Proteomic Classifier Research Articles

Machine Learning of Plasma Proteomics Classifies Diagnosis of Interstitial Lung Disease.

Rationale: Distinguishing connective tissue disease-associated interstitial lung disease (CTD-ILD) from idiopathic pulmonary fibrosis (IPF) can be clinically challenging. Objectives: To identify proteins that separate and classify patients with CTD-ILD and those with IPF. Methods: Four registries with 1,247 patients with IPF and 352 patients with CTD-ILD were included in analyses. Plasma samples were subjected to high-throughput proteomics assays. Protein features were prioritized using recursive feature elimination to construct a proteomic classifier. Multiple machine learning models, including support vector machine, LASSO (least absolute shrinkage and selection operator) regression, random forest, and imbalanced Random Forest, were trained and tested in independent cohorts. The validated models were used to classify each case iteratively in external datasets. Measurements and Main Results: A classifier with 37 proteins (proteomic classifier 37 [PC37]) was enriched in the biological process of bronchiole development and smooth muscle proliferation and immune responses. Four machine learning models used PC37 with sex and age score to generate continuous classification values. Receiver operating characteristic curve analyses of these scores demonstrated consistent areas under the curve of 0.85-0.90 in the test cohort and 0.94-0.96 in the single-sample dataset. Binary classification demonstrated 78.6-80.4% sensitivity and 76-84.4% specificity in the test cohort and 93.5-96.1% sensitivity and 69.5-77.6% specificity in the single-sample classification dataset. Composite analysis of all machine learning models confirmed 78.2% (194 of 248) accuracy in the test cohort and 82.9% (208 of 251) in the single-sample classification dataset. Conclusions: Multiple machine learning models trained with large cohort proteomic datasets consistently distinguished CTD-ILD from IPF. Many of the identified proteins are involved in immune pathways. We further developed a novel approach for single-sample classification, which could facilitate honing the differential diagnosis of ILD in challenging cases and improve clinical decision making.

A proteomic classifier panel for early screening of colorectal cancer: a case control study

BackgroundDiagnosis of colorectal cancer (CRC) during early stages can greatly improve patient outcome. Although technical advances in the field of genomics and proteomics have identified a number of candidate biomarkers for non-invasive screening and diagnosis, developing more sensitive and specific methods with improved cost-effectiveness and patient compliance has tremendous potential to help combat the disease.MethodsWe enrolled three cohorts of 479 subjects, including 226 CRC cases, 197 healthy controls, and 56 advanced precancerous lesions (APC). In the discovery cohort, we used quantitative mass spectrometry to measure the expression profile of plasma proteins and applied machine-learning to select candidate proteins. We then developed a targeted mass spectrometry assay to measure plasma concentrations of seven proteins and a logistic regression classifier to distinguish CRC from healthy subjects. The classifier was further validated using two independent cohorts.ResultsThe seven-protein panel consisted of leucine rich alpha-2-glycoprotein 1 (LRG1), complement C9 (C9), insulin-like growth factor binding protein 2 (IGFBP2), carnosine dipeptidase 1 (CNDP1), inter-alpha-trypsin inhibitor heavy chain 3 (ITIH3), serpin family A member 1 (SERPINA1), and alpha-1-acid glycoprotein 1 (ORM1). The panel classified CRC and healthy subjects with high accuracy, since the area under curve (AUC) of the training and testing cohort reached 0.954 and 0.958. The AUC of the two independent validation cohorts was 0.905 and 0.909. In one validation cohort, the panel had an overall sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of 89.9%, 81.8%, 89.2%, and 82.9%, respectively. In another blinded validation cohort, the panel classified CRC from healthy subjects with a sensitivity of 81.5%, specificity of 97.9%, and overall accuracy of 92.0%. Finally, the panel was able to detect APC with a sensitivity of 49%.ConclusionsThis seven-protein classifier is a clear improvement compared to previously published blood-based protein biomarkers for detecting early-stage CRC, and is of translational potential to develop into a clinically useful assay.

Role of CKD273 in Diabetic Nephropathy: A Systematic Review

Introduction: Diabetic Nephropathy (DN) is a major cause of increased morbidity and mortality in patients with Diabetes Mellitus (DM). The development of novel biomarkers for kidney disease, especially DN, is becoming a focus of medical research. Serum creatinine and urine albumin measurements had been the gold standards for the diagnosis of DN until the advent of newer markers like Cystatin C, Kidney Injury Molecule-1, etc. Currently, there is no non invasive investigation for the prediction of DN. Nevertheless, the field of proteomics is advancing at a rapid pace, such that urinary proteomic biomarkers have found a spot in the diagnosis of DN. The role of the urinary proteomic Chronic Kidney Disease 273 (CKD273) classifier, in particular, has been demonstrated in early prediction. Aim: To highlight the performance of the CKD273 classifier in the early prediction of DN. Materials and Methods: Pubmed and Scopus databases were extensively searched using the keywords diabetes, DM, renal disease or kidney disease, proteomics or peptide, urinary markers or biomarkers. Full-text original articles conducted on humans and written in English, published between 2014 and 2022, specifically prospective and retrospective observational studies, crosssectional studies, Randomised Controlled Trials (RCT), and casecontrol studies investigating the efficacy of the urinary biomarker CKD273 in the prediction of DN were included in the review. The literature search, strictly based on Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines, led to the identification of eight relevant studies, which were included in the present review. The quality and Risk of Bias (RoB) of the selected articles were assessed using the Newcastle Ottowa scale (NOS) and Risk Of Bias 2 (ROB2) tool. Results: The selected studies proved the efficacy of the urinary proteomic classifier CKD273 as a cost-effective biomarker in predicting the decline in renal function in both normoalbuminuric diabetics and DN patients. Conclusion: Based on the articles included in the present systematic review, it has been found that CKD273 plays a vital role in the early prediction of renal dysfunction in normoalbuminuric diabetics as well as in the detection of progressing renal disease in DN patients. It has also been found to be cost-effective compared to established markers such as estimated Glomerular Filtration Rate (eGFR), Urine Albumin Excretion Rate (UAER), etc.

