Discogenic low back pain (DLBP) results from the compromised functionality of a vertebral disc, primarily attributed to disc degeneration. Its characteristic clinical manifestation involves a progressively intensifying axial low back pain, notably exacerbated by extended periods of sitting and often devoid of radicular symptoms. Estimates suggest its prevalence ranges between 26% to 42% among individuals with chronic low back pain [1]. In degenerative disc disease, the nociceptive nerve fibers grow into normally aneural inner parts of the annulus, and sympathetic afferents also increase in the disc [2]. This mechanism underscores the substantial involvement of nerve and blood vessel infiltration, which significantly contributes to DLBP.

The diagnosis of DLBP relies on characteristic clinical symptoms, imaging studies, and provocation discography. Despite its invasive nature, provocation discography is the sole method of directly testing and confirming the concurrent pain associated with a specific disc. Additionally, following discography, computed tomography can offer insights into the morphological aspects of the disc, revealing the presence of annular tears or degenerative changes [3,4]. Provocation discography with appropriate controls can be a useful diagnostic method for evaluating DLBP [5]. These diagnostic tools pave the way for a multitude of treatments applied in clinical practice for managing DLBP. However, consensus among clinicians regarding the most effective treatment remains elusive. Of the diverse array of treatments available, intradiscal steroid injection (ISI) stands as an interventional treatment for DLBP. Despite being non-curative, its appeal lies in its minimal invasiveness and straightforward procedure, making it a favored option for patients averse to invasive surgical interventions. The efficacy of ISI remains a topic of debate within the medical community. Several prospective, double-blind, randomized controlled studies have been conducted, offering varying perspectives on the effectiveness of ISI in managing DLBP. Khot et al.'s findings revealed that, in their study, ISI did not demonstrate enhanced clinical outcomes at the one-year mark compared to the placebo group among patients experiencing discogenic low back pain [6]. Contrarily, Peng et al.'s study showcased the effectiveness of ISI in alleviating pain and enhancing the functional status of patients experiencing DLBP specifically associated with end plate degenerative changes [7]. Because of controversy about its therapeutic effect, selecting appropriate subjects for ISI who are predicted to have a good prognosis is important.

In our investigation, we delved into uncharted territory by exploring the correlation between clinical and radiological factors and the outcome of ISI in patients suffering from DLBP. Previous studies had not explored the spectrum of positive prognostic factors associated with ISI for this patient group.

1. Subjects

The study was approved by the institutional review board (IRB No, 2014AN0163). We conducted a retrospective analysis, gathering and scrutinizing data from May 2011 to December 2014. The inclusion criteria were as follows: chief complaint of axial low back pain with a duration of minimum of three months and failed conservative treatments. positive disc confirmed through provocation discography, and the intradiscal steroid injection performed after the diagnosis. Exclusion criteria comprised individuals with a history of prior spine surgery, spinal fractures, infections, tumors, predominant radiculopathy, or identified psychological issues. Patients who exhibited positive results for provocation discography across multiple disc levels were also excluded. Consequently, we could pinpoint the specific painful disc to a single level for each patient, facilitating the administration of ISI at the individualized, singular painful level.

2. Provocation discography protocol

We conducted pressure-controlled discography using a standard posterolateral approach with a 25-gauge spinal needle guided by C-arm imaging. Once the needle was accurately positioned at the disc's center, it was connected to the Automated Pressure-Controlled Discography system (APCD system, Cybermedic, Iksan, Korea) [8]. During the procedure, contrast medium was gradually injected into each disc at a controlled rate (0.01-0.05 cc/second) using the motorized syringe integrated into the APCD system while monitoring the injection pressure. Patients were prompted to articulate the nature, location, and intensity of pain experienced on a 0-10 visual analog scale (VAS) as they became aware of it. Their subjective experience of pain during discography was categorized as follows: no pain, dissimilar pain, similar pain, or concordant pain. Discs eliciting similar or concordant pain were classified as positive, while those inducing no pain or dissimilar pain were classified as negative. Throughout the procedure, intradiscal pressure was continuously recorded. We used the protocol of discography of Derby et al. [9,10]. Injection was continued until one of the following endpoints was reached: 1) evoked pain ≥ 7 VAS; 2) intradiscal pressure ≥ 80-100 psi; or 3) volume of injection ≥ 3.5 ml [5]. After the procedure, computed tomography (CT) was performed.

3. Radiographic and clinical data

We thoroughly examined the patients' medical records and carefully assessed the magnetic resonance images (MRIs). Clinical information such as gender, age, and the duration of pain was meticulously recorded. Pain persisting for > 6 months was categorized as chronic pain.

