Skip to content

Biomarkers

TL;DR — A live PubMed E-utilities sweep on 2026-08-30 found many candidate and within-sample “validation” studies, but no externally validated, clinically adopted blood, tissue, imaging or biomechanical test that diagnoses symptomatic DDD, identifies a painful level, predicts treatment response or monitors regeneration. Conventional MRI grade is reproducible but age-sensitive; quantitative MRI, serum matrix fragments, inflammatory mediators, proteomics and transcriptomic classifiers remain research tools (Pfirrmann 2001, PMID 11568697; Russo 2023, PMID 37247638; Khan 2017, PMID 29265416). Most biomarker studies are small, cross-sectional, spectrum-biased and lack external validation. The critical task is not discovery of another correlated signal, but prospective proof of incremental clinical utility.

Intended uses

Biomarker use Clinical question Required comparator
Susceptibility Who will develop degeneration? Age, genetics, exposures
Diagnostic Is degeneration present? Standardized imaging/pathology
Pain-source Which level/structure generates pain? Independent reference standard
Prognostic Who will worsen? Clinical/imaging model
Predictive Who benefits from treatment? Randomized interaction
Pharmacodynamic Did target biology change? Predefined dose-response
Surrogate Does marker capture clinical benefit? Multi-trial causal validation
Safety Is tissue injury occurring? Adjudicated adverse outcome

A marker can succeed for one use and fail for another.

Validation ladder

  1. Analytical validity: accurate and reproducible measurement.
  2. Biological association: differs by relevant state.
  3. Clinical validity: predicts diagnosis/prognosis externally.
  4. Incremental value: improves a clinical model.
  5. Clinical utility: changes decisions and outcomes.
  6. Surrogacy: treatment effect on marker predicts treatment effect on clinical outcomes.

Most DDD candidates remain at steps 1–2.

Conventional MRI markers

Marker Strength Limitation
Pfirrmann grade Reliable standardized ordinal morphology Age and reader effects (PMID 11568697)
Disc height Quantitative geometry Position/time of day
HIZ Possible annular fissure/inflammation Asymptomatic prevalence
Modic type Endplate/marrow phenotype Heterogeneous pain association
Endplate defect Mechanistic plausibility Detection/definition variation
Herniation morphology Anatomical localization Symptoms require concordance

Asymptomatic degeneration rises from 37% at age 20 to 96% at age 80, demonstrating poor stand-alone pain specificity (Brinjikji 2015, PMID 25430861).

Symptomatic-control meta-analysis shows several MRI features are associated with pain, but association is not individual-level classification (Brinjikji 2015, PMID 26359154).

Quantitative MRI

Technique Candidate substrate Validation gap
T2 mapping Water/collagen Protocol and pain specificity
T2* Collagen organization Field/sequence harmonization
T1ρ Proteoglycan/macromolecule Thresholds and availability
Sodium MRI Fixed charge density Signal, scan duration
Diffusion/ADC Water mobility Artifacts and overlap
DCE-MRI Endplate enhancement Contrast and modeling
MR spectroscopy Metabolites/lipids Reproducibility and voxel size

Systematic review found these techniques can detect compositional changes but studies use heterogeneous acquisition, segmentation and reference standards (Russo 2023, PMID 37247638).

T2 mapping correlates with Pfirrmann grade, yet grade-group distributions overlap (Stefanou 2023, PMID 39119364). Multi-parameter MRI studies remain small (Xiong 2018, PMID 29511438).

Sodium MRI is biologically attractive as a proteoglycan proxy, but recent clinical feasibility does not establish diagnostic utility (Nakahashi 2025, PMID 40522507).

Endplate perfusion

DCE-MRI can measure enhancement near endplates. A nine-patient, 45-disc study observed associations between enhancement pattern, level and degeneration grade (Muftuler 2015, PMID 25421547).

Before clinical use it needs:

  • test–retest reliability;
  • standardized regions and kinetic model;
  • correction for age/vascular disease;
  • prospective pain and progression prediction;
  • treatment interaction;
  • contrast safety and practicality.

Inflammatory blood markers

Candidate markers include CRP, IL-1 family, IL-6, IL-8, TNF and chemokines. A review found inconsistent associations between inflammatory biomarkers, low-back pain and disc degeneration (Khan 2017, PMID 29265416).

