Figure 1. Typical Modic changes type 1 (high signal intensity) in a lumbar segment.
Chronic low back pain (CLBP) affects approximately one billion people globally and remains among the leading causes of years lived with disability. For decades, CLBP was understood primarily through biomechanical models – focusing on disc herniation, spinal stenosis, and structural degeneration. However, this mechanical paradigm could not explain a troubling paradox: many patients with pathological imaging findings experience minimal pain, while others with subtle structural findings suffer from disabling symptoms.
Over the past 30 years, experts have increasingly turned toward the biopsychosocial explanatory model as the predominant paradigm.
The discovery of Modic changes (MC) provided a critical piece of this puzzle and challenged purely psychosocial explanations as the sole driver of chronic back pain. Perhaps for this reason, Modic changes were often neglected in clinical practice for many years. Only in the 2010s – when robust epidemiological studies demonstrated a strong correlation between MC and pain severity – did Modic research gain prominence in spine medicine.
Most recently, we have been witnessing yet another paradigm shift. Newer and more sophisticated microbiological and imaging technologies have provided compelling evidence that Modic Type 1 changes may be driven by low-virulence bacterial infection within the intervertebral disc. This so-called 'infection hypothesis' has significant clinical implications, as it opens the possibility of targeted antibiotic therapy for a substantial subset of CLBP patients in whom conventional treatment has proven ineffective.
Modic changes are focal areas of bone marrow oedema and inflammation within the vertebral endplates – the cartilaginous interface between the intervertebral disc and the vertebral body. Unlike universal age-related disc degeneration, Modic changes represent active pathology: inflammation, edema, and abnormal bone formation that correlate strongly with pain severity and functional disability.
Modic's original 1988 classification identified two distinct types based on MR signal characteristics:
MC-1 changes appear hypointense (dark) on T1-weighted MR and hyperintense (bright) on T2-weighted sequences, indicating edema and active inflammation. Type 1 changes are the most clinically significant and most painful subtype, associated with:
Figure 1. Typical Modic changes type 1 (high signal intensity) in a lumbar segment.
MC-2 changes appear hyperintense on both T1- and T2-weighted sequences, indicating fatty infiltration of bone marrow. Type 2 changes represent approximately 50–60% of all Modic cases and typically:
Clinical Significance of Type Transitions: An important observation is that MC-1 can progress to MC-2 over months to years, reflecting chronification of acute inflammation into permanent fatty degeneration. This progression supports early intervention in Type 1 disease before changes become irreversible. It can be difficult to distinguish between Type 1 and Type 2 on MR. In many cases, both Type 1 and Type 2 exist in the same disc, or Type 1 in one disc and Type 2 in adjacent discs.
Modic changes are present in approximately 30–45% of patients with CLBP compared to only 15% of asymptomatic adults. This strong association, combined with robust correlations between MC volume and pain severity, establishes MC as a significant and measurable marker for CLBP pathology.
Clinically, patients with MC-1 often report:
These clinical characteristics distinguish MC-related pain from mechanical causes – such as isolated disc herniation – or simple age-related degeneration.
In 2008, researchers from Denmark proposed a novel hypothesis: that Modic Type 1 changes might result from low-virulence bacterial infection within the intervertebral disc, rather than purely inflammatory or mechanical processes. This hypothesis arose from observations that certain bacterial species – particularly Cutibacterium acnes (formerly Propionibacterium acnes) – had been identified in disc tissue from patients with degenerative conditions.
Initial scepticism toward this hypothesis was justified. The fundamental question was whether bacteria found in disc samples represented true infection or contamination from skin flora during surgical sampling. This 'infection versus contamination' debate dominated the literature for over a decade.
Recent comprehensive reviews and methodologically rigorous studies have substantially resolved this debate in favour of true infection. A critical analysis by Gilligan et al. (2021) evaluated microbiological studies using stringent quality criteria, including:
When these stringent criteria are applied, the evidence becomes compelling.
Quantitative Microbiology: Studies employing correct quantitative methods identified bacteria in approximately 40% of disc samples (range 30–50%), with C. acnes present in 36% of total samples. These findings have been made across multiple independent research centers, providing strong support for the infection hypothesis.
Histological Visualization: Fluorescence in situ hybridization (FISH) studies – which directly visualize bacterial aggregates within tissue – have identified small, irregularly distributed bacterial colonies embedded in disc tissue, accompanied by host inflammatory cells. This direct visualization of bacterial biofilms within disc tissue essentially excludes contamination as an explanation.