A pathway activity-based proteomic classifier stratifies prostate tumors into two subtypes

Prostate cancer (PCa) is the second most common cancer in males worldwide. The risk stratification of PCa is mainly based on morphological examination. Here we analyzed the proteome of 667 tumor samples from 487 Chinese PCa patients and characterized 9576 protein groups by PulseDIA mass spectrometry. Then we developed a pathway activity-based classifier concerning 13 proteins from seven pathways, and dichotomized the PCa patients into two subtypes, namely PPS1 and PPS2. PPS1 is featured with enhanced innate immunity, while PPS2 with suppressed innate immunity. This classifier exhibited a correlation with PCa progression in our cohort and was further validated by two published transcriptome datasets. Notably, PPS2 was significantly correlated with poor biochemical recurrence (BCR)/metastasis-free survival (log-rank P-value < 0.05). The PPS2 was also featured with cell proliferation activation. Together, our study presents a novel pathway activity-based stratification scheme for PCa.

Prognosis and Personalized In Silico Prediction of Treatment Efficacy in Cardiovascular and Chronic Kidney Disease: A Proof-of-Concept Study.

(1) Background: Kidney and cardiovascular diseases are responsible for a large fraction of population morbidity and mortality. Early, targeted, personalized intervention represents the ideal approach to cope with this challenge. Proteomic/peptidomic changes are largely responsible for the onset and progression of these diseases and should hold information about the optimal means of treatment and prevention. (2) Methods: We investigated the prediction of renal or cardiovascular events using previously defined urinary peptidomic classifiers CKD273, HF2, and CAD160 in a cohort of 5585 subjects, in a retrospective study. (3) Results: We have demonstrated a highly significant prediction of events, with an HR of 2.59, 1.71, and 4.12 for HF, CAD, and CKD, respectively. We applied in silico treatment, implementing on each patient's urinary profile changes to the classifiers corresponding to exactly defined peptide abundance changes, following commonly used interventions (MRA, SGLT2i, DPP4i, ARB, GLP1RA, olive oil, and exercise), as defined in previous studies. Applying the proteomic classifiers after the in silico treatment indicated the individual benefits of specific interventions on a personalized level. (4) Conclusions: The in silico evaluation may provide information on the future impact of specific drugs and interventions on endpoints, opening the door to a precision-based medicine approach. An investigation into the extent of the benefit of this approach in a prospective clinical trial is warranted.

Impact of an integrated classifier using biomarkers, clinical and imaging factors on clinical decisions making for lung nodules.

An integrated classifier that utilizes plasma proteomic biomarker along with five clinical and imaging factors was previously shown to be potentially useful in lung nodule evaluation. This study evaluated the impact of the integrated proteomic classifier on management decisions in patients with a pretest probability of cancer (pCA) ≤50% in "real-world" clinical setting. Retrospective study examining patients with lung nodules who were evaluated using the integrated classifier as compared to standard clinical care during the same period, with at least 1-year follow-up. A total of 995 patients were evaluated for lung nodules over 1 year following the implementation of the integrated classifier with 17.3% prevalence of lung cancer. 231 patients met the study eligibility criteria; 102 (44.2%) were tested with the integrated classifier, while 129 (55.8%) did not. The median number of chest imaging studies was 2 [interquartile range (IQR), 1-2] in the integrated classifier arm and 2 [IQR, 1-3] in the non-integrated classifier arm (P=0.09). The median outpatient clinic visit was 2.00 (IQR, 1.00-3.00) in the integrated classifier arm and 2.00 (IQR, 2.00-3.00) in the non-integrated classifier (P=0.004). Fewer invasive procedures were pursued in the integrated classifier arm as compared to non-integrated classifier respectively (26.5% vs. 79.1%, P<0.001). All patients in the integrated classifier arm with post-pCA (likely benign n=39) had designated benign diagnosis at 1-year follow-up. In patients with lung nodules with a pCA ≤50%, use of the integrated classifier was associated with fewer invasive procedures and clinic visits without misclassifying patients with likely benign lung nodules results at 1-year follow-up.