Confounder Why important
Obesity Raises systemic inflammation
Smoking Vascular/inflammatory effects
Infection Can mimic Modic type 1
Arthritis/autoimmunity Non-disc cytokine source
Exercise/time of day Acute marker variation
Medication NSAID/steroid effects
Pain chronicity Central and systemic consequences

Systemic cytokines lack level specificity even if associated with disease.

Matrix turnover markers

Candidates include collagen neoepitopes, aggrecan fragments, COMP and metalloproteinases. Serum COMP and ADAMTS7 were studied as diagnostic candidates, but one case-control study cannot establish generalizable thresholds (Ding 2024, PMID 38528617).

Serum periostin correlated with MRI severity in a clinical study, requiring replication and adjustment for bone and other tissue sources (Morimoto 2024, PMID 38340176).

Ferritin and ACE have also been reported in association studies, but both are systemic, nonspecific analytes with many determinants (Guo 2022, PMID 36008563; Guo 2022, PMID 36055434).

Tissue biomarkers

Surgical disc tissue permits matrix, cell, cytokine and omic analysis but introduces severe spectrum bias. Patients undergoing surgery differ from asymptomatic imaging cases and nonoperative chronic pain.

MMP expression correlates with Pfirrmann grade in fusion samples, demonstrating biological coherence but not a noninvasive diagnostic test (Aripaka 2022, PMID 36248158).

Serglycin secreted by late-stage nucleus pulposus cells was proposed as a degeneration biomarker; translation requires measurable circulating/local signal and external clinical validation (Chen 2024, PMID 38167807).

Genomic markers

Common variants influence susceptibility, but effect sizes are too small and heterogeneous for a pain-level test. Candidate-gene reviews emphasize inconsistent replication (Mayer 2013, PMID 23537453).

Genomic use Current status
Population susceptibility Research associations
Early screening Not validated
Pain-source diagnosis No evidence
Procedure selection No evidence
Biologic-treatment selection No evidence

A chromosome 9 locus associated with Modic change shows phenotype-specific discovery is possible but not clinical prediction (Freidin 2019, PMID 30808802).

Transcriptomics and single-cell markers

Bulk and single-cell studies identify inflammatory, fibrotic, stress and progenitor-like states. A proposed SFRP4-secreted fibro-NPC subtype illustrates how single-cell discovery can generate a candidate marker, but surgical sampling, batch effects and cluster instability remain (Xu 2025, PMID 40770348).

Bioinformatic immune-infiltration and ferroptosis signatures frequently use public datasets with small sample sizes and risk of overfitting (Wang 2025, PMID 40993805; Zhou 2025, PMID 40042449).

Minimum replication requires an independent cohort, locked assay, predefined cutoff and clinical endpoint.

Proteomics and metabolomics

Omics can identify patterns not visible in single-analyte assays. Risks include batch effects, multiple testing, diet/medicine confounding and optimistic internal validation.

Stage Required practice
Discovery False-discovery control and full feature reporting
Model building Nested cross-validation
Locking Freeze features and coefficients
External validation New site/scanner/ancestry
Utility Decision-impact trial

An MRI-spectroscopy lipid peak linked to metabolomic/proteomic inflammatory pathways is hypothesis-generating; commentary and replication are needed before “biomarker” implies use (Vadalà 2025, PMID 41508583).

Biomechanical biomarkers

Candidates include segmental motion, stiffness, disc pressure, shear, finite-element stress and paraspinal activation. They may be closer to mechanical symptoms than morphology but depend on modeling assumptions and measurement conditions.

Kinematic MRI studies show changing disc morphology with position but have not established a validated pain threshold (Roberts 2021, PMID 33940491).

Pain phenotyping markers

Quantitative sensory testing, conditioned pain modulation, neuroimaging and psychosocial profiles may characterize pain processing. They are not disc-specific and should not be presented as structural diagnosis.

A useful model may combine:

  • disc/endplate structure;
  • root/facet/hip exclusions;
  • pain distribution;
  • sensory phenotype;
  • inflammation/metabolism;
  • psychological and sleep factors.

The model must outperform simpler clinical assessment in external prospective data.

