Skip to content

Biomarkers and imaging markers

TL;DR — No blood or imaging marker independently diagnoses HHD or dictates therapy. ECG-LVH is insensitive but prognostic; echo mass, geometry and strain are accessible but load/indexation dependent; CMR measures mass and fibrosis reproducibly but vendor and reference values limit portable thresholds (Pewsner 2007, PMID 17726091; Zhao 2024, PMID 39156132). REMODEL derived a two-marker staging approach using NT-proBNP 152 pg/mL and hs-troponin T 12.7 pg/mL, with dual elevation associated with HR 17.11, but external validation and treatment utility are absent (Sharp 2026, PMID 41771092). LV mass-to-strain ratio and CMR phenotypes predict outcomes in recent cohorts; they remain prognostic candidates, not validated treatment targets (Hwang 2025, PMID 40970541; Lertsiripatarajit 2026, PMID 41953289). Validation must demonstrate calibration, incremental decision value and outcome improvement—not merely group separation.

1. Biomarker roles should not be conflated

Role Required question Typical design
Diagnostic Does it identify a prespecified HHD phenotype? Blinded cross-sectional accuracy study
Differential Does it separate HHD from HCM/amyloid/athlete? Multidisease comparator cohort
Staging Does it order severity reproducibly? External prospective cohort
Prognostic Does it predict events beyond established risk? Longitudinal validation
Predictive Does treatment effect differ by marker? Prespecified trial interaction
Surrogate Does marker change mediate clinical benefit? Randomized mediation/validation

Most HHD marker studies are cross-sectional or prognostic. Very few are predictive; none has established a fully validated surrogate endpoint (Nwabuo 2020, PMID 32016791; González 2024, PMID 38084597).

2. ECG markers

Marker Strength Weakness
Sokolow–Lyon voltage Simple, historical Low sensitivity; body habitus effects
Cornell voltage/product Prognostic and trial-used Sex thresholds; electrical/anatomical mismatch
Peguero–Lo Presti Higher sensitivity in some cohorts External performance varies
Strain pattern High-risk electrical phenotype Not specific to hypertension
P-wave/interatrial block Atrial remodeling signal Rhythm/age and limited specificity

Across 21 studies and 5,608 hypertensive patients, negative likelihood ratios of common ECG-LVH criteria were approximately 0.85–0.91, insufficient to exclude anatomical LVH (Pewsner 2007, PMID 17726091).

In 13,960 paired ECG/echo records, automated ECG sensitivity was 30.7% and specificity 84.4% (Bressman 2020, PMID 32194027).

Despite this, serial Cornell-product reduction predicted fewer cardiovascular events in LIFE, showing prognostic information does not require anatomical equivalence (Okin 2004, PMID 15547161).

3. Echocardiographic mass and geometry

Measure Candidate use Sources of variability
LV-mass index LVH detection and regression Linear assumptions, indexation, observer
Relative wall thickness Geometry classification Cavity/loading dependence
LA volume index Cumulative filling-pressure/rhythm burden AF, mitral disease, obesity
EF Established systolic category Insensitive to early longitudinal dysfunction
Diastolic indices Filling-pressure probability Age, rhythm, load and multi-index algorithm

Mass/geometry predict outcomes but threshold prevalence changes with sex, body-size index and modality (Stewart 2018, PMID 30408469; de Simone 2002, PMID 12364349).

4. Strain

GLS often becomes abnormal before EF. It is directionally intuitive but varies by vendor, image quality, loading and sign convention.

An eight-study meta-analysis (1,140 hypertensive adults) found treatment-associated GLS improvement from −17.7% to −19.6%, alongside LV-mass index reduction; meta-regression linked GLS change to mass change but not SBP change (Tadic 2022, PMID 35102087).

In 1,600 serially imaged patients, LV mass-to-strain ratio outperformed mass or GLS alone for predicting LVH change (AUC 0.726 vs 0.690 and 0.600) and predicted cardiovascular death/HF hospitalization (Hwang 2025, PMID 40970541).

An earlier LV-strain risk score predicted outcomes in asymptomatic HHD, but external treatment utility remains untested (Saito 2016, PMID 27344417).