PREDICTION OF CORONARY ARTERY DISEASE USING URINARY PROTEOMICS

Objective: Coronary artery disease (CAD) is multifactorial, caused by complex pathophysiology, and contributes to a high burden of mortality worldwide. Urinary proteomic analyses may help to identify predictive biomarkers and provide insights into the pathogenesis of CAD. Design and method: Urinary proteome was analyzed in 965 participants using capillary electrophoresis coupled with mass spectrometry. A proteomic classifier was developed in a discovery cohort with 36 individuals with CAD and 36 matched controls using the support vector machine. The classifier was tested in a validation cohort with 115 individuals who progressed to CAD and 778 controls and compared with a previously developed CAD-associated classifier, ACSP75. The Framingham risk score was available in 737 participants. Bioinformatics analysis was performed based on the CAD-associated peptides. Results: The novel proteomic classifier was comprised of 160 urinary peptides, mainly related to collagen turnover, lipid metabolism, and inflammation. In the validation cohort, the classifier provided an AUC of 0.82 (95% CI: 0.78-0.87) for the CAD prediction in 8 years, superior to ACSP75 (AUC: 0.53, 95% CI: 0.47-0.60). On top of ACSP75, the addition of the novel classifier improved the AUC to 0.84 (95% CI: 0.80-0.89). In a multivariable Cox model, a 1-SD increment in the novel classifier was associated with CAD (HR: 1.54, 95% CI: 1.26-1.89, P&lt;0.0001) higher risk of CAD. The new classifier further improved the risk reclassification of CAD on top of the Framingham risk score (net reclassification index: 0.61, 95% CI: 0.25-0.95, P = 0.001). Conclusions: A novel urinary proteomic classifier related to collagen metabolism, lipids, and inflammation showed potential for the risk prediction of CAD. Urinary proteome provides an alternative approach to personalized prevention.

Prediction of coronary artery disease using urinary proteomics.

Coronary artery disease (CAD) is multifactorial, caused by complex pathophysiology, and contributes to a high burden of mortality worldwide. Urinary proteomic analyses may help to identify predictive biomarkers and provide insights into the pathogenesis of CAD. Urinary proteome was analysed in 965 participants using capillary electrophoresis coupled with mass spectrometry. A proteomic classifier was developed in a discovery cohort with 36 individuals with CAD and 36 matched controls using the support vector machine. The classifier was tested in a validation cohort with 115 individuals who progressed to CAD and 778 controls and compared with two previously developed CAD-associated classifiers, CAD238 and ACSP75. The Framingham and SCORE2 risk scores were available in 737 participants. Bioinformatic analysis was performed based on the CAD-associated peptides. The novel proteomic classifier was comprised of 160 urinary peptides, mainly related to collagen turnover, lipid metabolism, and inflammation. In the validation cohort, the classifier provided an area under the receiver operating characteristic curve (AUC) of 0.82 [95% confidence interval (CI): 0.78-0.87] for the CAD prediction in 8 years, superior to CAD238 (AUC: 0.71, 95% CI: 0.66-0.77) and ACSP75 (AUC: 0.53 and 95% CI: 0.47-0.60). On top of CAD238 and ACSP75, the addition of the novel classifier improved the AUC to 0.84 (95% CI: 0.80-0.89). In a multivariable Cox model, a 1-SD increment in the novel classifier was associated with a higher risk of CAD (HR: 1.54, 95% CI: 1.26-1.89, P < 0.0001). The new classifier further improved the risk reclassification of CAD on top of the Framingham or SCORE2 risk scores (net reclassification index: 0.61, 95% CI: 0.25-0.95, P = 0.001; 0.64, 95% CI: 0.28-0.98, P = 0.001, correspondingly). A novel urinary proteomic classifier related to collagen metabolism, lipids, and inflammation showed potential for the risk prediction of CAD. Urinary proteome provides an alternative approach to personalized prevention.

The urinary proteomic profile of fibromuscular dysplasia.

Fibromuscular dysplasia (FMD) is a non-atherosclerotic, non-inflammatory arterial disease of unclear origin, that may affect virtually all middle-sized arteries, mostly found in young or middle-aged women. More than half of patients have lesions in two or more arterial beds. While many cases are asymptomatic or associated with aspecific symptoms (mild hypertension, migraines, tinnitus), a number of patients develop severe complications such as resistant hypertension, carotid dissection, early stroke or mesenteric ischemia. Two subtypes of FMD have been described: focal and multifocal FMD. Focal FMD, characterized by a focal arterial stenosis of variable length is often difficult to differentiate from localized atherosclerosis, inflammatory or genetic arterial diseases. The hallmark of multifocal FMD, the so-called string of beads, is considered as almost pathognomonic. However, besides classical forms, many patients harbour less typical lesions such as presence of one or two non-circumferential beads, mild irregularities or less regular successive enlargements of arteries. Furthermore, lesions indistinguishable from string of beads are often found in extra-coronary arteries of patients who underwent a spontaneous coronary artery dissection (SCAD), leading some authors to consider SCAD as a form of FMD. Finally, a substantial proportion of patients with FMD harbour arterial aneurysms or dissections, and patients with such complications are sometimes considered to have atypical FMD, even in the absence of a string of beads or focal stenosis. In view of these elements, the development and validation of biomarkers, and particular proteomic biomarkers is of interest to unravel the pathophysiology, to facilitate diagnosis, improve classification and predict complications and evolution of FMD, SCAD and related diseases. Urinary proteomics appears as particularly promising in view of the predominantly renovascular expression of FMD, and of previous identification of proteomic signatures associated with both renal and cardiovascular diseases. In a recent pilot study, we have described a urinary proteomic classifier that allowed distinguishing with high accuracy patients with (mostly renal, multifocal) FMD, both from healthy controls and from patients with chronic kidney diseases and coronary heart disease. This classifier includes mostly collagen-derived peptides, which may suggest increased collagen turnover in affected arteries. The objectives of this presentation are to describe urinary proteomic profile associated with primary FMD, to report preliminary results obtained in patients with SCAD associated with FMD or not, and to discuss the potential consequences of such findings for the understanding, diagnosis and classification of these overlapping entities.

