Dear Editor,
We sincerely thank the authors for their thoughtful and technically informed commentary regarding our network meta-analysis (NMA) on minimally invasive and endoscopic lumbar fusion. We appreciate the opportunity to clarify our methodological considerations and contextualize our findings within this rapidly evolving field.
1
1. Technical Heterogeneity Within “Full-Endoscopic TLIF” We fully acknowledge that uniportal full-endoscopic TLIF (FE-TLIF) encompasses technically distinct approaches, including trans-Kambin (KLIF) and posterolateral variants. These differ in working corridor, decompression strategy, and cage geometry. Our NMA was designed to synthesize comparative clinical outcomes as currently reported in the literature, rather than to perform a procedural micro-classification based on surgical nuance. Importantly, the majority of included FE-TLIF cohorts described decompression-based fusion procedures rather than purely indirect KLIF constructs.
While biomechanical concerns regarding smaller cages in trans-Kambin approaches are valid, currently available aggregate clinical data do not demonstrate inferior structural outcomes (such as pseudarthrosis) attributable to FE-TLIF. We agree that future prospective studies should stratify endoscopic subtypes to refine these interpretations.
2. Transitivity and Indication Differences The importance of the transitivity assumption in NMA cannot be overstated. In our study, we formally evaluated potential effect modifiers, including age, disease level, cage material, and study design. We found no imbalance sufficient to invalidate indirect comparisons. Although trans-Kambin techniques may be selected for milder pathology in some centers, meta-regression did not identify disease level as a significant modifier of primary outcomes in our included cohorts. Based on available data, the transitivity assumption remains methodologically defensible.
3. Learning Curve and Study Quality We concur that the learning curve for endoscopic fusion is clinically relevant. As explicitly acknowledged in our manuscript, operative times were longer in FE-TLIF and BE-TLIF, reflecting early adoption patterns. However, complication rates were not significantly higher, and fusion rates remained comparable. If learning-curve bias systematically disadvantaged endoscopic techniques, one would expect inferior clinical signals, which were not observed. Regarding evidence quality, we utilized GRADE methodology and rated the certainty as low to very low, framing our conclusions as hypothesis-generating.
4. On the Question of Prematurity We respectfully suggest that evidence synthesis is not premature simply because high-level randomized data remain limited. In rapidly evolving fields, structured integration of available evidence serves to clarify current signals and guide future investigation. Our study functions as an analytic framework for hypothesis refinement rather than a definitive endpoint, highlighting short-term clinical advantages while calling for more rigorous, technique-stratified trials.
5. Complication Profiling We agree that complications such as exiting nerve dysesthesia in trans-Kambin approaches differ mechanistically from dural tears in other techniques. However, inconsistent reporting across studies precluded subtype-specific pooling. We strongly support the call for standardized complication classification in future comparative trials.
Conclusion
We are grateful to the authors for advancing this scientific dialogue. As surgical innovation progresses, rigorous synthesis of evolving data remains essential. We believe our conclusions are methodologically sound, appropriately cautious, and clinically informative.