Figure 2. Microscopic tissue sample stained in the flourescence in situ hybridization method. Biopsy tissue from herniated disc tissue retrieved during surgery. Biofilm (green) covered aggregates of Propionibacterium acnes bacteria. The close presence of polymorphonuclear leucocytes (lilac) indicates an inflammatory immune response.
Low-Bacterial-Load Concept: A key insight from recent studies is that chronic disc infections contain low numbers of bacteria irregularly distributed throughout the tissue – entirely different from acute pyogenic infections with high bacterial loads. Many earlier studies employing culture-based methods with high bacterial thresholds for 'infection' thus produced false-negative results, incorrectly classifying low-virulence infections as 'sterile'.
In summary, when appropriate microbiological methods are applied, low-virulence infection is frequently identified in patients with Modic Type 1, and these findings are reproducible across multiple independent laboratories worldwide.
The leading hypothesis proposes a four-step pathway:
Step 1 – Disc Disruption: Mechanical injury, herniation, or annular tear compromises disc integrity and permits bacterial entry.
Step 2 – Early Infection: Bacteria (primarily C. acnes) establish a low-virulence infection within herniated disc material. At this stage, the disc may or may not yet show Modic changes on imaging.
Step 3 – Inflammatory Response: The host immune system recognizes the bacterial infection and mobilizes an inflammatory response. This inflammation extends into adjacent vertebral bone, causing oedema and bone-marrow changes visualized as Modic Type 1 on MRI.
Step 4 – Chronic MC-1 Disease: Established chronic infection drives persistent inflammation, pain, and functional disability. Over time, without treatment, Type 1 changes may progress to Type 2 (fatty degeneration).
Figure 3. The development of Modic changes. Two associated pathways. Mechanical pathway: (0) normal disc; (1) age-related disc degeneration and the development of spondylosis; (2) end plate damage with development of Modic changes type 1. Infectious pathway: (3) disc degeneration with the development of a disc herniation; (4) invasion of bacteria, causing a disc infection; (5) end plate damage and bacteria with the development of Modic changes type 1.
This model explains several clinical observations:
Recent animal experimental studies provide direct microbiological documentation supporting the infection hypothesis. When live Propionibacterium acnes is inoculated into disc tissue in rabbits and rats, it initiates gradual development of disc degeneration, destruction of adjacent endplates, and vertebral oedema over weeks. MRI scanning of experimental animals showed development of typical Modic Type 1 changes, which directly correlated with radiological findings in humans. As a contrasting experiment, injection of Staphylococcus aureus bacteria produced nearly catastrophic destructive intradiscal disease. These findings demonstrate that P. acnes initiates disc-endplate destruction, and the associated pathological changes on MRI correspond to findings designated as Modic changes in humans. The animal experiments thus confirm that low-virulence bacterial infection can drive variable pathology, underscoring the importance of classifying 'discitis' into 'high-virulence' and 'low-virulence' forms based on the pathogenic bacterium's properties.
Magnetic resonance imaging is the definitive diagnostic method for identifying and classifying Modic changes. Standard MRI protocols include:
Importantly, the presence of Modic changes on MRI should be correlated with clinical symptoms. Many asymptomatic individuals have incidental MC on imaging, indicating that radiographic findings alone do not determine clinical significance.
Emerging research suggests that the volume and intensity of MC-1 signal on STIR sequences may better predict both pain severity and responsiveness to antibiotic therapy than simple presence/absence of changes.
A critical limitation in clinical practice is our inability to definitively identify which CLBP patients with MC-1 harbour bacterial infection versus other inflammatory mechanisms. Although approximately 40% of herniated disc samples in surgical studies contain bacteria, we still do not know:
Historically, management of MC has centered on conservative approaches. These remain appropriate first-line options for many patients:
Physical Therapy and Exercise: Traditional exercise programs provide modest benefits (20–30% symptomatic improvement) for selected patients, although they do not address underlying pathology. Notably, specialized exercise protocols do not appear to improve outcomes specifically in MC-1 disease.
Anti-inflammatory Medications: NSAIDs and other anti-inflammatory agents show limited efficacy in MC disease (approximately 30% of patients achieve meaningful relief).
Spinal Injections: Epidural steroid injections and facet joint injections provide temporary pain relief but do not treat underlying MC pathology. Intradiscal steroid injections show inconsistent results and raise theoretical concerns about steroid-induced immunosuppression in potentially infected tissue.
Other Approaches: Weight management, smoking cessation, sleep optimization, and cognitive-behavioral approaches address pain perception and function but not underlying disease mechanisms.