Abstract 662: Longitudinal blood-based proteomic testing in advanced non-small cell lung cancer

Abstract Motivation. A blood-based proteomic host immune classifier (HIC) that identifies a patient's disease state as Hot or Cold can aid prognostication and help guide treatment decisions in non-small cell lung cancer (NSCLC). The HIC Hot label is associated with 2-2.5 times longer overall survival (OS) depending on treatment type. While pretreatment HIC labels are informative, label flips over the course of therapy have been observed. We sought to quantify the prevalence of label flips, determine their impact on prognosis, and investigate the utility of longitudinal testing to monitor changes in disease state throughout therapy. Methods. In a real-world observational study (NCT03289780) enrolling over 3,700 patients with NSCLC at any stage and line of therapy to date, 267 of 1464 patients with advanced (Stage 3b/4) NSCLC with at least 1 year follow-up (f/u) had longitudinal testing at first f/u. Patient disease states were designated HH (Hot baseline and f/u), HC (Hot baseline, Cold f/u), CH (Cold baseline, Hot f/u) or CC (Cold baseline and f/u). Results. For patients receiving all therapy lines and regimens, 38 of 209 patients with baseline Hot test result (18%) flipped to Cold (HC), whereas 28 of 58 patients with baseline Cold, (48%) flipped (CH) (table). Patients with a baseline Hot result (HH+HC) had significantly longer OS compared to baseline Cold (CH+CC). Comparing patient groups based on f/u result (HH+CH vs. HC+CC) yielded an even greater separation. Similar trends were observed in the subgroup of patients receiving chemotherapy (n=124), with 18 patients flipping from Hot to Cold (20%), and 12 from Cold to Hot (36%) at f/u. Average time to f/u (TTF) was 6 months and did not vary significantly between groups. Conclusions. For some patients with NSCLC, disease states and prognosis can change in response to therapy. Longitudinal proteomic testing may be a viable method to monitor changes in disease state and host immune response to guide treatment. Survival of study subjectsGroupmOS [95% CI], monthsMean TTF (SD) [95% CI], monthsTotal (N=267)6.12 (3.70) [5.67-6.56]HH (N=171)NR [17.0-und]5.65 (3.50) [4.01-7.29]HC (N=38)11.6 [9.2-und]5.94 (3.31) [4.86-7.03]CH (N=20)NR [11.6-und]6.30 (3.91) [5.71-6.89]CC (N=38)10.2 [7.9-17.4]5.71 (3.19) [4.67-6.76]Total (N=267)HR [95% CI]P valueHH vs. HC1.81 [1.11-2.96]0.018HH vs. CH1.03 [0.47-2.24]0.946HH vs. CC2.56 [1.60-4.10]&amp;lt;0.0001HC vs. CH0.59 [0.25-1.40]0.234HC vs. CC1.40 [0.78-2.52]0.260CH vs. CC2.38 [1.02-5.54]0.045All Tx (N=267)mOS [95% CI], monthsHR [95% CI], P valueHH+HC (N=209)NR [16.6-und]0.59 [0.39-0.89], P=0.012CH+CC (N=58)12.9 [9.3-und]HH+CH (N=191)NR [17.5-und]0.47 [0.33-0.69], P&amp;lt;0.0001HC+CC (N=76)11.6 [9.3-16.4]Chemo (N=124)mOS [95% CI], monthsHR [95% CI], P valueHH+HC (N=91)NR [15.3-und]0.52 [0.31-0.88], P=0.015CH+CC (N=33)11.6 [7.9-17.9]HH+CH (N=85)NR [15.6-und]0.42 [0.25-0.70], P&amp;lt;0.001HC+CC (N=39)9.6 [7.9-16.4]All Tx, all treatment; Chemo, chemotherapy; CI, confidence interval; HR, hazard ratio; NR, not reached; SD, standard deviation; TTF, time to follow-up; und, undefined Citation Format: Eric Schaefer, R. Brian Mitchell, Jason Boyd, James Orsini, Nagaprasad Nagajothi, Savita Bidyasar, Mazen Khalil, Mitchell Haut, Ray Page, Kan Huang, John W. Dubay, Wallace Akerley. Longitudinal blood-based proteomic testing in advanced non-small cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 662.

