1. Introduction
Chronic kidney disease (CKD) is now recognized as a major public health problem and a strong determinant of cardiovascular prognosis. The KDIGO 2024 guideline defines CKD as abnormalities of kidney structure or function persisting for at least three months, with implications for health, and classifies it according to cause, glomerular filtration rate (GFR) category, and albuminuria category [
1]. This classification frames CKD as a continuum of risk, in which lower GFR and higher albuminuria identify patients at increased risk of both renal and extra-renal complications [
1,
2].
The burden of CKD extends well beyond kidney-related outcomes. Worldwide, all-stage CKD affects approximately 697.5 million people, with an estimated prevalence of 9.1%. It accounts for 1.2 million deaths and 35.8 million disability-adjusted life-years, while impaired kidney function contributes to a further 1.4 million cardiovascular disease-related deaths and 25.3 million cardiovascular disability-adjusted life-years. These figures place kidney dysfunction among the major contributors to global cardiovascular morbidity and mortality [
3].
Lower estimated glomerular filtration rate (eGFR) and higher albuminuria are consistently associated with adverse renal and cardiovascular outcomes, including kidney failure, acute kidney injury (AKI), all-cause and cardiovascular mortality, hospitalization, coronary heart disease, stroke, heart failure (HF), atrial fibrillation, and peripheral artery disease [
4]. Kidney dysfunction therefore cannot be interpreted only as a downstream consequence of cardiovascular disease (CVD) or hemodynamic instability. It may also contribute to cardiovascular risk through inflammation, endothelial dysfunction, vascular calcification, neurohormonal activation, and myocardial remodeling [
3,
4].
AKI adds another layer of cardiovascular risk and is often under-recognized after the acute event has resolved. In routine practice, AKI is diagnosed using serum creatinine (SCr), estimated GFR, and urine output. These functional markers have limited sensitivity and specificity, may require a baseline creatinine value that is not always available, and do not directly identify structural tubular injury. As a result, AKI may be misclassified or detected only after clinically relevant renal damage has occurred, particularly in patients with acute cardiovascular decompensation or critical illness [
5].
The prognostic consequences of AKI often persist after discharge. AKI has been associated with higher risks of cardiovascular mortality, major cardiovascular events, HF, acute myocardial infarction, and stroke [
6]. It is also linked to incident CKD, CKD progression, HF events, and all-cause mortality. Post-discharge assessment of kidney function recovery and proteinuria, particularly around three months after hospitalization, may add useful prognostic information and supports a structured approach to post-AKI risk stratification [
7].
These observations are particularly relevant in cardiorenal syndrome (CRS), a condition in which acute or chronic dysfunction of the heart or kidneys induces dysfunction in the other organ. The American Heart Association scientific statement describes cardiorenal interactions as the result of hemodynamic mechanisms, including reduced cardiac output and venous congestion, together with neurohormonal activation, inflammation, oxidative stress, endothelial dysfunction, anemia, and mineral metabolism disturbances [
8]. In this setting, SCr may fail to distinguish a functional change in filtration from structural tubular injury. Biomarkers that reflect tubular damage, altered filtration, fibrosis, remodeling, and mineral metabolism abnormalities may therefore refine early diagnosis and cardiovascular risk prediction. This narrative review summarizes the evidence supporting NGAL, KIM-1, cystatin C, pro-enkephalin, ST2, galectin-3, and FGF-23 as candidate tools for cardiorenal risk assessment beyond creatinine-based evaluation.
2. Narrative Review Methodology
This narrative review evaluates emerging renal and cardiorenal biomarkers in relation to early kidney injury detection, cardiorenal dysfunction, and cardiovascular risk stratification. The review focused on five biomarker domains: tubular injury markers, represented by neutrophil gelatinase-associated lipocalin (NGAL) and kidney injury molecule-1 (KIM-1); functional renal markers beyond SCr, including cystatin C and pro-enkephalin (PENK); fibrosis, inflammation, and remodeling markers, including soluble suppression of tumorigenicity-2 (sST2) and galectin-3; mineral metabolism-related biomarkers, particularly fibroblast growth factor-23 (FGF-23); and the potential clinical relevance of these markers in patients with CKD, AKI, HF, acute coronary syndromes, diabetes, hypertension, and cardiorenal syndrome.
