Bart Ten Brinke

Since the 1970s, radiostereometric analysis (RSA) has been used to measure (early) migration of orthopaedic implants. RSA is a highly accurate roentgen technique which enables us to determine implant micro motion of tenths of millimeters. Chapter 1 of this thesis gives a brief overview of the history of RSA and a short explanation of the technique behind implant migration analysis. RSA has extensively been used to measure migration of total hip and total knee arthroplasties. Precision and accuracy of the technique have been clearly described as well as the relation between early implant migration, measured using RSA in the first two years postoperatively, and long-term implant survival. However, it is unclear whether RSA, when applied in the upper extremity, is as accurate and precise as in the lower extremity. Further, it is uncertain whether the predictive value of early migration for the risk of future loosening also applies to total joint arthroplasty in the upper extremity. Therefore, the aim of this thesis is to investigate the role of RSA in the evaluation of orthopaedic implants in the upper extremity.

PART I – Accuracy and precision of RSA in the upper extremity

In Chapter 2 we systematically reviewed the literature in order to examine the accuracy and precision of RSA in shoulder, elbow, wrist and hand arthroplasty. Fourteen studies concerning the shoulder, four studies on the elbow and five studies on trapeziometacarpal (TMC) joint arthroplasty were included. Precision values for RSA in the shoulder varied between 0.06 – 0.88mm for translations and between 0.05 – 10.7° for rotations. In the elbow joint, precision varied between 0.05 – 0.34mm and 0.16 – 0.76° and in the TMC joint between 0.16–1.83mm and 11 – 124°. Accuracy data were not reported in included studies. Adherence to existing RSA guidelines was poor in nearly all studies, leading to heterogeneously reported data. This systematic review demonstrated that RSA is a highly precise technique to assess migration of orthopaedic implants in the upper extremity. However, precision of rotation measurement is less precise around the axis of symmetry in symmetrical implants as humeral head resurfacing implants (HHRI). Moreover, precision of RSA is poor when used in small joints as the TMC joint due to the limited size of the surrounding bone and poor three-dimensional spread markers.

PART II – RSA in the trapeziometacarpal joint

In Chapter 3 we performed an experimental study to assess the feasibility of RSA of a surface replacement (SR) TMC joint prosthesis. The TMC joint of five human hand specimens was replaced by the prosthesis. Of each hand, ten pairs of RSA radiographs were made. Implant migration relative to the surrounding bone during the time inbetween the different radiographs was assumed to be zero. As a result, measured migration between these RSA scenes was considered as the measurement error of the technique. ‘Accuracy of zero motion’ varied between 0.11 and 0.26mm for translations and between 1.47 and 3.72° for rotations. This study showed the in vitro feasibility of RSA in the TMC joint with high precision for translations. Lower precision for rotations was attributed to the close position of the markers relative to each other and to the prosthesis.

The aim of the clinical pilot study described in Chapter 4 was to evaluate in vivo migration patterns of the SR TMC joint prosthesis. Secondary aims were to assess patient-related outcomes (DASH, Nelson) and implant survival after five years. Mean translations of the prosthesis varied between 0.00 and 0.50 mm five years postoperatively. Mean rotation values varied between 0.3 and 2.3°, but high standard deviations indicated poor precision. Two patients underwent an early revision because of pain, without signs of loosening. Clinical outcomes in the other patients improved significantly.

Long-term follow-up of this cohort is described in Chapter 5. The purpose of this study was to determine long-term survival and clinical outcomes ten years after implantation of the SR TMC joint prosthesis and to evaluate implant migration during follow-up. Although no additional revisions took place, two patients had a loose implant at final follow-up. DASH and Nelson scores deteriorated clearly in these two patients. Long-term clinical outcomes remained excellent in patients with a stable implant in terms of pain, range of motion and strength. Precision of RSA of the TMC joint prosthesis was determined in a clinical setting using double examinations. High precision values were found for translations, ranging from 0.10 and 0.12 mm. Rotations could not be calculated because of a lack of stable and well three-dimensional spread markers. In this chapter we extensively discuss several technical challenges of RSA in small joints like the TMC joint. Possible solutions as marker configuration (MC) models and reversed migration calculation are discussed.

