Reproducibility of Three Dimensional Digital Preoperative Planning for the Osteosynthesis of Distal Radius Fractures

Reproducibility of Three Dimensional Digital Preoperative Planning for the Osteosynthesis of Distal Radius Fractures Yuichi Yoshii, 1 Takuya Kusakabe, 1 Kenichi Akita, 2 Wen Lin Tung, 3 Tomoo Ishii 1 1 Department of Orthopaedic Surgery, Tokyo Medical University Ibaraki Medical Center, Ami 300-0395, Japan, 2 LEXI Co., Ltd., Tokyo, Japan, 3 Department of Occupational Therapy, Ibaraki Prefectural University of Health Sciences, Ami, Japan Received 30 November 2016; accepted 6 April 2017 Published online 2 May 2017 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jor.23578 ABSTRACT: A three-dimensional (3D) digital preoperative planning system for the osteosynthesis of distal radius fractures was developed for clinical practice. To assess the usefulness of the 3D planning for osteosynthesis, we evaluated the reproducibility of the reduction shapes and selected implants in the patients with distal radius fractures. Twenty wrists of 20 distal radius fracture patients who underwent osteosynthesis using volar locking plates were evaluated. The 3D preoperative planning was performed prior to each surgery. Four surgeons conducted the surgeries. The surgeons performed the reduction and the placement of the plate while comparing images between the preoperative plan and fluoroscopy. Preoperative planning and postoperative reductions were compared by measuring volar tilt and radial inclination of the 3D images. Intra-class correlation coefficients (ICCs) of the volar tilt and radial inclination were evaluated. For the implant choices, the ICCs for the screw lengths between the preoperative plan and the actual choices were evaluated. The ICCs were 0.644 (p < 0.01) and 0.625 (p < 0.01) for the volar tilt and radial inclination in the 3D measurements, respectively. The planned size of plate was used in all of the patients. The ICC for the screw length between preoperative planning and actual choice was 0.860 (p < 0.01). Good reproducibility for the reduction shape and excellent reproducibility for the implant choices were achieved using 3D preoperative planning for distal radius fracture. Three-dimensional digital planning was useful to visualize the reduction process and choose a proper implant for distal radius fractures. ß 2017 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 35:2646 2651, 2017. Keywords: three dimensional; distal radius fracture; computed tomography; osteosynthesis; locking plate Volar locking plates have been used widely for the osteosynthesis of distal radius fractures. 1 4 The locking plate has been shown to provide more stable fixation than conventional non-locking plate/screw systems. The advantages of locking plates apply to cases of highly comminuted fractures, unstable metadiaphyseal segments, and osteoporotic fractures. 5 Since the directions of the screws are fixed and distributed three dimensionally, the surgeon needs to determine the placement of the plate carefully at the fracture site. It has been suggested that proper plate width or placement is necessary to provide satisfactory subchondral support. 6,7 Several complications following internal fixation of distal radius fractures with a volar locking-plate have been reported, 8 10 including extensor and flexor tendon ruptures, carpal tunnel syndrome, and screw penetration to the joint, etc. These complications can be avoided with appropriate planning for the reduction and implant choices. Generally, templates are used for the preoperative planning of fracture management. Although the filmless clinical environment is now a reality, 11 the development of digital preoperative planning for orthopaedic trauma surgery is lagging behind in current image browsing systems. This may result in a poorer agreement of the reduction shape and implant choices between the preoperative plans and the postoperative results. It may require to adapt to digitization and to visualize surgical planning in fracture management in Correspondence to: Yuichi Yoshii (T: 81-29-887-1161; F: 81-29-888-8303;E-mail: yy12721@yahoo.co.jp) # 2017 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. order to offer higher reproducibility of the planning. A 3D digital pre-operative planning system for the osteosynthesis of distal radius fractures was developed for clinical practice. In this study, we hypothesized that 3D digital planning would be useful to reproduce the reduction shape and the implant choices for fracture management. The objective of this study was to evaluate the reproducibility of 3D planning for the osteosynthesis of distal radius fractures by comparing pre- and post-operative reduction and implant choices. METHODS This is a case control study (level of evidence: III). The study protocol was approved by our Institutional Review Board. This study was registered as NCT02909647 at ClinicalTrials. gov. Twenty wrists of 20 distal radius fracture patients who underwent osteosynthesis using locking plates (13 females, 7 males; age range 23 87, mean age 59.0 years) were evaluated. Patients were excluded if they reported a previous history of traumatic injuries to the arm. According to the preoperative CT scans, the fractures were classified using the AO classification system. 