Oral Presentation 51st International Society for the Study of the Lumbar Spine Annual Meeting 2025

Serum inflammatory profiles associate with intradiscal propionic acid levels in chronic low back pain patients with Modic changes (115243)

Jiamin Zhou 1 , Po-Hung Wu 2 , Noah Bonnheim 1 , Conor ONeill 1 , Jeffrey C Lotz 1 , Thomas M Link 2 , Abel Torres-Espin 3 4 5 , Aaron J Fields 1
  1. Department of Orthopaedic Surgery, University of California San Francisco, San Francisco, California, USA
  2. Department of Radiology and Biomedical Imaging, University of California San Francisco, San Francisco, California, USA
  3. School of Public Health Sciences, University of Waterloo, Waterloo, Ontario, Canada
  4. Department of Neurological Surgery and Brain and Spinal Injury Center (BASIC), University of California San Francisco, San Francisco, California, USA
  5. Department of Physical Therapy, University of Alberta, Edmonton, Alberta, Canada

Introduction: Modic changes (MC) are associated with endplate damage and chronic low back pain (cLBP). However, the etiology of MC is unknown, which hampers development of effective treatments. Two MC etiologies have been proposed: disc infection with bacteria (e.g., Cutibacterium acnes) and autoimmunity. For example, discs adjacent to MC have varying levels of propionic acid (PA), a metabolite of C. acnes, and bone marrow from MC exhibits upregulated innate or adaptive immune cell signatures depending on the amount of C. acnes in the MC-adjacent disc. Here, we sought to advance understanding of these etiologies by testing if serum biomarkers of inflammation and immune response differ between subjects with variable intradiscal PA levels. 

Methods: Serum was prospectively collected from 200 cLBP patients and assayed for 40 cytokines using multiplexed ELISA. High-sensitivity C-reactive protein was quantified separately. Principal components analysis (PCA) was performed on log-transformed cytokine concentrations. 3T lumbar spine MRI was acquired on the day of serum collection and included sagittal T1- and T2-weighted sequences to diagnose MC: no MC (n=85 patients); MC1 (n=80 patients); MC2 (n=35 patients). Patients having both MC1/MC2 were classified as MC1. Single-voxel MR spectroscopy (MRS) was also performed (1-4 discs/patient). The area-under-the-curve of the PA peak was normalized to voxel size. Disc MRS was acquired blinded to MC status; thus, MRS data included normal and MC-adjacent discs. To account for this, the maximum PA level of all discs was determined per patient, and the median PA level for the cohort was used to stratify patients with “low” or “high” PA levels. Differences in principal component (PC) values between patients with low vs. high PA levels were tested using separate one-way ANOVAs by MC type. p<0.05 was considered significant.

Results: The first three PCs explained 67.1% of the variance in cytokine levels (Figure 1A), and the PCs were not associated with age, sex, or BMI. There were no differences in age, sex, or BMI between patients in each MC group (overall 85/114 male/female; ages 51.3±15.6 years; BMI 26.1±4.7 kg/m2). PC values were associated with intradiscal PA levels (Figure 1B). For patients with MC1, PC2 was significantly higher in those with low PA levels (p=0.044); for patients with MC2, PC3 was significantly higher in those with low PA levels (p=0.007).

Discussion: Serum biomarkers were significantly associated with intradiscal PA levels in patients with MC. PC2, which had positive loadings for cytokines involved in lymphocyte activation (e.g., GM-CSF, IL-2) and immune modulation (e.g., IL-10, TNF-β) and negative loadings for cytokines related to macrophage/monocyte activation (e.g., MCSF, MIG, MIP-1α), signifies a heightened adaptive immune response. Higher PC2 values in MC1 patients with the lowest intradiscal PA levels are consistent with a non-bacterial, auto-immune etiology. PC3 had loadings weighted toward cytokines involved in angiogenesis and extracellular matrix regulation (e.g., PDGF, TIMP-1/2). Higher PC3 values in MC2 patients with low PA levels suggest more active tissue remodeling and repair. These new findings underscore two different PA-dependent MC etiologies with distinct inflammatory profiles that potentially reflect unique and concurrent immune responses.

6736ab1f534f7-CytoFigure1-small.png