Journal of Biomechanics 32 (1999) 695}703 Correlation between pre-operative periprosthetic bone density and post-operative bone loss in THA can be explained by strain-adaptive remodelling J. Kerner, R. Huiskes*, G.H. van Lenthe, H. Weinans, B. van Rietbergen, C.A. Engh, A.A. Amis Biomechanics Section, Mechanical Engineering Department, Imperial College, London, UK Orthopaedic Research Laboratory, Institute of Orthopaedics, University of Nijmegen, Netherlands Anderson Orthopaedic Research Institute, Arlington, VA, USA Received 27 January 1999 Abstract Periprosthetic adaptive bone remodelling after total hip arthroplasty can be simulated in computer models, combining bone remodelling theory with "nite element analysis. Patient speci"c three-dimensional "nite element models of retrieved bone specimens from an earlier bone densitometry (DEXA) study were constructed and bone remodelling simulations performed. Results of the simulations were analysed both qualitatively and quantitatively. Patterns of predicted bone loss corresponded very well with the DEXA measurements on the retrievals. The amount of predicted bone loss, measured quantitatively by simulating DEXA on "nite element models, was found to be inversely correlated with the initial bone mineral content. It was concluded that the same clinically observed correlation can therefore be explained by mechanically induced remodelling. This "nding extends the applicability of numerical pre-clinical testing to the analysis of interaction between implant design and initial state of the bone.  1999 Elsevier Science Ltd. All rights reserved. Keywords: Bone remodelling (remodeling); Bone resorption; Computer simulation; Bone density; Adaptation 1. Introduction Periprosthetic bone loss is recognized as a common complication of total hip replacement (Jacobs et al., 1993). It is a problem requiring attention in the follow up of clinical cases and in the design process of new hip prostheses. In clinical practice it was observed predomi- nantly around cementless hip stems (Engh et al., 1987; Kiratli et al., 1996). Bone loss a!ects the strength of the femur and increases the risk of bone fracture. Femoral components without proper proximal support experi- ence higher loading and the risk of fatigue fracture of the stem is increased (Engh et al., 1990). Apart from a!ecting the longevity of the primary reconstruction, reduced * Corresponding author. Orthopaedic Research Laboratory, Insti- tute of Orthopaedics, University of Nijmegen, P.O. Box 9101, 6500 HB Nijmegen, Netherlands. Tel.: #31-24-3614476; fax: #31-24-3540555. E-mail address: r.huiskes@orthp.azn.nl (R. Huiskes) bone stock presents serious problems for revision sur- gery. For all the above reasons periprosthetic bone loss has been studied extensively. Earlier radiographic studies were soon followed by more sensitive and more precise dual energy X-ray absorptiometry (DEXA) investiga- tions (Engh et al., 1987,1990; Kiratli et al., 1991,1992; McCarthy et al., 1991; Kilgus et al., 1993). Adaptive bone remodelling due to stress shielding was identi"ed as one of the important causes of the resorp- tion. This was con"rmed by patient studies and animal experiments, which showed that the observed e!ects of implant design parameters can indeed be explained as consequences of stress shielding (Engh et al., 1987; Engh and Bobyn, 1988; Maistrelli et al., 1991). As a process controlled by mechanical stimuli, adaptive bone re- modelling can be simulated in computer models which combine bone remodelling theories with "nite element analysis (Cowin, 1993; Carter et al., 1987; Hart et al., 1984; Huiskes et al., 1987,1992). The practical value of these models was established in a number of studies 0021-9290/99/$ - see front matter  1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 1 - 9 2 9 0 ( 9 9 ) 0 0 0 4 1 - X