Gait analysis: An effective tool to mechanically monitor the bone regeneration of criticalsized defects in tissue engineering applications

RESEARCH ARTICLE Gait analysis: An effective tool to mechanically monitor the bone regeneration of criticalsized defects in tissue engineering applications Pablo Bla ´zquez-CarmonaID 1,2 *, Juan Mora-Macı ´as 2,3 , Juan MorgazID 4 , Marı ´a del Mar Granados 4 , Jaime Domı ´nguez 1,2 , Esther Reina-Romo 1,2 1Department of Mechanical and Manufacturing Engineering, Escuela Te ´cnica Superior de Ingenierı ´a, Universidad de Sevilla, Seville, Spain, 2Instituto de Biomedicina de Sevilla (IBiS), University of Seville, Seville, Spain, 3Department of Mining, Mechanical, Energy and Building Engineering, Escuela Te ´cnica Superior de Ingenierı ´a, University of Huelva, Huelva, Spain, 4Department of Animal Medicine and Surgery, Universidad de Co ´rdoba, Campus Universitario de Rabanales, Co ´rdoba, Spain *[email protected] Abstract Introduction Tissue engineering has emerged as an innovative approach to treat critical-size bone defects using biocompatible scaffolds, thus avoiding complex distraction surgeries or limited stock grafts. Continuous regeneration monitoring is essential in critical-size cases due to the frequent appearance of non-unions. This work evaluates the potential clinical use of gait analysis for the mechanical assessment of a tissue engineering regeneration as an alternative to the traditional and hardly conclusive manual or radiological follow-up. Materials and methods The 15-mm metatarsal fragment of eight female merino sheep was surgically replaced by a bioceramic scaffold stabilized with an external fixator. Gait tests were performed weekly by making the sheep walk on an instrumented gangway. The evolution of different kinematic and dynamic parameters was analyzed for all the animal’s limbs, as well as asymmetries between limbs. Finally, potential correlation in the recovery of the gait parameters was evaluated through the linear regression models. Results After surgery, the operated limb has an altered way of carrying body weight while walking. Its loading capacity was significantly reduced as the stance phases were shorter and less impulsive. The non-operated limbs compensated for this mobility deficit. All parameters were normalizing during the consolidation phase while the bone callus was simultaneously mineralizing. The results also showed high levels of asymmetry between the operated limb and its contralateral, which exceeded 150% when analyzing the impulse after surgery. Gait recovery significantly correlated between symmetrical limbs. PLOS ONE PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 1 / 18 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Bla ´zquez-Carmona P, Mora-Macı ´as J, Morgaz J, Granados MdM, Domı ´nguez J, ReinaRomo E (2023) Gait analysis: An effective tool to mechanically monitor the bone regeneration of critical-sized defects in tissue engineering applications. PLoS ONE 18(12): e0296510. https:// doi.org/10.1371/journal.pone.0296510 Editor: Yaodong Gu, Ningbo University, CHINA Received: September 24, 2023 Accepted: December 13, 2023 Published: December 29, 2023 Peer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal.pone.0296510 Copyright: ©2023 Bla ´zquez-Carmona et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Conclusions Gait analysis was presented as an effective, low-cost tool capable of mechanically predicting the regeneration of critical-size defects treated by tissue engineering, as comparing regeneration processes or novel scaffolds. Despite the progressive normalization as the callus mineralized, the bearing capacity reduction and the asymmetry of the operated limb were more significant than in other orthopedic alternatives. Introduction Bone is a highly dynamic hierarchical composite with homeostatic mechanisms in charge of its continuous remodeling and regeneration in the event of defects caused by tumors, traumas, or infections. Nevertheless, the regenerative capacity of hard tissue is not unlimited. These defects have a size upper limit from which this rigid organ does not heal spontaneously, usually leading to non-unions [1]. Thus, the treatment of critical-sized bone defects is still today one of the significant challenges for acute care surgeons. To address this issue, they must resort to other clinical strategies, including those based on the Ilizarov method (distraction osteogenesis) or the incorporation of substitutes. For instance, bone transport trusts on the long-standing distraction osteogenesis to gradually move an osteotomized surrounded bony fragment towards the position of the original defect while simultaneously forming a bone callus on the other side, the docking site [2–4]. However, this regeneration