A literature search was conducted in PubMed/MEDLINE, Scopus, and Web of Science for articles published from January 2015 to April 2026. The search combined Medical Subject Headings (MeSH) and free-text terms related to renal biomarkers, cardiorenal biomarkers, cardiovascular risk, cardiovascular outcomes, chronic kidney disease, acute kidney injury, cardiorenal syndrome, and the individual biomarkers discussed in this review. The main search terms included “renal biomarkers”, “kidney biomarkers”, “cardiorenal biomarkers”, “cardiovascular risk”, “cardiovascular disease”, “major adverse cardiovascular events”, “mortality”, “chronic kidney disease”, “acute kidney injury”, “cardiorenal syndrome”, “NGAL”, “neutrophil gelatinase-associated lipocalin”, “KIM-1”, “kidney injury molecule-1”, “cystatin C”, “pro-enkephalin”, “penKid”, “ST2”, “soluble ST2”, “galectin-3”, “fibrosis”, “cardiac remodeling”, “FGF-23”, “fibroblast growth factor-23”, “mineral metabolism”, and “vascular calcification”.
Reference lists of relevant original studies, systematic reviews, meta-analyses, consensus documents, and major cardiology and nephrology guidelines were also screened manually. Priority was given to studies examining associations between renal or cardiorenal biomarker elevation and clinically relevant cardiovascular outcomes, including HF hospitalization, acute coronary events, left ventricular hypertrophy, atherosclerotic burden, cardiorenal deterioration, cardiovascular mortality, and all-cause mortality.
Eligible sources included original research articles, observational cohort studies, randomized controlled trials, translational and mechanistic investigations, systematic reviews, meta-analyses, authoritative narrative reviews, clinical guidelines, and consensus statements from nephrology and cardiology societies. Articles published in English were prioritized. Case reports, editorials, letters, and conference abstracts were generally excluded, except when they offered mechanistic or conceptual information not adequately covered in full peer-reviewed articles.
Because this was a narrative review, no formal risk-of-bias assessment, quantitative quality scoring, or meta-analysis was performed. Study selection was guided by mechanistic relevance, clinical applicability, biological plausibility, and contribution to understanding the interface between renal injury, cardiorenal dysfunction, and cardiovascular risk. Particular attention was given to studies assessing biomarkers beyond conventional creatinine-based evaluation, especially those exploring their potential value for early diagnosis, prognostic stratification, phenotype identification, or individualized monitoring in high-risk cardiovascular and renal populations.
3. Pathophysiological Basis of the Cardiorenal Biomarker Axis
CRS refers to a bidirectional disorder in which acute or chronic dysfunction of the heart or kidneys contributes to dysfunction of the other organ. Its biology cannot be reduced to impaired perfusion alone. Hemodynamic stress, neurohormonal activation, inflammation, oxidative stress, endothelial dysfunction, fibrosis, and metabolic abnormalities interact across the heart–kidney axis and shape both renal and cardiovascular outcomes [
8,
9,
10,
11].
For clinical interpretation, CRS is commonly divided into five types according to the primary organ injury, the secondary organ dysfunction, and the acute or chronic time course of the process. This classification is useful because biomarker interpretation differs between acute cardiac decompensation with secondary kidney injury, chronic heart failure with progressive renal dysfunction, primary acute or chronic kidney disease with cardiovascular consequences, and systemic disorders that affect both organs simultaneously. The five CRS types are summarized in
Table 1.
The classical model begins with renal hypoperfusion. In HF, reduced cardiac output may decrease renal arterial perfusion, activating the renin–angiotensin–aldosterone system (RAAS), the sympathetic nervous system, and vasopressin release. These responses help maintain circulatory volume in the short term, but persistent activation promotes sodium and water retention, vasoconstriction, increased preload, congestion, and further cardiac dysfunction [
8,
9]. Reduced forward flow, however, does not fully explain CRS. Renal filtration depends on the pressure gradient between arterial inflow and venous outflow; therefore, elevated central venous pressure and renal venous congestion may reduce intrarenal blood flow and impair glomerular filtration even when systemic blood pressure is preserved [
8,
9,
10].
Venous congestion is an important contributor to kidney injury in CVD. Increased central venous and intra-abdominal pressures may reduce renal plasma flow, impair the filtration gradient, and contribute to oliguria and worsening renal function (WRF) in acute decompensated heart failure (ADHF) [
10,
11]. This helps explain why renal dysfunction may occur in both reduced and preserved ejection fraction heart failure, and why decongestion has clinical relevance beyond symptom relief [
8,
9].
