J Neurosurg Spine October 9, 2026
Choosing sagittal correction targets remains a central challenge in adult spinal deformity surgery.
This multicenter study evaluates lumbar alignment together with the relationship between T4 and L1.
It also examines how early junctional compensation can alter postoperative measurements.
The findings favor interpreting spinal shape in the context of fusion extent and caution against excessive correction.
Objective
Evaluate whether L1 pelvic angle (L1PA) and T4–L1 pelvic angle targets are associated with proximal junctional failure (PJF), accounting for early changes above the instrumented spine.
Methods
Retrospective analysis of a prospective multicenter adult deformity registry. Of 861 eligible patients, 661 had complete two-year follow-up and data. The principal comparisons involved 604 patients fused from the ilium to T2–4 (225 patients) or T9–11 (379 patients).
Six-week radiographs were assessed against published alignment targets. A virtual model estimated alignment without early proximal junctional compensation. PJF comprised proximal-extension revision or severe radiographic junctional kyphosis.
Main results
PJF occurred in 108/604 patients (17.9%); 50/604 (8.3%) underwent proximal-extension revision.
In upper-thoracic constructs, PJF rates were 25.6% with L1PA overcorrection, 12.9% within target, and 11.1% with undercorrection (P=0.029). This association persisted after demographic and comorbidity adjustment (P=0.028).
T4–L1 categories alone did not significantly distinguish failure rates in this subgroup (P=0.094).
In lower-thoracic constructs, removing simulated junctional compensation reclassified 175/379 patients (46.2%). The initial association between T4–L1 category and PJF then lost significance (P=0.148).
Interpretation
The most persuasive finding is that six-week alignment may already incorporate compensatory changes or evolving junctional failure. Consequently, a postoperative measurement can reflect both the surgical correction and the spine’s response to it.
The upper-thoracic findings raise concern about overcorrection relative to the studied L1PA formula. They do not establish intentional undercorrection as a superior treatment strategy. Evidence supporting particular combinations of lumbar and thoracic shape remains exploratory.
Limitations
Observational design limits causal inference, and only 76.8% of eligible patients contributed complete two-year data. Immediate postoperative radiographs were unavailable, requiring modeled reconstruction. Spinal flexibility was not directly measured, and patient-reported outcomes were not evaluated in this analysis.
Adjustment included age, sex, BMI and comorbidity burden, leaving potential residual confounding. Numerous authors disclosed industry relationships.
Clinical takeaway
Interpret L1PA and T4–L1 measurements alongside fusion extent and junctional behavior. The study supports caution with excessive lumbar correction and highlights the limitations of applying identical targets to upper- and lower-thoracic constructs.