process carries inherent risks, particularly concerning viscoelastic and structural alterations in the neighboring soft tissues (e.g., skin, tendons, or muscles) [5–8]. Another traditional solution lies in incorporating autogenic and allogenic bone grafts to encourage cell regenerative activity in the gap, mainly vascularized free fibular or Papineau open cancellous bone grafting [9–11]. Despite being the most suitable substitute concerning biocompatibility, osteoinductive and osteoconductive properties, the availability of harvested autografting tissue is limited, and the allograft’s growing clinical demand exceeds the donor stock [12,13]. Since not long ago, research for novel alternatives to tissue transfer has been ongoing through tissue engineering (TE) [14]. TE is focused on finding the standard substitute material and structure to effectively emulate the intricate microenvironment of native bone tissue and further promote regeneration [15,16]. In this field, additive manufacturing techniques have gained popularity for their versatility in building stable three-dimensional porous scaffolds with a broad range of internal microarchitectures to investigate numerically. This flexibility materialized in many scientific works currently focus on the optimization of the microstructure of scaffolds with different types of materials beyond the traditional biomedical metals, including bioceramics or polymers [16–18]. Although scaffolds have been implanted with reasonable success in the repair of small bone defects [19,20], their success with critical-sized ones remains limited [21,22]. After an orthopedic surgery of a critical-sized bone defect, continuous monitoring of the bone regeneration process is advisable, especially in the early weeks when bone infections or non-unions frequently appear. In this line, the progress of a bone healing process is typically monitored by plain film radiology and densitometric methods [23–25]. Beyond being accompanied by frequent exposure to radiation, the lack of continuity in this follow-up constantly results in qualitative conclusions and time lags in critical decision-making, including the need for reintervention or fixation removal. These delays occur since defect mineralization is primarily reflected in radiological images time after the real increase in bone callus stiffness [26,27]. Poor correlations of radiographic measurements with the tissue mechanical properties PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 2 / 18 Funding: This project was supported by the Ministerio de Ciencia e Innovacio ´n (Government of Spain) through the grant number PID2020113790RB-I00 awarded to JMM and ERR, https:// www.aei.gob.es/convocatorias/buscadorconvocatorias/proyectos-idi-2020-modalidadesretos-investigacion-generacion, and by the Junta de Andalucı ´a (FEDER-UHU, Programa Operativo FEDER de Andalucı ´a 2014-2020) through the grant number UHU-202058 awarded to JMM, https:// www.agenciaandaluzadelaenergia.es/es/ financiacion/incentivos-2017-2020/fondo-feder2014-2020. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript" was included in the cover letter of the reviewed manuscript. Competing interests: The authors have declared that no competing interests exist. have been consequently reported in the literature [27]. Thus, despite the development of scoring systems and indexes to evaluate and compare sets of radiographs [23,24], in practice, clinicians tend both to overestimate or underestimate a lower callus strength than it really is [28]. Other traditional in vivo manual methods, such as clinical symptoms (tenderness or pain while bearing weight) or the mobility review of the treated bone, trust the clinician’s ability to interpret examinations and do not ensure the proper ossification in any case [29]. In the tissue engineering field, the opacity of its base material, its small pore sizes, and the generally high apparent stiffness of the structure could hinder performing a realistic analysis of the naïve tissue formation inside from radiological or manual assessments, especially in those critical first post-operative weeks. In recent decades, engineers have been developing alternatives that offer quantitative mechanical and structural parameters to predict bone regeneration indirectly. Loads through the fixation, interfragmentary displacements, strains, or acoustic emissions collected by means of a wide variety of sensors allow estimating the bone callus stiffness in vivo and mechanically comparing surgical protocols and bone regeneration approaches [5,30–35]. However, most instrumented systems can only be coupled into external fixators and require complex and expensive acquisition equipment not widely available today in the clinical routine. In vibrational measurements, the treated bone is also