Non-hemodynamic mechanisms further intensify the cardiorenal loop. Endothelial dysfunction reduces nitric oxide bioavailability and favors oxidative stress, inflammation, vascular permeability, extracellular matrix remodeling, and a prothrombotic state [
1]. Inflammatory cytokines, including tumor necrosis factor-α, interleukin-1, and interleukin-6, may exert cardiodepressant effects, while sustained RAAS activation and reactive oxygen species contribute to tubular ischemia, myocardial remodeling, and progressive fibrosis [
8,
11]. This pathway diversity provides the biological rationale for biomarkers that reflect different components of cardiorenal injury, rather than relying only on SCr.
Conventional renal indices do not capture the full spectrum of kidney–heart injury. SCr rises after a substantial fall in GFR, may lag 48–72 h after AKI, and is influenced by age, sex, muscle mass, volume status, ethnicity, and drug exposure. It may therefore fail to separate functional renal changes from structural parenchymal injury. In CRS, biomarker interpretation is more informative when linked to dominant biological pathways, including filtration impairment, tubular injury, inflammation, oxidative stress, endothelial dysfunction, fibrosis, and mineral metabolism disturbance [
12].
A second clinically useful distinction is between biomarkers that mainly inform acute cardiorenal injury and those that are more closely related to chronic remodeling, fibrosis, mineral metabolism disturbance, and long-term cardiovascular risk. This distinction is not absolute, because several biomarkers may be relevant in both acute and chronic settings. However, organizing biomarkers according to the dominant phase of injury may help clinicians interpret whether the signal mainly reflects acute tubular stress, functional renal deterioration, incomplete recovery after AKI, chronic cardiorenal remodeling, or CKD-related cardiovascular risk. This phase-based interpretation is summarized in
Table 2.
Cystatin C and PENK reflect functional renal impairment beyond creatinine-based assessment. Cystatin C is less dependent on muscle mass than creatinine and has been associated with adverse outcomes in acute and chronic HF, although its concentration may still be affected by inflammation, thyroid disease, obesity, smoking, glucocorticoid exposure, and malignancy [
12,
13]. PENK A is inversely related to GFR and has been reported to predict AKI and adverse outcomes in acute HF, supporting its role as an emerging functional cardiorenal biomarker [
12].
Tubular injury biomarkers add a different layer of information. NGAL is released by neutrophils, renal tubular epithelial cells, and cardiomyocytes in response to inflammation and injury. It may rise earlier than creatinine and has been linked to AKI prediction in cardiac surgery and acute HF [
12,
13]. KIM-1 reflects proximal tubular epithelial injury and may help identify structural tubular damage when changes in creatinine are delayed or clinically ambiguous [
12]. Biomarkers of fibrosis and remodeling, including galectin-3 and sST2, have been associated with cardiac fibrosis, kidney disease progression, HF prognosis, and possibly diuretic resistance [
12,
13]. Acute CRS also involves RAAS and sympathetic activation, oxidative stress, inflammation, apoptosis, and mitochondrial injury, supporting the use of multi-biomarker panels rather than single-marker risk assessment [
14].
Mineral metabolism provides another link between kidney dysfunction and cardiovascular injury. In CKD, FGF-23 rises as estimated GFR declines and phosphate handling becomes impaired. In adults with CKD stages 2–4, higher FGF-23 has been associated with greater left ventricular mass index and long-term adverse outcomes, including all-cause mortality, atrial fibrillation, and congestive HF. However, left ventricular hypertrophy (LVH) appears to explain only part of these associations, suggesting that FGF-23 may reflect broader CKD-related cardiovascular risk rather than a single myocardial hypertrophy pathway [
15].
Inflammation and fibrosis markers further extend this profile. In CKD populations, higher growth differentiation factor-15 (GDF-15), galectin-3, and soluble ST2 have been independently associated with mortality, while GDF-15 has also been associated with HF events [
16]. These data support a multi-biomarker model in which tubular injury, filtration impairment, inflammation, fibrosis, remodeling, and mineral metabolism disturbance jointly characterize cardiorenal risk.
