Establishment of an Individualized Distal Junctional Kyphosis Risk Index following the Surgical Treatment of Adult Cervical Deformities.

Study Design: A retrospective review of a multicenter comprehensive cervical deformity (CD) database.

Objective: To develop a novel risk index specific to each patient to aid in patient counseling and surgical planning to minimize postop distal junctional kyphosis (DJK) occurrence.

Background: DJK is a radiographic finding identified after patients undergo instrumented spinal fusions which can result in sagittal spinal deformity, pain and disability, and potentially neurological compromise. DJK is considered multifactorial in nature and there is a lack of consensus on the true etiology of DJK.

Materials And Methods: CD patients with baseline (BL) and at least one-year postoperative radiographic follow-up were included. A patient-specific DJK score was created through use of unstandardized Beta weights of a multivariate regression model predicting DJK (end of fusion construct to the second distal vertebra change in this angle by <-10° from BL to postop).

Results: A total of 110 CD patients included (61 yr, 66.4% females, 28.8 kg/m 2 ). In all, 31.8% of these patients developed DJK (16.1% three males, 11.4% six males, 62.9% one-year). At BL, DJK patients were more frail and underwent combined approach more (both P <0.05). Multivariate model regression analysis identified individualized scores through creation of a DJK equation: -0.55+0.009 (BL inclination)-0.078 (preinflection)+5.9×10 -5 (BL lowest instrumented vertebra angle) + 0.43 (combine approach)-0.002 (BL TS-CL)-0.002 (BL pelvic tilt)-0.031 (BL C2 - C7) + 0.02 (∆T4-T12)+ 0.63 (osteoporosis)-0.03 (anterior approach)-0.036 (frail)-0.032 (3 column osteotomy). This equation has a 77.8% accuracy of predicting DJK. A score ≥81 predicted DJK with an accuracy of 89.3%. The BL reference equation correlated with two year outcomes of Numeric Rating Scales of Back percentage ( P =0.003), reoperation ( P =0.04), and minimal clinically importance differences for 5-dimension EuroQol questionnaire ( P =0.04).

Conclusions: This study proposes a novel risk index of DJK development that focuses on potentially modifiable surgical factors as well as established patient-related and radiographic determinants. The reference model created demonstrated strong correlations with relevant two-year outcome measures, including axial pain-related symptoms, occurrence of related reoperations, and the achievement of minimal clinically importance differences for 5-dimension EuroQol questionnaire.