A Research Study to Understand and Predict the Failure of Shoulder Fracture Fixations Using Computer Simulations (PHILOS-SIM)

A Research Study to Understand and Predict the Failure of Shoulder Fracture Fixations Using Computer Simulations

Multicentric Prospective Interventional Study on Predicting the Mechanical Failure of PHILOS Plate Fixations in Older Adult Patients With Proximal Humerus Fractures Using Computer Simulations

Fractures of the shoulder, the so-called proximal humeral fractures, can be treated with locking plates that have shown good results in clinical practice, but complications can occur. In order to further improve the treatment of proximal humerus fractures and decrease the rate of fixation failures, further research is necessary. As a first step, the reasons for potential implant failures need to be understood. This study has been initiated by scientists at the AO Research Institute Davos (ARI), Switzerland, which is the research center of the AO Foundation (www.aofoundation.org), an international non-profit organization led by surgeons specialized in fractures such as these. Researchers at the ARI have been developing a computational simulation tool to predict fixation failure and demonstrated its efficiency in laboratory conditions. This clinical study has been organised to validate this tool using patient data, by comparing the risk of mechanical failure predicted by the computer simulated model with the clinically observed fixation failure. In future, this computer simulation tool is expected to help surgeons to select the best fixation for a given patient and to develop improved implants.

Older adult patients with complex, displaced and/or unstable proximal humerus fractures treated with the PHILOS plate (DePuy Synthes Inc.) will be recruited. Detailed patient background information will be assessed. Surgical details will be acquired. Pre-, intra- and post-OP imaging data will be collected in frame of the standard treatment protocol. A post-OP CT scan, which is performed only occasionally as standard of care, will be acquired for all patients participating in this study. The de-identified medical images will be transferred via secure file transfer protocol (FTP) to the sponsor. Patient activity will be recorded in the first 6 post-OP weeks by means of two wearable activity trackers attached on the skin of the treated upper arm and the chest. The tracker containing the encoded data will be shipped back to the sponsor. According to the standard post-OP protocols, follow-up (FU) examinations will be performed at 6 weeks, 3 months and 6 months after treatment. Additionally, another FU visit will be required at 3 weeks after surgery, where the activity trackers will be exchanged. FU radiographs during the standard of care visits will be compared with post-OP radiographs to determine if any mechanical failure has occurred. Mechanical failure or no failure status of the fixation will be determined at 6 months following fracture treatment. The patient-specific fixation failure risk will be prospectively predicted by means of computer simulations created from the pre-OP and post-OP CT images and surgical details. The prediction accuracy of the simulations will be evaluated by comparing the predicted failure risk with the clinically observed failure / no failure outcome. As a second approach, patient-specific shoulder activity (estimated based on patient pre-injury activity or measured directly post-operatively) will be incorporated into the simulations sequentially and its effect on the prediction accuracy will be evaluated. The interventions in this study, i.e. the examinations that are beyond the current standard of care, will be i) the extra acquisition of the post-OP CT image in the cases that the treating surgeon is not requesting them, ii) the questionnaires on comorbidities and pre-injury activity, and grip strength measurements to estimate activity of the patient and iii) the post-OP activity measurement by means of wearable sensors.

humerus, fracture, locking plate, PHILOS, fixation failure risk, computer simulation, patient activity tracking

The following visits will be performed: direct post-OP, 3 weeks, 6 weeks, 3 months and 6 months after surgery. Surgical details will be assessed from the surgical notes. Medical images will be collected according to the standard of care. A post-OP CT scan will also be acquired when the treating surgeon requests it as per standard of care, or additionally, as a study specific imaging procedure for the rest of the participants. Patient activity will be estimated based on various factors, including questionnaires on demographics, activity level, comorbidity score and contralateral grip strength. The actual post-OP shoulder activity will be measured by motion tracking by means of sensors attached to upper arm of the treated side and on the chest for 6 weeks and by grip strength assessed at the treated arm at week 6. Fixation failure status will be determined 6 months post-OP. Imaging and details regarding the fixation failure event will be sent to ARI.

Fixation failure risk estimated from the structural capacity (strength) of the fixation will be predicted by the patient-specific computer simulation based on the collected: pre-OP CT images post-OP CT images surgical details.

Clinical occurrence of radiographically confirmed mechanical fixation failure within 6 months after PHILOS plate treatment (failure / no failure status). Mechanical failures are defined as: loss of reduction (≥ 15° increase of varus malposition between the immediate post-OP position and FU radiograph, and/or relative change (5 mm) of greater or lesser tuberosity compared to the immediate post-OP position) cut-out failure (≥ 5 mm distal displacement of the humeral head fragment between immediate post-OP and FU) head impaction (≥ 5 mm difference in outer plate edge and tangent of humeral head between immediate post-OP and FU), and/or secondary screw perforation (Perforation of one or more screws through the humeral head on the FU radiograph (and lack of screw perforation on the intra- or post-OP radiograph based on a direct post-OP AP and axial radiographs))

Fixation failure risk predicted by the simulation when activity information is included in the model, estimated based on categorization of the participants in low, average and high pre-injury activity level from the following questionnaires and assessments: Demographics Age (in years) Gender Height (in meters) Weight (in kilograms) Residential status Dominant arm Side and type of injury Grip strength assessed at the uninjured side Shoulder activity scale (Marx) Parker Mobility Scale Barthel index Charlson Comorbidity Score

Fixation failure risk predicted by the simulation when actual post-OP shoulder activity information is included in the model

Fixation failure risk predicted by the simulation when actual post-OP shoulder activity information is included in the model, based on the following data: Directly measured post-OP shoulder activity (range of motion, frequency, velocity) (up to 6 weeks) Grip strength assessed at the treated side (6 weeks)

Inclusion Criteria: Age≥ 50 years Radiologically confirmed closed fracture of the proximal humerus and treatment within 10 days after injury Any displaced or unstable 3- or 4-part fracture of the proximal humerus except isolated displaced fractures of the greater or lesser tuberosity Completed primary fracture treatment with a PHILOS plate with or without cement augmentation of the screws Ability to understand the content of the patient information / informed consent form Willingness and ability to participate in the clinical investigation including imaging, activity tracking and other FU procedures Signed informed consent Exclusion Criteria: Previous proximal humerus fracture on the ipsilateral side Splitting fracture of the humeral head or humeral head impression fracture Fibula grafting, bone block or any other non-cement augmentation of the PHILOS plate fixation Any serious fracture fixation issue (serious malreduction, too long screw or screw perforating through the humeral head, broken screw or implant) that are recognized on the immediate post-OP x-rays, independently of being re-operated or not Associated nerve or vessel injury Any severe systemic disease: class 4 - 5 of the American Society of Anesthesiologists (ASA) physical status classification Recent history of substance abuse (i.e., recreational drugs, alcohol) that would preclude reliable assessment Prisoner Participation in any other medical device or medicinal product study within the previous month that could influence the results of the present study Pregnancy Pacemaker

Varga P, Inzana JA, Gueorguiev B, Sudkamp NP, Windolf M. Validated computational framework for efficient systematic evaluation of osteoporotic fracture fixation in the proximal humerus. Med Eng Phys. 2018 Jul;57:29-39. doi: 10.1016/j.medengphy.2018.04.011. Epub 2018 May 1.

Varga P, Grunwald L, Inzana JA, Windolf M. Fatigue failure of plated osteoporotic proximal humerus fractures is predicted by the strain around the proximal screws. J Mech Behav Biomed Mater. 2017 Nov;75:68-74. doi: 10.1016/j.jmbbm.2017.07.004. Epub 2017 Jul 4.