The MRIs were classified based on several criteria: disc degeneration grade, Modic type endplate changes, and HIZ. The disc degenerative grade was assessed using the Pfirrmann 5-scale classification system [11]. In our classification, we categorized grades 4 and 5 of degeneration as severe degeneration. Modic-type end-plate changes were defined in alignment with Modic et al.'s criteria, utilizing a combination of T1- and T2-weighted images for assessment [12]. The high-intensity zone (HIZ) was identified as a specific area exhibiting increased signal intensity within the posterior annulus fibrosus of the degenerative disc, as visualized on sagittal T2-weighted MRI images [13]. Based on the modified Dallas discogram classification system, the axial images from post-discography lumbar CT scans were utilized to evaluate and categorize the degree of disc disruption and radial annular tears [14]. The grading for radial annular tears ranged from 0 for a normal state to 3, signifying contrast leakage into the outer third of the annulus. Grade 4 is an extension of grade 3, providing additional detail on a severe grade 3 tear. In grade 4, not only does the contrast extend into the outer 1/3 of the annulus, but it also demonstrates concentric spread around the disc. Grade 5 characterizes a complete rupture of the outer layer of the annulus, leading to contrast leaking out of the disc. Sagittal images from post-discography lumbar CT scans were employed to classify discogram types and stages of disc degeneration. These classifications, Adams discogram classification, encompassed cotton ball, lobular, irregular, fissured, and ruptured types, which were distinguished on the basis of consistently identifiable features in the shape and density of the radio-opaque shadow [15]. The measurements were conducted by two independent observers who were unaware of the patients' clinical details and discographic findings, ensuring an unbiased assessment.

4. Intradiscal steroid injection technique

Patients identified with positive discography were advised to undergo treatment involving intradiscal steroid injection (ISI). In the ISI procedure, 20 mg of Triamcinolone acetonide (totaling 0.5 ml) was carefully and gradually injected into the nucleus pulposus at the specified intervertebral disc, utilizing fluoroscopic guidance.

5. Outcome measures

To assess the efficacy of ISI, we utilized the VAS. Baseline data was collected prior to the procedure, and the clinical outcomes were measured using VAS scores at 1-, 3-, and 6-month intervals post-procedure. A successful outcome was a more than 50% reduction in the VAS scores.

6. Statistical analysis

The analysis of pain reduction post-ISI was conducted using the Wilcoxon signed-rank test. Univariate logistic regression analysis was performed on each variable to ascertain the crude odds ratios associated with a successful outcome. A subsequent multivariate logistic regression analysis was conducted following the initial univariate analysis. This analysis involved utilizing the aforementioned independent variables to determine their collective impact on predicting the successful outcome of ISI, with the successful outcome serving as the dependent variable. Statistical analysis was performed using SPSS Statistics version 20 software (IBM, New York, USA). Significance was determined when the P-value was less than 0.05.

1. Participants

In this study, ISIs were administered at a single disc level in a cohort of 60 patients. The participants, with a mean age of 41.45 years (standard deviation: 12.57 years), encompassed 40 males (66.7%). The average duration of symptoms was 24.43 months, ranging from 1 to 124 months. Detailed basic clinical characteristics are delineated in Table 1.

Table 1 . Baseline characteristics of the study subject (n=60).

Variables	Value	Baseline VAS
Sex
Male	40 (66.7)	6.48 ± 1.45
Female	20 (33.3)	6.25 ± 1.89
Age (years)	41.45 ± 12.54 (16-78)
18-40	31 (51.7)	6.40 ± 1.60
> 40	29 (48.3)	6.28 ± 1.83
Duration of pain (months)	24.43 ± 32.78 (1-124)
> 6	36 (60.0)	6.17 ± 1.34
3-6	24 (40.0)	6.40 ± 1.60
Injection level
L4/5	37 (56.6)	6.11 ± 1.66
L5/S1	23 (38.3)	6.87 ± 1.39

Values are presented as mean ± standard deviation (range), or number (%)..

2. Clinical outcome after intradiscal steroid injection

As assessed via medical records, the clinical outcome indicated a notable decrease in the VAS score from 6.40 ± 1.59 to 3.87 ± 1.94 at the 6-month evaluation, demonstrating statistical significance (Fig. 1). Among the patients, four experienced prolonged severe pain and consequently underwent surgical intervention after the ISI. Encouragingly, no major complications, such as nerve damage, epidural hematoma, or spondylodiscitis, were reported following the procedure.

Figure 1. Changes of the visual analog scale (VAS) after intradiscal steroid injection (ISS) at 1-, 3- and six months of follow-up assessment. Values are presented as mean ± standard deviation. *Significant differences (P < 0.05) from the baseline.

