ISSLS PRIZE IN BIOENGINEERING SCIENCE 2019: biomechanical changes in dynamic sagittal balance and lower limb compensatory strategies following realignment surgery in adult spinal deformity patients

Adult spinal deformity (ASD) is an increasingly prevalent and costly problem [1], often requiring a lifetime of medical treatment, including surgical intervention. Surgical correction of ASD seeks to restore sagittal balance, which refers to the ability to maintain a mechanically effective center of pressure (gravity line) [2] via postural control of the spine and lower extremities while upright, including both quiet standing and more dynamic motions of daily life. Prior pivotal work showed an association between patient-reported health status and sagittal imbalance based on a measure of the sagittal vertical axis (SVA) from standing radiography [3]. These associations are drawn from questionnaires for patient-reported outcomes (PROs) for pain, disability, and quality of life, plus quantitative measurements of spinal alignment in standing radiography. However, questionnaires are subjective, and assessing spinal alignment from static standing radiography as a proxy for sagittal balance does not consider the effect of postural control and lower extremity mechanics that contribute to sagittal balance during dynamic motions of daily living.

Static SVA does not completely capture a patient’s ability to actively stabilize their spine when performing dynamic movements. For example, gait analyses of sagittal balance reveal compensatory mechanisms recruited from the lower extremities [4, 5]. In addition, the effect of muscular fatigue from short walks can significantly worsen standing SVA [6], which indicates that in-clinic radiographs that inform surgical realignment may not accurately reflect the patient’s condition outside the clinic. Consequently, assessing postural stability during dynamic tasks may pave the way for a better understanding of postural compensatory mechanisms, the hip–spine complex, and risk of post-surgical mechanical complications, including proximal junctional kyphosis (PJK) and failure [5,7]. The incidence of PJK is reported to occur following 20–39% [8] of cases and often requires a subsequent revision surgery, creating an added burden on the patient, clinicians, and hospitals.

Routine clinical assessment of dynamic sagittal balance is impractical given the setup time, space requirements, and high technology cost. Simple dynamic functional tasks, like a sit-to-stand (STS) maneuver, do not require the same amount of space as gait analysis and are diagnostically appropriate given that these tasks are commonly arduous for ASD patients due to the necessary whole body balance and postural control. Using 3D depth mapping sensor technology, we have developed an efficient means for collecting in-clinic motion analysis data. Integrated mathematical methods for noise filtering and kinematic constraints provide high-fidelity joint-position tracking for biomechanical modeling, enabling reliable dynamic quantification of spine, hip, and knee biomechanics. We used this novel technology to longitudinally assess DSB from a cohort of 15 ASD patients undergoing multi-level spinal fusion, and to compare pre- and postoperative DSB data to 10 healthy controls. The purpose of this study was to define a set of objective biomechanical metrics that are representative of post-surgical outcomes and may, in the future, be used to predict risk of post-surgical PJK.

Our goal was to use a validated in-clinic functional assessment tool to identify biomechanical metrics that associate with post-surgical outcomes. Current measures for surgical planning include static radiographic alignment and patient-reported questionnaires for pain, disability, and health status. Our data highlight new biomechanical measurements of DSB and lower limb compensatory strategies that change with spinal realignment surgery in ASD patients: (1) reduced peak SVA and biomechanical loads and muscular forces on the lower lumbar spine, and (2) altered relative biomechanical contributions from the hip and knee indicating reduced compensatory behavior from the hip. Changes in pre- to post-surgical DSB data moved toward the values from the healthy controls with most of the post-surgical metrics becoming equivalent to controls.

Dynamic biomechanical assessment of sagittal balance and lower limb compensation may fill gaps between static radiographic spinal alignment and patient-reported health status. Sagittal imbalance is thought to encompass how spinal malalignment can cause unwanted skewness in one’s center of pressure. However, when considering compensatory alignment of the pelvis and lower limbs, a malaligned spine may not shift center of pressure [14], but, instead, initiate mechanically ineffective compensatory balance strategies in order to maintain the position of the center of pressure between the feet [15]. Specific results from our study show that during a dynamic task requiring balance and postural stability, ASD patients deviate from healthy controls with motions relating to higher torque on and muscle forces needed from the lower lumbar spine, more work from the hip and in turn less active work from the knees in order to maintain a stable position of the center of pressure. These metrics are shown to improve with spinal realignment surgery.

