Assessment of sagittal alignment in spinal deformity patients: a clinical utility analysis of global alignment and proportion (GAP) score

INTRODUCTION: Spinal deformities, including adult spinal deformities (ASD) and adolescent idiopathic scoliosis (AIS), can cause significant pain and disability. Radiographic parameters are measured in order to establish quantitative relationships among spinal alignment and health related quality of life metrics. Various classification systems including the Scoliosis Research Society (SRS)-Schwab and Global Alignment Proportion (GAP) score have been developed to correlate radiographic parameters with relevant clinical outcomes. While the SRS-Schwab Classification is both a coronal and sagittal categorization that correlates to health related quality of life metrics, the GAP score is a newly developed 13-point scale that focuses on sagittal alignment to predict mechanical failure postoperatively. The study aims to evaluate the utility of the GAP score as a method of evaluating radiographic correction and maintenance over one year. METHODS: This was a retrospective analysis of ASD and AIS patients who underwent operative intervention. Patients were evaluated for GAP score changes from preoperative to postoperative and then to a one year follow up. Of the 285 screened patients, 73 had suitable radiographs for measurements. This included 45 ASD patients, 32 of which had one year follow up radiographs, and 28 AIS patients, of which only 19 had one year follow up radiographs. Measurements were obtained by three independent readers. These included pelvic incidence (PI), pelvic tilt (PT), L1-S1 lumbar lordosis (L1-S1 LL), L4-S1 lumbar lordosis (L4-S1 LL), sacral slope (SS), global tilt (GT), and sagittal vertical axis (SVA). GAP score subcategories were calculated and summed to produce the GAP score. RESULTS: The 73 included patients had an average age of 41.8 ± 24.1, BMI of 26.2 ±6.3, and consisted of 51 females (69.9%). All but one radiographic parameters exhibited an Intra Class Correlation (ICC) of over 0.66. PI had an ICC of 0.59 which was likely due to the incomplete visibility of the femoral heads in few radiographs. Among ASD patients, from pre to postoperative, PI-LL significantly decreased from 18° to 5°, PT significantly decreased from 26° to 19°, L1-S1 LL significantly increased from 37° to 47°, L4-S1 LL significantly increased from 28° to 31°, GT significantly decreased from 37° to 25°, SVA significantly decreased from 120 mm to 64 mm, and total GAP score significantly decreased from 8 to 6. For the 32 patients with long term data, the GAP score changed significantly from 9 to 6 to 7 at one year follow up. Of the 32 patients, 14 patients experienced a worsened score at the one year follow up. Among AIS patients, PI-LL increased from 0° to 3°, PT remained 13°, L1-S1 LL decreased from 56° to 53°, L4-S1 LL significantly decreased from 37° to 31°, GT increased from 14° to 16°, SVA decreased from 25 mm to 28 mm, and total GAP score increased from 3 to 4. For the 19 with long term data, the GAP score changed significantly from 3 to 3 to 2 and 2 patients experienced a worsened score at the one year follow up. DISCUSSION: The patient population showed a slight improvement in GAP score among both ASD and AIS groups. The improvement is more pronounced when evaluating with the SRS-Schwab Classification. The GAP score does not emphasize the same radiographic parameters and therefore does not correlate well with the SRS-Schwab Classification. Many patients also experienced a worsened GAP score by the one year follow up which can be attributed to the adjacent superior or inferior region (PJK), thoracolumbar region (construct failure), or compensation at the pelvis through hip flexion. Additionally, GAP is potentially oversensitive to measurement error and rounding differences. Thus, the utility of the GAP score in evaluating spinal deformity patients is questionable and requires further adjustment and evaluation based on prospective studies