5. CMR mass and tissue characterization

Marker Tissue/phenotype Central limitation
Cine LV mass/volumes Geometry and function Access, indexing and reference cohort
LGE Focal scar/replacement fibrosis Diffuse fibrosis underdetected; etiology matters
Native T1 Composite tissue environment Platform and disease non-specificity
ECV Extracellular expansion Hematocrit, sequence and field-strength dependence
Interstitial volume ECV × myocardial volume Derived, body-size and volume dependent
Feature-tracking strain Mechanics Software and load dependence
Stress perfusion Microvascular ischemia CAD exclusion and technical requirements

CMR is the reference method for LV mass and offers tissue characterization unavailable to routine echo (Mavrogeni 2017, PMID 28535761; Tadic 2021, PMID 32170529).

6. Differential diagnosis performance

A 26-study meta-analysis found HHD lower than HCM in native T1 (Hedges g −0.469), ECV (−0.417), LV-mass index (−0.437) and maximal wall thickness (−2.076) (Zhao 2024, PMID 39156132).

In a 224-person CMR study, discrimination between HHD and HCM was modest for GLS (c=0.639), LV-mass index (0.643), native T1 (0.718) and LGE volume (0.680) (Neisius 2019, PMID 31433823).

A 314-person multiparametric study reported 82% sensitivity and 100% specificity for combined strain and atypical LGE in its HHD-versus-HCM comparison (Giusca 2021, PMID 34247623). The striking specificity is an internal cohort result, not a universal cutoff.

7. Fibrosis markers

Diffuse interstitial fibrosis is a plausible bridge from pressure load to stiffness, ischemia, electrical heterogeneity and HF, but each assay samples a different construct (González 2024, PMID 38084597).

Marker family Examples Specificity problem
CMR T1, ECV, interstitial volume Other cardiomyopathies, edema, CKD
Collagen synthesis PICP and related peptides Bone/systemic turnover
Collagen degradation MMP/TIMP-related signals Vascular and systemic sources
Inflammation CRP, cytokines Obesity/infection/comorbidity
Oxidative stress Multiple experimental analytes Assay and causal ambiguity

Reviews spanning pathology and noninvasive measurement emphasize that no circulating fibrosis marker is validated as an HHD diagnostic or treatment trigger (Cuspidi 2006, PMID 16263734; Weber 2004, PMID 15106793; González 2024, PMID 38084597).

8. Natriuretic peptide and troponin

Natriuretic peptides reflect myocardial wall stress; hs-troponin reflects cardiomyocyte injury. Both are prognostic across many cardiac and renal states and therefore cannot establish hypertensive attribution.

REMODEL used internally selected thresholds:

Marker state Event association
NT-proBNP <152 and hs-TnT <12.7 pg/mL Reference
One elevated HR 3.44 (95% CI 1.71–6.94)
Both elevated HR 17.11 (8.12–36.09)

Source: Sharp 2026 (PMID 41771092). Obesity can suppress natriuretic peptide, while CKD and AF can raise it; assay and population recalibration are essential (Obokata 2017, PMID 28381470; Redfield 2023, PMID 36917048).

PRECISE-HF demonstrates that covariate-conditioned recalibration is feasible and, in its study populations, outperformed universal rule-out and age-adjusted rule-in thresholds. Using 535,583 UK primary-care patients with an NT-proBNP measured for suspected HF (derivation n=374,909; internal validation n=160,674; HF recorded in 10% of the validation cohort), a gradient-boosted model combining NT-proBNP with age, sex, ethnicity, eGFR, BMI, systolic BP, anaemia, loop-diuretic prescription and history of AF, diabetes, myocardial infarction and COPD achieved AUROC 0.896 and Brier score 0.061. Its rule-out threshold had 90.1% sensitivity and 98.5% negative predictive value while ruling out 64.8% of patients; versus ESC thresholds it ruled out 144 additional patients per 1,000 at the cost of three missed diagnoses and produced 73 fewer false positives per 1,000 at rule-in. External validation in the Swedish REVOLUTION-HF cohort gave AUROC 0.757 and Brier score 0.163 (Docherty 2026, PMID 42663089).

