Spinopelvic sagittal imbalance is risky for development of proximal instrumented fracture after posterior instrumentation.

Backgrounds: Postoperative fracture of the upper instrumented vertebrae (UIV) is thought to be as fracture type of proximal junctional failure (PJF), which usually needs revision surgery for salvage. The purpose of this study was to investigate the inuences of spinopelvic parameters, such as the pelvic incidence (PI) angles, sacral slope (SS) angles, and pelvic tilt (PT), sagittal vertical axis (SVA), and proximal local kyphosis (PLK) angle on the development of fracture type of proximal junctional failure after posterior instrumentation. Methods


Introduction
Proximal instrumented fracture means fracture developed at the uppermost instrumented vertebrae (UIV) after posterior instrumentation. Proximal junctional failure (PJF) after instrumented spinal fusion is a well-known complication that makes clinical outcomes worse and usually needs revision surgery. The etiologies of PJF include a few categories: degenerative process, hardware loosening, or fracture induced.
Proximal junctional kyphosis over 20 o with symptoms by disc degeneration at upper segment is thought as degeneration type of PJF; a fracture at UIV or UIV + 1 is considered to be fracture type of PJF [1]. Therefore, proximal instrumented fracture is a part of PJF. Degenerative process of PJF usually takes years to develop, but fracture type of PJF generally happens within few months after primary index surgery, which makes patients feel more frustrated because they need to undergo a revision surgery within a short period of time.
Spinopelvic malalignment is associated with clinical outcomes of lumbar arthrodesis, and also contributes to the development of adjacent segment degeneration after instrumentation [2][3][4]. Older age, osteoporosis, UIV level at the thoracolumbar junction and greater preoperative sagittal vertical axis (SVA) were identi ed to be risk factors of PJF [5]. Spinopelvic parameters such as sacral slope (SS), pelvic tilting (PT), and pelvic incidence (PI) has been recognized to affect overall global sagittal balance [6].
Little literatures focused on the relationship between spinopelvic alignment and proximal instrumented fracture. We hypothesized that spinopelvic malalignment contributes to the development of proximal instrumented fracture. To exclude factors by surgical levels, age, and sex in the current study, the patients were limited to 1:3 matched case-control cohorts using the same instrumentation and the same fusion techniques. The purpose of this study was to clarify the relationship between spinopelvic radiographic parameters and proximal instrumented fracture after posterior thoracolumbar or lumbar instrumented fusion.

Materials And Methods
This study (No. 202000340B0) was approval from the institutional review board of our hospital. All participates were received and signed informed consent. Patients underwent posterior instrumented fusion for thoracolumbar or lumbar diseases between January 2007 and December 2017 were reviewed. Those patients who had undergone revision surgery for symptomatic instrumented fracture at UIV were enrolled as the study group ( Figure 1). As this was a 1:3 matched case control study, seventy-two patients were selected as a control group ( Figure 2). We searched for these patients who matched the study group by main diagnosis with same surgical levels by reviewing medical records form our spinal surgery database. Same gender and similar age were another criterion for selecting. The preoperative medical condition of these patients was reviewed using the weighted Charlson Comorbidity Index (CCI) [7] and the American Society of Anesthesiologists (ASA) physical status classi cation [8]. Demographic data including age, sex, body mass index (BMI), bone mineral density (BMD) with T score, and fused segments of all study subjects were collected from medical records. Clinical outcomes were evaluated using the visual analogue scale (VAS) of leg and back pain [9] and the Oswestry Disability Index [10. The radiographic parameters including the lumbar lordosis (LL), sacral slope (SS), pelvic tilt (PT), pelvic incidence (PI), proximal local kyphosis (PLK), coronal scoliosis (CS), and sagittal vertical axis (SVA). LL angle was measured using Cobb method of upper endplate of L1-S1; SS angle was measured the angle between the sacral endplate and a horizontal line; PT angle was measured by the line through midpoint of sacral plate and midpoint of femoral heads axis, and the vertical line; PI angle measured by the line through midpoint of sacral line and midpoint of femoral heads axis, and the line vertical to sacral plate; PLK angle measured using Cobb method of upper endplate of UIV+1 and lower endplate of UIV presenting in kyphotic angle; CS angle was measured the maximal scoliosis angle on coronal plain radiographs; SVA was measured as the distance between the C-7 plumb line and the superior posterior corner of the S-1 vertebral body in the lateral radiograph. The achievement of a successful harmony of spinopelvic realignment had been mentioned by Schwab et al and score for spinopelvic realignment achievement were recorded [11]. The ideal realignment objectives in the sagittal plane included SVA < 50mm, PT < 20° and LL = PI ± 10°. The patient with PT < 20° would get 1 point, if not, get nothing. In the same way, the patient with LL = PI ± 10°, would get 1 point, if not, get nothing; the patient with SVA < 50 mm, would get 1 point, if not, get nothing. Thus, the total score ranges from 0 to 3. The higher score represents the better achievement of spinopelvic harmony.

Statistical analysis
Chi-square test and Fisher's exact test were used for categorical variables. Mann-Whitney U test was used for continuous variables. A value of p < 0.05 was considered statistically signi cant.

Postoperative radiographic parameters
The study group had signi cant lower LL (25.