The improvement of pain (Δ VAS) after ISI according to the clinical and radiological factors.

Table 2 illustrates the change in VAS corresponding to various factors. The alteration in pain intensity (Δ VAS) from baseline to the 6-month follow-up was calculated. No significant disparities in baseline VAS were observed among the subgroups. However, at the 6-month mark post-ISI, patients experiencing pain for less than six months exhibited a notably greater Δ VAS compared to those with chronic pain, demonstrating statistical significance (P = 0.020). Additionally, the Δ VAS was significantly more pronounced in patients with Modic type I endplate changes (P = 0.048) and the presence of HIZ (P = 0.046). Conversely, age, sex, initial VAS scores, level of ISI, disc degeneration grade, modified Dallas discogram classification, and Adams classification exhibited no significant differences in Δ VAS.

Table 2 . The reduction of VAS (Δ VAS) between baseline and 6-month follow-up after intradiscal steroid injection according to clinical and radiological variables.

Variables	Value (%)	Δ VAS	P-value†
Sex			0.107
Male	40 (66.7)	2.67 ± 1.87
Female	20 (33.3)	2.2 ± 2.17
Age (years)	41.45 ± 12.54 (16-78)		0.290
18-40	31 (51.7)	2.67 ± 1.87
> 40	29 (48.3)	2.45 ± 1.92
Duration of pain (months)			0.02*
1-6	24 (40.0)	2.67 ± 1.87
> 6	36 (60.0)	2.03 ± 1.71
Injection level			0.823
L4/5	37 (56.6)	2.65 ± 1.99
L5/S1	23 (38.3)	2.67 ± 1.87
Modified Dallas discogram			0.681
Grade 2	1 (1.7)	1.00
Grade 3	2 (3.3)	2.67 ± 2.31
Grade 4	35 (58.3)	2.57 ± 1.86
Grade 5	21 (35.0)	2.95 ± 1.90
Adams classification			0.108
Lobular	2 (3.3)	0.50 ± 0.71
Irregular	2 (3.3)	0.50 ± 0.71
Fissured	35 (58.3)	2.80 ± 1.89
Ruptured	21 (35.0)	2.86 ± 1.77
Modic change			0.048*
Type 0	28 (46.7)	2.25 ± 2.05
Type I	15 (25.0)	2.73 ± 1.58
Type II	13 (21.7)	3.00 ± 1.73
Type III	4 (6.7)	4.25 ± 1.26
Disc degeneration grade			0.107
Grade 2	6 (10.0)	4.33 ± 2.42
Grade 3	19 (31.7)	2.32 ± 1.89
Grade 4	30 (50.0)	2.33 ± 1.65
Grade 5	5 (8.3)	4.00 ± 0.71
High intensity zone			0.046*
Present	33 (55.0)	2.67 ± 1.87
Absent	627 (45.0)	2.21 ± 1.76

Values are presented as mean ± standard deviation or number (%)..

VAS: visual analogue scale..

Δ VAS: VAS 6 months after ISI – baseline VAS..

^†The Mann-Whitney U test and the Kruskal-Wallis test..

*Statistically significant..

3. Associations with clinical and radiological characteristics and the successful outcome of ISI

Table 3 presents the associations between the successful outcome of ISI and patients' clinical and radiological characteristics. The initial univariate analysis highlighted significant differences concerning pain duration, Modic type I endplate changes, and the presence of HIZ in relation to the successful outcome of ISI. In the multivariate binary logistic regression analysis, where the dependent variable was established as a binary outcome indicating the presence of a successful outcome following ISI, three independent predictors were identified. These predictors included pain duration of < 6 months (odds ratio (OR), 8.70; 95% confidence interval (CI), 1.78 to 42.37), Modic type I endplate changes (OR, 3.38; 95% CI, 0.73 to 15.78), and the presence of HIZ (OR, 4.90; 95% CI, 1.36 to 17.74). Furthermore, there were indications of potential significance for mild initial VAS (OR, 9.36; 95% CI, 0.87 to 100.39) and severe disc degeneration (OR, 5.02; 95% CI, 0.93 to 27.11), displaying borderline significance (P = 0.065, P = 0.061, respectively). No significant associations were observed for the other variables.

Table 3 . Unadjusted and adjusted odds ratio (OR) of the successful response after intradiscal steroid injection.

