Can Thoracolumbar Deformity Improve if Surgery Just for Lumbar Stenois Syndrome in Degenerative Thoracolumbar Kyphosis Patients ?

Background: Fusion across thoracolumbar spine or not for degenerative thoracolumbar kyphosis (DTLK) in lumbar stenois syndrome (LSS) remains controversial. The inuencing factors for postoperative TLK in this group have not been determined yet. So the study was to explore whether DTLK could improve with only surgery for lumbar stenois syndrome LSS and identify inuencing factors on postoperative TLK. Methods: The study was performed from January 2016 to December 2018. 69 participants (25 male) diagnosed LSS with DTLK were enrolled and surgery was only for LSS. Radiological parameters included TLK, lumbar lordosis (LL), pelvic incidence (PI), pelvic tilt (PT), sacral slope (SS) and osteoporosis. Clinical outcomes were visual analogue scale (VAS) and Oswestry disability index (ODI). According to lower instrumented vertebrae (LIV) on L5 or S1, inter-group comparisons were performed between LIV on L5 (L5 group) and S1 (S1 group). Results: Demographics was well-matched between L5 and S1 group with a mean follow-up of 24.3±12.1 (m). TLK improved with a mean of 16.2±7.6 (°) (P <0.001). It was insignicant on radiological and clinical parameters between L5 and S1 groups except a larger PT in S1 group (P=0.046). VAS (P=0.787) and ODI (P=0.530) were both indifference between normal TLK and DTLK at last (P (cid:0) 0.05). Postoperative TLK was affected by osteoporosis and SS, the latter was determined by PI and PT. Osteoporosis was the risk factor for TLK correction (P=0.001, OR=9.58). Conclusions: DTLK get improved if suegery only performed for LSS. TLK and clinical outcomes are comparable between L5 and S1 groups. Severe osteoporosis can impede TLK correction. cases in junctional kyphosis at endpoint


Background
Degenerative thoracolumbar kyphosis (DTLK), as a kind of adult spinal deformity, is a common degenerative spinal disease in the elderly [(1)].
Lumbar spinal stenois syndrome (LSS) combined with DTLK, with gradually increassing exposure rate nowadays, can results to low back pain and lower extremities dysfunction [(2)], which is sometimes cured by surgery, and the majority of the series mind the symptom of LSS instead of DTLK [(3)]. There is no doubt of operating on responsible levels with severe LSS, disc protrusion or instability, while whether instruments implanting and deformity correction acrossing or referring to the region of TLK has been lling with controversary [ (4,5)].
It is reported numerous successes hava been achieved on TLK deformity through 1 to 3 grade osteotomy and long-segment instrument spanning across this region, while adverse outcomes have emerged such as extensive trauma and operating duration, huge expenses and even instrument-related infection [(6, 7)]. Recently, more and more experienced surgeons proposed the de nation of short-segment xation and the theory of precision therapy [(8, 9)], by which it was enough to desolve the chief clinical complaints rather than correcting all malformation into a resonable range. Shin et al. [(10)]suggested simple decompression surgery, not the disordered levels, was able to restore satis ed sagittal alignment in 70% of patients caused by LSS, but the point was not further identi ed on LSS combined with DTLK yet.
There is biomechanical interaction among focal sagittal parameters of spino-pelvic alignment, where thoracolumbar segments, as a bridge element, is probably affected by other regional parameters [(1, 11)]. Once S1 was chosen as lower instrumented vertebrae (LIV), the sacral slope (SS) and pelvic tilt (PT) was hardly to be changed since the pelvic incidence (PI) was almost xed in the adult where PI=PT+SS. So, the parameters may be signi cant when L5 was as LIV with more range of pevic rotation [(12, 13)]. While it is unclear whether TLK can be in uenced when L5 or S1 is chosen as LIV in short-segment fusion through biomechanical chain. Therefore, the study included patients diagnosed as LSS combined with DTLK, who were performed short-segment xation simply for LSS, where upper instrumented vertebrea (UIV) was lower than L2. A short-term follow-up was completed to identify (1) whether DTLK could improve without intervention to thoracolumbar region, (2) whether TLK was different with LIV on L5 or S1 and (3) the in uenced factors on TLK after surgery.

Patients enrollment
The single-center retrospective protocal was performed from January 2016 to December 2018. The participants diagnosed LSS combined with DTLK was enrolled. The study was approved by local institutional review board and all patients have signed informed consent.
For DTLK cases, posterior lumbar interbody fusion (PLIF) and posterior-lateral fusion (PLF) was applied mainly for lumbar segments with spinal stenosis, disc hernia and instability, which was instrumented in-situ or with grade 1 to 3 osteotomy. All patients was operated by one senior surgeon.

