Changes in Spinal Instability After Combined Therapy of Conventional Radiotherapy and Bone Modifying Agents for Painful Vertebral Bone Metastases

Precise assessment of spinal instability is critical at the beginning and after radiotherapy for selection of the treatment and evaluating the effectiveness of radiotherapy. We investigated changes of spinal instability after radiotherapy and examined potential risk factors for the difference of the outcome of spinal instability for painful spinal metastases. We evaluated 81 patients who received radiotherapy for painful vertebral metastases in our institution between 2012 and 2016. The pain at the vertebrae was assessed. Radiological responses of irradiated vertebrae were assessed by computed tomography. Spinal instability was assessed by Spinal Instability Neoplastic Score (SINS). Follow-up assessments were done at the start of radiotherapy and at 1, 2, 3, 4, and 6 months after radiotherapy. At each of one to six months, pain disappeared in 62%, 84%, 93%, 98%, and 100% of patients. The median SINS were 8, 7, 6, 5, 5, and 4 at the beginning of radiotherapy and after 1, 2, 3, 4, and 6 months, respectively, which signicantly decreased over time (P < 0.001). Multivariate analysis revealed that PLISE was the only risk factor for spinal instability at one month. In conclusion, spinal instability signicantly improved over time after radiotherapy. Clinicians should take attention to PLISE in the radiotherapy of vertebral metastases.


Introduction
In patients with bone metastases, spinal metastases are common, and approximately 60-70% of patients with advanced cancer develop them 1,2 . Spinal metastases frequently impair quality of life (QOL) of patients by spinal instability, which cause pain and malignant spinal cord compression (MSCC) 1, 2 . In review articles, the predictors of spinal instability included spine location, tumor size, bone quality (BQ), and spinal deformity 3,4 . However, few studies have used validated assessment tools for precise evaluation and classi cation of the spinal instability 5,6 .
In 2010, the Spinal Oncology Study Group (SOSG) formalized the Spinal Instability Neoplastic Score (SINS) to assess spinal instability 7 . The SINS evaluate spinal instability by adding together six individual component scores: spine location, pain, bone quality (BQ), radiographic alignment, vertebral body collapse (VBC), and posterolateral involvement of spinal elements (PLISE). Since its introduction, several authors reported its usefulness for diagnosing spinal instability and the value of the pretreatment SINS for prediction of the occurrence of VCF, pain response, and need for irradiation after RT [8][9][10][11] .
The conventional RT is the most common treatment for spinal SREs. However, the rate of occurrence and progression of new VBC after RT has not been fully investigated 12,13 . Furthermore, to the best of our knowledge, no studies have focused on the evaluation of time course of spinal instability after RT and potential risk factors for differences in spinal instability in patients with painful vertebral bone metastases without paralysis by MSCC by SINS. We therefore investigated changes of spinal instability for up to 6 months after RT and examined potential risk factors for the difference of the outcome of spinal instability for painful spinal metastases without paralysis using SINS.

Inclusion criteria
The records of patients who received RT for palliation of painful vertebral bone metastases with administration of BMAs either denosumab or zoledronic acid in our institution between July 2012 and June 2016 were retrospectively evaluated. Patients were excluded if they had received previous surgery or RT to the same irradiated vertebrae, recurrence in the same irradiated vertebrae, clinical MSCC, sacral lesions, and those who were followed-up for less than one month.

Treatment
All patients underwent RT. Choice of regimen of RT dose fractionation was determined by the treating radiation oncologist. They were 8 Gy in 1 patient, 20 Gy in 9 patients, 27 Gy in 1 patient, 30 Gy in 61 patients, and 40 Gy in 9 patients. Systemic anticancer agents (endocrine therapy, molecular targeted therapy, and cytotoxic chemotherapy) were administered to 56 patients (69%) after RT. All patients were treated conservatively with bracing in some patients.

Pain assessment
The pain at the time of movement (mechanical pain) at metastatic vertebrae was assessed. Follow-up assessments of pain was performed at the beginning of RT and at 1, 2, 3, 4, and 6 months after RT.