Predictive biomarkers

Prediction requires treatment-by-marker interaction in randomized data. Improvement among biomarker-positive treated patients is insufficient because the marker may be prognostic.

Candidate Needed test
Modic 1 vs 2 for BVN ablation Randomized interaction, not subgroup response
Cytokines for antibiotics Replicated prespecified interaction
Disc volume for cell therapy Mediation and surrogate validation
Pfirrmann grade for PRP Predefined differential efficacy
Endplate perfusion for cells Treatment interaction and safety

Exploratory cytokine analyses from the AIM antibiotic study remain hypothesis-generating after the overall strategy failed (Bråten 2023, PMID 37252109; Grotle 2020, PMID 32546490).

Machine learning

Automated Pfirrmann models can reproduce reader labels, but this is a measurement task, not pain diagnosis (Gao 2021, PMID 33094867; Liawrungrueang 2024, PMID 39659374).

Model reports should include calibration, external validation, decision curves, missing data, subgroup performance and prospective impact. Accuracy without prevalence-calibrated predictive values is misleading.

Minimum biomarker study

Domain Standard
Intended use Declared before analysis
Population Consecutive and clinically relevant
Assay Blinded, reproducible, quality controlled
Reference Independent and prespecified
Sample size Events-per-parameter justified
Validation External, not random split only
Comparator Standard clinical model
Output Calibration, discrimination, net benefit
Transparency Protocol, code/features, funding

Evidence deepening: discriminating findings (2026-08-30)

The added evidence below was selected to change interpretation, not merely increase citation count. Each result is kept within its studied phenotype and design.

Evidence Quantified or mechanistic finding Consequence for interpretation
Analysis of miRNA-199-5p expression levels in serum samples of patients with lumbar disc degeneration (Akdeniz 2023, PMID 38063113) A serum study evaluated miR-199-5p expression against lumbar degeneration severity. Association and group separation do not establish individual diagnostic calibration.
Serum CXCL12/SDF-1 level is positively related with lumbar intervertebral disc degeneration and clinical severity (Er 2020, PMID 31852328) Serum CXCL12/SDF-1 correlated positively with lumbar degeneration and clinical severity in a clinical cohort. A candidate inflammatory marker requires adjustment for systemic inflammation and external replication.
Intervertebral Disc Elastography to Relate Shear Modulus and Relaxometry in Compression and Bending (Davis 2023, PMID 37732250) Disc elastography related shear modulus to MR relaxometry under compression and bending. A mechanically interpretable biomarker may complement composition, but acquisition and loading must be standardized.
Spatial geometric and magnetic resonance signal intensity changes with advancing stages of nucleus pulposus degeneration (Yang 2017, PMID 29162082) Spatial geometry and MR signal changed across advancing nucleus-pulposus degeneration stages. Region-aware measures may outperform a whole-disc mean.
Association Between the FokI and ApaI Polymorphisms in the Vitamin D Receptor Gene and Intervertebral Disc Degeneration: A Systematic Review and Meta-Analysis (Pabalan 2017, PMID 27797588) A meta-analysis assessed FokI and ApaI vitamin-D-receptor polymorphisms against degeneration. Candidate-gene associations require ancestry-aware replication and should not be used clinically.
MRI-based deep learning radiomics model for automated classification of disc degeneration in the lumbar spine (Patil 2026, PMID 42218311) A deep-learning radiomics model used 218 symptomatic patients and thousands of imaging features before feature reduction. Internal classification accuracy is not clinical utility; external calibration and asymptomatic controls are required.

Controversy carried forward

These additions narrow several claims but do not create a diagnostic gold standard. Where an imaging, molecular or treatment-response signal conflicts with sham-controlled, longitudinal or population evidence, the conflict is retained as a selection and transportability problem rather than resolved by vote.