Discussion
The incidence of PJF after surgery for patients with adult spinal deformities might range from 2% to 20%, and this wide range can be attributed mostly to the heterogeneity of the study populations, surgical method, and the difference in the de nition of PJF [12]. Yagi et al. observed 113 patients underwent surgical correction for their adult spinal deformities by method of propensity-matched comparison and concluded that low bone-mineral density is a signi cant risk factor to develop PJF [13]. Park et al. studied 63 cases underwent posterior instrumented fusion with their UIV at thoracolumbar junction and concluded that older than 70 years old and osteoporosis would increase incidence of PJF [14]. Because primary osteoporosis is positively related to age, it is easy to understand that patient with older age had higher incidence to develop PJF. In the current study, although the mean age surgery was similar, the study group had a signi cantly lower BMD than that in the control group, which explained that osteoporosis was really positively related to development of PJF. Furthermore, higher CCI leaded to lower BMD was found in the present study. The same phenomenon was also demonstrated by Bartels et al.: higher CCI and lower BMD were positively correlated at the same age [15].
Scheer et al. suggested that the UIV at T8 or below level would increase the risk to develop PJF [16]. Similarly, Shu ebarger et al. recommended that fusion should stop at T10 or above level but not at thoracolumbar junction in adult spinal deformity surgery to decrease the incidence of PJF [17]. In contrast, some studies did not support that the UIV at thoracolumbar junction was a risk factor to develop PJF [18,19]. In the current study, of 24 cases of the study group, 19 patients' UIV located at T11-L2; the other ve cases' UIV were at T10 and L3. But we could not conclude whether the UIV at TLJ was risky to develop PJK by our research, because these cases did not have the same diagnosis to determine their surgical level. In our practice, the selection of UIV usually chooses neutral and stable vertebrae on coronal plane, and the adjacent segment on the sagittal plane is healthy without obvious disc degeneration, and most cases' UIV in our study located at TLJ. However, another three times matched patients at TLJ in the control group were not found to be developed PJF, we would not suggest that every case should be always instrumented fusion to mid-thoracic level to prevent PJF.
Spinopelvic parameters have been studied in patients with adjacent segment disease after lumbar or thoracolumbar fusion. The main spinopelvic parameters include LL, SL, PT, and PI. Signi cant less LL was thought to be a risk factor to accelerate adjacent segment degeneration by Djursovic et al. [20]. In contrast, Lai et al found that less LL did not increase the incidence of adjacent segment disease after lumbar fusion [21]. In the current study, LL of the control group was signi cantly lordotic than that of the study group in both preoperative and postoperative radiographs. However, it was still di cult to determine the appropriate LL for individual person to prevent PJF. A lower PI has been indicated a risk factor to generate adjacent segment degeneration [22]. But there was no difference in PI between both groups in our study; we could not conclude the e cacy of PI on the development of PJF. Schwab et al. rst described the importance of PI-LL mismatch on the development of adjacent segment degeneration [23]. Generally, PI minus LL less than 10 degrees is believed to be ideal. Yagi et al. Smith et al. all con rmed larger PI-LL mismatch leaded to PJF [1,13]. Yang et al. concluded that a larger differences in PI-LL mismatch by surgical correction could decrease risk of PJF [24]. Similarly, we also found the PI-LL of the control group was lower than that of the study group, which were statistically signi cant differences in both preoperative and postoperative radiographs. Because each patient's PI will not be changed by surgery, therefore, it is necessary to correct LL to reduce PI-LL mismatch.
In this study, the authors found larger proximally segmental local kyphosis angle resulted in the development of fracture type of PJF in both preoperative (the control group: -9.20 o , the study group: 2.98 o , p< 0.001) and postoperative (the control group: -9.13 o , the study group: 5.3 o , p<0.001) radiographs.
Generally, segmental angle at TLJ should be 0 o in normal alignment. When proximal segmental angle becomes kyphotic, increase in the moment force of body weight will apply to the UIV. Under condition of osteoporosis, this increased force might induce fracture at UIV. Therefore, extending a few more segments of instrumentation proximally or cement augmentation at UIV might be considered when facing proximal segment kyphosis of planned UIV. In addition, some surgeon proposed additional surgical techniques to prevent junctional problems after instrumented fusion. Viswanathan et al. proposed a hybrid method with combining pedicle screw-rod construct and sub-laminar banding could provide biomechanical advantages to prevent proximal junctional stress [25]. Rodriguez-Fontan et al.
proposed another easier method which used Mersilene tape to stabilize the spinal process between UIV and UIV + 1 or 2 to achieve prevention of PJK [26]. However, these two methods had been clinically approved their e cacy to prevent the development of PJF.
Indeed, this study still had some limitations. First, this was a retrospective study, therefore, the patients were subjective to inconsistence in surgeon's experiences in selecting UIV level. Second, the number of the study group was still small, although 24 cases with of developed instrumented fracture were relatively larger in a series when comparing to previous literatures. Third, is was not possible to analyze all risk factors for PJF that were previously reported. Fourth, no case of UIV at mid-thoracic or upper thoracic was included in this study. Further prospective longitudinal studies are needed to identify the proper surgical method and selecting criteria for patients having a degenerative lumbar or thoracolumbar spine with spinopelvic imbalance.

Conclusions
Pre-and postoperative lower LL, and higher PI-LL/PLK were also signi cantly associated with development of proximal instrumented fracture. Therefore, obtaining appropriate LL and correcting PLK should be done at surgery to improve spinopelvic sagittal imbalance. The results suggest that the achivement of appropriate LL and PI-LL prevents proximal instrumented fracture after posterior instrumented fusion.

Declarations -Ethics approval
This study was performed after obtaining approval from the institutional review board of Chang Gung Memorial Hospital (No. 202000340B0). All methods were performed in accordance with Declaration of Helsinki.
-Consent to participate All participates were received and signed informed consent.
-Consent for publication Not applicable -Availability of data and materials.
All the necessary information is contained in the manuscript. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

-Competing interests
The authors have declared that there are no competing interests.

-Funding
This article did not receive any funding.

-Authors' contributions
Liao JC designed the study, collected the data, wrote the manuscript, and participated in revising the manuscript. Chen WJ designed the study and also participated in revising the manuscript.      T11-S1 L1-S1 L2-S1 L3-S1