12 The surgeries were performed by four trainees (residents and fellows) under the supervision of a single hand surgeon within 2 week of injury. Preoperative Planning Three-dimensional digital preoperative planning and a surgical simulation were performed in order to determine the reduction and the placement of implants in addition to the plate/screw size prior to surgery. Reduction and placement of the implants were simulated using software newly-developed by the authors (ZedTrauma, LEXI Co., Ltd. Tokyo, Japan). This is based on 3D planning software (ZedView, LEXI Co., Ltd. Tokyo, Japan), to which we added a reduction and implant choice simulation program. Planning using this software is based on digital imaging and communications in 2646

3D PLANNING FOR THE DISTAL RADIUS FRACTURE 2647 medicine (DICOM) data from CT scans. All patients had preand post-operative CT scans in order to compare the planned and post-operative reduction shape and placement of the implant. The CT comprised contiguous sections of 1 mm thickness. The software allows the surgeons to (i) visualize the fracture displacement; (ii) simulate repositioning of the fragments; (iii) place the plate and screws; (iv) adjust the sizes of the implants; and (v) check the shape after the reduction and implant placement by measuring the anatomical shape (Figs. 1 and 2). After importing the DICOM images to the software, a 3D image of the distal radius was made. Each distal radius fracture was segmented according to the main fracture fragments using the cut function. Each fragment was repositioned in accordance with fracture lines. Reduction of the fragment was performed to regain the volar tilt and the radial inclination, and less than 2 mm of step-off for the intra-articular displacement referring to the healthy side wrist X-ray. Simulation of the implantation of the volar locking plate used 3D templates with variable sizes of plates and screws. The Stellar II locking plate (HOYA Technosurgical, Inc., Tokyo, Japan) was used in this study. This plate system has small, medium, and large sizes for plate width, and short and long sizes for plate length. Screw lengths from 10 to 24 mm for the distal (2.4 mm diameter) and 10 to 20 mm for the proximal (2.6 mm diameter) are available. Computer aided design models of different-sized implants were installed in the software. The plate size was chosen to cover the distal fragment maximally and to not exceeded the width of the distal radius. After the planning, osteosynthesis was performed under general anesthesia. During the surgery, the operator performed the reduction and the placement of the plate while comparing images between the preoperative plan and fluoroscopy. The surgeons tried to reproduce the planned position of the implant by checking the distances from the margin of the implant to the margin of the radius under fluoroscopy. The screw sizes were determined by intraoperative measurement in reference to the preoperative plan. Evaluations To evaluate the reproducibility of the 3D planning, preoperative planning and postoperative reductions were compared by measuring the volar tilt and radial inclination of the 3D images in the plan group (Fig. 3). The axis of the radius was adjusted. The angle between a line from the dorsal edge to the volar edge of the radius and the line perpendicular to the longitudinal axis of the radius was measured as 3D volar tilt. The angle between a line from the radial styloid tip to the ulnar aspect of the distal radius and the line perpendicular to the longitudinal axis of the radius was measured as 3D radial inclination. These radiological parameters were measured by a single hand surgeon, who was one of the inventors of the software. In addition, the planned sizes of implant were compared with the actual sizes used at operation. The screw holes were numbered from 1 to 8 for the distal (there were seven holes in small and medium plates, and eight holes in a large plate), and from one to three for the proximal. The screw lengths actually chosen were recorded according to the screw number. Statistical Analysis The results are expressed as mean standard deviation. Intra-class correlation coefficient (ICC) values of the 3D measurements (volar tilt and radial inclination) between the pre-operative plan and the post-operative 3D images were evaluated. For accuracy of the implant choices, the ICCs for the screw lengths between the preoperative plan and the actual choices were evaluated. An ICC less than 0.40 was considered poor reproducibility, between 0.40 and 0.59 was Figure 1. Example image of fracture reduction. After importing the DICOM images to the software, a 3D image of the distal radius was made. The distal radius was segmented according to the main fracture fragments using the cut function. Each fragment was repositioned in accordance with fracture lines. After repositioning the fragments, the bone shape was checked three dimensionally.

