Pain relief after CT-guided pars injections in lumbar spondylolysis: analysis of MRI findings and CT-contrast distribution

Georg Wilhelm Kajdi, Sophia S. Goller, Christoph Germann, Christoph J. Laux, Reto SutterEuropean 2025 · DOI

TL;DR

This study investigates predictors of successful pain relief following CT-guided pars injections for lumbar spondylolysis. They found that the presence of isthmic bone marrow edema (BME) on MRI is a significant predictor of positive outcomes, while the distribution of contrast during the injection is not. This insight can help clinicians better select patients for this treatment and manage expectations.

What Problem Does This Paper Address?

Lumbar spondylolysis, a stress fracture in the vertebral arch, is a common cause of chronic lower back pain (CLBP), particularly in adults. CT-guided pars injections are a treatment option, but understanding which patients are most likely to benefit has been challenging. Specifically, researchers wanted to know if imaging findings on pre-injection MRI or the way contrast agent spreads during the injection could predict successful pain relief. This knowledge gap hinders personalized treatment strategies and patient counseling.

Methods

The researchers conducted a retrospective study on 134 patients with bilateral lumbar spondylolysis and CLBP who received CT-guided pars injections. Pain relief was assessed using the numerical rating scale (NRS) at baseline, 15 minutes post-injection, and one month post-injection. Successful pain relief was defined as a 50% or greater reduction in NRS pain score.

Two radiologists independently reviewed pre-injection lumbar MRI scans and CT scans of the injection procedure. They assessed various MRI findings, including isthmic bone marrow edema (BME), disc degeneration, facet joint degeneration, and soft tissue edema. For the CT scans, they analyzed the pattern of contrast distribution during the injection, categorizing it as peri-defect, intra-defect, or trans-defect (reaching the epidural space).

Statistical analysis, including logistic regression, was used to identify imaging predictors of successful pain relief. Inter-reader agreement was also assessed to ensure the reliability of the imaging assessments.

Datasets

The study utilized data from 134 patients who underwent CT-guided lumbar pars injections between July 2015 and July 2024 at a single institution. The data included patient demographics, pain scores (NRS), pre-injection MRI scans, and CT scans of the injection procedure. No new datasets were introduced.

Key Insights and Ideas

  • Isthmic BME is a strong predictor: The presence of bone marrow edema around the pars defect on MRI is a significant indicator of successful pain relief. This suggests that active inflammation or healing processes might be more responsive to the injection.
  • Contrast distribution is irrelevant: Surprisingly, how the contrast agent spread during the injection (peri-defect, intra-defect, or trans-defect) did not correlate with pain relief. This challenges the notion that precise contrast placement is critical for success in this specific procedure.
  • Pars injections offer meaningful relief: The study confirms that CT-guided pars injections can provide substantial pain relief for a significant portion of patients with CLBP due to spondylolysis.

Key Results

  • Overall Pain Relief: 55% of patients reported successful pain relief (≥ 50% pain reduction) at the 1-month follow-up. The average pain reduction was 34% at 15 minutes and 48% at 1 month post-injection.
  • BME as a Predictor: Patients with isthmic BME were 6-9 times more likely to experience successful pain relief at 15 minutes post-injection and 2-3 times more likely at 1 month post-injection.
  • Contrast Distribution: CT contrast distribution did not correlate with pain relief at any time point (p ≥ 0.27).
  • Other MRI Findings: While some other MRI findings (uni-segmental degeneration, soft tissue edema/synovial cysts) were associated with immediate pain relief, only isthmic BME remained a significant predictor at the 1-month follow-up.

Limitations and Future Work

The authors acknowledge several limitations:

  • Short-term outcome study: The study focused on immediate and 1-month outcomes, and longer-term efficacy is not addressed.
  • Lack of a control group: The absence of a saline injection control group limits definitive conclusions about the placebo effect.
  • Concomitant treatments: The use of physiotherapy and oral pain medication during the follow-up period was not systematically documented, potentially confounding the results.
  • MRI vs. CT for bony changes: While MRI was used for most findings, CT might offer a more precise depiction of discrete degenerative bony changes.

Future work could involve prospective, randomized controlled trials with longer follow-up periods and standardized documentation of concomitant treatments. Further investigation into the biological mechanisms underlying the predictive value of isthmic BME could also be beneficial. The authors also suggest caution in solely relying on BME for treatment decisions, as its absence does not preclude success.

Summary updated

Abstract

Question Imaging predictors of successful pain relief in lumbar spondylolysis patients with CLBP receiving bilateral pars injections are unknown. Findings Isthmic BME on MRI was a significant predictor of successful pain relief immediately and 1-month post-injection. CT-contrast distribution did not influence pain relief. Clinical relevance CT-guided pars injections offer successful pain relief in spondylolysis patients with CLBP. Isthmic BME on MRI is a significant predictor for successful pain relief, whereas CT-contrast distribution during injection is not.
BibTeX
@inproceedings{kajdi2025pain,
  author = {Georg Wilhelm Kajdi and Sophia S. Goller and Christoph Germann and Christoph J. Laux and Reto Sutter},
  title = {{Pain relief after CT-guided pars injections in lumbar spondylolysis: analysis of MRI findings and CT-contrast distribution}},
  booktitle = {European},
  year = {2025},
  doi = {10.1007/s00330-025-11903-8},
}
Figure