Clinically, the cardiorenal biomarker axis should therefore be interpreted in two complementary ways: first, according to the dominant biological pathway reflected by the biomarker, and second, according to the acute, subacute, or chronic phase of cardiorenal disease. NGAL and KIM-1 are most useful for identifying tubular injury, particularly in acute or unstable settings. Cystatin C and PENK refine functional renal assessment when creatinine is delayed or confounded. sST2 and galectin-3 mainly inform inflammatory, fibrotic, and remodeling pathways, while FGF-23 reflects the mineral metabolism axis that links CKD to vascular calcification, LVH, and long-term cardiovascular risk (
Figure 1).
7. FGF-23 and the Mineral Metabolism–Cardiovascular Axis
FGF-23 is a bone-derived phosphaturic hormone produced mainly by osteocytes and osteoblasts. It regulates phosphate and vitamin D metabolism through the FGF receptor–Klotho axis. In the kidney, FGF-23 downregulates sodium–phosphate cotransporters in proximal tubules, increases phosphaturia, suppresses 1,25-dihydroxyvitamin D synthesis, and modulates parathyroid hormone secretion [
46,
47]. In CKD, FGF-23 is one of the earliest detectable abnormalities of mineral metabolism. Its concentration rises progressively as GFR declines, initially as an adaptive response to maintain phosphate balance. With advancing CKD, reduced nephron mass, Klotho deficiency, and phosphate retention contribute to FGF-23 resistance and to the broader phenotype of CKD-mineral and bone disorder (CKD-MBD) [
46,
47].
The cardiovascular relevance of FGF-23 lies in its connection with myocardial and vascular injury. Elevated FGF-23 has been associated with LVH, cardiac fibrosis, arterial stiffness, atrial fibrillation, atherosclerosis, and higher cardiovascular and all-cause mortality [
48]. Mechanistic data indicate that FGF-23 may promote cardiomyocyte hypertrophy through FGFR4/calcineurin/NFAT signaling, independently of Klotho. It may also interact with inflammation, RAAS activation, and sodium handling, although the relative contribution of these pathways in clinical disease remains difficult to separate [
46,
48].
FGF-23 is also linked to the vascular phenotype of CKD. Higher FGF-23 has been positively associated with arterial calcification, carotid intima–media thickness, and pulse wave velocity, whereas Klotho shows inverse associations with calcification and arterial thickness [
49]. CKD-related vascular calcification is further promoted by hyperphosphatemia, hypercalcemia, oxidative stress, inflammation, and loss of calcification inhibitors. These processes favor osteogenic transformation of vascular smooth muscle cells and contribute to arterial stiffness [
50]. FGF-23 therefore differs from tubular injury, filtration, and fibrosis biomarkers because it reflects a mineral metabolism pathway through which kidney dysfunction may contribute to vascular calcification, myocardial hypertrophy, and cardiovascular risk.
The fibrosis-, remodeling-, and mineral metabolism-related biomarkers that complement tubular and functional renal markers are summarized in
Table 5.
8. Clinical Applications: Toward Biomarker-Guided Cardiovascular Risk Stratification
Clinical application of renal and cardiorenal biomarkers should start from a pragmatic premise: creatinine, eGFR, albuminuria, natriuretic peptides, and troponins remain necessary, but they do not fully characterize cardiorenal risk. In CRS, conventional markers may be delayed, nonspecific, or difficult to interpret because renal dysfunction, congestion, inflammation, and treatment-related hemodynamic changes often coexist. Biomarkers are therefore most useful when interpreted in relation to specific clinical questions, such as whether renal worsening during decongestion reflects structural tubular injury or functional hemodynamic change, whether persistent congestion is contributing to renal impairment, or whether a patient is at imminent risk of AKI [
51].
A biomarker-guided approach may be particularly relevant in high-risk cardiovascular populations, including patients with HF, ACS, CKD, diabetes, hypertension, and AKI. In acute CRS, laboratory panels may be more informative than isolated markers. In patients with acute HF at risk of type 1 CRS, cystatin C, changes in SCr, KIM-1, NGAL, and TIMP-2 × IGFBP7 may support earlier identification of AKI. Conversely, in patients with AKI at risk of acute HF or type 3 CRS, high-sensitivity troponin, BNP, and N-terminal pro-B-type natriuretic peptide (NT-proBNP) may help detect cardiac involvement [
52].