The most significant therapeutic advance in MC disease has been the demonstration of antibiotic efficacy in treating Modic Type 1 disease. This approach is based on the bacterial infection hypothesis and represents a fundamental shift from symptom relief to pathology-directed treatment.
Two randomized controlled trials have demonstrated statistically and clinically significant benefits of prolonged oral antibiotic therapy in MC-1 patients:
Albert et al. (2013) RCT: 162 patients with MC-1 received either amoxicillin-clavulanate (1500–3000 mg daily) or placebo for 100 days. Results showed:
Bråtan et al. (2019) RCT: 118 MC-1 patients received either amoxicillin (2250 mg daily) or placebo for 90 days. Results showed statistically significant improvements in disability and pain, but with somewhat less robust improvements compared to Albert et al. However, subsequent analysis of Bråtan's data has shown that a smaller subgroup of MC-1 patients actually achieved quite substantial long-term improvement despite the suboptimal dosing of active antibiotic in the treatment group.
Recent Subgroup Analysis (Kristoffersen et al., 2020): Re-analysis of Bråtan data using refined imaging criteria (STIR signal intensity) identified a substantial patient subgroup (STIR-3 classification) achieving remarkable improvement:
Across RCTs and open-label studies, approximately 50–60% of MC-1 patients demonstrate meaningful improvement (≥30% pain reduction) with antibiotic therapy, compared to 10–20% in placebo groups. This represents genuine efficacy far exceeding conventional treatments. However, the fact that not all MC-1 patients respond underscores the need for better patient selection criteria.
Antibiotic therapy is generally well-tolerated in the short-term (3–6 month) courses used for MC treatment. Common side effects include:
Building on evidence for systemic antibiotic efficacy, next-generation approaches explore direct drug delivery to the infected disc. The most advanced is PP353 (intradiscal linezolid formulation), currently in clinical trials. The rationale for intradiscal delivery includes:
Clinical data from early randomized trials are favorable. However, additional larger studies must be completed to fully evaluate results. Intradiscal antibiotic therapy represents a promising new treatment possibility and may prove valuable for carefully selected future MC-1 patients.
BVNA is a minimally invasive procedure for patients with chronic low back pain associated with Modic Type 1 changes. Under imaging guidance, a radiofrequency catheter is introduced to ablate (destroy) the basivertebral nerves, which transmit pain signals from the degenerative intervertebral disc. Early results show promising pain relief without invasive surgery.
Important limitation: BVNA is newly FDA-approved and currently available only in the United States. European patients await approval and market access.
Surgical intervention (typically fusion) is rarely indicated for MC disease alone and should only be pursued after conservative and medical management have been exhausted:
Modic Type 1 changes represent active vertebral bone inflammation strongly associated with pain severity
Disc prolapse tissue in CLBP patients often contains low-virulence bacterial infection (particularly C. acnes)
Bacterial infection likely contributes to MC-1 changes in a subset of patients
Antibiotic therapy significantly reduces pain and disability in approximately 50–60% of MC-1 patients
Clinical progress with antibiotic therapy persists at 12-month follow-up
These unanswered questions do not remove the evidence for antibiotic efficacy in MC-1, but they highlight the need for continued intensive research efforts.
Gilligan, C. J., Cohen, S. P., Fischetti, V. A., Hirsch, J. A., & Czaplewski, L. G. (2021). Chronic low back pain, bacterial infection and treatment with antibiotics. The Spine Journal, 21(6), 903–914. https://doi.org/10.1016/j.spinee.2021.02.013
Manniche, C., & Jordan, A. (2016). 10 years of research: from ignoring Modic changes to considerations regarding treatment and prevention of low-grade disc infections. Future Science OA, 2(2), FSO117. https://doi.org/10.4155/fsoa-2016-0017
Manniche, C., & O'Neill, S. (2019). New insights link low-virulent disc infections to the etiology of severe disc degeneration and Modic changes. Future Science OA, 5(5), FSO389. https://doi.org/10.2144/fsoa-2019-0022
Manniche, C., & Hall, G. M. (2021). Chronic low back pain, Modic changes and low-grade virulent infection: efficacy of antibiotic treatment. Future Science OA, 7(7), FSO703. https://doi.org/10.2144/fsoa-2021-0026
© 2026 Claus Manniche, MD. All rights reserved. This review is based on published peer-reviewed literature and clinical expertise. It is provided for informational purposes and should not replace professional medical advice. Always consult a qualified healthcare provider regarding diagnosis and treatment of back pain or Modic changes. This article reflects scientific understanding as of August 2026. Generated: August 31, 2026