required to be subcutaneous [23,32]. As another option, gait analysis has been used to follow up on kinematics [36], electromyographic (EGM) [37], or load-bearing parameters [38–40] potentially modified by orthopedic surgery. During kinematics tests and EGM, measurements offer complete information regarding the position, orientation, and electrical activity of many body segments at the expense of the time-consuming tasks of attaching skin markers or electrodes to the patients’ bodies [36,37]. Among all the potentially collectible biomechanical parameters, some authors have developed algorithms to optimize the best set that identifies significant differences in gait patterns under the specific conditions of each study [41,42]. By cons, the ground reaction force (GRF) measurement allows analyzing force-time curves collected by load-bearing platforms during the patient’s stance phases. The simplicity and the possibility of directly examining these data in vivo with no complex data post-processing make it the most extrapolated-toclinic method to study the dynamic evolution of the lower limbs after orthopedic treatments or surgeries. GRF has also been reported to reflect the recovery of apparent stiffness at the defect site until the full functional recovery in fracture healing and distraction osteogenesis processes [38–40]. It also makes it possible to compare the impact of surgery on bearing capacity between orthopedic treatments, as well as the speed of recovery in each case. In addition, clinical decisions based on GRF data are proven more reliable than those taken from kinematic or EGM parameters [43,44]. As far as the authors are concerned, gait analysis has not been tested in bone tissue engineering applications on lower limbs, and its usefulness in assessing the proper regeneration of bony defects treated with scaffolds remains unknown. In this line, this work aims to verify the usefulness of gait analysis in vivo for assessing the bone regeneration of weight-bearing critical-sized defects treated with bioceramic scaffolds externally stabilized. The proposed monitoring methodology is clinicians-friendly through low-cost devices whose collected data do not require complex post-processing tasks. This work also delved into the potential correlation between the different analyzed gait parameters in the operated and non-operated limbs throughout the TE mineralization phase as a means to simplify gait analysis in real clinical scenarios. Besides, the recovery mode of the studied parameters, including data from the contralateral limb, will be compared with other equivalent orthopedic treatments on the same bone model (e.g., bone transport) to investigate the functional advantages of each regeneration process. PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 3 / 18 Materials and methods Animals and bone model TE experiments were performed in vivo on eight right-back ovine metatarsi of adult female Merino sheep (n = 8). A control group of 3 non-operated animals, randomly selected, were included in the study to contrast results with the experimental group directly (n = 3). Therefore, experiments were carried out in a total of 11 animals. Sample size was selected according to the standard deviation and statistical tests found in a previous study [39]. This study and its protocol were previously approved by the Animal Ethics of the University of Co ´rdoba (Protocol Number: 2021PI/21) and guaranteed during surgical interventions and experimental phases the animal’s welfare in strict accordance with the ARRIVE guidelines, the European (2010/63/UE) and national (RD 1201/2005) regulations on animal experimentation, thus avoiding stressful situations. A completed copy of the Full ARRIVE 2.0 Guidelines checklist is provided in the S1 Checklist of the Supporting Information. The selected mammalian model has numerous advantages. Sheep are docile, low-priced, and have a bone composition and body weight comparable to humans [45–47]. These similarities allow for easier extrapolation of conclusions to real clinical cases. The metatarsus was chosen as bone mode due to its thin layer of surrounding soft tissues, finding exclusively periosteal tissue on the lateral and medial sides. This anatomical advantage allowed a safer and more accessible tissue engineering surgery. The animals were obtained from a farm for research and were marked on the wool to avoid confounders. They were healthy, with an adequate vaccination and deworming protocol. Sheep with an adequate metatarsal length were selected for the study (>12 cm). They were stabled 15 days before the start in the research center facilities for an adaptation period where all experiments were carried out. Their facilities were adapted to provide for their needs, including shade, comfortable lying