Radiological parameters and clinical outcomes
Radiological parameters included kyphosis apex, thoracic kyphosis (TK), TLK, lumbar lordosis (LL), PI, PT and SS. TK was the angle between upper endplate of T5 and lower endplate of T12; TLK was the angle between upper endplate of T10 and lower endplate of L2; LL was between upper endplate of L1 and upper endplate of S1. The de nation for PI, PT and SS was shown in Fig 1 (Fig 1). Osteoporosis was determined by X-ray, MRI and surgical records, where severe osteoporosis was manifested by decreased bone density, thinned trabecular bone, biconcave or wedge-shaped changes in vertebrae, which was together evaluated during operation [(14)].
Clinical outcomes were evaluated by visual analogue scale (VAS) and Oswestry disability index (ODI). VAS ranged from 0 to 10 and a higher score implied more severe pain. ODI re ected disability on lumbar spine function and quality of life (0-50 score) and higher index represented more disability. All outcomes were measured before operation (baseline) and at follow-up endpoint.
Furthermore, according to LIV on L5 or S1, inter-group comparisons were performed between LIV on L5 (L5 group) and S1 (S1 group).

Statistical analysis
The measurement data was depicted as mean± standard deviation. The dichotomous between groups were analyzed by χ 2 test. Independent sample t test was used for inter-group measurement data and Paired t test was applied between baseline and endpoint, while Wilcoxon test was used for ordinal data. Pearson or Spearman correlation analysis were used among parameters and between TLK and demographics. Multiple linear regression and logistic regression were for determining in uencing factors of TLK. SPSS 22.0 (IBMC, Armonk, New York, USA) the software for statistical analysis and P <0.05 was signi cant difference.