2648 YOSHII ET AL. Figure 2. Example image of volar locking plate placement. The plate size was chosen when it covered the distal fragment maximally and did not exceed the width of the distal radius. The longest possible screws that did not penetrate the extensor compartment or radiocarpal joint were determined. This was checked with axial, coronal, and sagittal views of the repositioned bone outline. considered moderate, between 0.60 and 0.74 was considered good, and between 0.75 and 1.00 was considered excellent.13 To compare the screw lengths between the preoperative plan and the actual choice, the paired t-test was used. p values less than 0.05 were considered significant. All analyses were performed using SPSS version 13.0 (SPSS, Chicago, IL). RESULTS There was one patient with A2 (extraarticular, simple, and impacted) fracture, two patients with A3 (extraarticular, multifragmentary) fracture, one patient with C1 (complete articular, articular simple, metaphysial simple) fracture, eight patients with C2 (complete articular, articular simple, metaphysial multifragmentary) fracture, and eight patients with C3 (complete articular, articular multifragmentary) fracture. The correlations of the 3D measurements are shown in Figure 4. Before the 3D planning, the mean values of volar tilt and radial inclination were 10.6 18.2 and 8.5 10.9. In the 3D measurements, the results of volar tilt were 11.4 2.7 and 10.8 2.8 in the preoperative planning and postoperative 3D images, respectively. The results of radial inclination were 21.6 3.9 and 21.3 3.0 in the preoperative planning and postoperative 3D images, respectively. The ICCs were 0.644 (good, p < 0.01) and 0.625 (good, p < 0.01) for the volar tilt and radial inclination in the 3D measurements, respectively. The planned size of plate was used in all of the plan group patients. Small plates were chosen for ten patients, medium for eight patients, and large for two patients. The results of screw choices are shown in Figures 5 and 6. Since the distal 8 screw hole was used in only two patients who needed a large plate, the distal 8 screw length data were excluded from the analysis. The differences between preoperative plan and actual choice (¼actual choice plan) are shown in Figure 5a and b. Shorter screws were chosen at operation for the distal 1 and 5 screw holes compared to other distal screw holes. There was a significant difference between preoperative plan and actual choice for the distal 5 screw hole (p ¼ 0.032). The ICC of the Figure 3. 3D measurement of the volar tilt and radial inclination. (a) 3D image of pre-operative plan. (b) 3D image of post-operation. The angle between a line from the dorsal edge to the volar edge of the radius and the line perpendicular to the longitudinal axis of the radius was measured as 3D volar tilt. The angle between a line from the radial styloid tip to the ulnar aspect of the distal radius and the line perpendicular to the longitudinal axis of the radius was measured as 3D radial inclination. The screw holes were numbered from 1 to 8 for distal, and from 1 to 3 for proximal.

3D PLANNING FOR THE DISTAL RADIUS FRACTURE 2649 Figure 4. Correlations for the 3D measurements. (a) Correlations of the volar tilt. (b) Correlations of the radial inclination. The ICCs were 0.644 (good, p < 0.01) and 0.625 (good, p < 0.01) for the volar tilt and radial inclination in the 3D measurements, respectively. screw choice was 0.860 (excellent, p < 0.001). No screws penetrated into either the extensor compartments or the radiocarpal/distal radioulnar joints. DISCUSSION The development of digital preoperative planning for orthopedic trauma surgery is a long way behind digital planning for joint arthroplasty. In most cases, classical tracing paper or simple measurements on an image viewer for X-ray or CT scans are still in clinical use. The benefits of the conventional method are simplicity, familiarity, low cost, and less exposure to radiation. However, it is difficult to visualize which direction to move the fragments, and which screw size to choose prior to surgery. A precise preoperative plan is necessary for good postoperative outcomes. Three-dimensional imaging methods for planning in trauma surgery were recently developed using CT scanning. 14 16 In addition, computerized virtual surgery planning has been increasingly applied in various orthopedic procedures. Digital preoperative planning was developed mainly in the field of hip and knee joint replacement. 17 19 There are several software programs for planning joint replacement (i.e., Orthoplanner, Zed Hip, & Knee). In fracture surgery, 3D pre-operative planning is useful for viewing and measuring the fracture displacement. 20 Adding manipulation would enable a virtual operation to be completed, including the testing for fit of possible implants. Thus, a 3D digital preoperative planning system for fracture surgery was developed. This system allowed us to recognize fracture fragments, reposition the fragments, place the implant, adjust the implant size, and measure the angle or the distance of the reduction shape three-dimensionally. In addition, a function to display the bone-implant contact area was added. Correction of the anatomical shape and proper implant selection and positioning are closely related to clinical outcomes in distal radius fracture management. In the AAOS guideline, 21 the suggested indications for operative fixation of a distal radius fracture are radial shortening >3 mm, dorsal tilt >10, or intraarticular displacement or step-off >2 mm. In the radiographic evaluation, normal radial inclination is Figure 5. Results of screw choices. (a) Average lengths of the screws in the preoperative plan and the actual choices. The gray bar indicates the average length of the screws in the pre-operative plan. The black bar indicates the average length of the screws in the actual choices. There was a significant difference between preoperative plan and actual choice of screw length for the distal 5 screw hole (p ¼ 0.032) (b) The differences between the preoperative plan and the actual choices. Relatively shorter screws were chosen at operation for the distal 1 and 5 screw holes compared to the preoperative plan.

