Micro-CT imaging of surgical screw tightening in human trabecular bone E. Perilli 1, M. Ryan 1, R. Ab-Lazid 1, J.J. Costi 1, K.J. Reynolds 1 1 Medical Device Research Institute, School of Computer Science, Engineering & Mathematics, Flinders University, Adelaide, South Australia, Australia, Aims Bone fracture fixation in orthopaedic surgery occurs with the use of implants, such as bone screws. Screws are tightened, typically manually, with the level of tightening depending on surgeons manual feel. However, this subjective insertion torque control can lead to overtighening and screw stripping 1. Previously our group has shown that stripping torque can be predicted based on the torque measured at screw head contact 2. The question remains however, how much tightening should be applied?, particularly in trabecular bone. Bone quality, in terms of bone density and micro-architecture, plays a major role in determining the mechanical strength of the bone-screw construct 3. The mechanical strength of the construct is experimentally assessed by applying a tensile force until the screw strips from the bone, giving pullout strength (F Pullout). However, inconsistent findings exist in the literature about the relative importance of bone quality and applied insertion torque on F Pullout. The work presented here is part of a larger ongoing study in our group 4,5, which combines micro-ct imaging of surgical screw insertion into human femoral heads, experimental pullout testing, and stepwise imaging of screw tightening, performed using a computer-controlled micro-mechanical testing device that fits a micro-ct scanner. According to our study using an automated micro-mechanical screw-insertion device on human femoral heads (Fig.1), bone specimens having a more plate-like trabecular structure and higher bone volume fraction exhibit higher insertion torque measured at screw head contact, compared to those with a more rod-like structure and lower density 4 (Fig.2). It could be expected that these specimens also exhibit higher pullout strength 5. The 1 st part of this study investigates which, among measurements of bone micro-architecture and applied insertion torque (T Insert), has the strongest correlation with F Pullout 5. In the 2 rd part of this study, currently ongoing, we aim at performing stepwise micro-ct imaging of screw insertion. That is, visualizing the deformation of the trabecular bone at each rotation step with increased tightening torque. 1. Determining which, among measurements of bone microarchitecture and applied T Insert, has the strongest correlation with F Pullout 1. Method: 1.1 Human bone samples. Femoral heads (n=46) obtained from hip replacement surgery, part of a previous study, were used 4,5. Specimens were kept moistened with saline solution and stored at -20 C until testing. Prior to testing, specimens were thawed overnight in a refrigerator. 1.2 Computer-controlled screw tightening beyond head contact.
One screw insertion was performed for each femoral head, at the centre of the cut surface. An aluminium cancellous screw (length 45.5mm, outer thread diameter 7.0mm), replicating an orthopeadic screw (Cat.no.7110-7050, Smith and Nephew, London, UK), was used instead of the original titanium, to avoid artifacts during micro-ct scanning. The screw was first inserted through a pilot hole. Then the bone-screw assembly was transferred to a micro-mechanical test device, specially designed and constructed in Flinders University 4,5, for automated insertion of the screw beyond head contact, performed inside the micro-ct scanner (SkyScan model 1076, SkyScan, Kontich, Belgium) (Fig.1a). Our group has previously demonstrated that it is possible to predict the torque at which the screw will strip (T Max), based on measurement of T Insert at screw head contact, through a linear relationship 2. Once inside the micro-ct scanner, each screw was automatically tightened to a torque level ranging between 55% and 99% of the predicted maximum torque (T Max) 5. This range corresponds to a wide variation of screw insertion torque values reported in the literature 5. Upon completion of screw tightening, a micro-ct scan of the bone-screw construct was performed, without removing the bone- screw construct from the scanner (Fig.1a). 1.3 Micro-CT imaging and morphometric analysis Scanner settings: 17.3µm pixel size, X-ray source 100kVp, 80mA, rotation step 0.4, exposure time 885ms, 4 frames averaging, 180 rotation, 1.0mm-thick Al filter 5,6. For each specimen, 900 cross-section images were reconstructed (total height 15.6mm), each image was 35x35mm in size, centered over the inserted screw and saved in 8 bit grayscale format (software NRecon, V1.6.6.0, SkyScan) (Fig.1b). Extraction of the volume of interest (VOI) for bone morphometric analysis: from the stack of cross-sections, a VOI consisting of a trabecular bone annulus (height 12mm, outer diameter 14mm, inner diameter 7.5mm, annulus thickness 3.25mm) surrounding the screw was extracted (software CTAnalyser,V1.11,SkyScan) (Fig.2). This region has been found to contain the majority of the deformed trabeculae after screw insertion 7. Then images were binarised (uniform thresholding) and the following morphometric parameters were calculated over the VOI (CTAnalyser): bone volume fraction (BV/TV), bone surface density (BS/TV), structure model index (SMI), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp) and trabecular number (Tb.N) 5. (a) (b) Femoral head Figure 1: (a) Automated micro-mechanical screw insertion device mounted in the micro-ct scanner. A micro-ct scan was performed of the screw-bone construct (17.4µm/pixel), after completion of the screw insertion. (b) Micro-CT cross-section images of a femoral head, with screw. 1.4 Biomechanical testing: Screw pullout Screw pullout tests were then performed using a materials testing machine (Instron 8511 Ltd., High Wycombe, UK), at a pullout rate of 5 mm/min. The screw displacement and the maximum tensile force (F Pullout) exerted on the load cell were recorded 5.
