The fate of thoracolumbar surgeries in patients with Parkinson’s disease, and analysis of risk factors for revision surgeries

Parkinson’s disease (PD) emerges a second common neurodegenerative condition after Alzheimer disease, characterized by tremors, bradykinesia, rigidity and postural instability [1]. In addition, patients with PD usually present with poor bone quality, severe muscular dysfunction and abnormal posture. These sequelae might result in spinal pathology and result in spine operations for some patients with scoliosis, kyphosis, osteoporotic fracture, or degenerative spondylosis [2]. Comparing with the general population, it would be expected that spinal surgeries in PD patients would be prone to fail, necessitating revision surgery. This is the dilemma facing the spine surgeon when treating PD patients with spinal disease. A few reports have documented that outcomes for PD patients after spine surgeries are not satisfactory because of a high complication rate. Koller et al. described a series of 23 patients that had a 35% reoperation rate [3]. Bourghli et al. reported 12 PD patients with long fusion, and half the patients eventually underwent revision [4]. Kimura et al. reported a multicenter retrospective study which contained 67 PD patients who underwent lumbar surgery, divided into three groups (laminectomy, fusion, corrective surgery) according to their surgical methods [5]. The results showed the surgical failure rate was higher in the fusion group and in the corrective surgery group than in the laminectomy only group; and a small preoperative lumbar lordosis angle was associated with a failure of the initial surgery. The risk factors for revision spinal surgery in PD patients undergoing instrumented thoracolumbar surgery are not well defined due to the small study series with limited number of patients. Because the data concerning about risk factors for PD patients underwent spine surgery are still limited, the optimal guideline for surgically treating this kind of patient remains inconsistent. In this study, we reviewed the clinical and radiographic results of PD patients who underwent thoracolumbar or lumbar instrumented fusion procedures at our department, and analyzed the risk factors leading to revision surgery. The purpose of the study was to identify factors leading to surgical failure in PD patients who underwent instrumented spine surgery, and to develop surgical strategy for this kind of patients.

After obtaining the approval from the institutional review board, we retrospectively reviewed patients with PD who underwent spinal surgeries between October 2004 and April 2015 at our institute. The inclusion criteria was PD patients who were underwent decompression with instrumented thoracolumbar or lumbar arthrodesis. The main pathology for surgery was spinal stenosis with lumbar or thoracolumbar instability, including degenerative lumbar spondylolisthesis, degenerative lumbar scoliosis, degenerative lumbar kypho-scoliosis, degenerative thoracolumbar kypho-scoliosis, and adjacent segment instability after previous surgery. We excluded index surgeries for infection or tumor. In the present study, all patients underwent conventional open surgeries, either posterior approach only or anterior combined with posterior approach.

Preoperative plain radiographs (lateral, anteroposterior, and flexion-extension) and magnetic resonance imaging were used to assess the spondylolisthesis, degenerative coronal or sagittal deformity, and spinal stenosis. Postoperative plain radiographs and magnetic resonance imaging were used to evaluate new fractures, implant-related complications and adjacent stenosis or instability.

Recent meta-analysis have shown a rising prevalence of PD with age (all per 100,000): 428 for ages 60–69 years; 425 for ages 65–74 years; 1087 for ages 70–79 years; and 1903 in those older than age 80 [8]. In addition, advancement in medical care and public health has led to a rapidly growing geriatric population, those of which continue to lead active lives well into their eighth and ninth decade. As such, the number of PD patients with complicated spine problems who need to be treated will increase.

PD patients present with decreased bone quality and postural dysfunction. A meta-analysis conducted by Zhao et al. concluded that PD patients were at higher risk of osteoporosis than healthy controls (OR 1.18, 95% CI: 1.09–1.27) [9]. Another meta-analysis measuring the same parameters found similar results (OR 2.61, 95% CI: 1.69–4.03) [10]. In addition, other studies suggest that the relationship between PD and low bone density is associated with the H&Y stage, and to the duration of disease evolution [11–14]. Several mechanisms may contribute to PD-related bone loss, including weight loss, immobilization, Vitamin D levels, L-dopa therapy and dietary deficiency [15]. In our study, the revision group was composed of a significantly higher ratio of patients with osteoporosis and those with severe modified H&Y stage scores. Indeed, modified H&Y stage was the only independent risk factor for revision observed in our study, which suggests these two risk factors were positively associated or confounding.

Sagittal deformities of PD including camptocormia and antecollis, can be contributed to by muscular rigidity, axial dystonia, weakness caused by myopathy, body scheme defects due to centrally impaired proprioception, and by structural changes in the spine [16]. Oh et al. found 42% of PD patients had significant sagittal mal-alignment using the threshold of SVA > 50 mm, and 51% of those with PD had spino-pelvic mismatch (PI-LL > 10°) [17]. Bissolotti et al. analyzed 31 consecutive PD patients in a cross-sectional study, focusing on sagittal alignment [18]. Although the anatomical parameter PI was similar, the functional PT appeared to be increased and SS decreased when compared to the healthy adult cohort. In another retrospective study of 175 PD cases, they found male gender, longer disease duration, higher H&Y class, and a low plumb line-L3 distance were negative factors for spinal imbalance and risk of falling [19]. In our study, patients with PD in both the revision and non-revision group had abnormal lumbo-pelvic radiographic parameters compared with the general population (decreased LL, increased PT and decreased SS), although this could be related to age and degenerative change as well. Given these reports, spinal surgery would be a more difficult task in PD patients, than in those without PD.