Open questions

  • Can any biomarker identify painful versus painless degeneration within the same MRI grade? (Khan 2017, PMID 29265416)
  • Which quantitative MRI measure is reproducible across scanners and predicts longitudinal outcomes? (Russo 2023, PMID 37247638)
  • Does endplate perfusion predict cell-therapy survival or response? (Muftuler 2015, PMID 25421547)
  • Can multi-omic signatures survive locked external validation? (Xu 2025, PMID 40770348)
  • What marker–treatment interactions are strong enough to alter care? (Bråten 2023, PMID 37252109)

References

  1. Pfirrmann CW, Metzdorf A, Zanetti M, et al. Magnetic resonance classification of lumbar intervertebral disc degeneration. Spine. 2001;26(17):1873-8. PMID 11568697
  2. Russo F, Ambrosio L, Giannarelli E, et al. Innovative quantitative magnetic resonance tools to detect early intervertebral disc degeneration changes: a systematic review. The spine journal : official journal of the North American Spine Society. 2023;23(10):1435-1450. PMID 37247638
  3. Khan AN, Jacobsen HE, Khan J, et al. Inflammatory biomarkers of low back pain and disc degeneration: a review. Annals of the New York Academy of Sciences. 2017;1410(1):68-84. PMID 29265416
  4. Brinjikji W, Luetmer PH, Comstock B, et al. Systematic literature review of imaging features of spinal degeneration in asymptomatic populations. AJNR. American journal of neuroradiology. 2015;36(4):811-6. PMID 25430861
  5. Brinjikji W, Diehn FE, Jarvik JG, et al. MRI Findings of Disc Degeneration are More Prevalent in Adults with Low Back Pain than in Asymptomatic Controls: A Systematic Review and Meta-Analysis. AJNR. American journal of neuroradiology. 2015;36(12):2394-9. PMID 26359154
  6. Stefanou DE, Velonakis G, Karavasilis E, et al. T2 mapping of lumbar intervertebral disc: quantitative evaluation of degeneration in relation to Pfirrmann grading system and a template for intervertebral disc segmentation. Hippokratia. 2023;27(3):75-81. PMID 39119364
  7. Xiong X, Zhou Z, Figini M, et al. Multi-parameter evaluation of lumbar intervertebral disc degeneration using quantitative magnetic resonance imaging techniques. American journal of translational research. 2018;10(2):444-454. PMID 29511438
  8. Nakahashi Y, Saida T, Yoshida M, et al. Evaluation of lumbar intervertebral disc degeneration using 23Na-MRI in clinical settings. Skeletal radiology. 2025;54(12):2725-2734. PMID 40522507
  9. Muftuler LT, Jarman JP, Yu HJ, et al. Association between intervertebral disc degeneration and endplate perfusion studied by DCE-MRI. European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society. 2015;24(4):679-85. PMID 25421547
  10. Ding JY, Yan X, Zhang RJ, et al. Diagnostic value of serum COMP and ADAMTS7 for intervertebral disc degeneration. European journal of medical research. 2024;29(1):196. PMID 38528617
  11. Morimoto T, Kobayashi T, Ito H, et al. Serum periostin levels correlate with severity of intervertebral disc degeneration. European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society. 2024;33(5):2007-2013. PMID 38340176
  12. Guo Y, Li C, Shen B, et al. Is intervertebral disc degeneration associated with reduction in serum ferritin? European spine journal : official publication of the European Spine Society, the European Spinal Deformity Society, and the European Section of the Cervical Spine Research Society. 2022;31(11):2950-2959. PMID 36008563
  13. Guo Y, Guo K, Hu T, et al. Correlation between serum angiotensin-converting enzyme (ACE) levels and intervertebral disc degeneration. Peptides. 2022;157:170867. PMID 36055434
  14. Aripaka SS, Bech-Azeddine R, Jørgensen LM, et al. The expression of metalloproteinases in the lumbar disc correlates strongly with Pfirrmann MRI grades in lumbar spinal fusion patients. Brain & spine. 2022;2:100872. PMID 36248158
  15. Chen F, Lei L, Chen S, et al. Serglycin secreted by late-stage nucleus pulposus cells is a biomarker of intervertebral disc degeneration. Nature communications. 2024;15(1):47. PMID 38167807