1.5 Statistical analysis Pearson s correlation coefficients were used to evaluate the relationships F Pullout vs. microarchitectural parameters, and F Pullout vs. T Insert. Figure 2: VOI of trabecular bone (white colour) surrounding the screw (blue colour). Bone with high BV/TV and plate-like structure (left), bone with low BV/TV and rod-like structure (right) 5. 1. Results F Pullout showed the strongest correlation with T insert (R= 0.88, p<0.001), followed by SMI (R= - 0.81, p<0.001) and BV/TV (R= 0.73, p<0.001). F Pullout exhibited moderate correlations with the remaining micro-architectural parameters (BS/TV, Tb.Th, Tb.Sp and Tb.N), with R ranging from 0.31 to 0.46. Fig.3 shows scatter plots and best fit lines for F Pullout vs.smi, and F Pullout vs. T Insert, respectively. Figure 3: Scatter plots with best fit lines for F Pullout vs., T Insert and F pullout vs. SMI 5. 1. Conclusion Our findings confirm that F Pullout is significantly influenced both by the level of applied insertion torque, as well as the quality of the host bone 5. The F Pullout of cancellous screws exhibited the strongest correlations with T Insert, followed by the micro-architectural parameters SMI and BV/TV. Furthermore, the correlations between F Pullout and T Insert remained significant even after normalizing T Insert by BV/TV and SMI (data not shown) 5. This suggests that T Insert is the primary factor affecting F Pullout in trabecular bone and that increases in the applied T Insert will lead to increases of F Pullout. However, excessive tightening can lead to bone yielding and beyond that, to screw stripping. The level of tightening torque after which bone yielding occurs, and how F Pullout will be affected once this level is reached, should be examined in detail in future studies on cancellous bone screws in human bone 5.
2. Towards stepwise micro-ct imaging of screw insertion 2. Method Five excised human femoral heads were used (see point 1.1 above). Cancellous lag screws, commonly used for fixation of femoral neck fractures, were inserted to head contact (see point 1.2). The insertion torque measured at head contact was used to predict stripping torque (T max) 2. Screws were incrementally tightened within the micro-ct scanner, using the novel testing device (Fig.1) 4,5. For each femoral head, micro-ct scans (pixel size 17.4 µm, see point 1.3) were performed at 6 time points, with incremental tightening torque from head contact to T max (stripping torque), without removing the construct from the scanner. 2D and 3D images were reconstructed at each time point (software NRecon, SkyScan) and visually inspected (software Dataviewer, SkyScan) to identify trabecular deformation around the screw threads. 2. Results The insertion tests were performed in conjunction with stepwise micro-ct image acquisition. Preliminary visual image analysis demonstrated that trabecular deformation was evident around the top surface of the screw threads (Fig.4).. Figure 4: 3D slices through the bone (white colour) and screw (grey), at head contact (a) and post-failure (b). A perforation through trabeculae is visible at the screw threads (arrows). 2. Conclusion We present preliminary results of the novel screw insertion device aimed at visualization of the stepwise deformation of the peri-implant bone at the micro-structural level, at varying insertion torque levels. This has the potential to allow quantification of the micro-strains induced in the peri-implant bone by use of digital volume correlation or finite element analysis in future. The study, which is currently being performed, will allow a better understanding of the coupling between insertion torque and micro-structural failure mechanics.
References: 1. Stoesz MJ, Gustafson PA, Patel BV, Jastifer JR, Chess JL, Surgeon perception of cancellous screw fixation, J Orthop Trauma, 28(1):e1-7, 2014 2. Reynolds KJ, Cleek TM, Mohtar AA, Hearn TC, Predicting cancellous bone failure during screw insertion J Biomech, 46(6):1207-10, 2013 3. Basler SE, Traxler J, Müller R, van Lenthe GH, Peri-implant bone microstructure determines dynamic implant cut-out Med Eng Phys, 35:1442 1449, 2013 4. Ab-Lazid R, Perilli E, Ryan MK, Costi JJ, Reynolds KJ, Does cancellous screw insertion torque depend on bone mineral density and/or microarchitecture? J Biomech, 47(2):347-53, 2014 5. Ab-Lazid R, Perilli E, Ryan MK, Costi JJ, Reynolds KJ, Pullout strength of cancellous screws in human femoral heads depends on applied insertion torque, trabecular bone microarchitecture and areal bone mineral density, J Mech Behav Biomed Mater, 40:354-61, 2014 6. Perilli E, Briggs AM, Kantor S, Codrington J, Wark JD, Parkinson IH, Fazzalari NL, Failure strength of human vertebrae: prediction using bone mineral density measured by DXA and bone volume by micro-ct, Bone, 50(6):1416-25, 2014 7. Wirth AJ, Goldhahn J, Flaig C, Arbenz P, Müller R, van Lenthe GH, Implant stability is affected by local bone microstructural quality, Bone, 49,473 478, 2011 Acknowledgements This work was funded by an Australian NHMRC Grant (Grant ID 595933).