Prior studies demonstrate improvement in functional performance after spinal realignment surgery. A heightened effort to understand gait motion behavior in ASD patients has emerged [5], and one study has shown results supporting improved kinematic metrics in ASD patients at 1 year and 2 years following surgery [16]. Main findings include increased thoracopelvic, hip, and knee range of motion during the gait cycle. While these findings demonstrate improvement in gait after surgery, the ability to integrate these findings into clinical practice to support more effective outcomes remains a challenge. Although gait is the most studied dynamic functional and is often difficult for ASD patients, gait may be less practical for routine clinical assessment using motion analysis. Other dynamic functional tasks requiring active postural stability may be more challenging than gait for pre-surgical ASD patients due to heightened loading on the spine and lower limbs.

Our study utilized an STS maneuver for its practical implementation within a clinic room and requiring adequate postural stability to effectively rise from a stable seated position to a stable standing position. STS is often employed as a clinically meaningful functional task for spine patients. Other studies that have assessed STS on spine function primarily focus on its relevance as functional test for low back pain patients [17‐22] and often include a different form of the test (e.g., 5-time STS [22, 23]). In studies that have assessed sagittal plane mechanics of STS in low back pain patients, most only distinguish degrees of motion at the trunk, hip, and knee and are limited with a cross-sectional study design. The few studies that have measured joint kinetics during STS in low back pain patients show linkages between compensatory lower limb biomechanics and relatively higher loads on the spine [17, 24]. Our present study uniquely examines STS in ASD patients whose lower limb compensatory biomechanics are likely a compensatory response to poor postural control from spinal malalignment; however, the lower limb compensatory biomechanics could be exacerbating poor outcomes by increasing loads on the spine.

We found that biomechanical standing strategies from the STS task improve following spinal realignment surgery. Our results support that pre-surgical ASD patients employ a ‘quasi-static leaning’ strategy to transition from a stable seated to stable standing position [25]. Using this leaning strategy, the subject leans as far forward as possible to move their torso directly above their feet, before engaging their knees and hips to rise. This essentially recruits postural compensatory mechanics to maintain a passively stable position of one’s center of pressure. Our results showing post-surgical improvements in dynamic peak SVA, lumbar torque, extensor muscle force, and hip torque and energy suggest that subjects become less reliant on the quasi-static lean strategy (Fig. 4). Furthermore, subjects that are exercising more strength from the knees following surgery may support potential benefits from leg strengthening exercises for improved outcomes.

Our results allude to potential predictors for PJK risk. The rates of revision surgery due to PJK/PJF following multi-level fusion in ASD patients are high, and a lack of understanding regarding risks persists. Some known risk factors for PJK that relate to possibly excessive loading on the spine include large changes in radiographic SVA [26], poor paraspinal muscle quality [27], and neuromuscular comorbidities [7]. ASD patients with ambitious realignments and overcorrection are also at a higher risk of post-surgical PJK/PJF [28]. One patient subject from our present study experienced a PJK complication requiring revision surgery, and their longitudinal DSB behaviors indicate that a lack of improvement in peak SVA and changes in relative contributions of torque and energy in the hip and knee may be risk factors for PJK (Fig. 3). A better understanding of how patient movement may place excessive loads on the spine would help develop pre- and post-surgical rehabilitation strategies to improve postoperative DSB and prevent overloading the vulnerable region between unfused and (newly) fused spinal segments.

Limitations for this current work include, firstly, our relatively small sample size. The purpose of this study was to establish how DSB changes following realignment surgery to create a number of clinically meaningful metrics for deciphering outcomes from DSB assessments. Future work will collect a larger sample to begin exploring associations between patient-reported outcome scores and radiographic alignment data with DSB metrics.

Other limitations regard possible constraints from the biomechanical modeling. One is how we normalized the DSB biomechanical metrics for mass and stature. It is important that we normalized data so that DSB metrics were comparable between subjects regardless of body size, but it creates values and units that can be difficult to understand. Another limitation from the modeling being the spine is represented as primarily one segment. Future work aims to improve the model to distinguish thoracic motion from lumbar. Lastly, age differences between our patient subjects and healthy controls could confound our analyses. However, one study found loads on the lower lumbar spine during STS to be similar between relatively younger and older individuals provided both were asymptomatic [29].

In conclusion, our study utilizes a novel in-clinic approach for measuring dynamic biomechanical metrics of sagittal balance through the spine, hip, and knee. This technology enabled us to track changes in sagittal balance biomechanics following spinal realignment surgery in ASD patients and introduce new objective outcomes. Longitudinal changes in dynamic sagittal balance following successful multi-level spinal realignment indicate changes in spinal biomechanics reducing forces acting on the lower lumbar spine along with changes in lower limb dynamics matching that of healthy controls. Furthermore, potential risk factors for PJK may include inadequate improvement in peak SVA and relative ratios of hip and knee torque and energy. Our results suggest that surgical planning should consider how pre- and post-surgical rehabilitation could improve sagittal balance and improve outcomes.