  16. Mayer JE, Iatridis JC, Chan D, et al. Genetic polymorphisms associated with intervertebral disc degeneration. The spine journal : official journal of the North American Spine Society. 2013;13(3):299-317. PMID 23537453
  17. Freidin M, Kraatari M, Skarp S, et al. Genome-wide meta-analysis identifies genetic locus on chromosome 9 associated with Modic changes. Journal of medical genetics. 2019;56(7):420-426. PMID 30808802
  18. Xu Y, Xie Z, Gu S, et al. Fibro-NPC: a pathogenic subtype identified at single-cell resolution with secreted SFRP4 as a biomarker in intervertebral disc degeneration. Journal of translational medicine. 2025;23(1):867. PMID 40770348
  19. Wang Y, Hu B, Tian L, et al. Comprehensive profiling of immune cell infiltration and biomarker identification in intervertebral disc degeneration. Journal of orthopaedic surgery and research. 2025;20(1):829. PMID 40993805
  20. Zhou Y, Wang K, Ren M, et al. Identification and functional validation of ACSL1 as a biomarker regulating ferroptosis in nucleus pulposus cell. Bioscience reports. 2025;45(4):215-31. PMID 40042449
  21. Vadalà G, Ambrosio L, Russo F, et al. A Commentary on "Magnetic Resonance Spectroscopy Lipids Peak May Serve as a Potential Biomarker for Back Pain in Intervertebral Disc Degeneration: An Integrative Metabolomics and Proteomics Study Investigating the Role of the Lipid Droplets-Interleukin-17 Inflammatory Axis". Neurospine. 2025;22(4):934-936. PMID 41508583
  22. Roberts S, Gardner C, Jiang Z, et al. Analysis of trends in lumbar disc degeneration using kinematic MRI. Clinical imaging. 2021;79:136-141. PMID 33940491
  23. Bråten LCH, Gjefsen E, Gervin K, et al. Cytokine Patterns as Predictors of Antibiotic Treatment Effect in Chronic Low Back Pain with Modic Changes: Subgroup Analyses of a Randomized Trial (AIM Study). Journal of pain research. 2023;16:1713-1724. PMID 37252109
  24. Grotle M, Bråten LC, Brox JI, et al. Cost-utility analysis of antibiotic treatment in patients with chronic low back pain and Modic changes: results from a randomised, placebo-controlled trial in Norway (the AIM study). BMJ open. 2020;10(6):e035461. PMID 32546490
  25. Gao F, Liu S, Zhang X, et al. Automated Grading of Lumbar Disc Degeneration Using a Push-Pull Regularization Network Based on MRI. Journal of magnetic resonance imaging : JMRI. 2021;53(3):799-806. PMID 33094867
  26. Liawrungrueang W, Cholamjiak W, Sarasombath P, et al. Artificial Intelligence Classification for Detecting and Grading Lumbar Intervertebral Disc Degeneration. Spine surgery and related research. 2024;8(6):552-559. PMID 39659374
  27. Akdeniz FT, Barut Z. Analysis of miRNA-199-5p expression levels in serum samples of patients with lumbar disc degeneration. Cellular and molecular biology (Noisy-le-Grand, France). 2023;69(12):83-87. PMID 38063113
  28. Er ZJ, Yin CF, Wang WJ, Chen XJ. Serum CXCL12/SDF-1 level is positively related with lumbar intervertebral disc degeneration and clinical severity. Innate immunity. 2020;26(5):341-350. PMID 31852328
  29. Davis ZR, Gossett PC, Wilson RL, Kim W, Mei Y, Butz KD, et al. Intervertebral Disc Elastography to Relate Shear Modulus and Relaxometry in Compression and Bending. bioRxiv : the preprint server for biology. 2023:2023.09.01.555817. PMID 37732250
  30. Yang SH, Espinoza Orías AA, Pan CC, Senoo I, Andersson GBJ, An HS, et al. Spatial geometric and magnetic resonance signal intensity changes with advancing stages of nucleus pulposus degeneration. BMC musculoskeletal disorders. 2017;18(1):473. PMID 29162082
  31. Pabalan N, Tabangay L, Jarjanazi H, Vieira LA, Dos Santos AA, Barbosa CP, et al. Association Between the FokI and ApaI Polymorphisms in the Vitamin D Receptor Gene and Intervertebral Disc Degeneration: A Systematic Review and Meta-Analysis. Genetic testing and molecular biomarkers. 2017;21(1):24-32. PMID 27797588
  32. Patil S, Gandhi OH, Karabacak M, Carr MT, Margetis K. MRI-based deep learning radiomics model for automated classification of disc degeneration in the lumbar spine. Scientific reports. 2026;16(1):24695. PMID 42218311