The Interplay of Endplate Bone Quality,Cage Geometry,and Segmental Biomechanics in Minimally Invasive Transforaminal Lumbar Interbody Fusion

To the Editor,

We read with great interest the article by Liao et al regarding the protective role of endplate Hounsfield units (HU) against cage subsidence (CS) in minimally invasive transforaminal lumbar interbody fusion (MI-TLIF). ¹ The authors provide a compelling site-specific HU threshold for bone quality assessment. However, we believe that the clinical applicability of this 221 HU threshold should be interpreted alongside critical surgical variables, specifically cage geometry, construct density, and the unique biomechanical demands of the lumbosacral junction. ² Besides, while the cut-off value for endplate HU was associated with the risk of subsidence in patients with L1 HU <117, it did not serve as an objective definition of a sclerotic endplate or as a universal reference for all MI-TLIF patients.

In support of this perspective, our prospective randomized clinical trial published in Neurosurgery demonstrated that cage shape (banana-shaped vs straight) significantly influences radiological outcomes, including restoration of disc height and segmental lordosis. ³ Liao et al utilized a uniform banana-shaped cage, which is typically designed to reside on the stronger anterior apophyseal ring to maximize lordotic correction. However, the authors followed the method by Okano et al to measure the endplate HU, which was used to predict subsidence after stand-alone lateral lumbar cage fusion. ⁴ The mechanical footprint and peak pressure distribution differ inherently between cage designs. Straight cages, often positioned mid-laterally, may interact differently with the endplate’s spatial bone density distribution. Therefore, a “protective” HU value measured within a 5-mm region of interest (ROI) may require recalibration depending on the specific cage design and its intended contact points.

Furthermore, the L5-S1 segment warrants a separate discussion. This level is the cornerstone of distal lumbar lordosis and global sagittal alignment, yet it faces the highest shear forces and reported nonunion rates—up to 27%—due to the significant sacral slope. ⁵ While the authors included L5-S1 in their cohort, the mechanical interaction between endplate sclerosis and these extreme axial-shear stresses might necessitate a higher HU threshold for stabilization at this junction compared to the upper lumbar levels.

Additionally, the risk of CS is heavily influenced by the number of cages used. ⁶ Biomechanical finite-element models have shown that single-cage TLIF constructs result in significantly higher peak stresses on the bone-implant interface and pedicle screws compared to bilateral constructs. ⁷ Since a single cage has a smaller total contact area, it potentially negates the protective effect of moderate endplate sclerosis. Integration of cage number (unilateral vs bilateral) into the predictive model would provide a more robust framework for risk-stratification.

In conclusion, while site-specific HU is a vital predictor, CS is a multifactorial phenomenon. We suggest that future models should integrate bone quality (HU) with cage geometry (shape and number) and segmental level (L5-S1) to optimize surgical planning and patient outcomes in MI-TLIF.