Modic changes: updated insights into back pain, MRI diagnosis, surgical outcomes, and emerging treatments

Background and definitions of Modic changes

MRI classification, grading (MCG), and clinical meaning

MC are vertebral bone‑marrow signal changes adjacent to the endplate of a degenerating disc. Classic MRI phenotypes are: MC1 (T1 low/T2 high ± STIR hyperintensity; fibrovascular/inflammatory marrow), MC2 (T1 high/T2 iso–high; fatty remodeling), and MC3 (T1/T2 low; sclerosis). Natural history commonly includes MC1→MC2 conversion with occasional progression to sclerosis. 
Beyond type, extent matters: the Modic–Udby definition and Modic Change Grade (MCG) categorize vertical involvement—A <25% (minor), B 25–50% (major), C >50% of vertebral body height/volume—enhancing reliability and clinical utility.

Epidemiology and types of Modic changes

Prevalence, phenotypes, and progression in spine degeneration

Prevalence increases with age and overall degenerative burden, with a predilection for lower lumbar levels and the anterior vertebral third. Population‑based cohorts (e.g., ISSLS Prize–winning twin work) link MC to subsequent episodes of severe, disabling low back pain independent of other degenerative features. 

Clinically, MC occur more frequently in symptomatic cohorts than in asymptomatic controls. Interpret MC within the degenerative cascade—endplate defects, disc height and signal loss, facet arthropathy, sagittal alignment, and degenerative spinal deformity. Mixed type MCs and interconversion has been described.

Imaging and measurement of Modic changes—from labels to quantitative burden

MRI protocols, lesion burden (MCG), and quantitative methods

MRI protocols should include T1‑ and T2‑weighted sequences with STIR to capture marrow edema. To limit misclassification, standardize sequences and strict structured reporting should be emphasized. Quantitative MRI can refine phenotyping: proton‑density fat fraction (PDFF) differentiates fatty from fibrovascular marrow, and ultrashort‑echo‑time (UTE) MRI improves CEP visualization. Machine‑learning/radiomics are emerging for automated MC detection and extent estimation, enabling reproducible multicentre research and pragmatic registries.
Descriptive reporting template: “L4–5: MC1, MCG‑B, anterior two‑thirds; CEP irregularity; endplate defect present.” This captures type, burden (extent), topography, and the endplate context most relevant to biomechanics and pain.

Pathobiology of Modic changes—endplate–marrow biology and pain mechanisms

Endplate–marrow biology, degeneration models, and bacterial debate

The etiology of MC is likely multifactorial. A consistent observation across cohorts is that MC occur only in the presence of disc degeneration, which strongly implicates degeneration as the primary driver. A unifying mechanobiologic model posits that endplate microdamage impairs cartilage endplate (CEP) transport and permits leakage of disc-derived danger signals into adjacent vertebral marrow. The resultant immune–fibrovascular response manifests as MC1 (heightened turnover/inflammatory signalling), followed by fatty remodelling (MC2) and, in a subset, sclerosis (MC3). Nociception likely arises from chemical and mechanical stimulation of nociceptors within the injured endplate–marrow complex.

The role of bacteria in MC remains uncertain. While some biopsy series report positive sampling (often Cutibacterium acnes) in up to ~70% of samples, most of these data derive from disc tissue—the body’s largest avascular structure—rather than from intraosseous vertebral marrow, which is the actual location of MC. Given the frequent detection of skin commensals, perioperative contamination is a plausible explanation for many “positive” findings. To date, no study has conclusively demonstrated bacteria within the bony vertebra at the MC site, leaving infection as an unproven—and likely minority—pathway relative to degeneration-driven mechanisms.

Clinical Associations with low back pain and disability

Evidence linking Modic changes to symptoms and long-term outcomes 

Longitudinal population studies identify MC/endplate damage as independent predictors of severe, disabling LBP episodes. In specialty cohorts, MC1 and higher‑burden lesions correlate more strongly with pain; effect sizes vary with sampling, imaging standards, and analytic approach. Interestingly, in long-term cohort studies (+10 years) of patients with baseline chronic LBP, the presence of MC at baseline did not portend worse outcomes; patients with MC actually demonstrated less disability and fewer sick‑leave days at long‑term follow‑up—reinforcing that label ≠ destiny and that MC biology is dynamic.

Nonoperative management of Modic changes

Conservative care, antibiotics, and basivertebral nerve ablation (BVNA)

Conservative care—education, graded exercise, and analgesia/NSAIDs—remains first‑line for nonspecific LBP and suffices for many patients with incidental or low‑burden MC. Antibiotics: an initial single‑centre RCT suggested benefit of prolonged amoxicillin‑clavulanate in post‑herniation MC1, but the larger multicentre AIM RCT did not demonstrate clinically important benefit; routine antibiotics are not recommended absent clear infection. BVNA: studies including long‑term follow‑up support durable improvements in pain and function for carefully selected patients with vertebrogenic pain and lumbar MC1/2.

