Age-Adjusted PI-LL Versus L1 Pelvic Angle for Discriminating Proximal Junctional Kyphosis in Lower Thoracic Fusion for Adult Spinal Deformity
In brief
Below-target L1 pelvic angle linked to 51% kyphosis rate after fusion
In a retrospective study of 373 adults undergoing spinal fusion, 51% of patients with a below-target L1 pelvic angle developed proximal junctional kyphosis, compared with 32% without one. A below-target age-adjusted pelvic incidence minus lumbar lordosis was not associated with higher rates, although both measures discriminated similarly in adjusted models. When targets conflict, L1 pelvic angle may be the more informative boundary.
- Journal
- Neurosurgery (Q1)
- Published
- 12 November 2026
- Study design
- Cohort / observational study
- Evidence level
- Level 3, Low (CEBM 3b)
- Authors
- Se-Jun Park, Tae Soo Shin, Jin-Sung Park, Dong-Ho Kang, Jun-Seok Oh, Chong-Suh Lee
- PMID
- 43018276
- DOI
- 10.1227/neu.0000000000004296
Why clinicians should know about it
- Picked for Spine Surgery (paper of the day, 16 November 2026): Compares alignment targets for junctional kyphosis after deformity fusion
Abstract
BACKGROUND AND OBJECTIVES: Proximal junctional kyphosis (PJK) is a prevalent complication after adult spinal deformity surgery. Age-adjusted pelvic incidence minus lumbar lordosis (PI-LL) and L1 pelvic angle (L1PA) have both been proposed as lumbar correction targets, but their relative discriminative value for PJK has not been directly compared. METHODS: We retrospectively analyzed 373 patients who underwent fusion to the sacrum for adult spinal deformity with an uppermost instrumented vertebra between T9 and L1. PJK was defined as a proximal junctional angle of 20° or greater at 2 years or revision for proximal junctional complications. Each offset was the measured postoperative value minus its target value ([age -55]/2 + 3 for age-adjusted PI-LL; 0.5 × pelvic incidence -21 for L1PA), with negative values indicating a measurement below target. Directional agreement was assessed, and discrimination was evaluated with unadjusted restricted cubic spline and nested multivariable models adjusted for 10 clinical confounders, comparing areas under the receiver operating characteristic curve by the DeLong test. RESULTS: PJK developed in 133 patients (35.7%), and the 2 offsets correlated moderately (r = 0.535). Among 243 patients with a negative age-adjusted PI-LL offset, 181 (74.5%) did not have a negative L1PA offset; the reverse occurred in only 9 of 71 (12.7%) (P < .001). A negative L1PA offset was associated with PJK (50.7% vs 32.1%, P = .005), whereas a negative age-adjusted PI-LL offset was not (37.9% vs 31.5%, P = .271). Adding either offset improved discrimination over the base model (areas under the receiver operating characteristic curve 0.678-0.713 and 0.721, P = .031 and .017), but the 2 models did not differ (P = .643). CONCLUSION: The 2 indices discriminated PJK comparably, but only a negative L1PA offset identified patients at increased risk. When the 2 targets conflict, the L1PA target can be used as the more informative boundary.
Abstract as published, via PubMed.
For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.