Early-stage lumbar paraspinal muscle injury: endoscopic versus transforaminal lumbar interbody fusion: a retrospective comparative analysis.

Objective: While previous studies frequently reported the clinical efficacy and minimal invasiveness of endoscopic lumbar interbody fusion (Endo-LIF) for lumbar spinal stenosis (LSS), few existing studies quantitatively measure early-stage postoperative paraspinal muscle injury. This study aimed to preliminarily quantify differences in early postoperative paraspinal muscle changes between Endo-LIF and transforaminal lumbar interbody fusion (TLIF) for single-level LSS. The observed alterations, if substantiated in future studies, might offer considerations for tailoring postoperative rehabilitation.

Methods: This retrospective cohort included 90 severe LSS patients undergoing single-level fusion, allocated to Endo-LIF group (n = 48) or TLIF group (n = 42). Comprehensive data encompassed demographics, operative metrics, patient-reported outcomes (PROs) including Oswestry Disability Index (ODI) and Visual Analogue Scale (VAS), and acute paraspinal muscle trauma biomarkers ((creatine kinase (CK), C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR)). The cross-sectional area (CSA) of paraspinal musculature (multifidus and erector spinae) was quantified at the index instrumented level using axial T2-weighted magnetic resonance imaging (MRI), with measurements obtained preoperatively and during early follow-up (FU). Muscle boundaries were delineated using semi-automated tools (ITK-SNAP v4.0.2) with manual correction, and CSA values were calculated via MATLAB-based custom algorithms.

Results: Preoperative demographics, leg/back pain VAS, ODI, and inflammatory markers were comparable between Endo-LIF and TLIF groups. At 3 days postoperatively, Endo-LIF demonstrated superior back VAS and ODI (P < 0.001) but comparable leg pain VAS. Both groups achieved significant PRO improvements (P < 0.001). TLIF had significantly higher low back myofascitis incidence (P < 0.001). Endo-LIF showed significantly reduced blood loss (58.3[IQR 50, 75] days vs. 214.3[IQR 150, 250] mL) and shorter hospitalization (5.2 [IQR 5, 6] days vs. 7.1 [IQR 7, 8] days) (both P < 0.001), but longer operative time and greater fluoroscopy use (P < 0.001). Complication rates were similar (P = 0.27). CRP and CK levels at postoperative day 1 were significantly higher in TLIF (P < 0.001). Long-term follow-up revealed no significant intergroup differences in PROs (all P > 0.01). Postoperative paraspinal muscle CSA decreased in both cohorts, with a more pronounced reduction observed in the TLIF group, statistically associated with surgical approach and smoking status (both P < 0.001).

Conclusions: Compared with TLIF, Endo-LIF demonstrated relatively early back pain relief, reduced intraoperative blood loss, shorter hospital stays, and lower levels of acute muscle injury markers. These potential benefits were counterbalanced by longer operative durations and greater reliance on fluoroscopy. Both approaches achieved largely comparable long-term functional outcomes with similar safety profiles. Paraspinal muscle CSA measurements suggested a comparatively lesser degree of early muscle injury with Endo-LIF versus TLIF.