areas, and secure fencing to keep them unstressed and safe. In the operated group, a 15-mm critical-size bone segment of right metatarsus was surgically replaced by a subject-specific bioceramic scaffold. For comparative reasons, this bone defect size was selected based on the size in other ovine bone regeneration experiments in the literature [4,5,39]. The defect was externally stabilized by a modular Ilizarov-type external fixator. This external solution minimizes body invasion compared to internal plates and splits. Hence, they reduce the soft tissue damage, leading to an early recovery of the animals’ mobility and a lower impact on the analyzed gait parameters in the weeks immediately following the intervention. When the experimental part of the study concluded, the animals were euthanized by an overdose of sodium pentobarbital IV Euthasol1. The following subsections cover the design and manufacturing process of the scaffold, the steps of the orthopedic surgery, details about the in vivo gait experiments, and the radiological assessment. In the “Surgery procedure” subsection, all relevant steps taken to ameliorate animal suffering during the research are also detailed. Design and fabrication of the scaffold The scaffold was designed with the subject-specific geometry of the bone fragment to be replaced. For this, computed tomography was performed on the right hindlimb of the sheep before surgery (voxel size 0.12 x 0.12 x 0.60 mm). A multiplanar hard tissue thresholding was applied to each metatarsal scan using the software InVesalius1(Renato Archer Information Technology Center, Amarais, Brazil), which enabled to generate a 3D bone reconstruction from the image stack, as shown in Fig 1A. Afterward, an intermediate 13-mm bone segment was sliced from the metatarsal geometry employing the solid modeling CAD software SpaceClaim1(SpaceClaim Corporation, Concord, MA, USA), and its inner medullary cavity was filled. Two building modifications were also carried out on this initial geometry to improve PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 4 / 18 the scaffold stability in the defect and speed up the regenerative response (see Fig 1A). Firstly, a coupling cylinder (4 mm Ø; 2 mm length) was added over one end face to immobilize the structure in vivo in the distal bone marrow. This coupler would prevent the scaffold from rotating or moving during the animals’ daily activity, which could interfere with the proper growth of the naïve tissue. At the other extreme, the solid was hollowed out by another cylinder (4 mm Ø; 10 mm length) for grafting, as further explained. A robocasting device (3-D Inks Still-water1, Tulsa, Oklahoma, USA), an extrusion-based 3Dprinting technique, was selected. It worked by depositing a 45 vol% hydroxyapatite slurry forming a ceramic network of perpendicularly oriented layers of bars, as illustrated in Fig 1B. This material was chosen since bioceramics are proven to enhance naïve tissue growth by regulating osteoblast proliferation and differentiation while the structure is reabsorbed [48]. From previous works in the literature, the concentration of colloidal suspensions in the 3D-printing ink (45 vol%) was optimized to obtain the suitable viscoelastic properties for an effective deposition and assembly while ensuring the proper mechanical performance of the sintered structure in vivo [49–51]. An example of the final geometry of the patient-specific scaffold is provided in the S1 File of the Supporting Information. The printing nozzle diameter, the pore size, or the layer overlap were also numerically optimized in previous works to maximize cell diffusion and proliferation while ensuring the mechanical integrity of the structure under the ovine physiological loads [17]. The final microarchitecture had a porosity of 59.3%, a 560.8 μm pore size, and a specific surface area of 5768.9 m -1 [17]. After drying at room temperature, the organic components of the implant were eliminated under heating at 400˚C for 1 hour. They were finally sintered at 1300˚C for 2 hours to compact the paste-like scaffold. The chemical sterilization of the structures was achieved using formaldehyde at 60˚C and relative humidity of 80%. Fig 1. Steps in the design of the bone tissue engineering surgery. (A) Design of the bioceramic scaffold from the 3D geometry of the ovine metatarsus reconstructed by computed tomography scans. (B) 3D-printing of the hydroxyapatite structure using roboscasting. (C) Harvesting of the spongy grafting tissue from the lateral side of the contralateral humerus. (D) Implantation of the scaffold. https://doi.org/10.1371/journal.pone.0296510.g001 PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 5 / 18 Surgical procedure Before surgery, Amoxiciline 15 mg/kg Clamoxyl 1 IM and Meloxicam 0.2 mg/kg