Load response of an osseointegrated implant used in the treatment of unilateral transfemoral amputation: An early implant loosening case study
Introduction
Osseointegrated implants overcome a number of complications associated with traditional socket-fitted prostheses for the treatment of lower limb amputation, including residual limb pain and skin irritation at residual stump-socket interface, which can occur in up to 62% of recipients (Hagberg and Brånemark, 2001; Lyon et al., 2000; MacLachlan and Pamela, 2001; Mak et al., 2001). Clinical studies have shown that, compared to socket-fitted prostheses, transfemoral osseointegration reduces tissue damage and residual pain (Al Muderis et al., 2016b; Branemark et al., 2001; Sullivan et al., 2003), while improving patient activity levels, and facilitating larger range of hip joint movement (Brånemark et al., 2014; Hagberg et al., 2008, Hagberg et al., 2005). In two-year follow-up studies by Brånemark et al., 2014 and Hagberg et al., 2008, implant survival rates were 92% in 51 patients and 94% in 18 patients, respectively. Nebergall et al. (2012) used radiostereometric analysis in 51 patients to show no significant implant migration or rotation at ten-year follow-up. In contrast, a systematic review of outcomes of osseointegrated prostheses for transfemoral amputation reported a maximum revision rate of 67%, with osseointegrated implants loosening in up to 6% of patients, and periprosthetic fracture occurring in up to 9% of cases (van Eck and McGough, 2015). In a study of 51 amputees with osseointegrated implants, mechanical failure of the prosthesis was observed in up to 29% of cases (Brånemark et al., 2019), including damage to the abutment and/or abutment screw that forms the connection between the intramedullary stem and the prosthetic lower limb. Failures have also been reported to the intermedullary stem, with four incidents described across 77 patients (Al Muderis et al., 2016b; Zimel et al., 2016).
Transfemoral amputees with osseointegrated implants typically undergo post-operative rehabilitation for up to six months to encourage integration of bone into the implant prior to ambulation (Hagberg and Branemark, 2009). In the first phase of rehabilitation, a controlled loading regime is applied to incrementally stimulate bone mineralization and strength, while minimizing risk of implant overloading. Core and limb strengthening exercises are commonly performed in conjunction with axial weight bearing using a short training prosthesis, and static loads of increasing magnitude gradually applied until the amputee is able to support half their bodyweight without significant pain (Al Muderis et al., 2017b; Hagberg et al., 2005; Hagberg and Branemark, 2009). Some protocols also adopt eccentric prosthesis loading exercises to encourage bone growth in directions other than axial (Hagberg and Branemark, 2009). In the second phase of rehabilitation, prosthetic limbs are fitted, and parallel bars and crutches employed to limit load transmission through the entire prosthesis. The patient may repeat the static weight bearing tasks, and progressively introduce different phases of the gait cycle until they are able to ambulate unaided (Leijendekkers et al., 2017). At present, the loading patterns at the bone-implant interface during rehabilitation tasks such as static weight bearing, and normal walking are poorly understood. As a consequence, over-emphasis on functional rehabilitation tasks that under-load the bone in the residual limb may not adequately prepare the patient for the dynamic loads during walking, and may ultimately result early implant loosening or failure (Huiskes et al., 1987; Turner et al., 1986).
Finite element modeling has been used in several studies to investigate stress-strain distributions and long-term remodelling of bone surrounding osseointegrated transfemoral implants (Helgason et al., 2009; Lee et al., 2008; Stenlund et al., 2017; Tomaszewski et al., 2010). While these previous studies have been used to describe risk of implant failure, implant loading was achieved using a resultant external force and moment applied directly to the prosthesis. However, non-knee-spanning muscles contribute to the linear and angular acceleration of the knee as well as knee joint-contact loading via dynamic coupling (Zajac et al., 2002); replacing the forces of individual muscles by an equivalent knee load may ultimately underestimate prosthesis force response (Blemker et al., 2007; Herzog et al., 2003). The study by Lee et al., 2008 showed distinctive differences in bone-implant interface stresses between weight-bearing and walking; however, the femur did not incorporate subject-specific bone geometry or material properties, which would have significantly altered the magnitude and distribution of the stress predictions.
