Oral Presentation 50th International Society for the Study of the Lumbar Spine Annual Meeting 2024

Introduction: Disc degeneration is associated with chronic low back pain (cLBP) in some individuals, but is also typical of normal ageing. Currently, there are no techniques for differentiating age-related from pathologic MRI findings of disc degeneration. Here, we propose a new method for differentiating physiologic from pathologic disc ageing that highlights a novel phenotype of early disc degeneration in young cLBP patients.

Methods: Prospectively collected T1ρ-weighted magnetic resonance imaging (MRI) and spectroscopy (MRS) data from 206 cLBP patients (>3 month duration; 49.0±14.8 years old; 111 female, 95 male) and 53 asymptomatic controls (43.9±13.3 years old; 26 female, 27 male) were used. The mean T1ρ relaxation time in the nucleus pulposus (NP-T1ρ), which correlates with disc glycosaminoglycan content [1], was used as a biomarker of disc biochemical composition. For each lumbar disc in the cLBP group, we calculated the T1ρ Z-score: Z_i=(T1ρ_i–T1ρ_ctl,i)/SD_ctl,i, where T1ρ_i is the measured NP-T1ρ value, and T1ρ_ctl,i and SD_ctl,i are the regression-predicted (age and level specific) mean NP-T1ρ and standard deviation (SD) values from the control sample [2]. The Z-score quantifies age-adjusted deviation in disc composition: Z=-1 implies that the measured NP-T1ρ value is 1 SD lower than the mean of asymptomatic age-matched controls.

Ratios of acidic (alanine, lactate, propionate) and structural (carbohydrate/collagen, proteoglycan) MRS markers can predict discography-concordant pain with high accuracy [3]. Here, we used an established MRS-based diagnostic test (Nociscan, Aclarion) to predict whether each disc (n=670 discs from 134 cLBP patients) was likely to be positive/negative for discogenic pain (MRS+/-). Then, logistic regression adjusted for age, sex, BMI, and Pfirrmann grade was used to evaluate whether the Z-score was associated with an MRS+ test (suggestive of discogenic pain).

Results: There was wide variation in Z-scores (range: -3.0–4.4, n=1022 discs) consistent with high levels of biochemical and structural heterogeneity (NP-T1ρ range: 31.5–188.8 ms, Pfirrmann grades: 1–5). On average, NP-T1ρ values in cLBP patients were 0.2 SD below normal (mean Z-score:-0.2±1.1). The Z-score was not associated with age in the upper lumbar spine (p=0.18) but was associated with age in the lower lumbar spine (p<.0001, Fig. 1). In the lower lumbar spine, younger individuals had lower Z-scores; for example, 56% of patients below age 50 (62/111) exhibited Z<-1 at L5-S1, compared with only 17% of patients over age 50 (16/96). Lower Z-scores were significantly associated with an MRS+ test in logistic regression adjusted for age, sex, and BMI (OR, 95%CI: 2.0, 1.6–2.6, p<.0001). The Z-score remained statistically associated (p=.002) with MRS+ after adjusting for Pfirrmann grade, which was not associated with MRS+ in the adjusted models (p=.11).

Discussion: The T1ρ Z-score enables assessment of age-relative levels of biochemical disc degeneration and could thus help distinguish physiologic from pathologic disc ageing. In cLBP patients, lower Z-scores were associated with a higher likelihood of discogenic pain independent of structural degeneration as measured by Pfirrmann grade. We observed low Z-scores in the lower lumbar spines of many young cLBP patients, which suggests a previously un-reported phenotype and topic for future studies to isolate mechanisms and risk factors.