Revision surgery for proximal junctional failure: A single-center analysis

Abstract

Background

Proximal junctional kyphosis (PJK) is a radiographic complication following adult spinal deformity (ASD) surgery due to degeneration of mobile segments adjacent to fused spine. Proximal junctional failure (PJF) represents PJK with structural failure, neurologic deficit, or mechanical instability warranting revision with extension of fusion above the uppermost instrumented vertebra (UIV). This study investigates the clinical presentation, mechanisms of failure, revision strategies, and outcomes for ASD patients who develop PJF after instrumented fusion to the pelvis.

Methods

Fifty-four ASD patients who developed PJF after a posterior instrumented fusion to the pelvis at a single institution from 2009 to 2021 were analyzed. PJF was defined by radiographic PJK with (1) UIV or UIV+1 fracture, UIV screw pullout, or soft-tissue posterior ligamentous disruption, and (2) neurological deficit at presentation.

Results

The cohort was stratified into upper thoracic (UT, 10 patients, T2-T6), lower thoracic (LT, 35 patients, T8-T11), and lumbar (L, 9 patients, L1-L3) spine UIV groups based on index surgery. Patients developed PJF at a median of 14 months (mean 18 ± 16, range: 1–78) after their index surgery. Neurological deficits at presentation included radiculopathy (61 %), myelopathy (48 %), motor deficits (33 %), and bowel or bladder incontinence (9 %). Mechanisms of PJF were vertebral fracture and screw pullout (UT: 50 %, LT: 80 %, L: 89 %, P < 0.001) or soft-tissue disruption (UT: 50 %, LT: 20 %, L: 11 %, P = 0.089) at the UIV. Revision surgery commonly involved posterior column osteotomies (63 %) rather than three-column osteotomies (9 %). Of patients in the UT group, 40 % were extended above the cervicothoracic junction. In the LT and L groups, 91 % and 89 % of patients were extended to the UT and LT spine, respectively. Median follow-up for the cohort after revision for PJF was 24 months (range: 2–89). A total of 26 patients (48 %) required a second revision surgery (median 14 months, range: 1–50), 16 of whom (28 %) were revised for recurrent PJF. Patient-specific and radiographic risk factors for recurrent PJF could not be elucidated.

Conclusion

In this series of ASD patients, after revision for PJF, recurrent PJF was the most common complication requiring another revision. Junctional failures tended to be vertebral body fracture and screw pullout in the LT and L spine and soft tissue disruption in the UT spine. Most revisions involved posterior column osteotomies with proximal extension across the thoracolumbar junction or apex of thoracic kyphosis (e.g., L to LT, LT to UT); notably, nearly half of UT failures were not extended to the cervical spine. Future research is warranted to elucidate risk factors for recurrent PJF and potential preventative strategies.