Two limits keep this from resolving the HHD biomarker question. The target condition is diagnosed HF, not HHD attribution or pre-HF remodeling, and the drop in discrimination on external validation shows the calibration is setting-dependent. What it does establish is a template REMODEL has not yet followed: probabilistic thresholds conditioned on the exact covariates — age, kidney function, BMI, AF — that distort natriuretic peptides in hypertensive populations.

9. Treatment-response markers

Study Marker change Interpretation
LIFE ECG Lower Cornell product Associated with fewer events; not randomized mediator
LIFE echo Lower LV-mass index Associated with fewer events; serial substudy
STEP Less incident ECG-LVH Did not explain most intensive-treatment benefit
REVERSE-LVH Greater interstitial-volume fall with ARNI Phase 2, no clinical outcomes

Sources: Okin 2004 (PMID 15547161), Devereux 2004 (PMID 15547162), Deng 2023 (PMID 37259845), Lee 2025 (PMID 40739095).

10. Emerging multi-marker and machine-learning models

Recent studies combine ECG, echo, CMR and biomarkers, but development performance commonly exceeds external performance. Differential diagnosis, prognosis and treatment response are different tasks and require separate validation.

Required reporting Why
Locked outcome/case definition Prevent label leakage
External site and time validation Test transportability
Calibration and confidence intervals Avoid AUC-only reporting
Missing-data pathway Reflect clinical deployment
Comparison with simple model Establish incremental value
Decision-curve or trial utility Show consequences of use
Subgroup performance Detect inequity by sex/ancestry/obesity/CKD

11. Candidate-marker grading

Marker Analytical validity Clinical validity Clinical utility
ECG Cornell product High Prognostic Used in trials; not sole treatment target
Echo LV mass Moderate–high Prognostic Supports organ-damage assessment
GLS Improving Prognostic cohorts/meta-analysis No treat-to-GLS trial
CMR ECV/T1 High within platform Group/prognostic association No universal threshold
NT-proBNP + hs-TnT Established assays Single HHD derivation cohort No marker-triggered trial
Fibrosis peptides Variable Inconsistent None established

12. Minimum validation pathway

  1. Define HHD independently of the marker under test.
  2. Prespecify assay/imaging acquisition and thresholds.
  3. Validate in a new geography, ancestry mix and care setting.
  4. Report calibration, discrimination and reclassification.
  5. Test whether the marker changes a decision.
  6. Randomize the marker-guided strategy.
  7. Measure clinical benefit, harm, cost and equity.

Open questions

  • Will REMODEL thresholds calibrate across assays, obesity, CKD, age and ancestry? (Sharp 2026, PMID 41771092)
  • Which fibrosis metric—ECV, interstitial volume, LGE or a circulating panel—best predicts a modifiable event pathway? (González 2024, PMID 38084597; Lee 2025, PMID 40739095)
  • Can mass-to-strain ratio improve decisions beyond separate LV mass, GLS and clinical risk? (Hwang 2025, PMID 40970541)
  • What is the smallest interoperable HHD imaging dataset across echo and CMR? (Ismail 2023, PMID 37176563)
  • Would probabilistic, covariate-conditioned thresholds outperform the fixed REMODEL cutoffs for HHD staging, as they do for HF diagnosis? (Sharp 2026, PMID 41771092; Docherty 2026, PMID 42663089)

