Surgical interventions to correct adult spinal deformities (ASD) consist of the rigid fixation of multiple intervertebral joints and involve a high risk of mechanical complications. One of the most common affects the proximal junction between vertebral segments (both those treated and those not treated by the operation), which worsens the curvature that the spine should have in relation to the pelvis to maintain a balanced body posture. Proximal junction failure generates more pain and disability and requires a second operation in more than 20% of cases. 

Hence, a research group, led by UPF, has investigated how to improve the accuracy of clinical indices to measure the risk of the occurrence of proximal junction failureafter spine surgery. The research has enjoyed the collaboration of Hospital del Mar, the Vall d’Hebron Research Institute (VHIR), the Galeazzi Orthopaedic Institute in Milan (IRCCS Galeazzi), and the Shulthess Klinik in Zurich (Switzerland).

The research has managed to raise the accuracy level of these indices to above 95%, essentially based on the incorporation of new biomechanical descriptors, referring to the mechanical conditions of the patients’ musculoskeletal apparatus. Until now, these indices only took into account geometric descriptors of vertebral alignment.

The findings of this study have been published recently in the journal Spine, in an article entitled Proximal Junction Failure in Spine Surgery: Integrating Geometrical and Biomechanical Global Descriptors Improves GAP Score-Based Assessment. At UPF, the research has been conducted by the SIMBIOsys (Simulation, Imaging and Modelling for Biomedical Systems) research group of the BCN MedTech Research Unit of the Department of Information and Communication Technologies (DTIC). This article has been written by Morteza Rasouligandomani (main author) and Alejandro del Arco, spine surgeon at Hospital del Mar (second author). Also co-authoring the article are Jérôme Noailly and Miguel Ángel González Ballester, two of the co-directors of the SIMBIOsys group; Ferran Pellisé, head of the Spinal Unit at Vall d'Hebron University Hospital and of the Spine Research Unit of the VHIR; and Fabio Galbusera, of the Shulthess Klinik in Zurich (previously linked to the IRCCS, Galeazzi Orthopaedic Institute in Milan).

Between 36 and 59% of people who undergo surgery to correct adult spine deformities suffer some mechanical complication, especially during the first year after the intervention.

The results of this research may contribute to improving surgical techniques to operate on adult spine deformities, various abnormalities that can cause chronic pain or decreased mobility among patients. One of the major risks of these operations, especially of so-called vertebral instrumentation (one of the most frequent surgical techniques) are mechanical complications. Between 36 and 59% of people who undergo this intervention suffer them. Age is a risk factor that favours the occurrence of these complications.

Proximal junction failure is one of the most common mechanical complications following spine operations

Among these mechanical complications, the most common is precisely proximal junction failure (PJF). PJF affects between 1.4 and 35% of people operated on to correct deformities of the spine and, in addition, its occurrence is early onset (between the first three and six months after the operation). PJF can lead disruptions of the vertebral ligaments or joints. It can cause severe pain and mobility problems and, in extreme cases, total (paraplegia) or partial (paraparesis) loss of leg mobility.

Taking into account the severe effects that PJF entails for the health of patients and the complexity of repeat surgeries to correct it, the optimal solution is to prevent proximal junction failure, that is, to adopt measures so that people who undergo surgery for ASD do not develop it. In this regard, it is essential to improve the accuracy of clinical indices that predict the likelihood that a patient will suffer PJF after a surgical intervention. 

This was the goal of the research team led by UPF’s BCNMedTech Research Unit, which has identified new biomechanical descriptors to complement current indices to measure the risk of PJF onset. So far, one of the most advanced indices to do so is the so-called GAP (Global Alignment and Proportion) score, which essentially assesses this risk based on the alignment of the column and the sagittal balance. In a situation of sagittal balance, there is a harmonic relationship between the curvature of the spine and the pelvis. The problem is that the GAP score is not sufficiently accurate to differentiate the risk of the occurrence of PJF from other mechanical complications of the spine, and the efficiency of this index is still a bone of contention in the evaluation of some clinical studies. 

Biomechanical descriptors, on patients’ mechanical conditions, may improve the accuracy of risk assessment indices 

To correct these shortcomings, the research team proposes adding biomechanical descriptors, referring to patients’ biological and physiological characteristics, to the geometric descriptors (of sagittal alignment and balance), currently used by the GAP score. These biomechanical descriptors include body mass index (BMI) and bending moment (BM). A bending moment is essentially a force that causes something to bend. Thus, in this case, the maximum bending moment refers to the maximum force that the spine could withstand due to effects of body mass or poorly compensated by the muscles.

From the study, “we have determined that the combination of geometric descriptors of the alignment and of the sagittal balance of the spine with biomechanical descriptors is the best way to predict the risk of the occurrence of proximal junction failure”, explains Morteza Rasouligandomani, lead author of the study (UPF). 

Dr. Alejandro del Arco, spine surgeon at Hospital del Mar and second co-author of the study, adds: "The importance of this work is to understand the spine, the human body, which goes beyond a static, two-dimensional structure that must be aligned and it is a dynamic structure must be in equilibrium. We have gone beyond angular and geometric measurements with the calculation of forces, tensions and deformities, which are truly responsible for the stability or failure of a structure. At the same time, we have opened up the possibility of creating a simulation tool to improve surgical planning”.

The results of the research will enable improving the GAP score, “the most efficient surgical planning tool in adult scoliosis surgery and one of the most used worldwide”, assures Dr. Ferran Pellisé, head of the Spinal Unit at Vall d'Hebron University Hospital and of the Spine Research Unit at VHIR. “The results of this new publication once again validate and further reinforce the proposal for surgical planning in adult vertebral deformity made and published in October 2017, by the European Spine Study Group, based at the Vall d’Hebron University Hospital”, Pellisé adds.  

In this latest research, the new descriptors, which will help improve the GAP score, have been tested on a sample of 112 patients aged between 50 and 75 years, 57 of whom suffered from proximal junction failure and 55 were healthy patients (control group). Three-dimensional images of their spine were taken using biplanar EOS radiographs (with two X-ray sources). 
 
The researchers have shown that biomechanical descriptors such as the maximum bending moment (BM) can increase the mean accuracy of clinical indices to predict proximal junction failure by more than 90% (with a minimum accuracy not below 83%). The study concludes that an index that uses only biomechanical descriptors, like BM, is more likely (with a mean prediction of 92.16% and a minimum prediction of 83%) to predict proximal junction failure than the GAP score, based solely on geometric descriptors (mean prediction of 87.39% and minimum prediction of 77%). However, the study finds that the best solution is to combine geometric and biomechanical descriptors in a single clinical index, since, in this case, the accuracy of the calculation of the risk of PJF achieves a mean of 95.02%, with a minimum prediction of 87%.