Surgical implications of Modic changes

Impact on fusion, discectomy, spinal stenosis, and patient selection

Operative indications are usually driven by neural compression, instability, deformity, and global sagittal balance rather than MC alone. Systematic reviews and meta‑analyses indicate that preoperative MC do not consistently impact outcomes after lumbar discectomy, fusion, or disc arthroplasty. Some series suggest nuanced effects—e.g., early fusion kinetics or implant–endplate interface behaviour may vary with lesion extent/type—but data are heterogeneous. In lumbar spinal stenosis, lesion extent carries prognostic value: major MC are associated with inferior 2‑year VAS‑back/leg and ODI compared with minor MC, despite similar baselines.

Practical algorithms for MRI reading and checklists for clinical decision-making

Structured reporting checklists and treatment pathways

A) MRI reading & reporting checklist

• Modic type (1/2/3) at each level
• Extent: MCG‑A (<25%), MCG‑B (25–50%), MCG‑C (>50%)
• Topography: anterior/middle/posterior third; uni‑/bilateral across midline
• Cartilage endplate (CEP) integrity and bony endplate defects
• Adjacent disc grade; foraminal/central stenosis; facet arthropathy
• Optional quantitative markers: PDFF (%), STIR edema burden

B) Clinic pathway (simplified)

• Chronic axial LBP with MC1/2 at concordant levels → consider vertebrogenic pain phenotype
• Failed conservative care → evaluate BVNA candidacy (inclusion/exclusion per RCT criteria)
• Antibiotics only within trials or with compelling infectious evidence
• For surgical candidates, incorporate MCG into shared decision‑making; set expectations when major MC are present

Quick evidence tables on Modic changes (summary)

MRI types, histopathology, burden grades, and clinical implications

Modic type — MRI signature — Histopathology — Clinical notes

1. Type 1 — T1 low / T2 high ± STIR edema — Fibrovascular marrow, high turnover — Often more symptomatic; dynamic; can convert to Type 2
2. Type 2 — T1 high / T2 iso–high — Fatty marrow replacement/remodeling — May be less painful on average; prior Type 1 history common
3. Type 3 — T1 low / T2 low — Sclerosis — Denotes chronicity; limited near‑term reversibility

MCG Grade — Vertical extent — Typical implications

• A: <25% of vertebral body height/volume — Often incidental/small burden; monitor + conservative care
• B: 25–50% — Higher baseline disability; closer follow‑up; consider targeted options if refractory
• C: >50% — High burden; careful counselling pre‑op; risk of persistent pain; integrate vertebrogenic strategies

 

The research frontier— future research on Modic changes (2025–2030)

Precision phenotyping, quantitative MRI, immunomodulation, and surgical decision tools

• Phenotyping MC1 into bacterial vs. sterile subtypes (culture/omics, immunophenotyping, imaging correlates)
• Quantitative MRI endpoints (PDFF, UTE‑CEP) and automated extent scoring for trials/registries
• Osteoimmunomodulation and endplate‑targeted repair strategies; biomaterials for CEP micro‑repair
• BVNA optimization: imaging selection rules by MCG/topography; health‑economic evaluation across systems
• Prospective surgical decision tools combining MCG with CEP/endplate metrics
• Natural‑history cohorts (dense imaging triggers) and extension to cervical/thoracic MC

 

Clinical key takeaways

Modic changes represent a dynamic endplate–marrow process that co-evolves with disc degeneration and can play a significant role in vertebrogenic back pain. For clinicians, the extent of the lesion—captured by the Modic Change Grade (MCG)—is often more predictive of disability and outcomes than type classification alone. In nonoperative care, education, exercise, and analgesia remain the foundation, while basivertebral nerve ablation has emerged as a valuable option for appropriately selected patients with MC1/2 after failing conservative therapy. Routine use of antibiotics is not supported outside of trials.

In surgery, the presence of Modic changes rarely defines the indication on its own, yet high-burden lesions (particularly MCG-B and C) can guide decision-making, shape prognostic conversations, and help set patient expectations—especially in lumbar fusion and spinal stenosis cases. Looking forward, advances in precision phenotyping, quantitative MRI, and endplate-targeted repair strategies promise to refine diagnosis and allow for more tailored interventions. For now, integrating MCG into both reporting and clinical discussions provides a practical framework to bridge imaging with patient care.

 

Figures?

Text 4 here Text 4