The most common long-term complication of adult degenerative scoliosis patients who receive corrective surgery is ASD23,24. According to previous studies, the reasons and risk factors for ASD after long segmental spinal surgery remain controversial. Some reports have indicated that aging and biomechanical changes in the fused segment were risk factors10,25,26, but unrandomized controlled patient characteristics, preoperative surgical data, and mixed short and long segmental spinal fusions generated doubtful results.
Whether the possibility of ASD increases as more motion segments are included in spinal fusion remains unclear. Dehnokhalaji et al.27 found that the fusion level was not a significant factor in contributing ASD, but only the distal intervertebral disc was compared. Faldini et al. proposed that clinical ASD could not be decided by levels of fusion because different operation methods were made: radiculopathy in short fusion group with more properly aligned spine versus spinal deformity related back pain in long fusion group28. Gillet et al. included patients with five or more fusion levels and found that the risk of ASD was not increased 29. However, in our studies, the number of levels fused was 4.83± 0.23 in ASD group and 4.74± 0.21 in non-ASD group without significant difference suggested that the adjacent degeneration was not affected by the fusion levels.
The rib cage provides a stabilizing effect and may balance the detrimental effect of long-level fusion in the development of ASD. These findings may be attributed to the bracing effect of the rib cage. Thus, the recommended length of fusion should be extended to the thoracic area when treating patients with ASD proximal to a prior fusion29. In our study, the upper instrumented vertebrae above T12 was 7 patients (23.3%) in ASD group, and 5 (16.7%) in non-ASD group without significant difference which means extending the fusion level to the thoracic spine did not alter the risk of ASD; thus, “rib cage protection” had no effect because sagittal alignment has a more profound influence than structural anatomical protection.
The potential for variations in spinal curves is associated with two coordinating pelvic positional parameters. PIA (morphological) is calculated as the algebraic sum of the SSA (positional) and the PTA (positional) which means SSA and PTA could adjust reversely to keep PIA constant3031. Faldini et al analyzed changes in spino-pelvic parameters after surgery of high-grade lumbar isthmic spondylolisthesis in 41 patients and found that failure of instrumentation was significantly correlated with higher post-operative PTA32. In contrast, we found that lower PIA patients may be a contributing factor to the development of ASD after long spinal fusion (Table 2). The different finding maybe resulted from different diagnosis and operations. Many studies25,33–38 have focused on how to correct the sagittal profile, including increased LLA, reduced normal C7 plumb line and PTA, but our results showed that there was no significant significant difference between the incidence of ASD with the correction of all spinopelvic parameters (Table 3). Only preoperative, postoperative and follow-up PIA were responsible for the progression of ASD. The PIA, or pelvic base angle, is a useful descriptive terminology and an extremely important parameter for determining the global spinal balance of an individual39,40. The PIA thus implies the relative angle of sacral plate with respect to the femoral head and the amount of lumbar lordosis required to maintain an erect posture39–42. In summary, higher PIA was always accompanied by higher LLA due to higher SSA, which decreased the adjacent segment facet pressure and reduced the energetic consumption of erect muscle, which is compatible with our results. The SSA and LLA in non-ASD group were larger than in ASD group in pre-, post-operative and follow-up measurements. The accepted postoperative sagittal profile did not indicate proper preservation of an erect back muscle. One study43 found significant geometrical reductions of erector spinae by approximately 26% and 14% at the L5-S1 and L4-L5 levels, respectively, after posterior lumbar surgery. Another study44 found that fusion generated a 12% reduction in the total multifidus muscle force during erect standing, and 10.5% reductions were produced during 20° flexion. In our study, muscle weakness may be one of the major causes of adjacent level degeneration.
Although the PIA is an anatomical indicator of sagittal balance and is simple to estimate, the correlation between the PIA and ASD has not been determined. Our study revealed that the lower preoperative PIA contributed to the development of ASD after long segmental fusion maybe result from the suspicious of lower amount of back erect muscle43,44. A decreased LLA resulting from the lesser postoperative and final follow-up PIA may lead to an anterior shift of the upper body trunk4346. The patient would have to spend more energy in maintaining sagittal balance, which may cause the adjacent level to sustain more compressive load47. Eventually, the back muscle fatigues, and ASD develops. Preservation of back muscle plays a more important role than anatomical correction in preventing ASD in low PIA patients.
Limitations of the study
A prospective study with the same allocation criteria and more patients should be included to establish a more powerful design model. Geographic results, such as magnetic resonance scanning of erector spinae, should be compared to indicate the final results. Studies conducted in multiple centers could decrease the selection bias.