The objective of this study was twofold. Firstly, to develop a patient-specific multi-body musculoskeletal model of the residual limb of an amputee with an osseointegrated prosthesis in the period before the implant loosened due to a periprosthetic fracture, and secondly, to use this model to evaluate loading patterns at the bone-implant interface during static weight-bearing with both a short training prosthesis and the entire prosthetic limb, as well as walking with the prosthetic limb. The findings of this case study will help in the development of personalised rehabilitation regimens to improve bone-implant integration and minimize early implant loosening.
Section snippets
Subject recruitment and imaging
One male subject (age: 59 years, weight: 83 kg, height: 1.90 m) with an above-knee amputation due to an automobile accident was recruited 12 months following primary surgery to implant a transfemoral osseointegrated prosthesis. The implant was an Osseointegrated Prosthetic Limb (OPL; Permedica s.p.a, Milan, Italy) with a cross-screw design (Al Muderis et al., 2018), connected to a Genium microprocessor controlled knee (Otto Bock 3B1), a pylon with in-built torsion adaptor (Otto Bock 2R21) and
Results
The largest muscle contribution to support the residual limb during standing was from adductor brevis, adductor longus, gluteus medius (anterior) and iliacus, with each generating between 62.1 and 248.9 N (Table 2). In comparison, there were greater forces generated by particular hip-spanning muscles during the stance phase of walking, with peak forces greater than 500 N for gluteus maximus, gluteus medius, rectus femoris, iliacus and psoas (Table 2). The resultant force at the abutment of the
Discussion
The objective of this case study was to employ patient-specific multi-body musculoskeletal modeling to evaluate the load response of an osseointegrated transfemoral implant prior to its failure due to loosening. This was achieved by quantifying forces, stresses and strains generated during static weight bearing with a training prosthesis and full lower limb prosthesis, as well as during over-ground walking with the lower limb prosthesis. While post-operative static-weight bearing tasks are
Contributors
Dale Robinson: Modeling and simulations, data analysis, manuscript writing.
Lauren Safai: Modeling and simulations.
Vahidreza Harandi: Data curation, including motion analysis experiments.
Mark Graf: Conceptualization of project.
Eduardo Cofré Lizama: Data curation, including motion analysis experiments.
Peter Lee: Funding of project.
Mary P Galea: Data curation, including motion analysis experiments.
Fary Khan: Funding of project.
Kwong Ming Tse: Modeling and simulations.
David Ackland:
Declaration of competing interest
None.
References (53)
- et al.
Sensitivity of model predictions of muscle function to changes in moment arms and muscle–tendon properties: a Monte-Carlo analysis
J. Biomech.
(2012) - et al.
Risk of failure during gait for direct skeletal attachment of a femoral prosthesis: a finite element study
Med. Eng. Phys.
(2009) - et al.
Adaptive bone-remodeling theory applied to prosthetic-design analysis
J. Biomech.
(1987) - et al.
FE stress analysis of the interface between the bone and an osseointegrated implant for amputees–implications to refine the rehabilitation program
Clin. Biomech.
(2008) - et al.
Effect of mechanical stimuli on skeletal regeneration around implants
Bone
(2007) - et al.
Skin disorders in amputees
J. Am. Acad. Dermatol.
(2000) - et al.
Strain energy in the femoral neck during exercise
J. Biomech.
(2014) - et al.
Dependence of yield strain of human trabecular bone on anatomic site
J. Biomech.
(2001) - et al.
Trabecular bone modulus-density relationships depend on anatomic site
J. Biomech.
(2003) - et al.
Trabecular bone microdamage and microstructural stresses under uniaxial compression
J. Biomech.
(2005)