Radiological assessment
The status of vertebral bone was evaluated by CT (Aquilion, Canon) at 120 kV and slice thickness of 5 mm. All images were viewed with routine bone window settings (window level 200HU, window width 2000HU) with axial, coronal, and sagittal plane. Two authors (E.N., S.S.) evaluated the CT images, and any disagreements were resolved by consensus. We evaluated the individual component scores of SINS: spine location, BQ, VBC, radiographic alignment, and PLISE (Table 2). Follow-up assessments of radiological evaluations was performed at the beginning of RT and at 1, 2, 3, 4, and 6 months after RT.
BQ is classi ed as lytic, mixed, or blastic. Lytic lesions that have been successfully treated by RT subsequently appear normal or sclerotic on CT due to the reparative process called re-ossi cation [14][15][16] .
Radiological responses of irradiated vertebrae with lytic and mixed lesions after RT were assessed as follows; blastic change is de ned as complete ll-in or sclerosis of initially lytic or mixed lesion and mixed change is de ned as development of a sclerotic rim or partial ll-in or sclerosis of initially lytic lesion. VBC was de ned as reduction in vertebral body height compared to the height of upper and lower vertebral bodies. The degree of collapse was scored as 3 (> 50% collapse), 2 (< 50% collapse), 1 (no collapse with > 50% body involved), or 0 (none of the above) based on the criteria of SINS. The development of a new VBC in patients without VBC at the beginning of RT and progression of VBC in patients with VBC at the beginning of RT were evaluated.

Spinal instability
For assessing spinal instability, we used SINS. The SINS evaluate spinal instability by adding together six individual component scores: spine location, pain, BQ, radiographic alignment, VBC, and PLISE (Table 2) 7 . The minimum score is 0, and the maximum is 18. The total score is divided in three categories of stability: stable (0-6 points), potentially unstable (7-12 points), and unstable (13-18 points). In this study, we divided them in two categories: stable (< 7) and unstable (≥7).

Statistical analysis
Overall survival (OS) was estimated by Kaplan-Meier method. OS of patients with stable (SINS < 7) or unstable (SINS ≥7) at the beginning of RT were evaluated by Kaplan-Meier method and compared with the log-rank test.
The rate of progression of collapse was estimated using the Kaplan-Meier method. To assess risk factors for new VBC, progression of collapse, and spinal instability at one month after RT in patients of spinal instability (SINS ≥ 7) at the time of the RT, clinical data were assessed, including the following: age, gender, primary cancer site, radiation site, chemotherapy after RT, overall dose (RT), and/or SINS and some of components of SINS (BQ, radiographic spinal alignment, VBC, and PLISE).
Univariate analysis was performed using chi-square test and multivariate analysis was performed using logistic regression. For all analyses, associations were considered signi cant if the associated P value was < 0.05. All statistical analyses were performed with the statistical computing software BellCurve for Excel (Social Survey Research Information Co., Tokyo, Japan).
This study was approved by the Ethical Review Board of our hospital and conducted in accordance with the World Medical Association Declaration of Helsinki.

Overall survival
Some patients died of the disease in the follow up period. The number of evaluated patients was 81, 68, 56, 48, and 36 at 1, 2, 3, 4, and 6 months, respectively. At three and six months after RT, the OS rates were 78% and 56% ( Figure 1A). At three and six months after RT, the OS rates were 86% and 79% in patients with stable spine and 76% and 51% in patients with unstable spine, respectively ( Figure 1B). There was no association between the OS and SINS (stable (< 7) or unstable (≥7)) (P = 0.97).
New VBC occurred in 6 patients (7%). It occurred no patient of no collapse with ≤ 50% body involved and 6 patients (27%) of no collapse with > 50% body involved at the beginning of RT. Its degree was less than 50% collapse, which occurred at 1 month after RT without no further collapse. Univariate analysis revealed that BQ (lytic lesion) and PLISE were the risk factors for new VBC (Table 4). Multivariate analysis revealed no factor for new VBC.
In patients of < 50% collapse at the beginning of RT, the collapse progressed in 30 patients (60%) till 1 to 4 months (median 1 month) In six patients the collapse progressed to be > 50% collapse. In patients of > 50% collapse at the beginning of RT, it progressed till 2 months (median 1 month) in 3 patients (75%). The rate of progression of collapse was 35%, 67%, 84%, 92%, and 92% at 1, 2, 3, 4, and 6 months, respectively ( Figure 2). Univariate analysis revealed that PLISE was the only risk factor for progression of collapse at one month (Table 5). Multivariate analysis revealed that PLISE was the only risk factor for progression of collapse at one month (RR, 5.4; 95% CI, 1.08 to 27.72; P < 0.05).
At the beginning of RT, destruction of posterolateral elements of the spine was seen in 19 patients (24%); unilateral in 16 patients (20%) and bilateral in 3 patients (4%). In 6 patients, it was repaired by recalci cation after RT.
The score of SINS increased in 8 patients (10%) by the progression of the collapse and/or occurrence of deformity. The grade of instability was advanced in all but one patient in whom potentially unstable became unstable (1.2%).
Univariate analysis revealed that overall dose (< 30Gy), BQ (lytic lesion), and PLISE were the risk factors for spinal instability at one month (Table 7). Multivariate analysis revealed that PLISE was the only risk factor for spinal instability at one month (RR, 6.3; 95% CI, 1.05 to 37.6; P < 0.05). At one month spinal instability was seen in 89% and 52% of the patients with and without PLISE, respectively, which was signi cant (P < 0.01).