2650 YOSHII ET AL. Figure 6. Correlation of screw choices between the preoperative plan and the actual choices. The ICC of the screw choice was 0.860 (p < 0.01). reported as a mean of 25.4 with a standard deviation of 2.2. Normal volar tilt is reported as a mean of 14.5 and a standard deviation of 4.3. 22 The reduction shape was close to the normal range of radiographic parameters in this study. There were good correlations between pre- and post-operative reduction in 3D measurements. When simulating the repositioning, the distal radius was reconstructed by moving and rotating the fragments according to the fracture line. These process may be helpful to understand which fragments need to reposition. The number and the lengths of the distal locking screws are important to ensure adequate fixation strength and maintain fracture fixation. Sometimes, it is difficult to determine the length of distal locking screws with a depth gauge due to thin cortex of the bone or comminution of the fracture fragments. Using the preoperative planning, it was easy to determine the screw sizes even when there were comminuted fragments at the dorsal cortex of the radius. The length of distal locking screws is important because it relates with the fixation strength and extensor complications. 23 It has been suggested that locked unicortical distal screws of at least 75% of the length of bicortical screws offer construct stiffness similar to bicortical fixation. 24 At the same time, there is a risk of penetration to the extensor compartment when choosing longer screws. In this study, we found that the distal locking screws in the most radial position (distal 1 and 5) were about 3.5 4.4 mm shorter than the screws in the middle of the plate (distal 2, 3, 6). The screws in the most ulnar position (distal 4, 7) were about 0.4 1 mm shorter than the screws in the middle of the plate. From these results, it may be possible to determine the length of all distal locking screws after deciding one screw length. Some limitations of the present study should be noted. First, the sample size of this study was small. This is because, we wanted to grasp the problems of this new method before proceeding to a larger study. It may be necessary to clarify the benefit of using 3D planning by evaluating clinical outcomes in future studies. Second, this protocol requires CT scanning. The radiation exposure level with CT (2.0 mgy) is about eight times higher than with regular posteroanterior and lateral view X-rays (0.25 mgy). There may be some concerns about radiation exposure. It has been reported that specific protocols combining filters and image post-processing software may solve the radiation dose problem. 25 Finally, the reproducibility of the bone shape was confined to moderate in this study. There is room for improvement. This may be due to the learning process associated with a new technique. In addition, there is still no standard for the 3D reduction shape. This may be another reason for the moderate reproducibility. It is necessary to develop a method to compare 2D images of X-rays and 3D images of the CT scan. In addition, we need to define the proper implant size for the bone shape. In conclusion, a 3D pre-operative planning system was developed that could be used clinically to plan reduction and choose implants of distal radius fractures. Three-dimensional digital planning was useful to visualize the reduction process and to choose proper implants for distal radius fractures. AUTHORS CONTRIBUTIONS YY contributed in research design, analysis of the data, write manuscript, TK contributed in acquisition and analysis of the data, write manuscript, KA contributed in acquisition and analysis of the data, write manuscript, WT contributed in analysis of the data, write manuscript, TI contributed in research design, write manuscript. All authors were fully involved in the study and approved final version of this manuscript. CONFLICTS OF INTEREST There was no financial support for this study. One of the authors (KA) is a paid employee at LEXI Co., Ltd. The software described in this study has been developed through a collaborative study between Tokyo Medical University Ibaraki Medical Center and LEXI Co., Ltd. KA wrote the program used in this study. YY suggested the function necessary for the program. There is no financial relationship between Tokyo Medical University Ibaraki Medical Center and LEXI Co., Ltd. REFERENCES 1. Orbay JL, Touhami A. 2006. Current concepts in volar fixedangle fixation of unstable distal radius fractures. Clin Orthop Relat Res 445:58 67. 2. Anglen J, Kyle RF, Marsh JL, et al. 2009. Locking plates for extremity fractures. J Am Acad Orthop Surg 17:465 472. 3. Gehrmann SV, Windolf J, Kaufmann RA. 2008. Distal radius fracture management in elderly patients: a literature review. J Hand Surg Am 33:421 429. 4. Chen NC, Jupiter JB. 2007. Management of distal radial fractures. J Bone Joint Surg Am 89:2051 2062. 5. Greiwe RM, Archdeacon MT. 2007. Locking plate technology: current concepts. J Knee Surg 20:50 55.