Variable	Unadjusted OR (95% CI)	P-value†	Adjusted OR (95% CI)	P-value‡
Sex
Male	1.833 (0.620-5.423)	0.273	1.262 (0.105-15.203)	0.855
Female	1.000		1.000
Age (years)
18-40	1.292 (0.466-3.582)	0.622	1.988 (0.408-9.693)	0.395
> 40	1.000		1.000
Initial VAS
0-6	1.500 (0.539-4.171)	0.437	9.361 (0.873-100.390)	0.065
7-10	1.000		1.000
Pain duration (months)
1-6	4.200 (1.347-13.093)	0.013*	8.695 (1.784-42.373)	0.007*
> 6	1.000		1.00
Injection level
L4/5	0.831 (0.292-2.364)	0.729	0.331 (0.034-3.237)	0.342
L5/S1	1.000		1.000
Modified Dallas discogram
Grade 2	1.000		1.000
Grade 3	0.000	1.000	0.001	1.000
Grade 4	1.455 (0.112-18.956)	0.775	0.000	1.000
Grade 5	0.856 (0.280-2.614)	0.784	0.000	1.000
Adams classification
Lobular	1.000		1.000
Irregular	0.000	0.999	7.086	1.000
Fissured	2.108 (0.698-6.364)	0.186	5.903	0.999
Ruptured	0.909 (0.214-4.143)	0.174	2.007	0.999
Modic change
Type 0	1.000		1.000
Type I	2.308 (0.626-8.513)	0.039*	3.382 (0.725-15.783)	0.021*
Type II	1.346 (0.360-5.036)	0.659	0.442 (0.0075-2.608)	0.367
Type III	3.462 (0.320-37.473)	0.307	4.869 (0.411-40.523)	0.999
Disc degeneration grade
Grade 1, 2, 3	1.000		1.000
Grade 4, 5	1.253 (0.441-3.562)	0.673	5.017 (0.928-27.106)	0.061
High intensity zone
Present	2.893 (0.976-8.578)	0.045*	4.902 (1.355-17.739)	0.015*
Absent	1.000		1.000

CI: confidence interval; OR: odds ratio..

^†Unadjusted odds ratios by logistic regression analysis..

^‡Adjusted odds ratios by multivariate logistic regression analysis..

*Statistically significant..

Indeed, the intervertebral disc undergoes morphological and biochemical alterations associated with age. As individuals age, therés a decline in proteoglycans and water content within the nucleus pulposus. This decrease contributes to a compromised ability to efficiently transfer axial stress across the disc, resulting in diminished hydrodynamic capabilities, particularly in distributing stress to the outer annulus fibrosus [16,17]. The reduction in proteoglycans and water content, particularly within the nucleus pulposus due to aging, can contribute to the emergence of localized tissue damage like internal disc disruption or annular tears. These damages allow nociceptive nerve fibers and blood vessels to penetrate the annulus's typically non-nervous inner regions. This infiltration can lead to the sensitization of nociceptors, triggered by various inflammatory mediators, thereby eliciting an inflammatory response within the disc. Previous studies have provided evidence of elevated levels of inflammatory mediators like interleukin (IL)-1β, IL-6, IL-8, and tumor necrosis factor-alpha (TNF-α) within degenerated intervertebral discs. These inflammatory mediators are thought to play a pivotal role in the development of pain among patients suffering from DLBP. Their heightened presence within the degenerated disc is associated with the pathophysiology and progression of pain in these individuals [18,19]. Annular tears may also progress the leakage of the inflammatory mediators to the adjacent epidural structures, leading to DLBP [20]. The inflammatory reaction, whether initiated by direct chemical irritation or an indirect autoimmune response to the nucleus pulposus, is considered a primary source of pain in DLBP. These inflammatory processes, stemming from various mechanisms, are widely acknowledged as significant contributors to the pain experienced by individuals suffering from DLBP [21,22].

Indeed, prior research has elucidated the mechanism behind the anti-inflammatory effect of corticosteroids. These agents exert direct inhibition on multiple fronts: they can impede various phospholipases and the transcription of cytokines and metalloproteases. Additionally, corticosteroids can stabilize lysosomal and other cellular membranes by hindering the recruitment of leukocytes at inflammatory sites [23]. Corticosteroids exert their influence by binding to their receptors, subsequently impeding the presentation of adhesive molecules on endothelial surfaces. This action can modulate the inflammatory response. Leveraging these theoretical underpinnings that link inflammation to the pathogenesis of DLBP and the anti-inflammatory effects of corticosteroids, ISI emerges as a viable treatment option for individuals grappling with DLBP.

