Study Design

This single-blind randomized controlled trial was conducted in accordance with the Consolidated Standards of Reporting Trials (CONSORT) guidelines.

CONSORT-Compliant Trial Presentation

Trial design: single-blind, parallel-group randomized controlled trial (RCT) with 1:1 allocation to experimental (HILT + collagenase) or control (sham laser + collagenase) groups.

Participants: adults aged 40–70 years with lumbar disc herniation (LDH) confirmed by magnetic resonance imaging (MRI), persistent symptoms ≥ 6 months, and failure of ≥ 3 months of conservative treatment. Exclusion criteria included spinal instability, infection, or collagenase allergy.

Allocation consisted of block randomization (block size = 4) by disc herniation level (L4–L5/L5–S1) using computer-generated sequences; allocation concealment was performed via sequentially numbered, sealed opaque envelopes.

Blinding: participants and outcome assessors were blinded to group assignment. Sham laser therapy (identical device, no energy output) maintained blinding.

Interventions: both groups received standard collagenase chemonucleolysis. The experimental group added HILT (LTS-1500, 15 W, 10 J/cm2, five daily sessions); the control group received sham laser.

Primary outcomes included VAS pain scores; secondary: Oswestry Disability Index (ODI), straight-leg-raising angle, Short Form-36 (SF-36). Assessments were conducted at baseline, 1 week, 1 month, 3 months, and 6 months postoperatively.

Throughout the study, outcome assessors and statisticians remained blinded to group allocation, with emergency unblinding protocols established for severe adverse events. The study was carried out at the Department of Pain Management, The First People’s Hospital of Changzhou (a tertiary medical center with ISO 15189-certified laboratory facilities) between October 2023 and October 2024. Follow-up assessments were completed by February 2025. The trial was registered prior to participant enrollment at the Medical Research Registration and Filing Information System (www.medicalresearch.org.cn) (registration number: MR-32-25-022787), adhering to World Health Organization (WHO) International Clinical Trials Registry Platform standards.

Patients diagnosed with LDH were enrolled in this clinical investigation. Figure 1 illustrates the patient selection and randomization process. All patients were fully informed about the study protocol and provided written informed consent prior to participation. Participants were randomly assigned to either the experimental or control group using a computer-generated random number Table (1:1 allocation ratio), stratified by age (40–55 years versus 56–70 years). Allocation concealment was maintained through sequentially numbered, sealed opaque envelopes prepared by an independent statistician. The envelopes were opened only after baseline assessments to prevent selection bias.

Fig. 1

Flow diagram of the clinical trial

Throughout the study, both patients and outcome assessors remained blinded to group allocation. An independent reviewer responsible for outcome evaluation was not involved in intervention delivery or data collection. To maintain blinding, the intervention team was separate from the outcome assessment team. Unblinding occurred only in cases of medical emergencies and was documented in compliance with ethical guidelines.

Blinded outcome assessors conducted standardized evaluations at five predetermined time points: preoperative baseline, postoperative week 1, postoperative month 1, postoperative month 3, and postoperative month 6. Data collection procedures followed CONSORT guidelines and utilized validated instruments, including the visual analog scale (VAS), Oswestry Disability Index (ODI), and SF-36 Quality of Life Questionnaire, to ensure methodological rigor. This institution has obtained the rights to use the ODI and VAS scores.

The primary outcome was the pain severity measured by the visual analog scale (VAS; 0–10) [17], validated in prior LDH studies (test–retest reliability: intraclass correlation coefficients [ICC] = 0.89). Secondary outcomes included functional improvement assessed via the Oswestry Disability Index (ODI; Cronbach’s α = 0.87) [18]; Neurological recovery evaluated through the straight-leg raise test (interrater agreement: κ = 0.78). Based on senior clinicians’ expertise and relevant literature, we defined the minimal clinically important difference (MCID) thresholds as: 3 points for visual analog scale (VAS) scores and 15 points for Oswestry Disability Index (ODI).

Patient Eligibility

Diagnoses were confirmed by board-certified pain medicine physicians with ≥ 5 years of clinical experience in lumbar disc herniation (LDH) management. Diagnostic criteria incorporated both MRI findings and clinical evaluation results.

Inclusion criteria included: (1) patients meeting the diagnostic criteria for lumbar disc herniation (LDH) (International Classification of Diseases [ICD]-10: M51.1; ICD-11: FA50.0 or FA50.1). (2) Patients with a disease course exceeding 1 month and having undergone at least 2 weeks of conservative therapy. (3) Patients between the ages of 40 and 70 years with a (4) body mass index (BMI) 18.5–30 kg/m2 (excluding obesity or malnutrition as confounders). (5) Patients had to have the ability to comprehend the study protocol and provide written informed consent (using a standardized international template). The diagnostic criteria for lumbar disc herniation are as follows: (1) radiating pain in the lower limb, with pain location corresponding to the affected nerve distribution area; (2) abnormal sensation in the lower limb, with reduced superficial sensation in the corresponding affected nerve distribution area of the skin; (3) positive findings in the straight leg raising test, straight leg raising reinforcement test, contralateral straight leg raising test, or sciatic nerve traction test; (4) weakened tendon reflex compared with the healthy side; (5) decreased muscle strength; and (6) presence of disc herniation and nerve compression confirmed by lumbar spine MRI or computed tomography (CT), which aligns with the symptoms and signs of the affected nerve [15]. Diagnosis of lumbar disc herniation requires meeting three out of the first five criteria, in addition to the sixth criteria (recommended level A, evidence level 1a).

