Response to the Letter to the Editor on “Perioperative GLP-1 Receptor Agonist Use is Associated with Reduced Revisions and Complications Following ACDF: A Propensity-Matched Analysis”

We would like to thank the authors for the opportunity to respond to their letter regarding our recent study, “Perioperative GLP-1 Receptor Agonist Use is Associated with Reduced Revisions and Complications Following ACDF: A Propensity-Matched Analysis. ¹ ” We appreciate their insight and thoughtful feedback on the findings of our work.

The authors raise the point that radiographic fusion does not invariably correlate with patient-reported or functional outcomes after ACDF. We agree that pseudarthrosis represents an imperfect surrogate and that radiographic union alone cannot fully capture clinical success. ² As noted, large database studies are inherently limited in their ability to incorporate validated patient-reported outcome measures such as pain scores or the Neck Disability Index. However, our study was not designed to interpret radiographic fusion in isolation. In addition to pseudarthrosis, we evaluated several clinically meaningful endpoints, including revision surgery, readmissions, dysphagia, and healthcare utilization. More importantly, the observed reduction in pseudarthrosis among GLP-1RA users was accompanied by concordant decreases in revision procedures and postoperative resource utilization, suggesting that the association was not confined to an imaging surrogate alone. Because revision surgery represents a clinically driven decision that typically reflects symptomatic failure rather than radiographic findings alone, ³ these parallel findings provide indirect support that the differences observed may have clinical relevance. We therefore interpret pseudarthrosis in this context as one component of a broader pattern of outcomes rather than as a standalone measure of success.

The authors also emphasize the time-dependent biology of bone healing and the possibility that GLP-1RAs may exert phase-specific effects during different stages of fusion. We agree that distinguishing between preoperative initiation, early continuation, and delayed postoperative initiation would be valuable in defining mechanistic pathways. However, our exposure definition was specifically designed to evaluate the effect of sustained perioperative therapy rather than the biologic impact of a single initiation time point. In our study, GLP-1RA users were defined as patients with documented prescriptions in both the 6-month period prior to surgery AND the 6-month period following surgery, ⁴ thereby capturing individuals maintained on continuous therapy across the fusion period rather than those with isolated or newly initiated postoperative use. This approach was intended to represent a stable metabolic treatment state during the period in which fusion biology occurs and to minimize misclassification related to intermittent or reactive prescribing patterns.

The authors further raise an important question regarding the potential interaction between GLP-1RA therapy and established fusion-enhancing techniques. We agree that understanding whether GLP-1RAs provide additive benefits alongside autograft, allograft, or biologic augmentation—or potentially reduce the need for more aggressive fusion strategies—would have meaningful implications for surgical planning and value-based care. Importantly, we did not intend to suggest that GLP-1RAs replace established fusion strategies. Rather, as discussed in our manuscript, the biologic effects of GLP-1RAs on osteoblastic activity and inflammation make it plausible that these agents may complement existing fusion-promoting methods. ⁵ Nonetheless, this hypothesis remains exploratory and requires prospective validation. ⁶ Our findings should therefore be interpreted as identifying a potential systemic modifier of fusion biology across heterogeneous real-world surgical practices, rather than as evidence supporting substitution for grafting or biologic augmentation.

We also appreciate the authors’ comment regarding the generalizability of our findings to non-diabetic populations. We agree that the pleiotropic effects of GLP-1RAs on inflammation, body composition, and bone metabolism may extend beyond patients with type 2 diabetes mellitus (T2DM), ⁷ and that determining applicability across broader metabolic phenotypes represents an important future direction. However, the restriction to patients with T2DM was intentional and methodologically necessary to isolate the clinical association of GLP-1RA therapy itself rather than the metabolic differences between fundamentally dissimilar patient populations. ⁸ By limiting the cohort to individuals with the same underlying disease and applying propensity matching across demographic and medical comorbidities, our design minimized confounding related to indication, baseline metabolic status, and access to therapy. ⁹ Including non-diabetic patients would have introduced substantial heterogeneity in baseline physiology, glycemic control, and prescribing patterns, making it difficult to discern whether observed differences were attributable to the medication or to underlying metabolic variation. Accordingly, our study was designed to evaluate GLP-1RA therapy as a potential modifier of surgical outcomes within a metabolically comparable population, prioritizing internal validity over broad generalizability. The findings should therefore be interpreted as evidence specific to diabetic patients undergoing ACDF, rather than as proof of universal benefit across all patient groups.

We thank the authors again for their thoughtful commentary and constructive dialogue. Our findings should be interpreted as supportive evidence of a potential association between sustained perioperative GLP-1RA therapy and improved ACDF outcomes within a metabolically comparable diabetic population. While mechanistic pathways and broader applicability remain to be clarified, we believe our study contributes to the growing body of literature exploring metabolic optimization as a modifiable risk factor in spine surgery. Future prospective and mechanistic studies will be essential to clarify optimal exposure timing, biologic effects, and broader applicability across metabolic phenotypes.