Carragee et al. conducted a study involving 75 subjects without significant low back pain, assessing the impact of discography on disc degeneration progression [24]. They compared the degenerative changes in lumbar discs subjected to injections ten years prior with the same disc levels in matched subjects who did not undergo discography. This investigation aimed to elucidate the potential accelerated progression of degenerative findings associated with the discography procedure despite the numerous advantages of ISI as a minimally invasive treatment option. Their research highlighted a significant outcome: discography led to accelerated disc degeneration observed over a 10-year follow-up compared to the control group. Moreover, an additional study reported potential histologic alterations associated with intradiscal steroid injection (ISI), including intervertebral disc calcification [25]. Therefore, when we decide to perform ISI, it is important that we should consider who will be a good responder for ISI.

No prior study has thoroughly examined diverse clinical and radiological factors to predict the outcomes following ISI among DLBP patients. Our investigation revealed that pain relief notably favored patients with Modic type I endplate changes, the presence of a high-intensity zone (HIZ), and those without chronic pain. Specifically, individuals exhibiting Modic type I changes experienced superior pain improvement compared to other types, consistent with earlier research. Notably, Buttermann et al. have also emphasized that patients displaying Modic type I changes on MRI are most likely to benefit from ISI in the short term [26]. This could be because Modic I change is associated with an active inflammatory process of the vertebral end plate by proinflammatory cytokines [27]. So, the anti-inflammatory effect of ISI may be better than other Modic types, representing a more chronic end plate process.

Our research findings also indicated that patients exhibiting a HIZ experienced more significant pain relief following ISI compared to those without HIZ. Our study characterized the HIZ as a concentrated area displaying heightened signal intensity within the posterior annulus fibrosus, observed in the sagittal T2-weighted image [13]. Previous studies proposed that the HIZ is associated with painful outer annular disruption, and histology of the HIZ lesion showed the ingrowth of the vascularized granulation tissue [28]. In their study, Dongfeng et al. observed a significant presence of TNF-α positive cells and some CD68-positive cells within the HIZ. These findings strongly suggested a connection between HIZ and the inflammatory response occurring within painful intervertebral discs [29]. Though the precise mechanism remains unclear, a HIZ on MRI is believed to stimulate neurovascular ingrowth and an associated inflammatory reaction, contributing to DLBP. This theoretical premise suggests a close association between patients exhibiting HIZ and an inflammatory process. Consequently, individuals with DLBP and detectable HIZ on MRI might exhibit a higher susceptibility to ISI than those lacking HIZ, owing to the perceived link with inflammation.

Indeed, our study evidenced substantial improvement among individuals experiencing pain for less than six months. This observation aligns with the recognized significance of inflammation in the initial phases of the condition. Furthermore, chronic pain's impact on somatosensory processing is noteworthy. Chronic neuropathic disorders have been associated with heightened excitability, diminished inhibition, and structural reorganization within the spinal cord, influencing sensory responses. These elements collectively elucidate the marked improvement observed in patients enduring pain for less than six months [30]. In our findings, chronic patients exhibited potentially reduced responsiveness to ISI compared to those with shorter-term pain. The three significant outcomes of our study collectively suggest that ISI effectiveness is notably pronounced in patients experiencing DLBP primarily linked to inflammatory pathways.

Our study bears limitations that necessitate a cautious interpretation of the findings. It was retrospective and had a small sample size, potentially limiting the generalizability of the results. The concurrent use of physical therapy in most patients makes isolating the specific impact of intradiscal steroid injection (ISI) challenging. Moreover, our assessment relied solely on the Visual Analog Scale (VAS), neglecting broader disability measures like the Oswestry Disability Index, which could have offered a more comprehensive evaluation of treatment outcomes. Additionally, the predominant focus on L4/5 and L5/S1 discs restricts the extrapolation of our conclusions to higher-level discs. The assessment solely at the 6-month mark limited our ability to discern the long-term benefits of ISI, and potential variability in morphological analysis may have influenced result accuracy, necessitating caution in interpreting our study's conclusions.

Our study represents a pioneering effort, being the first to delve into the various favorable prognostic factors linked to successful ISI outcomes in DLBP patients. Despite the ongoing controversy surrounding ISI efficacy and its potential contribution to disc degeneration aggravation, our findings offer a crucial insight. They suggest the feasibility of selecting patients with a higher likelihood of positive ISI outcomes, presenting an alternative treatment avenue for individuals unresponsive to conservative therapies. This approach balances risks and benefits and minimizes potential complications arising from unnecessary discography and ISI procedures. Moving forward, robustly designed prospective studies are imperative to validate and refine this correlation, paving the way for more informed treatment decisions in DLBP management.

This work was supported by Institute of Information & communications Technology Planning & Evaluation (IITP) grant funded by the Korea government (MSIT) (No. 2022000218, Development of XR twin-based training content technology for rehabilitation).

No potential conflict of interest relevant to this article was reported.