Exclusion criteria included: (1) lumbar spondylolisthesis (Meyerding grade ≥ II), congenital/degenerative bony spinal stenosis (sagittal diameter < 10mm), vertebral fracture (Genant semi-quantitative grade ≥ 1), spondylolysis, or scoliosis (Cobb angle > 10°); (2) spinal infections (e.g., tuberculosis and discitis), tumors (primary/metastatic), or inflammatory diseases (ankylosing spondylitis and rheumatoid arthritis); (3) cauda equina syndrome (urinary retention, saddle anesthesia, or anal sphincter dysfunction); (4) inability to comply with prolonged bed rest; (5) cardiopulmonary diseases: New York Heart Association (NYHA) Class III–IV heart failure, chronic obstructive pulmonary disease (COPD) GOLD Stage ≥ 3; (6) uncontrolled metabolic disorders: hemoglobin (Hb)A1c > 8%, thyroid dysfunction (thyroid-stimulating hormone [TSH] < 0.4 or > 4.0 mIU/L); (7) coagulopathy (international normalized ratio [INR] > 1.5, platelets < 100 × 10⁹/L) or long-term anticoagulant use (warfarin, aspirin, or clopidogrel); (8) severe mental illness or positive suicide risk screening); (9) spinal surgery or interventional procedures (e.g., epidural injections and radiofrequency ablation) within the past 6 months; (10) tattoos or pigmented lesions in/near the treatment area (interfering with device application or imaging); (11) patients deemed legally incompetent (guardians cannot sign consent on their behalf); (12) pregnancy or lactation; (13) severe osteoporosis (T-score ≤ −3.0 or recent fragility fracture); and (14) concurrent participation in other clinical trials (requires documented washout period).

Definition of the Comorbidity Population

Participants were classified into the comorbidity subgroup if they had ≥ 2 chronic conditions (e.g., diabetes and cardiovascular disease) based on the Charlson Comorbidity Index.

Administration and Procedure

All patients underwent combined intradiscal and extradiscal collagenase injection therapy, with 120 IU administered intradiscally and 400 IU extradiscally. All procedures were performed by senior pain management specialists.

All surgical procedures were standardized as follows: preoperative preparation involved positioning patients in the prone position with continuous electrocardiographic monitoring and secured intravenous access for emergency management. CT imaging was utilized to localize the target intervertebral disc (responsible level) and determine the puncture site via surface anatomical landmarks. Under CT guidance, puncture needles were precisely advanced into both the target disc and anterior epidural space. Discography was then performed by injecting 0.5–1 mL contrast agent into the disc to verify pain provocation consistency with original symptoms. A diagnostic test injection of 2 mL local anesthetic–saline mixture into the anterior epidural space preceded a 15-min observation period. Prior to therapeutic administration, needle positioning was reconfirmed via CT, followed by sequential injections of 120 IU collagenase intradiscally and 400 IU extradiscally under real-time imaging guidance. Postoperative protocol mandated strict bed rest for 5 days, with subsequent graduated ambulation initiated under physician supervision to ensure procedural efficacy and patient safety (Fig. 2). Adverse events were prospectively monitored, including transient erythema at laser sites (n = 4 experimental group), self-limiting headaches (n = 2 control group), and mild paraspinal spasms (n = 3 experimental group). Reported adverse events were mild and group-specific: in the experimental group, four patients developed transient erythema at laser irradiation sites (localized skin redness without warmth or pain), and three experienced mild paraspinal spasms (brief lumbar muscle tightness without functional limitation). In the control group, two patients reported self-limiting headaches (mild, nonthrobbing pain without nausea or dizziness). Management included close clinical observation: erythema was addressed with local cold compresses (3× per day, 10 min/session); spasms and headaches were managed with rest and reassurance. No pharmacological interventions (e.g., analgesics or antiinflammatories) were required. All events resolved within 72 h without pharmacological intervention or protocol deviation. No severe adverse reactions (e.g., burns and neurological deficits) occurred in either group. Treatment compliance was 93.2% in the intervention group (monitored via medication logs) and 89.5% in the control group (assessed by attendance records). Dropout rates were 8.0% (n = 6) and 6.7% (n = 5) in the intervention and control groups, respectively. Adverse events occurred in 14.7% (n = 11) of intervention patients and 5.3% (n = 4) of controls (P = 0.08), with no severe complications reported.