A NOVEL URINARY BIOMARKER PREDICTS 1 YEAR MORTALITY AFTER DISCHARGE FROM INTENSIVE CARE

Objective: The urinary proteome reflects molecular drivers of disease. The aim of this study was to construct a urinary proteomic biomarker predicting 1 year post ICU mortality. Design and method: In 1243 patients, the urinary proteome was measured on ICU admission, using capillary electrophoresis couple with mass spectrometry along with clinical variables, circulating biomarkers (BNP, hsTnT, active ADM, and NGAL) and urinary albumin. Methods included support vector modelling to construct the classifier, Cox regression, the integrated discrimination (IDI) and net reclassification (NRI) improvement and area under the curve (AUC) to assess predictive accuracy, and Proteasix and protein-proteome interactome analyses. Results: In the discovery (deaths/survivors, 70/299) and test (175/699) datasets, the new classifier ACM128, mainly consisting of collagen fragments, yielding AUCs of 0.755 (95% CI, 0.708–0.798) and 0.688 (0.656–0.719), respectively. While accounting for study site, clinical risk factors, hazard ratios in 1243 patients were 2.41 (2.00–2.91) for ACM128 (+1 SD), 1.24 (1.16–1.32) for the Charlson Comorbidity Index (+1 point), and &gt; = 1.19 (P &lt; = 0.022) for other biomarkers (+1 SD). ACM128 improved (P &lt; = 0.0001) IDI (&gt;= + 0.50), NRI (&gt;= + 53.7) and AUC (&gt;= + 0.037) over and beyond clinical risk indicators and other biomarkers. Interactome mapping, using parental proteins derived from sequenced peptides included in ACM128 and in-silico predicted proteases, including/excluding urinary collagen fragments (63/35 peptides), revealed as top molecular pathways protein digestion and absorption, lysosomal activity and apoptosis. Conclusions: The urinary proteomic classifier ACM128 predicts the 1-year post ICU mortality over and beyond clinical risk factors and other biomarkers and revealed molecular pathways potentially contributing to a fatal outcome.

Assessment of Integrated Classifier’s Ability to Distinguish Benign From Malignant Lung Nodules: Extended Analyses and 2-Year Follow-Up Results of the PANOPTIC (Pulmonary Nodule Plasma Proteomic Classifier) Trial

Assessment of Integrated Classifier’s Ability to Distinguish Benign From Malignant Lung Nodules: Extended Analyses and 2-Year Follow-Up Results of the PANOPTIC (Pulmonary Nodule Plasma Proteomic Classifier) Trial

Definition and Independent Validation of a Proteomic-Classifier in Ovarian Cancer.

Simple SummaryThe heterogeneity of epithelial ovarian cancer and its associated molecular biological characteristics are continuously integrated in the development of therapy guidelines. In a next step, future therapy recommendations might also be able to focus on the patient’s systemic status, not only the tumor’s molecular pattern. Therefore, new methods to identify and validate host-related biomarkers need to be established. Using mass spectrometry, we developed and independently validated a blood-based proteomic classifier, stratifying epithelial ovarian cancer patients into good and poor survival groups. We also determined an age dependence of the prognostic performance of this classifier and its association with important biological processes. This work highlights that, just like molecular markers of the tumor itself, the systemic condition of a patient (partly reflected in proteomic patterns) also influences survival and therapy response and could therefore be integrated into future processes of therapy planning.Mass-spectrometry-based analyses have identified a variety of candidate protein biomarkers that might be crucial for epithelial ovarian cancer (EOC) development and therapy response. Comprehensive validation studies of the biological and clinical implications of proteomics are needed to advance them toward clinical use. Using the Deep MALDI method of mass spectrometry, we developed and independently validated (development cohort: n = 199, validation cohort: n = 135) a blood-based proteomic classifier, stratifying EOC patients into good and poor survival groups. We also determined an age dependency of the prognostic performance of this classifier, and our protein set enrichment analysis showed that the good and poor proteomic phenotypes were associated with, respectively, lower and higher levels of complement activation, inflammatory response, and acute phase reactants. This work highlights that, just like molecular markers of the tumor itself, the systemic condition of a patient (partly reflected in proteomic patterns) also influences survival and therapy response in a subset of ovarian cancer patients and could therefore be integrated into future processes of therapy planning.

MO048MULTICENTRE PROSPECTIVE VALIDATION OF THE URINARY PEPTIDOME-BASED CLASSIFIER CKD273 AS A PREDICTOR OF RENAL FUNCTION DECLINE IN SUBJECTS WITH TYPE 2 DIABETES