3D PLANNING FOR THE DISTAL RADIUS FRACTURE 2651 6. Orbay JL, Fernandez DL. 2004. Volar fixed-angle plate fixation for unstable distal radius fractures in the elderly patient. J Hand Surg Am 29:96 102. 7. Fok MW, Klausmeyer MA, Fernandez DL, et al. 2013. Volar plate fixation of intra-articular distal radius fractures: a retrospective study. J Wrist Surg 2:247 254. 8. Arora R, Lutz M, Hennerbichler A, et al. 2007. Complications following internal fixation of unstable distal radius fracture with a palmar locking-plate. J Orthop Trauma 21:316 322. 9. Soong M, Earp BE, Bishop G, et al. 2011. Volar locking plate implant prominence and flexor tendon rupture. J Bone Joint Surg Am 93:328 335. 10. Al-Rashid M, Theivendran K, Craigen MA. 2006. Delayed ruptures of the extensor tendon secondary to the use of volar locking compression plates for distal radial fractures. J Bone Joint Surg Br 88:1610 1612. 11. Choplin RH, Boehme JM, 2nd, Maynard CD. 1992. Picture archiving and communication systems: an overview. Radiographics 12:127 129. 12. Mueller ME, Nazarian S, Koch P. 1987. AO classification of fractures. Berlin: Springer-Verlag. 13. Cicchetti Domenic V. 1994. Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology. Psychol Assessment 6:284 290. 14. Doornberg JN, Rademakers MV, van den Bekerom MP, et al. 2011. Two-dimensional and three-dimensional computed tomography for the classification and characterisation of tibial plateau fractures. Injury 42:1416 1425. 15. Xia S, Zhang Y, Wang X, et al. 2015. Computerized virtual surgery planning for ORIF of proximal humeral fractures. Orthopedics 38:e428 e433. 16. Brown GA, Firoozbakhsh K, Gehlert RJ. 2001. Threedimensional CT modeling versus traditional radiology techniques in treatment of acetabular fractures. Iowa Orthop J 21:20 24. 17. Sariali E, Mauprivez R, Khiami F, et al. 2012. Accuracy of the preoperative planning for cementless total hip arthroplasty. A randomised comparison between three-dimensional computerised planning and conventional templating. Orthop Traumatol Surg Res 98:151 158. 18. Viceconti M, Lattanzi R, Antonietti B, et al. 2003. CT-based surgical planning software improves the accuracy of total hip replacement preoperative planning. Med Eng Phys 25:371 377. 19. Levine B, Fabi D, Deirmengian C. 2010. Digital templating in primary total hip and knee arthroplasty. Orthopedics 33:797. 20. Fadero PE, Shah M. 2014. Three dimensional (3D) modelling and surgical planning in trauma and orthopaedics. Surgeon 12:328 333. 21. Lichtman DM, Bindra RR, Boyer MI, et al. 2011. American Academy of Orthopaedic Surgeons clinical practice guideline on: the treatment of distal radius fractures. J Bone Joint Surg Am 93:775 778. 22. Mann FA, Wilson AJ, Gilula LA. 1992. Radiographic evaluation of the wrist: what does the hand surgeon want to know? Radiology 184:15 24. 23. Zenke Y, Sakai A, Oshige T, et al. 2013. Extensor pollicis longus tendon ruptures after the use of volar locking plates for distal radius fractures. Hand Surg 18:169 173. 24. Wall LB, Brodt MD, Silva MJ, et al. 2012. The effects of screw length on stability of simulated osteoporotic distal radius fractures fixed with volar locking plates. J Hand Surg Am 37:446 453. 25. Henckel J, Richards R, Lozhkin K, et al. 2006. Very low-dose computed tomography for planning and outcome measurement in knee replacement. The imperial knee protocol. J Bone Joint Surg Br 88:1513 1518.