Fig. 2

(a) CT-guided intradiscal and extradiscal injection; (b) intervertebral discography and spinal canalography

Interventions

Both groups received conventional therapy to ensure ethical equivalence and reflect real-world clinical practice. This approach minimized confounding from differential access to baseline care and allowed isolation of the collagenase intervention effect while maintaining patient safety and compliance with standard-of-care guidelines. The medication regimen consisted of: flurbiprofen axetil (50 mg every 12 h; manufactured by Yuanda Medical Nutrition Science [Wuhan] Co., Ltd.), mannitol (250 mL every 12 h; manufactured by Baxter Healthcare [Shanghai] Co., Ltd.), and omeprazole (20 mg twice daily; manufactured by Hainan Hailin Pharmaceutical Co., Ltd.)

Group I: the Control Group

The control group received sham laser irradiation with identical treatment parameters to the experimental group, including matched session duration (4–6 min) and daily application frequency over the 5-day intervention period. Sham procedures were administered using the same LTS-1500 device configured in placebo mode, maintaining identical auditory/visual feedback while disabling actual energy output, thus ensuring effective blinding through equivalent treatment experience without photobiomodulation effects.

Group II: the Experimental Group

The high-intensity laser therapy (HILT) system employed in this investigation was the LTS-1500 model (DJO Global), a Class IV therapeutic laser device delivering dual-wavelength output (890/910 nm) with maximum power capacity of 15 W. When configured at full-power continuous wave (CW) mode (15 W output), this system generates an energy flux density of 900 J/min at the target tissue interface. Through standardized beam collimation covering an irradiation field approximating conventional compact disc dimensions (diameter 12 cm), the delivered energy density reaches 10 J/cm2 per application, aligning with World Association for Laser Therapy (WALT) safety guidelines for deep tissue photobiomodulation.

Both patient groups received identical laser therapy protocols administered by a single certified laser therapist. Treatment commenced on the first postoperative day, with daily sessions initiated at 8:00 AM. All patients underwent irradiation of standardized anatomical fields, with session duration (4–6 min) adjusted according to body size. Treatments were delivered consecutively over 5 days in a dedicated therapy room, accommodating only one patient at a time owing to equipment limitations. Following each session, adverse events related to laser therapy were assessed and documented: absence of reactions was recorded as “none,” while observed reactions were described in detail with remedial measures implemented. Patients declining further treatment during hospitalization were classified as trial withdrawals. Posttrial analyses included withdrawal rate, protocol completion rate, and adverse event incidence.

Statistical Analysis

A priori sample size calculation was performed using G*Power 3.1. Based on pilot data, we assumed a large effect size (Cohen’s d = 0.8) for pain reduction (VAS) between groups, given the limited sample size. With α = 0.05, power = 0.80, and two-tailed testing, the calculation required 26 participants per group. To account for potential attrition, we enrolled 30 patients per group (total N = 60). Variables were controlled by adjusting for potential confounders (e.g., age, comorbidity status, and preoperative analgesic use) in multivariable regression models, while interaction terms evaluated effect modification by age and comorbidity. Baseline homogeneity between groups was measured using standardized mean differences (SMD), with an SMD threshold of < 0.1 indicating minimal between-group differences across all covariates (age, gender, BMI, herniation level, and symptom duration). Intention-to-treat (ITT) analysis was applied, retaining all 60 randomized participants in their originally assigned groups for statistical evaluation, regardless of treatment compliance, protocol deviations, or dropout status. Missing outcome data were managed using multiple imputation (five datasets) with baseline characteristics and prior assessments as predictors to preserve the integrity of the ITT principle. Treatment compliance rates were 86.7% in the experimental group (26/30 completed all sessions) and 83.3% in the control group (25/30). Dropout rates were 13.3% (n = 4) in the experimental group and 16.7% (n = 5) in the control group. Adverse effects were mild and self-limiting, occurring in six experimental patients and two controls, with no severe complications reported. Quantitative data analysis adhered to rigorous statistical protocols: normality assessment was conducted via Shapiro–Wilk testing, complemented by Levene’s test for homoscedasticity verification. For parametric datasets demonstrating Gaussian distribution and variance homogeneity, independent Student’s t-tests were employed with results expressed as mean ± standard deviation (SD). Continuous outcomes were analyzed using nonparametric Mann–Whitney U tests owing to small sample size and nonnormal distributions. Categorical outcomes were assessed with Fisher’s exact tests. Effect sizes were reported as Cohen’s d or odds ratios with 95% confidence intervals. A two-tailed α-level of 0.05 defined statistical significance, with Bonferroni correction applied for multiple comparisons where appropriate. All computations were executed in SPSS Statistics version 20.0 (IBM Corporation, Armonk, NY, USA) using validated macroscripts to ensure analytical reproducibility. Treatment compliance was 86.7% in the intervention group (26/30 completed all sessions) and 83.3% in the control group (25/30). Dropout rates were 13.3% (n = 4) and 16.7% (n = 5) in the intervention and control groups, respectively. Adverse events occurred in six intervention patients (e.g., mild dizziness) and two controls (P = 0.25).