Abstract Background and Aims In the PRIORITY study, it was recently demonstrated that the urinary peptidome-based classifier CKD273 was associated with increased risk for progression to microalbuminuria. As a prespecified secondary outcome, we aim to evaluate the classifier CKD273 as a determinant of relative reductions in eGFR (CKD-EPI) of 30% and 40% from baseline, at one timepoint without requirements of confirmation. Method The ‘Proteomic prediction and Renin angiotensin aldosterone system Inhibition prevention Of early diabetic nephRopathy In TYpe 2 diabetic patients with normoalbuminuria trial’ (PRIORITY) is the first prospective observational study to evaluate the early detection of diabetic kidney disease in subjects with type 2 diabetes (T2D) and normoalbuminuria using the CKD273 classifier. Setting 1775 subjects from 15 European sites with a mean follow-up time of 2.6 years (minimum of 7 days and a maximum of 4.3 years). Patients Subjects with T2D, normoalbuminuria and estimated glomerular filtration rate (eGFR) ≥ 45 ml/min/1.73m2. Participants were stratified into high- or low-risk groups based on their CKD273 score in a urine sample at screening (high-risk defined as score &amp;gt; 0.154). Results In total, 12 % (n = 216) of the subjects had a high-risk proteomic pattern. Mean (SD) baseline eGFR was 88 (15) ml/min/1.73m2 in the low-risk group and 81 (17) ml/min/1.73m2 in the high-risk group (p &amp;lt; 0.01). Baseline median (interquartile range) urinary albumin to creatinine ratio (UACR) was 5 (3-8) mg/g and 7 (4-12) mg/g in the low-risk and high-risk groups, respectively (p &amp;lt; 0.01). A 30 % reduction in eGFR from baseline was seen in 42 (19.4 %) subjects in the high-risk group as compared to 62 (3.9 %) in the low-risk group (p &amp;lt; 0.0001). In an unadjusted Cox-model the hazard ratio (HR) for the high-risk group was 5.7, 95 % confidence interval (CI) (3.9 to 8.5; p&amp;lt;0.0001). After adjustment for baseline eGFR and UACR, the HR was 5.2, 95 % CI (3.4 to 7.8; p&amp;lt;0.0001). A 40 % reduction in eGFR was seen in 15 (6.9 %) subjects in the high-risk group whereas 22 (1.4 %) in the low-risk group developed this endpoint (p&amp;lt;0.0001). In an unadjusted Cox-model the HR for the high-risk group was 5.0, 95 % CI (2.6 to 9.6; p&amp;lt;0.0001). After adjustment for baseline eGFR and UACR, the HR was 4.8, 95 % CI (2.4 to 9.7; p&amp;lt;0.0001). Conclusion In normoalbuminuric subjects with T2D, the urinary proteomic classifier CKD273 predicts renal function decline of 30 % and 40 %, independent of baseline eGFR and albuminuria.

Randomized, phase II study of ficlatuzumab with or without cetuximab in patients with pan-refractory, recurrent/metastatic (R/M) head and neck squamous cell carcinoma (HNSCC).

TPS6594 Background: Patients with pan-refractory R/M HNSCC, with clinical resistance to cytotoxic therapy, anti-EGFR molecular targeting, and immunotherapy, have poor survival. An established tumor-intrinsic resistance mechanism to cetuximab, an anti-EGFR IgG1 monoclonal antibody (mAb), is activation of the hepatocyte growth factor (HGF)/cMet pathway, which converges with the EGFR network at both the PI3K/Akt and MAPK nodes allowing for reciprocal compensation. Moreover, over-expression of HGF in the tumor microenvironment is immunosuppressive. Convergent data suggest that HGF/cMet pathway inhibition concurrent with EGFR blockade may overcome cetuximab resistance. We previously reported a Phase I study of ficlatuzumab, a humanized anti-HGF IgG1 mAb, with cetuximab in cetuximab-resistant R/M HNSCC. The combination showed promising safety, overall response rate (ORR) and progression-free survival (PFS). Preliminary biomarker analyses showed that high circulating cMet was associated with poor PFS whereas serum Veristrat, a proteomic classifier associated with worse prognosis in the setting of anti-EGFR monotherapy, was not. An increase in total peripheral T cells, particularly the CD8+ subset, was associated with treatment response while progression was associated with expansion of a unique myeloid population. We designed a follow-on randomized phase II trial evaluating ficlatuzumab with or without cetuximab in pan-refractory, R/M HNSCC with signaling and immune correlatives. Methods: This is a multicenter phase II trial with a randomized, non-comparative, two-arm design (ficlatuzumab 20 mg/kg with or without cetuximab 500 mg/m2 every 2 weeks) in patients with pan-refractory R/M HNSCC. Key eligibility criteria include: R/M HNSCC; cetuximab resistance (progression during or within 6 months of cetuximab-radiation or palliative cetuximab); platinum resistance; prior exposure to anti-PD1 mAb; ECOG 0-1; consent to baseline research biopsy. The primary objective is to evaluate the efficacy of each arm as measured by PFS. To test the hypothesis that either regimen improves historical PFS from 2 to 3.33 months requires 66 eligible patients. Key secondary endpoints are ORR and survival. Mechanistic biomarkers include tumor HGF/cMet pathway activation, tumor and peripheral immune profiles, soluble cMet, and serum Veristrat. Thirty-five of 66 subjects have enrolled at 6 centers. A Bayesian continuous monitoring rule for futility has not been triggered for either arm. Clinical trial information: NCT03422536 .

Proteomic Tissue-Based Classifier for Early Prediction of Prostate Cancer Progression.

