Response to Letter to the Editor Regarding “3D-Printed Titanium: Game-Changer for Standalone Lateral Lumbar Interbody Fusion? An Analysis of Risk Factors for Revision Surgery”

We thank the authors for their interest in our recent publication, “3D-Printed Titanium: Game-Changer for Standalone Lateral Lumbar Interbody Fusion? An Analysis of Risk Factors for Revision Surgery”. ¹ We appreciate the opportunity to respond to the methodological concerns raised.

We agree that retrospective studies are inherently subject to potential bias. To mitigate this, we applied a multivariable Poisson regression and a hierarchical mixed-effects Poisson model including surgeon ID as a random effect. The random intercept variance was negligible, suggesting minimal confounding by individual surgical practice patterns. Table 1 in our manuscript provided an overview of patient demographics and surgical characteristics. Regarding surgical complexity, we collected and compared data on the indication for SA-LLIF, including adjacent segment disease, status post laminectomy, and deformity. None of these differed between patients treated with PEEK vs 3D-printed titanium. Factors that were different or trended toward significance (e.g., number and location of LLIF levels and length of follow-up) were included as covariates in our multivariable analyses.

Model diagnostics of the multivariable Poisson regression (Table 3 in the manuscript) demonstrated excellent fit (Pearson χ²(228) = 188.2, P = 0.97) with a dispersion ratio of 0.83, indicating no evidence of overdispersion and supporting the validity of the Poisson model. Robust standard errors were applied throughout, and findings remained consistent across sensitivity analyses. While unmeasured confounding cannot be entirely excluded, a quantitative sensitivity analysis using the E-value method ² indicated that such a confounder would need to be associated with both implant type and revision risk by a risk ratio of at least 3.26 each, above and beyond measured covariates, to fully explain the observed association. To further minimize the risk of overfitting, our models were restricted to a small set of prespecified covariates. Together, these findings suggest that the model demonstrated both good fit and reasonable robustness to unmeasured confounding.

We acknowledge that case complexity and patient characteristics could theoretically influence implant selection and surgical technique. However, this critique does not apply to our specific cohort. During the study period, three surgeons consistently used 3D-printed titanium in all patients, whereas one surgeon exclusively used PEEK cages. Although this introduces the limitation of comparing surgeons A-C vs surgeon D, all four surgeons were highly experienced in LLIF surgery well before the start of the observation period (September 2018) and used the same mini-open technique. Nevertheless, this remains an important limitation, which is explicitly acknowledged in our study.

The distinction between early (<2 years) and late (>2 years) revision patterns was intentional and clearly described in both the Results and Discussion sections. The absence of a difference in revisions performed two years or more after the index procedure indicates that the divergence in revision rates between implants occurs within the first 2 years, which is consistent with the expected timing of cage subsidence vs fusion maturation, when implant material and design exert their greatest influence. Importantly, this finding does not imply that there is no long-term difference between implants; rather, it reflects that the early difference persists over time, as fewer early revisions in the 3D-printed titanium group translate into a sustained lower cumulative revision rate. Numerous biomechanical and basic science studies have demonstrated enhanced osseointegration with 3D-printed titanium compared with PEEK, attributed to its porous microstructure, osteoblast adhesion, and a modulus of elasticity closer to that of native bone.^3-5

We agree that future prospective studies incorporating PROMs, imaging, and cost-effectiveness analyses are needed to further clarify the role of 3D-printed titanium implants, ideally within randomized controlled designs. Cost-effectiveness, however, is region-specific. In the United States, where revision decompression or fusion procedures carry substantial direct and indirect costs, an implant associated with a 50% relative and 10% absolute risk reduction in such revisions would be expected to result in meaningful cost savings. Beyond economics, we believe surgeons have an ethical obligation to provide the best possible care for their patients.

Our study represents the largest single-institution comparison of implant materials in standalone LLIF and was designed to evaluate revision surgery as an objective, clinically meaningful outcome. We appreciate the thoughtful engagement of our colleagues and welcome continued discussion on this topic.