Metacam 1 IV were supplied to the animals to prevent infections and excessive inflammation. Surgery was carried out under general anesthesia induced by detomidine 20 μg/kg Detogesic 1 IV and morphine 0.2 mg/kg Morfina B.Braun 1 IM and was maintained with the inhalational anesthetic isoflurane IsoVet11–1.2% transported in 100% oxygen. The sheep’s body temperature, blood pressure, oxygen saturation, expired fraction of carbon dioxide and electrocardiograms were constantly monitored during the intervention. The animal was placed in right lateral decubitus for a medial approach. The right hindlimb was shaved, and its skin was aseptically prepared using chlorhexidine and an antiseptic alcohol-based solution. An 8-cm incision was then performed using an electrical scalp to expose, after a careful separation of the periosteal tissue, the underlying metatarsal segment to replace. Before osteotomizing, an Ilizarov-type external fixator was implanted in the metatarsus to keep the resulting unconnected bone fragments aligned. This stabilizing frame comprises two aluminum rigs interconnected by metallic bars and fixed to the bone through a total of six drilled 4-mm ØSchanz pins. More information about the external fixation design and its mechanical properties is provided in Bla ´zquez-Carmona et al. [52]. Two parallel transversal osteotomies 15 mm apart from each other were made in the intermediate part using an oscillating saw. As a result, a critical-size 15 mm defect was created in each animal. Adding bone morphogenetic proteins or grafted tissue to the scaffold is an extended strategy used in tissue engineering experiments to avoid a lack of spontaneous healing in critical-size defects [21,22]. In this line, cancellous bone autograft from the lateral side of the contralateral humerus head was immediately harvested through a 7-cm incision using Volkmann spoons, as shown in Fig 1C. This spongy tissue was inserted in the inner hole of the scaffold, working as a cell-seeding vehicle to accelerate osteogenesis. The scaffold was then implanted into the defect and fixed to the distal fragment by introducing the coupler into the bone marrow, as illustrated in Fig 1D. Since the scaffold’s main body was 13 mm long, the remaining 2 mm proximal gap was filled with the remaining cancellous grafting tissue. After a quick radiological analysis to verify the scaffold implantation and the non-joint invasion of the fixator’s pins, the operated sheep were recovered with oxygen under assistance. During the first five days after surgery, analgesia was also provided (meloxicam 0.1 mg/kg SC Metacam 1 and buprenorphine 0.03 mg/kg IM Bupaq1) according to ruminant pain scales. During the research, external fixators were periodically checked, and the skin-pin transitions were cleaned with chlorhexidine Desinclor1. Gait analysis Gait tests consisted of making the sheep walk over a guided circular gangway containing a wireless pressure-sensitive platform Pasco PS-2141 1 (PASCO, Roseville, CA, USA), as can be seen in Fig 2A. The load platform (35 x 35 cm) continuously measures the dynamic vertical GRF in the animals’ stance phases (sampling rate of 50 Hz), the largest component of the total GRF [53]. The platform was embedded inside the walking gait to isolate vibrational noises and enable measuring ‘‘steady-state” walking. Beyond the operated right ipsilateral hindlimb (IH), limb gait conditions were also analyzed to assess potential compensation mechanisms for loss of bearing capacity after tissue engineering intervention: the ipsilateral forelimb (IF), the contralateral hindlimb (CH), and the contralateral forelimb (CF). Fig 2B shows a scheme of these gait parameters on a control GRF curve. Prior to surgery, an acclimation period was also needed for each animal since they are naturally gregarious animals, thus being averse to walking in isolated settings [54]. After surgery and a latency and recovery period of 7 days, weekly measurement sessions were carried out by recording 7–10 stance phases per limb, thus PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 6 / 18 analyzing the data as the daily average. Each gait phase for every limb was determined by analyzing the collected data, starting from the time-point when an increase in ground reaction force was recorded until the time-point at which it returned to 0 N. As inclusion criteria, acquired stance phases performed above or below amble speed (2–4 km/hr) were visually discarded. Moreover, data in which the animal interrupted its march at any point of the gangway were not included in the analysis either. During the experimental tests, a supervisor meticulously wrote down the specific time-points at which each limb executed a valid full stance phase on the load platform to further associate each GRF curve