Although ~40% of screen-detected prostate cancers (PCa) are indolent, advanced-stage PCa is a lethal disease with 5-year survival rates around 29%. Identification of biomarkers for early detection of aggressive disease is a key challenge. Starting with 52 candidate biomarkers, selected from existing PCa genomics datasets and known PCa driver genes, we used targeted mass spectrometry to quantify proteins that significantly differed in primary tumors from PCa patients treated with radical prostatectomy (RP) across three study outcomes: (i) metastasis ≥1-year post-RP, (ii) biochemical recurrence ≥1-year post-RP, and (iii) no progression after ≥10 years post-RP. Sixteen proteins that differed significantly in an initial set of 105 samples were evaluated in the entire cohort (n = 338). A five-protein classifier which combined FOLH1, KLK3, TGFB1, SPARC, and CAMKK2 with existing clinical and pathological standard of care variables demonstrated significant improvement in predicting distant metastasis, achieving an area under the receiver-operating characteristic curve of 0.92 (0.86, 0.99, p = 0.001) and a negative predictive value of 92% in the training/testing analysis. This classifier has the potential to stratify patients based on risk of aggressive, metastatic PCa that will require early intervention compared to low risk patients who could be managed through active surveillance.

The novel urinary proteomic classifier HF1 has similar diagnostic and prognostic utility to BNP in heart failure.

AimsMeasurement of B‐type natriuretic peptide (BNP) or N‐terminal pro‐BNP is recommended as part of the diagnostic workup of patients with suspected heart failure (HF). We evaluated the diagnostic and prognostic utility of the novel urinary proteomic classifier HF1, compared with BNP, in HF. HF1 consists of 85 unique urinary peptide fragments thought, mainly, to reflect collagen turnover.Methods and resultsWe performed urinary proteome analysis using capillary electrophoresis coupled with mass spectrometry in 829 participants. Of these, 622 had HF (504 had chronic HF and 118 acute HF) and 207 were controls (62 coronary heart disease patients without HF and 145 healthy controls). The area under the receiver operating characteristic (ROC) curve (AUC) using HF1 for the diagnosis of HF (cases vs. controls) was 0.94 (95% CI, 0.92–0.96). This compared with an AUC for BNP of 0.98 (95% CI, 0.97–0.99). Adding HF1 to BNP increased the AUC to 0.99 (0.98–0.99), P < 0.001, and led to a net reclassification improvement of 0.67 (95% CI, 0.54–0.77), P < 0.001. Among 433 HF patients followed up for a median of 989 days, we observed 186 deaths. HF1 had poorer predictive value to BNP for all‐cause mortality and did not add prognostic information when combined with BNP.ConclusionsThe urinary proteomic classifier HF1 performed as well, diagnostically, as BNP and provided incremental diagnostic information when added to BNP. HF1 had less prognostic utility than BNP.

Early detection of diabetic kidney disease by urinary proteomics and subsequent intervention with spironolactone to delay progression (PRIORITY): a prospective observational study and embedded randomised placebo-controlled trial

Microalbuminuria is an early sign of kidney disease in people with diabetes and indicates increased risk of cardiovascular disease. We tested whether a urinary proteomic risk classifier (CKD273) score was associated with development of microalbuminuria and whether progression to microalbuminuria could be prevented with the mineralocorticoid receptor antagonist spironolactone. In this multicentre, prospective, observational study with embedded randomised controlled trial (PRIORITY), we recruited people with type 2 diabetes, normal urinary albumin excretion, and preserved renal function from 15 specialist centres in ten European countries. All participants (observational cohort) were tested with the CKD273 classifier and classified as high risk (CKD273 classifier score >0·154) or low risk (≤0·154). Participants who were classified as high risk were entered into a randomised controlled trial and randomly assigned (1:1), by use of an interactive web-response system, to receive spironolactone 25 mg once daily or matched placebo (trial cohort). The primary endpoint was development of confirmed microalbuminuria in all individuals with available data (observational cohort). Secondary endpoints included reduction in incidence of microalbuminuria with spironolactone (trial cohort, intention-to-treat population) and association between CKD273 risk score and measures of impaired renal function based on estimated glomerular filtration rate (eGFR; observational cohort). Adverse events (particularly gynaecomastia and hyperkalaemia) and serious adverse events were recorded for the intention-to-treat population (trial cohort). This study is registered with the EU Clinical Trials Register (EudraCT 20120-004523-4) and ClinicalTrials.gov (NCT02040441) and is completed. Between March 25, 2014, and Sept 30, 2018, we enrolled and followed-up 1775 participants (observational cohort), 1559 (88%) of 1775 participants had a low-risk urinary proteomic pattern and 216 (12%) had a high-risk pattern, of whom 209 were included in the trial cohort and assigned to spironolactone (n=102) or placebo (n=107). The overall median follow-up time was 2·51 years (IQR 2·0-3·0). Progression to microalbuminuria was seen in 61 (28%) of 216 high-risk participants and 139 (9%) of 1559 low-risk participants (hazard ratio [HR] 2·48, 95% CI 1·80-3·42; p<0·0001, after adjustment for baseline variables of age, sex, HbA1c, systolic blood pressure, retinopathy, urine albumin-to-creatinine ratio [UACR], and eGFR). Development of impaired renal function (eGFR <60 mL/min per 1·73 m2) was seen in 48 (26%) of 184 high-risk participants and 119 (8%) of 1423 low-risk participants (HR 3·50; 95% CI 2·50-4·90, after adjustment for baseline variables). A 30% decrease in eGFR from baseline (post-hoc endpoint) was seen in 42 (19%) of 216 high-risk participants and 62 (4%) of 1559 low-risk participants (HR 5·15, 95% CI 3·41-7·76; p<0·0001, after adjustment for basline eGFR and UACR). In the intention-to-treat trial cohort, development of microalbuminuria was seen in 35 (33%) of 107 in the placebo group and 26 (25%) of 102 in the spironolactone group (HR 0·81, 95% CI 0·49-1·34; p=0·41). In the safety analysis (intention-to-treat trial cohort), events of plasma potassium concentrations of more than 5·5 mmol/L were seen in 13 (13%) of 102 participants in the spironolactone group and four (4%) of 107 participants in the placebo group, and gynaecomastia was seen in three (3%) participants in the spironolactone group and none in the placebo group. One patient died in the placebo group due to a cardiac event (considered possibly related to study drug) and one patient died in the spironolactone group due to cancer, deemed unrelated to study drug. In people with type 2 diabetes and normoalbuminuria, a high-risk score from the urinary proteomic classifier CKD273 was associated with an increased risk of progression to microalbuminuria over a median of 2·5 years, independent of clinical characteristics. However, spironolactone did not prevent progression to microalbuminuria in high-risk patients. European Union Seventh Framework Programme.