Case 1
A case of 77 years old female of lung cancer patient is shown in Figure 4. At the beginning of RT, < 50% collapse of vertebral body and destruction of bilateral pedicles was seen, as score of 11 of SINS ( Figure   4A, B). At 1 month after RT, pain continued and VBC progressed with occurrence of malalignment (kyphosis). No re-calci cation was seen, as score of 13 of SINS ( Figure 4C, D). At 2 months after RT, pain disappeared and no re-calci cation was seen, as score of 10 of SINS ( Figure 4E, F). At 3 months after RT, re-calci cation (partial clerosis of initially lytic lesion) was seen in vertebral body as judged to be achieved mixed change. Then, the total SINS score was 9 ( Figure 4G, H). At 4 months after RT, complete ll-in and sclerosis of initially lytic was seen in vertebral body, which was judged as achieved blastic change. Blastic change was achieved in both facets. Then, the total SINS score was 5 ( Figure 4I, J).

Discussion
In this study, we showed the improvement of spinal instability by combination therapy of RT and BMAs by disappearance of pain and re-calci cation, though VBC and malalignment progressed in some patients. Previous studies showed that pain decreased in 71-75% of patients at three months after RT 17,18 . In this study, pain disappeared in 93% of patients at three months after RT. Lytic lesions that have been successfully treated by RT subsequently appear normal or sclerotic on CT due to the reparative process called re-ossi cation [14][15][16] . In this study, the improvement of BQ was obtained by re-calci cation which could be facilitated by combined therapy of RT and BMAs.
On the other hand, the instability of spine can increase by the progression of VBC. Although conventional RT is the most commonly used for spinal SREs, the occurrence of new VBC has not been fully investigated. Rief et al. reported the occurrence of new VBC as 2% in various cancer type 12 . Lee et al.
reported the occurrence of new VBC as 9% in colorectal cancer 13 . But the investigation of occurrence of VBC in these studies limited to all patient. In this study, the occurred of new VBC were seen in 7% among all patients, but when evaluated among the patients with no collapse, it occurred just in patients with > 50% body involved at the beginning of RT with high rate of 27%.
In patients with VBC at the beginning of RT, the collapse progressed in 61% of patients after RT. It occurred 1 month after RT and stop within 2 months in most patients with collapse progression rate of 84% and 92% at 3 and 6 months, respectively. Multivariate analysis revealed that PLISE was the only risk factor for progression of the collapse. Then, clinicians should be aware of the need of close monitoring of progression of the collapse in patients with PLISE.
Precise assessment of spinal instability is critical at the start of RT to decide patients who require surgical intervention as well as assessing it after RT to evaluate the effectiveness of RT 3, 4, 7 . There are several tools for evaluating the spinal instability 3,4 . However, none of these tools has been completely validated or widely used in a clinical setting. Several studies utilizing Taneichi score reported that patients who were classi ed unstable prior to radiotherapy, 17-19% and 24-32% of patients were classi ed as stable at three and six months after RT, respectively 5,19 . However, this tool is limited to be utilized for lytic thoracolumbar lesion. In 2013, the American Academy of Orthopedic Surgeons introduced SINS as a classi cation system of spinal instability in an instructional course lecture for general practitioners 20 .
In this study, we rst used SINS as an assessment tool of spinal instability in the course of RT. Although the SINS increased in 10% of patients, the degree of stability was advanced in all but one patient (1.2%).
The median SINS showed a signi cant decrease over time (P < 0.001).
We showed that PLISE was the only risk factor for both progression of vertebral collapse and continuous spinal instability. In vertebral bone, posterolateral elements of the spine is one of the most important factors for spinal stability. Then, clinicians should pay attention to not only VBC but also PLISE.
Limitations include a small sample size of only 81 patients. This will be due to the selection of patients who received close monitoring by CT with combined therapy of RT and BMAs and common in the study for patients with bone metastases, given their relatively short survival.
In conclusion, although progression of vertebral collapse and malalignment occurred, disappearance of pain and re-calci cation were obtained after RT which led to the spinal stability in patients with painful vertebral bone metastases. PLISE was the risk factor for both progression of vertebral body collapse and continuous spinal instability. Then, clinicians should pay attention to not only VBC but also PLISE.
Declarations EN, TK, and TO organized the study. EN, SS, and YS treated the patients. EN, RN, ST, and KS collected and analyzed data.

Funding
All authors have no funding in this study.

Competing interests
The author(s) declare no competing interests.

Data availability
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.