ACCELERATING OSSEOINTEGRATION WITH BIOELECTRONIC IMPLANTS: A DOUBLE-BLIND PAIRED-LIMB OVINE STUDY

Clinical evidence for externally delivered electrical/mechanical stimulation in bone healing is mixed: proponents describe large benefits, while others report no difference. This may be because inconsistent targeting, anatomical differences, and physical phenomena (e.g., attenuation, reflection) mean cells experience highly variable stimulation doses. This study aimed to overcome this by developing implants that stimulate directly at the bone-implant interface. A bioelectronic implant (Fig. 1a; Ø8×16 mm) was developed to provide controlled electromechanical stimulation at the bone-implant interface. These implants had a rough titanium fixation surface (Ra: 75 µm) akin to contemporary cementless arthroplasty devices, but had electronics and sensors sealed inside. Eight were inserted into surgical defects (Fig. 1b; Ø8 mm) in the femoral condyles of four skeletally mature ewes (License P16F4AA0A). Half were activated intraoperatively by a mobile phone app such that they delivered electromechanical stimulus postoperatively (active) and half were left off (passive), with a double-blind paired-limb study design. Samples were retrieved after six weeks. Bone growth was analysed using fluorochrome histomorphometry (day 7: oxytetracycline; 21: alizarin red; 42: calcein blue). Microcomputed tomography (µCT) and scanning electron microscopy with energy dispersive X-ray spectroscopy (SEM/EDX) assessed mineralisation. Histology (H&E) was analysed by two blinded reviewers. Temperature was measured daily with onboard sensing. Bone growth at the implant interface was 58% faster with stimulation between days 7–21 (Fig. 1c; 5.4±1.2 vs. 3.5±0.9 µm/day; p=0.001) and 47% faster between days 21–42 (Fig. 1c; 3.3±0.6 vs. 2.3±0.7 µm/day; p=0.005). Remodeling within three trabeculae of the interface increased by 14% between days 7–42 (p = 0.017); no differences were observed beyond this, confirming effects were localised. µCT revealed mineralised bone within 100 µm of all implants, which was consistently high for the active samples (Fig. 1e&g vs. 1f&i); SEM/EDX confirmed mineralisation (Ca:P ratio ~1.7). Histology (Fig. 1h vs. 1j) found moderate-to-strong osseointegration around all active implants, but only half the passive controls. Temperature sensing indicated a beneficial short-sharp inflammatory response for the active implants relative to controls (Fig. 1d; p<0.05). Controlled, locally-targeted electromechanical stimulation from a bioelectronic implant leads to osseointegration that is >50% faster and twice as reliable as contemporary clinical technology. This first-of-its-kind technological advance could lead to a more digital future for orthopaedics whereby implant fixation is controlled and sensed postoperatively. Acknowledgements: UKRI (EP/W524323/1, EP/X52556X/1, EP/R042721/1), Wellcome Trust (223797/Z/21/Z), Imperial (DT-prime, FILM). For any figures or tables, please contact the authors directly.