A novel urinary biomarker predicts 1-year mortality after discharge from intensive care

RationaleThe urinary proteome reflects molecular drivers of disease.ObjectivesTo construct a urinary proteomic biomarker predicting 1-year post-ICU mortality.MethodsIn 1243 patients, the urinary proteome was measured on ICU admission, using capillary electrophoresis coupled with mass spectrometry along with clinical variables, circulating biomarkers (BNP, hsTnT, active ADM, and NGAL), and urinary albumin. Methods included support vector modeling to construct the classifier, Cox regression, the integrated discrimination (IDI), and net reclassification (NRI) improvement, and area under the curve (AUC) to assess predictive accuracy, and Proteasix and protein-proteome interactome analyses.Measurements and main resultsIn the discovery (deaths/survivors, 70/299) and test (175/699) datasets, the new classifier ACM128, mainly consisting of collagen fragments, yielding AUCs of 0.755 (95% CI, 0.708–0.798) and 0.688 (0.656–0.719), respectively. While accounting for study site and clinical risk factors, hazard ratios in 1243 patients were 2.41 (2.00–2.91) for ACM128 (+ 1 SD), 1.24 (1.16–1.32) for the Charlson Comorbidity Index (+ 1 point), and ≥ 1.19 (P ≤ 0.022) for other biomarkers (+ 1 SD). ACM128 improved (P ≤ 0.0001) IDI (≥ + 0.50), NRI (≥ + 53.7), and AUC (≥ + 0.037) over and beyond clinical risk indicators and other biomarkers. Interactome mapping, using parental proteins derived from sequenced peptides included in ACM128 and in silico predicted proteases, including/excluding urinary collagen fragments (63/35 peptides), revealed as top molecular pathways protein digestion and absorption, lysosomal activity, and apoptosis.ConclusionsThe urinary proteomic classifier ACM128 predicts the 1-year post-ICU mortality over and beyond clinical risk factors and other biomarkers and revealed molecular pathways potentially contributing to a fatal outcome.

The urinary proteomics classifier chronic kidney disease 273 predicts cardiovascular outcome in patients with chronic kidney disease.

The urinary proteomic classifier chronic kidney disease 273 (CKD273) is predictive for the development and progression of chronic kidney disease (CKD) and/or albuminuria in type 2 diabetes. This study evaluates its role in the prediction of cardiovascular (CV) events in patients with CKD Stages G1-G5. We applied the CKD273 classifier in a cohort of 451 patients with CKD Stages G1-G5 followed prospectively for a median of 5.5 years. Primary endpoints were all-cause mortality, CV mortality and the composite of non-fatal and fatal CV events (CVEs). In multivariate Cox regression models adjusting for age, sex, prevalent diabetes and CV history, the CKD273 classifier at baseline was significantly associated with total mortality and time to fatal or non-fatal CVE, but not CV mortality. Because of a significant interaction between CKD273 and CV history (P = 0.018) and CKD stages (P = 0.002), a stratified analysis was performed. In the fully adjusted models, CKD273 classifier was a strong and independent predictor of fatal or non-fatal CVE only in the subgroup of patients with CKD Stages G1-G3b and without a history of CV disease. In those patients, the highest tertile of CKD273 was associated with a >10-fold increased risk as compared with the lowest tertile. The urinary CKD273 classifier provides additional independent information regarding the CV risk in patients with early CKD stage and a blank CV history. Determination of CKD273 scores on a random urine sample may improve the efficacy of intensified surveillance and preventive strategies by selecting patients who potentially will benefit most from early risk management.