Results
A total of 69 DTLK patients (25 male) was included with a mean follow-up of 24.3±12.1 (m), the age and body mass index (BMI) were respectively 68.9±9.2 (55-84) (y) and 26.1±3.5 (kg/m 2 ). The most operated segments was L2-L5 (29.0%), followed by L3-L5 (24.6%) and L3-S1 (14.5%). There were no signi cances in gender, age and BMI between L5 group and S1 group, so was osteoporosis (P 0.05). The number of operated segments was larger in S1 group (P <0.05) ( Table 1). SS between the two time points (P 0.05). In addition, VAS and ODI both decreased (1.9±1.6 and 7.9±6.3, respectively) at last (both P <0.001) ( Table 2). At baseline, there were no signi cances on radiological parameters between L5 and S1 group, so were VAS and ODI (P <0.05). These parameters were also comparable at the endpoint between groups except a larger PT (22.5±8.6) in S1 group (P=0.046). In L5 group, TK (20.7±12.1) and TLK (16.5±7.3) decreased at nal compared to baseline (P=0.041 and P=0.001, respectively) while the others kept stable, the same situation in S1 group (P=0.047 and P=0.012, respectively) ( Table 3) (Fig 2).     Although increasing requirement on body shape, it may be undserved to perform orthopdics with huge cost, enlarged invasion while lower effectiveness, especially for the middle-aged and elderly [(16)]. Due to the shorter rest-survival expectation, self-regulation or conservative treatment is usually adopted for the most. However, as the most common degenerative spinal disease in this group, LSS can induce lower extremities pain and intermittent claudication with poor quality of life and there is no doubt that surgical treatment on responsible segment is chosen for most cases with severe LSS [ (10)]. The population diagnosed LSS combined with DTLK is non-ignorable based on magnanimous LSS cases, where they often wonder whether DTLK needs to be corrected since the chief complaints are mainly caused by LSS. It is also too confused for many spinal surgeons to give the identi ed answer.
Surgical correction with osteotomy and long-segment fusion on TLK has proven to be effective, leading to superior body appearance, clinical and radiographic outcomes, especially proper alignment is restored [ (7,17)]. However, surgical treatment spanning across malformed TLK remains challenging as demonstrated by revisions (9.0-17.6%) and adverse events such as extensive tissue stripping, burden expenses and rod fracture [ (18, 19)]. With the concept of minimum of invasion and maximum of e cacy, the superiority of precision therapy and short-level instrument are promoting, particularly with more profound knowledge of ASD, quali ed implants with better biocompatibility and Young's modulus, improved technology and perioperative management [(20)]. Once lower complications ratio and effective quality of life is still achieved with decompression and xation only on responsible level, it will be signi cant in saving marvelous wealth for society, government and themselves. This study was retrospectively performed on cases with LSS and DTLK, where the series underwent PLIF or PIF only for LSS, not referring to TLK. In total, it showed effective outcomes although there were 4 patients (5.8%) in instability on upper adjacent segment and 2 cases (2.9%) in proximal junctional kyphosis while with no symptom at endpoint (data not shown), which proved the concept of "precison xation" for DTLK ones was appropriate.
Patients with LSS appear with lumbar extension and improve with trunk fexion, where they can take severely stooped posture aiming pain relief [ (18) (10)]. In addition, the paraspinal muscles got stronger by cooperating with functional exercise after surgery, which probably reduced TLK [ (24)].
Sagittal alignment, biomechanism and clinical outcomes has been discussed when L5 or S1 was respectively chosen as LIV. Yasuda et al.
[ (25)] found fusion to L5 was conducted for selected ASD patients with better ODI and less complex deformity contrasted with S1 as LIV. Choi et al. [(13)] identi ed S1 double screws are a viable option for sacropelvic xation in ASD patients when L5 pedicle screw xation was not possible, and S1 group achieved better reconstruction on LL while restricted pelvis rotation. In this study, the only difference between L5 and S1 group was whether L5-S1 was etiological segment. The clinical outcomes were comparable in L5 and S1 at endpoint but the pelvic retroversion compared to L5 group, which was not consistent to Yasuda et al. while agreed with Choi et al. Although the restrictied motion of L5-S1 in S1 group, the compensary of proximal segment played an important role, which made comparable results on alignments between the 2 groups [ (26)]. In addition, both groups acquired decreased TLK and TK at last, mainly because the removal of compression and enlargement of lumbar canal, consequently the stooped posture for pain-relief was corrected [(2, 17)], which was in line with Fuji et al.
There are interaction among alignment parameters but TLK seems like estranged from others. It is considered that the abnormality of the transition site is affected by many aspects. Thoracolumbar region locates in the transitional part of both anatomical and biomechanical structure with large shear force, where vertebrae or intervertebral discs involved in the region are prone to wedge-shaped. Then, brosis of the anteriot ligament force it closer between adjacent vertebrae, especially in the elderly with heavy work [(27)]. TLK was affected by SS, which was determined by PI and PT. The pelvis is critical factor for sagittal alignment and is responsible for retroversion. Based on the relation SS = PI-PT, the amount of pelvic rotation can even approximate PI with sacral endplate horizontal, so individuals with high PI have wider range of adaptation. This movement is very signi cant, corresponding to an increase of PT correlated to back pain and disability [(28, 29)].Therefore, the retroversion of pelvis, in association with proximal spine forward leaning and even dorsal hin and knee fexion, analogue to sitting and resulting to capacity over-consumption. However, no relevance between TLK and clinical outcomes suggests that TLK puts little impact on qualitt of life in short-term and adequate decompression on LSS is fundamental method.
Age-related osteoporosis, with sparse trabecular and fragile cortex, can induce thoracolumbar spine deformities, resulting in kyphosis, shortened budy length and trunk anteversion, especially based on dorsally mismatched elastic modulus cause by instrument. Yagi et al. The study rstly identi es the e cacy on DTLK improvement by only performing PLIF on LSS, which can proposal a new mentality on treatment DTLK and provide evidence for surgeon for preoperative conversation. The result puts a further explaination on the interaction between the focal deformity and whole spino-pelvic alignment. It also emphasizes the superiority of shorter-level xation, supplementing and riching the theory of precision treatment. There were some limitations to be mentioned. Firstly, the sample in both groups is so small and a prospective, larger cohort with longer follow up will strength the conclusion. It can not provide surgical strategy for DTLK with LSS or stability in upper lumbar level (T12-L1 or L1-L2) where the instrument and fusion is inevitable. In addition, the results is not suitable for coronal deformity such as degenerative lumbar scoliosis or rigid TLK such as fracture-derived kyphosis, ankylosing spondylitis or Scheuermann's disease. Availability of data and material: The data sets used and/or analyzed during the current study are available as an additional supporting les.

Competing interests:
The authors declare that they have no competing interests.  Diagram of spino-pelvic radiological parameters. (A) TK was the angle between upper endplate of T5 and lower endplate of T12. TLK was the angle between upper endplate of T10 and lower endplate of L2. LL was between upper endplate of L1 and upper endplate of S1. (B) PT was the angle between plumb line and the center of the femoral head to midpoint of upper endplate of S1. PI was the vertical line passing through the midpoint of upper endplate on S1, then second line connecting midpoint of the upper endplate on S1 and the femoral head and the angle between the second line and vertical line. SS was the angle between upper endplate of S1 and horizontal line.