671 - Identifying Thoracic Compensation and Predicting Reciprocal Thoracic K...

General Session: Lumbar

Presented by: B. Diebo - View Audio/Video Presentation (Members Only)


T. Protopsaltis(1), B. Diebo(2), R. Lafage(2), J. Smith(3), D. Sciubba(4), P. Passias(1), E. Klineberg(5), D. Burton(6), C. Ames(7), C. Shaffrey(3), S. Bess(8), R. Hart(9), F. Schwab(2), V. Lafage(2), International Spine Study Group

(1) NYU Hospital for Joint Diseases, New York, NY, United States
(2) Hospital for Special Surgery, New York, NY, United States
(3) University of Virginia, Charlottesville, NH, United States
(4) Johns Hopkins Hospital, Baltimore, MD, United States
(5) UC Davis Health System, Sacramento, CA, United States
(6) The University of Kansas Hospital, Kansas City, KS, United States
(7) UCSF, San Francisco, CA, United States
(8) Rocky Mountain Scoliosis & Spine, Denver, CO, United States
(9) OHSU, Portland, OR, United States


Purpose: To compare patients with postoperative reciprocal thoracic kyphosis (RK) and those who maintain thoracic alignment (MT) and to identify thoracic compensation to determine if it predicts the reciprocal change and PJK.

Introduction: Adult spinal deformity requires pts to recruit compensatory mechanisms like pelvic retroversion and knee flexion. However, thoracic hypokyphosis is a less recognized and poorly defined compensation.

Methods: 219 pts undergoing TLD correction were included with fusions to the pelvis and UIV of T9-L1. Pts were divided into those with postop reciprocal thoracic kyphosis (RK: Δ unfused TK ≥15°) and those who maintained thoracic alignment (MT). Thoracic compensation was defined as theoretical thoracic kyphosis (tTK) minus preop TK. The tTK was calculated from LL=2(PI+TK) + 10 where LL was PI+10 for PI≤40°, PI=PI for PI 40°-70° and PI-10 for PI≥70°.

Results: For RK (n=117), the mean change in unfused TK was 21.7° and the mean PJK was 17.6° vs 6.1° and 5.7° (p< 0.001) for MT (n=102). RK and MT were similar in age, BMI, gender, and comorbidities. RK had larger preop PI-LL mismatch (30.7 vs 23.6 p=0.008) and less preop TK (22.3 vs 30.6 p< 0.001), otherwise SVA, PT and TPA were similar. RK pts had more PI-LL correction (29.8 vs 17.3, p< 0.001) and more preop thoracic compensation (29.9 vs 20.0, p< 0.001). There were no differences in preop HRQOL except RK had worse SRS appearance (2.2 vs 2.5, p=0.005). Using a logistic regression model, the only predictor for RK was more thoracic compensation. Postoperatively the RK and MT groups were well aligned by all SRS Schwab modifiers. The RK group had 76 patients with PJK and 39 without There were no differences between these pts in thoracic compensation, PI-LL correction, or preop/postop alignment except PT (Figure). HRQOL were not different for any group at 6wk and 1y. Both younger RK and older (>65y) had larger thoracic compensation and more correction than MT young and old. The tTK was similar to the postop TK for all groups.

Conclusions: Postoperative reciprocal thoracic kyphosis can be anticipated and incorporated into preop planning of thoracolumbar deformity correction by calculating tTK. Reciprocal thoracic kyphosis was predicted by the magnitude of preoperative thoracic compensation.

PJK Compensation