The spino-pelvic matching is reported an important predictor of QOL. A study conducted by Smith et al.  compared patients with improved PI-LL, PT, and SVA with incorrected ones and found that a better spino-pelvic parameters after orthopedics lead to a significantly satisfied life. Based on the correlation between the sagittal sequence and QOL, Schwab et al. regarded SVA < 40 mm, PI-LL < ± 10° and PT < 20° as the target of sagittal sequence alignment. However, Gelb et al.  confirmed that the threshold for SVA of 8.0 cm is appropriate for asymptomatic people over 60s, suggesting the strictness of SVA and PI-LL matching in elderly should be relaxed . With a less physiological LL in middle-aged and elderly relative to the young, the PI and LL was in a mismatched tendency and thus a constant threshold of parameters is unreasonable for various age groups. DTLK is an outcome with dual effects of compensation and decompensation  and there was probably significance on sagittal spino-pelvic sequence in DTLK with LSS compared to non-deformed LSS. This study explored the importance of spino-pelvic parameters on DTLK underwent short-segment fixation with more than 2-year follow-up. It was believed that the postoperative QOL was related to PI-LL and PT, but not to SVA.
Lafage et al. believed that DTLK with LSS tended to be performed surgery as the increased function demangding for the elderly . They included 773 patients (mean of 53.7 years) and found the ideal spino-pelvic parameters (SVA, PT, and PI-LL) increased with the age, where the elderly had more LL loss and trunk forward tilt, which is necessary to be considered during surgical reconstruction.
Long-segment fixation can obtain satisfactory local correction and sequential reconstruction, but there are also many disadvantages such as large cost, long operation and recovery time, even probably with unsatisfied outcomes [17, 18]. In recent decades, the concept of short-segment fixation on adult spinal deformity has become more mature. Wang et al.  implanted short-segment fixation in degenerative scoliosis and found though the sagittal and coronal Cobb angle adjacent to UIV increased, it made no effects on VAS and ODI. Kim et al.  compared short-segment fusion (UIV at T9-L1) and long-segment fusion (UIV at T1-T6) with a 2-year follow-up and believed the latter required longer operation time and hospital stay while the rate of complications and clinical outcomes was comprable. In our data, the UIV was selected lower than L2 and LIV was uniformly fixed at L5 since the fixation of S1 could affect pelvic rotation with more selection bias. Although the fusion range escaping from TLK, the postoperative TLK was still reduced and the sagittal parameters tended to be "reasonable". At first, preoperative LSS and nerve root compression forced a decrease of LL and compensate trunk anterior tilt, which was corrected by sufficient decompression after surgery [21, 22]. Then, more anterior version of pelvis, improved PI-LL matching and proximal TK compensation were acquired with the fixation and 1–2 grade of osteotomy on L4-S1. In addition, the increased strength of the paraspinal muscles in the two years by functional exercises may also improve the sagittal sequence .
In this study, the final PI-LL matching and whole balance got improved while with no significance on PT, indicating that the pelvic rotation was not obvious in DTLK group according to the preoperative PT (20.1 ± 10.5°) nearly in a normal range. There were limited cases acquired all corrction of three parameters, reflecting that the sagittal parameters are not improved simultaneously. Where, the sagittal sequence rebuilt sometimes ignored the whole balance and the reconstruction of the sequence was probably not equivalent to sagittal balance .
Berjano et al.  described the sagittal compensation mechanism and observed three stages of the whole imbalance: equilibrium state, compensate balance and compensate imbalance. Jackson et al.  pointed out the whole imbalance was a common cause in patients with low back pain and suggested the correction of sagittal balance could improve the symptom. However, Hasegawa et al.  believed that the postoperative QOL had nothing to do with the balance state while close to the muscle tension, which was in accordance with the point of this study. According to the theory of Barrey et al., the majority of DTLK patients were in a state of compensate balance. Although some cases were in imbalance after the surgery, the tension of paraspinal muscle was relaxed and with no difference in QOL compared to ones with equilibrium state [8, 28].
PI is the only morphological parameter and has a strong correlation with LL. Initially, PI-LL < 10° was determined as the standard range, roughly estimating the reasonable LL and the degree of orthopedics . However, the threshold of PI-LL was dynamically adjusted with the differences of ethnics and age, where Lafage proposed that the ideal PI-LL was more than 17.0° for people over 75 years old. Physiologically, middle-aged and elderly patients were with intervertebral disc dehydration and height collapse with a tendency of LL losing. According to Hasegawa et al.  with the ideal relationship LL = 32.9 + 0.60 × PI−0.23 × age, the trend of PI-LL was close to this study and our series believed the mismatch of PI-LL would affect the clinical efficacy. Zhang et al.  included spinal deformed patients (mean of 65y) with fusion and showed that the PI-LL < 10° group had a lower ODI score and rate of complications.
PT was a sensitive parameter of lumbar spino-pelvic mismatch and more studes realized that pelvic rotation could affect surgery efficacy. Lazennec et al.  confirmed that patients suffering from persistent low back pain after lumbar surgery had decreased SS and increased PT (P = 0.020) and the PT was almost twice that of normal ones. Therefore, not removing the risks for pelvic retroversion could cause lose of LL, similar to sitting posture, which is considered in retroversion state, discomfort could derive from the long-term overstretching of the back muscles terminating on the iliac crest. There was no consensus on ideal threshold of PT and some believed PT < 20° was the fixed range, while Vialle et al.  reducing the formula of PT = 0.37 × PI−7 was widely accepted. Kim et al.  conducted the relationship between lumbar spine balance and clinical outcomes after PLIF and found that the improvement of VAS and ODI was correlated to the correction of PT but not to other spino-pelvic parameters. After lumbar spine fusion, the LL motion sharply reduced and the range of PI-LL was resricted, where the sequence adjustment was mainly depended on the proximal thoracic spine and the rotation of pelvis. Thus, PT was crucial in the evaluation of postoperative clinical efficacy.
There were some limitations. Firstly, it was a series study only with pre- and post-operative contrustion, while the control group composed of non-deformed ones and long-segment fixation ones were not included. Then, the choice of LIV at S1 may have an impact on pelvic rotation, but this study did not include ones with S1 as LIV for stratification analysis. Finally, postoperative complications such as proximal junctional kyphosis and adjacent segment degeneration were not regarded, while the comprehensive outcomes and longer follow-up would strength the intensity of conclusion.