal pro-B-type natriuretic peptide levels were not associated with mortality in the multivariate model that included secretoneurin measurements, and N-terminal pro-B-type natriuretic peptide did not improve patient classification on top of age. Secretoneurin levels were also associated with hospital mortality after adjusting for other risk factors and improved patient classification in cohort #2. In both cohorts, the optimal cutoff for secretoneurin levels at ICU admission to predict hospital mortality was ≈ 175 pmol/L, and higher levels were associated with mortality also when adjusting for Simplified Acute Physiology Score II and Sequential Organ Failure Assessment scores. Secretoneurin levels provide incremental information to established risk indices for the prediction of mortality and shock in critically ill patients with severe infections.

Predicting 5-Year Survival Status of Patients with Breast Cancer based on Supervised Wavelet Method

ObjectivesClassification of breast cancer patients into different risk classes is very important in clinical applications. It is estimated that the advent of high-dimensional gene expression data could improve patient classification. In this study, a new method for transforming the high-dimensional gene expression data in a low-dimensional space based on wavelet transform (WT) is presented. MethodsThe proposed method was applied to three publicly available microarray data sets. After dimensionality reduction using supervised wavelet, a predictive support vector machine (SVM) model was built upon the reduced dimensional space. In addition, the proposed method was compared with the supervised principal component analysis (PCA). ResultsThe performance of supervised wavelet and supervised PCA based on selected genes were better than the signature genes identified in the other studies. Furthermore, the supervised wavelet method generally performed better than the supervised PCA for predicting the 5-year survival status of patients with breast cancer based on microarray data. In addition, the proposed method had a relatively acceptable performance compared with the other studies. ConclusionThe results suggest the possibility of developing a new tool using wavelets for the dimension reduction of microarray data sets in the classification framework.

Biological clustering supports both “Dutch” and “British” hypotheses of asthma and chronic obstructive pulmonary disease

BackgroundAsthma and chronic obstructive pulmonary disease (COPD) are heterogeneous diseases.ObjectiveWe sought to determine, in terms of their sputum cellular and mediator profiles, the extent to which they represent distinct or overlapping conditions supporting either the “British” or “Dutch” hypotheses of airway disease pathogenesis.MethodsWe compared the clinical and physiological characteristics and sputum mediators between 86 subjects with severe asthma and 75 with moderate-to-severe COPD. Biological subgroups were determined using factor and cluster analyses on 18 sputum cytokines. The subgroups were validated on independent severe asthma (n = 166) and COPD (n = 58) cohorts. Two techniques were used to assign the validation subjects to subgroups: linear discriminant analysis, or the best identified discriminator (single cytokine) in combination with subject disease status (asthma or COPD).ResultsDiscriminant analysis distinguished severe asthma from COPD completely using a combination of clinical and biological variables. Factor and cluster analyses of the sputum cytokine profiles revealed 3 biological clusters: cluster 1: asthma predominant, eosinophilic, high TH2 cytokines; cluster 2: asthma and COPD overlap, neutrophilic; cluster 3: COPD predominant, mixed eosinophilic and neutrophilic. Validation subjects were classified into 3 subgroups using discriminant analysis, or disease status with a binary assessment of sputum IL-1β expression. Sputum cellular and cytokine profiles of the validation subgroups were similar to the subgroups from the test study.ConclusionsSputum cytokine profiling can determine distinct and overlapping groups of subjects with asthma and COPD, supporting both the British and Dutch hypotheses. These findings may contribute to improved patient classification to enable stratified medicine.

Cancer diagnosis by nuclear morphometry using spatial information

Methods for extracting quantitative information regarding nuclear morphology from histopathology images have been long used to aid pathologists in determining the degree of differentiation in numerous malignancies. Most methods currently in use, however, employ the naïve Bayes approach to classify a set of nuclear measurements extracted from one patient. Hence, the statistical dependency between the samples (nuclear measurements) is often not directly taken into account. Here we describe a method that makes use of statistical dependency between samples in thyroid tissue to improve patient classification accuracies with respect to standard naïve Bayes approaches. We report results in two sample diagnostic challenges.