The rationale for this approach reflects the biology of kidney injury. SCr and urine output are functional markers and may miss subclinical AKI, in which tubular epithelial injury precedes measurable decline in glomerular filtration. Newer biomarkers capture different injury domains: cystatin C and PENK reflect filtration changes; NGAL, KIM-1, N-acetyl-β-D-glucosaminidase (NAG), and L-FABP reflect tubular damage; TIMP-2 and IGFBP7 reflect tubular stress and cell-cycle arrest; interleukin-18 (IL-18) and interleukin-9 (IL-9) reflect intrarenal inflammation; and CCL14 may identify persistent severe AKI [
53].
In practice, these markers could complement conventional assessment by helping assign patients to a dominant cardiorenal phenotype, such as tubular injury, filtration impairment, congestion, myocardial injury, fibrosis/remodeling, inflammation, or mineral metabolism disturbance. Such phenotyping may support earlier renal injury detection, prediction of CKD progression, identification of high-risk HF patients, assessment of CRS, cardiovascular risk stratification, and monitoring of treatment response. Interpretation must remain contextual. Temporary creatinine increases during decongestion do not always indicate structural kidney injury, while troponin and NT-proBNP may be chronically elevated in renal dysfunction, even outside ACS or acute HF [
54].
Beyond early diagnosis and prognosis, the reviewed biomarkers may also have indirect relevance for clinical management. However, this should be interpreted cautiously, because most available evidence supports biomarker use for risk stratification, phenotype recognition, and monitoring rather than for definitive biomarker-guided therapeutic decisions. These interpretations may help clinicians decide which patients require closer follow-up, repeated renal and cardiac biomarker assessment, more careful evaluation of congestion and renal function, or earlier multidisciplinary cardiorenal assessment (
Table 6).
In CKD, multi-biomarker strategies may improve cardiovascular risk prediction by integrating signals from myocardial stress, tubular injury, inflammation, extracellular matrix remodeling, and vascular calcification. A protein panel combining NT-proBNP, KIM-1, osteopontin, and TIMP-1 has been shown to predict incident cardiovascular events and to identify higher-risk patients across CKD stages [
55]. This supports the concept that panels reflecting several biological pathways may be more informative than isolated renal or cardiac markers when selecting patients who may require closer monitoring or intensified prevention.
In ACS, biomarker profiling illustrates the overlap between kidney dysfunction and cardiovascular risk. Inflammatory, angiogenic, and kidney-related biomarkers tend to increase as kidney function declines. Among these markers, endothelial cell-specific molecule-1, FGF-23, and KIM-1 have been associated with major adverse cardiovascular events and death, while FGF-23 appears to retain independent prognostic value after adjustment for other biomarker signals [
56]. This finding is consistent with the relevance of mineral metabolism and inflammatory–vascular pathways in ACS, including in patients with impaired kidney function.
Post-AKI follow-up is another setting in which biomarker-guided risk stratification may be clinically useful. Vascular biomarker panels measured after hospitalization can identify phenotypes related to vascular injury, repair, and lower-risk profiles. Patients with a vascular injury phenotype appear to have higher subsequent HF risk, and biomarker data may improve prediction of HF or death beyond clinical variables alone [
57]. These findings support the value of post-discharge reassessment, especially in patients whose renal function appears to recover but who remain at increased cardiovascular risk.
Biomarker-guided strategies should also be integrated into the broader cardiovascular–kidney–metabolic framework. Albuminuria remains central for screening and monitoring kidney and cardiovascular risk in diabetes, hypertension, and CVD, while NT-proBNP and high-sensitivity troponin support HF detection, ACS diagnosis, and long-term cardiovascular risk stratification. These markers require careful interpretation: albuminuria is biologically variable, NT-proBNP is influenced by age, CKD, atrial fibrillation, and obesity, and troponin may reflect chronic myocardial injury rather than acute plaque-related ischemia [
58]. A practical workflow would therefore begin with identification of a high-risk patient, followed by conventional assessment, targeted biomarker testing, dominant phenotype assignment, risk categorization, and individualized monitoring or treatment intensity.
This proposed biomarker-guided framework for cardiovascular risk stratification is illustrated in
Figure 2, integrating conventional assessment with targeted biomarker panels to identify dominant cardiorenal phenotypes and guide individualized monitoring and treatment intensity.