to their respective limb. The ground reaction force data from their recorded stance phases is provided in the S2 File of the Supporting Information. As shown in Fig 2B, four main dynamic and kinematic parameters were analyzed in every collected stance phase per limb: the maximum ground reaction force (GRF peak ), the average ground reaction force in the stance phase (GRF mean ), the contact time with the ground (t c ), and the relative impulse (Imp) calculated as the area under the GRF normalized by t c to decouple the effect of gait speed. All the above parameters are standard in gait analysis experiments with load platforms [24,38,39]. They were additionally normalized by the body weight of each specimen (BW) to remove body-size dependence. Consequently, the weight of the animal was controlled throughout the entire experimentation period, especially during the first weeks after surgery when an appetite and a slight weight loss were commonly suffered by sheep. Furthermore, gait asymmetry between ipsilateral and contralateral limbs was assessed for each previous gait parameter according to the formulation employed in previous works [39,55]. This index for the parameter xcan be calculated using Eq 1. Asymmetryð%Þ ¼ 100 �j XIXC 0:5�ðXIþXCÞj ð1Þ where X I and X C are the daily mean values of each gait parameter for the ipsilateral and the contralateral limbs, respectively. The asymmetry index should be close to zero both for hindand forelimbs in healthy specimens without any orthopedic pathology. Fig 2. Gait tests and analyzed parameters. (A) One sheep of the study walking in the instrumented gangway containing the wireless load platform. (B) scheme of the measured gait parameters over a healthy ground reaction force curve (GRF) normalized by the bodyweight of the sheep. https://doi.org/10.1371/journal.pone.0296510.g002 PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 7 / 18 Statistical analysis was also performed using the software MATLAB 1 (The MathWorks Inc., Natick, MA, USA) to assess the potential dependency between all analyzed gait parameters of the four limbs. Precisely, the goodness-of-fit of linear regression models was measured through the coefficient of determination, R-squared and p-value. The coefficient of determination R-squared is traditionally interpreted as the percentage of variance in one variable predicted or explained by the other [56]. In this field, authors tend to interpret R-squared over 0.49 and 0.81 as a strong and very strong correlation, respectively [5,38,39]. The p-values for the coefficients indicate whether these previous relationships are or not statistically significant. Thus, a low p-value (<0.01) would confirm that one gait parameter is significantly dependent of another. X-ray follow-up In parallel with the gait analysis, the regeneration progress was verified through monthly xrays of the operated metatarsal. These images were taken to clinically ensure the non-appearance of non-unions and to qualitatively associate changes detected in gait conditions with the level of ossification at the different regeneration stages. Results Gait parameters normalize over healing time After a week of post-surgical recovery and latency, the sheep were able to walk appropriately on the instrumented gate. In general, the weight of the animals decreased from 57.90 ±9.62 kg after surgery to 53.04 ±6.34 kg before the sacrifice. Fig 3 compares the evolution of the GRF, normalized by the body weight (% BW), during gait tests at different time points of the consolidation phase, specifically weeks 3, 10, and 30 after surgery. For one experimental test in these weeks of one of the specimens, the curves show the means (dotted black lines) and standard deviations of the force of the 7–8 stance phases of all limbs: ipsilateral and contralateral forelimbs (IF and CF, blue curves) and hindlimbs (IH and CH, red curves). Note that the operated metatarsal in which the bioceramic scaffold replaced a critical-size bone fragment was the ipsilateral hindlimb (IH). The healthy control group data are also included (green curves). Comparing with these control curves, a remarkable alteration in gait conditions and limbs’ loading capacity of the animal was found in the first weeks after surgery. First, the percentage of body weight carried by the operated IH was reduced by approximately 10%. Furthermore, similarly to humans, the control vertical force data of the sheep’s hind limbs during a gait cycle has a very particular M-shape with four increasing and decreasing intervals. This behavior is fundamentally due to initial movements and accelerations of the animals’ center of mass in the body weight distribution with the forelimbs, as well as the appearance of a propulsive action before push-off. This M-shape completely disappeared