Proximal junctional degeneration and failure: a novel classification and clinical implications

BACKGROUND CONTEXT: Proximal junctional problems occur at high rates after adult spinal fusion, and are often identified as proximal junctional kyphosis/failure (PJK/PJF). Existing classifications are descriptive, but not necessarily correlated with mechanism or clinical course. PURPOSE: The aim of this study is to propose a new classification system that will include different types of degenerative changes and failures related to the proximal junction, and to analyze the clinical and radiographic characteristics, and revision rates. STUDY DESIGN/SETTING: Retrospective radiographic review. PATIENT SAMPLE: Patients with posterior fusion of ≥3 levels and upper instrumented level (UIV) at or distal to T8, between 2018-2021, and with minimum 2-year follow-up were identified from a single-center multi-surgeon database. Revisions for nondegenerative reasons (e.g. infections) were excluded. OUTCOME MEASURES: Radiographic and clinical findings, and rates of revision were compared between different types of proximal junctional degeneration. METHODS: The proposed classification system identified 4 patterns: Type 1: symmetrical collapse involving multiple levels cranial to the UIV, Type 2: single adjacent level collapse with bony erosion ± screw penetration into disc space, Type 3: fracture, and Type 4: spondylolisthesis (Figure 1). RESULTS: A total of 150 patients (F/M:100/50) were included (mean age 65.1±9.8 years and follow-up 3.2±1 years); 92 (61%) patients developed degenerative changes in the proximal junctional region, and were classified as Type 1 (32, 21%), Type 2 (15,10%), Type 3 (22, 15%) and Type 4 (12, 8%). Demographic, preoperative radiographic, and postoperative radiographic variables were compared between different types of degeneration. Greater preop SVA was associated with Types 3 and 4, greater preop PI-LL mismatch with Type 4 and greater postop thoracic kyphosis with Types 2 and 3. All results in parentheses are for types 1, 2, 3 and 4, respectively. Mean time to revision was shorter for Type 3 cases (1.9, 2.1, 0.9, 2.1 years, p=0.004). Rates for developing PJK were 31%, 67%, 46% and 33% (p=0.125), and for requiring revision due to PJK/PJF were 63%, 73%, 73% and 83% (p=0.564), and for developing neurologic deficit were 9%, 20%, 32% and 17% (p=0.221). CONCLUSIONS: This novel comprehensive classification system defines different modes of degeneration at the proximal junction, without reliance on angular definitions of PJK/PJF. Different types have different clinical courses (fractures have significantly shorter time to revision). Pre/postop radiographic characteristics are associated with specific types of failure (i.e. spondylolisthesis had the highest preop SVA). Future studies with larger sample sizes are needed to validate this novel classification. FDA Device/Drug Status: This abstract does not discuss or include any applicable devices or drugs.