PART III – RSA in the elbow joint

Concerning the elbow joint, only a few short-term RSA studies have been published and knowledge about the influence of early migration on long-term outcomes is lacking. The aim of chapter 6 was to evaluate long-term survival of sixteen Instrumented Bone Preserving (IBP) total elbow prostheses. This implant was designed to preserve intercondylar bone and to improve initial fixation and stability. To investigate the relation between early migration and long-term outcomes, previously published short-term migration values were compared between revised and non-revised implants. Further, long-term migration and clinical outcomes (Elbow Function Assessment (EFA), Broberg and Morrey Elbow Functional Rating Index (EFRI), Oxford Elbow Score (OES) and Visual Analog Scale (VAS) for pain) of the IBP prostheses were reported with a minimal follow-up time of ten years. Ten-year survival was 75%, decreasing to 63% after fourteen years of follow-up. Known short-term migration values did not differ between revised and non-revised implants, however the number of patients in this study was not sufficient to draw conclusions about the potential relation between early migration and long-term outcomes. Long-term migration values could be calculated in four patients and varied widely. Clinical outcomes worsened substantially over time in almost all patients. Several patients with loose prostheses at final follow-up refrained from revision surgery, indicating that implant revision might not be the right endpoint in studies investigating the relation between early migration and long-term loosening in total elbow arthroplasty.

PART IV – RSA in the shoulder joint

In Chapter 7 we evaluated fixation and migration patterns and clinical outcomes of the Simpliciti stemless shoulder system. Stemless humeral implants have been developed to overcome stem-related complications like fractures and complex revisions. Based on theoretical advantages and promising short-term results, stemless implants gained popularity and the number of stemless implants will soon surpass stemmed arthroplasty. In our study, patient reported outcome measures improved significantly in all patients at two-year follow-up, in accordance with existing literature. Fixation of stemless humeral components completely relies on cancellous metaphyseal bone. This may potentially lead to suboptimal fixation affecting long-term implant survival. We found that twenty out of 24 implants stabilized within 12 months postoperatively, but that four implants showed continuous migration between 12 and 24 months and did not stabilize during follow-up. Although a relation between increased early migration and long-term failure has not been demonstrated so far, the level of migration found in this study raises concern and critical assessment of long-term outcomes in stemless shoulder arthroplasty is of vital importance.

PART V – General discussion

In chapter 8 we discuss the main findings of this thesis. These days, assessment of early migration of orthopaedic implants in upper limb arthroplasty is sporadically performed. In recent years however, the ability of highly precise migration measurement of upper extremity implants has been demonstrated, particularly for TSA. Although the number of RSA studies is not sufficient to investigate the relation between early migration and long-term outcomes, several clues for the predictive value of early migration in TSA have been described. Based on current literature and data from this thesis, we suggest routine migration assessment using RSA of all new shoulder implants prior to market release. Further investigation of the predictive value of RSA in TSA requires an increase in the number of studies evaluating short-term migration and long-term follow-up. Further standardization of reporting migration outcomes is essential to facilitate comparison and pooling of data. In this chapter we discuss several factors that should be taken into account in the further evaluation of migration patterns in the upper extremity, including biomechanical differences between upper and lower limb joints, the influence of rheumatoid arthritis (RA) on migration and the importance of determining the right endpoint for implant failure. Notwithstanding high accuracy and precision in TEA, TWA and TMC joint arthroplasty, concerns about the feasibility of RSA have to do with the small size of the surrounding bone. Before further RSA studies will be performed using model-based RSA, we recommend to examine the feasibility of newly developed techniques as CT-based RSA in small joints of the upper extremity.