References

  1. Pewsner D, et al. Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review. BMJ. 2007;335:711. PMID 17726091
  2. Zhao Q, et al. Cardiac magnetic resonance imaging for discrimination of hypertensive heart disease and hypertrophic cardiomyopathy: a systematic review and meta-analysis. Front Cardiovasc Med. 2024;11:1421013. PMID 39156132
  3. Sharp A, et al. Role of natriuretic peptides and cardiac troponins in staging hypertensive heart disease: the REMODEL study. Eur J Heart Fail. 2026. PMID 41771092
  4. Hwang IC, et al. Left Ventricular Mass-to-Strain Ratio to Predict Change in Left Ventricular Hypertrophy and Prognosis in Hypertensive Heart Disease. J Am Heart Assoc. 2025;14:e042032. PMID 40970541
  5. Lertsiripatarajit P, et al. Cardiac magnetic resonance assessment of left ventricular phenotypes and prognostic implications in hypertensive heart disease. Clin Hypertens. 2026;32:e14. PMID 41953289
  6. Nwabuo CC, et al. Pathophysiology of Hypertensive Heart Disease: Beyond Left Ventricular Hypertrophy. Curr Hypertens Rep. 2020;22:11. PMID 32016791
  7. González A, et al. Myocardial Interstitial Fibrosis in Hypertensive Heart Disease: From Mechanisms to Clinical Management. Hypertension. 2024;81:218-228. PMID 38084597
  8. Bressman M, et al. Determination of Sensitivity and Specificity of Electrocardiography for Left Ventricular Hypertrophy in a Large, Diverse Patient Population. Am J Med. 2020;133:e495-e500. PMID 32194027
  9. Okin PM, et al. Regression of electrocardiographic left ventricular hypertrophy during antihypertensive treatment and the prediction of major cardiovascular events. JAMA. 2004;292:2343-9. PMID 15547161
  10. Stewart MH, et al. Prognostic Implications of Left Ventricular Hypertrophy. Prog Cardiovasc Dis. 2018;61:446-455. PMID 30408469
  11. de Simone G, et al. Prognosis of inappropriate left ventricular mass in hypertension: the MAVI Study. Hypertension. 2002;40:470-6. PMID 12364349
  12. Tadic M, et al. Effect of long-term antihypertensive therapy on myocardial strain: a meta-analysis. J Hypertens. 2022;40:641-647. PMID 35102087
  13. Saito M, et al. Prognostic Implications of LV Strain Risk Score in Asymptomatic Patients With Hypertensive Heart Disease. JACC Cardiovasc Imaging. 2016;9:911-21. PMID 27344417
  14. Mavrogeni S, et al. The emerging role of Cardiovascular Magnetic Resonance in the evaluation of hypertensive heart disease. BMC Cardiovasc Disord. 2017;17:132. PMID 28535761
  15. Tadic M, et al. Comprehensive assessment of hypertensive heart disease: cardiac magnetic resonance in focus. Heart Fail Rev. 2021;26:1383-1390. PMID 32170529
  16. Neisius U, et al. Cardiovascular magnetic resonance feature tracking strain analysis for discrimination between hypertensive heart disease and hypertrophic cardiomyopathy. PLoS One. 2019;14:e0221061. PMID 31433823
  17. Giusca S, et al. Multi-parametric assessment of left ventricular hypertrophy using late gadolinium enhancement, T1 mapping and strain-encoded cardiovascular magnetic resonance. J Cardiovasc Magn Reson. 2021;23:92. PMID 34247623
  18. Cuspidi C, et al. Hypertensive myocardial fibrosis. Nephrol Dial Transplant. 2006;21:20-3. PMID 16263734
  19. Weber KT. Fibrosis in hypertensive heart disease: focus on cardiac fibroblasts. J Hypertens. 2004;22:47-50. PMID 15106793
  20. Obokata M, et al. Evidence Supporting the Existence of a Distinct Obese Phenotype of Heart Failure With Preserved Ejection Fraction. Circulation. 2017;136:6-19. PMID 28381470
  21. Redfield MM, et al. Heart Failure With Preserved Ejection Fraction: A Review. JAMA. 2023;329:827-838. PMID 36917048
  22. Devereux RB, et al. Prognostic significance of left ventricular mass change during treatment of hypertension. JAMA. 2004;292:2350-6. PMID 15547162
  23. Deng Y, et al. Intensive Blood Pressure Lowering Improves Left Ventricular Hypertrophy in Older Patients with Hypertension: The STEP Trial. Hypertension. 2023;80:1834-1842. PMID 37259845
  24. Lee V, et al. Effects of sacubitril/valsartan on hypertensive heart disease: the REVERSE-LVH randomized phase 2 trial. Nat Commun. 2025;16:6981. PMID 40739095
  25. Ismail TF, et al. Hypertensive Heart Disease-The Imaging Perspective. J Clin Med. 2023;12. PMID 37176563
  26. Docherty KF, et al. Personalized evaluation of NT-proBNP and clinical characteristics in suspected heart failure: the PRECISE-HF study. Eur Heart J. 2026;:ehag686. PMID 42663089