The previous studies reported the revision rates in PD patients who had undergone spinal surgeries were from 21 to 86% [2–4, 7]. In our study, the revision rate was 29%, and the most common reasons for revision surgeries included hardware failure, instrumented fracture and compression fracture. All of these could be related to poor bone quality or progressive postural abnormality, a similar conclusion drawn by Babat et al., who reported that of 14 patients that underwent spine surgery, 12 (86%) required additional surgery, undergoing a total of 31 reoperations [20].

To improve postoperative satisfaction, some authors suggested that restoration of spinopelvic balance is paramount. Koller et al. described a series of 23 PD patients, in which 78% were satisfied or very satisfied despite a high rate of reoperations (35%); and proposed that the reconstruction of physiological lumbar lordosis and lumo-pelvic parameters was the key to prevent failure of surgery [3]. Postoperative or follow-up sagittal imbalance (C7-sagittal center vertical line; C7-SVL > 10 cm) had a significantly increased rate of revision surgery (p = 0.031). Bourghli et al. also echo this concept [4]. They concluded that long posterior instrumentation and fusion, from T2 to the pelvis, can restore sagittal and frontal imbalance, providing good clinical and radiographic results over the intermediate term with a high rate of satisfaction, despite 17% proximal junctional kyphosis rate and 50% revision rate.

So far, no ideal radiographic objectives allow surgeons to follow a PD patient’s course peri-operatively. We incorporated the concept of Schwab et al. to establish a scoring system, thus enable us to evaluate peri-operative spino-pelvic realignment achievement. Schwab et al. concluded that the following parameters during surgical intervention can achieve successful patient-specific spinopelvic realignment in the sagittal plane [7]. First of all, global spinal realignment should attempt to achieve a postoperative SVA < 50 mm, to attenuate the feeling of “falling forward.” Second, a PT < 20° is required during efficient ambulation. Finally, LL = PI ±9° may be used to achieve patient-specific alignment. It is evident that the goal of ideal spino-pelvic alignment cannot be obtained in all cases because of a number of limitations. We analyzed not only the spino-pelvic parameters peri-operatively, but also the objective score of spino-pelvic realignment achievement (score 0, 1, 2), to assess whether the realignment was necessary. Because this is a retrospective study and the SVA was not obtained for all patients, we did not use SVA as part of scoring system variables. There was no significant difference in preoperative or postoperative lumbo-pelvic radiographic parameters and score for spino-pelvic realignment achievement. Although Kaplan-Meier analysis showed no significance, we did notice a trend that lower scores led to earlier revisions. These findings suggest surgeons should maintain spino-pelvic harmony as much as possible.

Appropriate selection of surgical indications and awareness of possible risk factors may improve the outcomes of spinal surgeries. Increasing numbers of recent reports noticed the importance of the nature of PD itself on surgical outcomes. Moon et al. reported 20 patients with short lumbar fusions with poor surgical outcomes due to the progressive nature of PD [2]. Therefore, medical or surgical management of PD itself was also important to improve the outcomes of spinal surgery. Schroeder et al. reported 96 patients underwent lumbar spine surgery, but only 63 patients underwent instrumented surgeries [21]; the risk factors for additional surgery (p < 0.05) included an H&Y stage ≥3, a history of diabetes mellitus, treatment for osteoporosis, a combined anterior and posterior approach, the use of bone morphogenetic proteins (BMP), and the use of a spine interbody device. In a series of 48 PD patients with spinal deformity described by Bouyer et al., the rate of surgical revision was 42% and a well-balanced pre-operative condition was the only factor associated with optimal results [22]. In our study, the risk factors for revision surgery included modified H&Y stage ≥3, osteoporosis and a corrective osteotomy during surgery. In the binary logistic regression analysis, the modified H&Y stage ≥3 (p < 0.001) was the only independent risk factor. In other words, aggressive control of PD before and after surgery is necessary to prevent surgical complications. The Kaplan-Meier analysis revealed a trend for earlier revision in those with extensive surgical correction, such as corrective osteotomy, long instrumentation (surgical levels > 3), surgery extending to the thoracic spine, and a lower score of spino-pelvic realignment achievement.

There were a number of limitations in this study. The main limitation of this retrospective study was the limited number of participants, although the current study had the largest number of PD patients who underwent instrumented thoracolumbar surgery at a single institute to date. Second, variables including SVA and PD medication history could not be completely obtained due to the retrospective nature of the study. Third, the present study was a retrospective review of a heterogeneous patient population with different levels of spine disease, which might have bias and confound results.

A modified H&Y stage score ≥ 3 was the only independent risk factor leading to revision surgery. For a PD patient planning for elective thoracolumbar surgery, aggressive control of PD before and after surgery is necessary to prevent surgical complications. Corrective osteotomy or longer instrumentation also led to earlier revision. An optimal spino-pelvic realignment after surgery could reduce the occurrence of revision.