9. Current Limitations and Future Directions
Renal and cardiorenal biomarkers have strong biological rationale and increasing clinical support, but they are not yet ready for universal routine implementation. One limitation is that many clinical definitions still depend on SCr, eGFR, and urine output, although these markers are delayed, functional, and nonspecific. In AKI, creatinine may remain unchanged during early tubular injury and may rise only after substantial functional decline. Urine output is also difficult to interpret, as it can be influenced by hemodynamics, diuretic exposure, and transient physiological responses [
5,
59]. This creates a diagnostic gap between structural kidney injury and conventional criteria, particularly in cardiovascular patients exposed to congestion, hypoperfusion, contrast agents, nephrotoxic drugs, or critical illness.
Novel biomarkers have their own limitations. NGAL, KIM-1, L-FABP, TIMP-2 × IGFBP7, cystatin C, PENK, and inflammatory markers may improve early detection and phenotyping, but their diagnostic performance varies according to clinical setting, assay platform, sampling time, biological matrix, and patient population [
59]. Several non-renal factors can also influence biomarker concentrations, including inflammation, sepsis, diabetes, obesity, malignancy, corticosteroid exposure, thyroid dysfunction, muscle mass, proteinuria, and CKD stage. These influences reduce specificity and complicate interpretation. For many biomarkers, accepted cut-off values are still lacking, and assay standardization remains incomplete [
5,
59,
60]. Cost and availability also add barriers, especially for multi-marker platforms, omics-based diagnostics, and point-of-care tools.
Clinical actionability remains another major challenge. A biomarker may improve diagnosis or prognosis without proving that biomarker-guided management improves patient outcomes. For routine adoption, biomarkers must show that they can change clinical decisions, improve patient-centered outcomes, and remain cost-effective in prospective studies. At present, the strongest evidence supports their use for risk stratification and phenotyping, while outcome-driven interventional data remain limited [
5,
60].
Future research should focus on prospective validation cohorts, harmonized assays, pre-analytical standardization, and clinically meaningful thresholds adjusted for age, sex, kidney function, and comorbidity burden. Multi-biomarker panels should be incorporated into risk models that combine clinical variables, cardiac and renal biomarkers, imaging findings, and phenotype-based assessment. Multi-omics approaches, including genomics, proteomics, metabolomics, and transcriptomics, may improve disease classification and support more precise cardiorenal phenotyping. Artificial intelligence may also help integrate complex biomarker and clinical data for real-time risk prediction and individualized treatment selection [
60,
61]. However, AI-based cardiovascular prediction models still require external validation, transparency, reproducibility, and bias assessment before clinical deployment. Many published models remain at high risk of bias and lack independent external validation [
61]. The next step is therefore not simply to identify additional biomarkers, but to translate biologically meaningful signals into standardized, interpretable, and clinically useful pathways.
10. Conclusions
Emerging renal and cardiorenal biomarkers may refine cardiovascular risk assessment beyond SCr and creatinine-based eGFR. Conventional renal markers remain necessary in routine practice, but they often reflect functional decline after injury has already developed and provide limited information on pathway-specific cardiorenal damage. In this context, NGAL and KIM-1 may help identify tubular stress and proximal tubular injury before overt functional deterioration becomes apparent. Cystatin C and PENK add information on renal functional impairment, particularly when creatinine is influenced by muscle mass, frailty, or acute hemodynamic changes. sST2 and galectin-3 reflect inflammation, fibrosis, and cardiac or renal remodeling, while FGF-23 captures the mineral metabolism pathway linking CKD to vascular calcification, LVH, arterial stiffness, and cardiovascular risk.
These biomarkers support a more integrated view of cardiorenal disease, in which tubular injury, altered filtration, congestion, inflammation, fibrosis, and mineral metabolism disturbance are assessed as related biological domains rather than isolated abnormalities. Their most plausible clinical value lies in earlier injury detection, cardiorenal phenotyping, identification of high-risk cardiovascular patients, and individualized monitoring. Routine implementation, however, remains limited by assay variability, uncertain thresholds, biological confounding, cost, and the lack of definitive evidence that biomarker-guided management improves clinical outcomes. Before these tools can be incorporated more broadly into clinical decision-making, prospective validation, assay standardization, and clinically actionable multi-biomarker risk models are needed.