Response to Letter to the Editor for “Femoral Bone Mineral Density Shows Stronger Correlation With Pedicle Screw Insertional Torque than Lumbar Bone Mineral Density or Hounsfield Units: A Retrospective in Vivo Study”

Abstract

We thank the authors for their insightful comments regarding our study evaluating the relationship between bone mineral density (BMD), Hounsfield units (HU), and pedicle screw insertional torque. While trajectory-specific CT-based assessments represent an important and evolving approach to evaluating local bone characteristics, existing literature has generally demonstrated only moderate correlations between DEXA or conventional HU measurements and insertional torque.

In our study, we extended prior work by evaluating both systemic and vertebra-specific metrics, including femoral and lumbar DEXA, as well as two HU measurements: L1 HU and HU at the instrumented vertebra. Among these, femoral BMD demonstrated the strongest correlation with insertional torque, while vertebral HU showed improved association compared to L1 HU but remained inferior to femoral BMD.

Although trajectory-specific HU measurements have shown promising results, prior studies often involve different screw trajectories, such as cortical bone trajectory techniques, which may limit direct comparability to conventional pedicle screw fixation. Furthermore, these methods require standardized measurement protocols that have not yet been widely established.

Importantly, screw fixation strength is influenced by both bone density and bone quality. As DEXA and HU primarily assess bone density, their moderate correlations may reflect the contribution of additional biological and structural factors. Ongoing studies incorporating laboratory and clinical variables may further clarify these relationships.

We agree that trajectory-specific approaches hold promise and look forward to future investigations that integrate both systemic and local assessments to optimize surgical planning.

The have been a number of papers investigating the correlation between DEXA or HU and screw insertional torque. A majority of papers showed only moderate correlations^1,3,5. As highlighted, DEXA-based measurements do not directly capture localized bone characteristics at the screw–bone interface. In particular, lumbar DEXA is not reliable in degenerative lumbar spine disease, because osteophytes, endplate sclerosis, facet hypertrophy, and vascular calcifications can artifactually elevate areal BMD measurements ⁶ .

We also agree that trajectory-specific CT-derived measurements may provide more localized information regarding fixation strength.

However, we would like to clarify that our study extended beyond a simple comparison of systemic BMD measures. In addition to lumbar and femoral DEXA, we evaluated two CT-derived Hounsfield unit (HU) metrics ¹ : HU at L1, a commonly used standardized reference, and ² HU measured at the vertebral body where pedicle screws were inserted ⁴ . In our cohort, femoral BMD demonstrated the strongest correlation with insertional torque (r = 0.557). Although HU at the instrumented vertebra showed a stronger association with insertional torque (r = 0.452) than L1 HU (r = 0.431), it remained inferior to femoral BMD.

The authors cited a study by Matsukawa et al ² , which demonstrated that regional HU values along the actual screw trajectory correlated strongly with insertional torque. However, it should be noted that this study evaluated cortical bone trajectory (CBT) screws rather than conventional pedicle screws, and therefore the screw trajectory and bone engagement characteristics differ. As such, these findings should be interpreted with caution when extrapolating to standard pedicle screw fixation.

Screw fixation strength is determined by a combination of bone density and bone quality. Both DEXA and CT-derived HU primarily reflect bone density and therefore may not fully capture other important determinants of fixation strength. This may explain the moderate correlations observed in our study. In this context, we are currently conducting further investigations incorporating additional factors beyond bone density, including laboratory markers and patient-specific clinical variables, to better characterize bone quality and its relationship to fixation strength.

While trajectory-specific HU measurements have shown promising results in prior studies^2,7, these techniques are highly dependent on measurement protocols, region-of-interest definition, and imaging workflows, and have not yet been standardized across institutions. Therefore, the generalizability of these findings remains to be established.

We agree that future studies incorporating trajectory-specific CT metrics will be valuable. We commend the authors for their ongoing prospective cohort study comparing lumbar and femoral DEXA T-scores with trajectory-specific CT HU, and we look forward with interest to the results of this important work. We appreciate the authors’ contribution in advancing this important discussion.