after surgery, reporting a single maximum GRF for punctual contact with the ground. In parallel, the rest of the limbs carried slightly more GRF during their stance phases at the beginning of the regeneration process (Fig 3, week 3): 11.2% of the BW above control in IF, 21.7% in CF, and 17.59% in CH. Nonetheless, the shape of the force curves was not drastically modified as the operated limb’s. Fig 3 also shows that all limbs normalized the distribution of body weight and the bearing capacity of their limbs to healthy data throughout the consolidation weeks. For instance, the operated IH limb reported values similar to those of the control group at week 30, with 40% BW compared to 44% in non-operated sheep. However, the GRF curve of the IH limb did not regain the healthy M-shape in the medium term. Simultaneous to this recovery, the x-rays of Fig 3 show a gradual ossification of the critical-size bone defect and the gradual formation of a tissue bridge that consistently mineralized, connecting the original proximal and distal bone fragments. PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 8 / 18 Similar trends were found when comparing the evolution of all the gait parameters analyzed for all the animals. The complete set of experimental data for these parameters is provided in the S3 File of the Supporting Information. For example, Fig 4A and 4B analyze the evolution of the GRF peak and GRF mean in the four limbs: operated IH limb (red empty marker), IF limb (blue empty marker), CH limb (red filled marker), and CF (blue filled marker). The means of the control data for the foreand hindlimbs are also represented as dotted lines. Both parameters significantly reduced the GRF levels after the TE surgery. Specifically, the replacement of the bone fragment by the bioceramic scaffold reduced the GRF peak to 15–23% of the BW (57– 163.46 N depending on the animal) and the GRF mean to 11–16% (39.52–95.13 N) when the control values were around 42% and 28% BW, respectively. These parameters took around 160 days to return to healthy values. The contralateral limb followed the inverse trend starting from 44–62% of the BW for the GRF peak parameter (242.97–329.67 N depending on the animal) and from 30–39% of the BW for the GRF mean (158.42–224.49 N). Likewise, the forelimbs also initially carried more body weight than their corresponding control during gait, 65% and 44% of the BW (around 370 and 250 N) for GRF peak and GRF mean , respectively. As reflected in Fig 4C, the regeneration process also impacted the t c . Although the stepping time of a sheep is, on average, 0.58 s in hindlimbs and 0.52 s in forelimbs, depending on the specimen, the intervened limbs made contact with the ground between 0.25–0.47 s. In contrast, other limb’s Fig 3. Evolution of the ground reaction force curves over the weeks after surgery. Ground reaction force curves (GRF) over the stance phases of the ipsilateral forelimb (IF), the operated right ipsilateral hindlimb (IH), the contralateral forelimb (CF), and the contralateral hindlimb (CH). Weeks after surgery and x-rays are included as a reference for the degree of ossification at said time-point. Control ground reaction force data is included for all limbs (green curves). https://doi.org/10.1371/journal.pone.0296510.g003 PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 9 / 18 13. Brydone AS, Meek D, Maclaine S. Bone grafting, orthopaedic biomaterials, and the clinical need for bone engineering. Proc Inst Mech Eng H. 2010 Dec; 224(12):1329–43. https://doi.org/10.1243/ 09544119JEIM770 PMID: 21287823. 14. Berthiaume F, Maguire TJ, Yarmush ML. Tissue engineering and regenerative medicine: history, progress, and challenges. Annu Rev Chem Biomol Eng. 2011; 2:403–30. https://doi.org/10.1146/annurevchembioeng-061010-114257 PMID: 22432625. 15. Lanza R, Langer R, Vacanti J P, Atala A. Principles of tissue engineering. Academic Press; 2020. 16. Bose S, Sarkar N, Banerjee D. Effects of PCL, PEG and PLGA polymers on curcumin release from calcium phosphate matrix for in vitro and in vivo bone regeneration. Mater Today Chem. 2018 Jun; 8:110– 120. https://doi.org/10.1016/j.mtchem.2018.03.005 Epub 2018 Apr 14. PMID: 30480167; PMCID: PMC6251318. 17. Bla ´zquez-Carmona P, Sanz-Herrera JA, Martı ´nez-Va ´zquez FJ, Domı ´nguez J, Reina-Romo E. Structural optimization of 3D-printed patient-specific ceramic scaffolds for in vivo bone regeneration in loadbearing defects. J Mech Behav Biomed Mater. 2021 Sep; 121:104613. https://doi.org/10.1016/j.jmbbm. 2021.104613 Epub 2021 Jun 8. PMID: 34126507. 18. Sohn HS, Oh JK. Review of bone graft and bone substitutes with an emphasis on fracture surgeries. Biomater Res. 2019 Mar 14; 23:9. https://doi.org/10.1186/s40824-019-0157-y PMID: 30915231; PMCID: PMC6417250. 19. Castellani C, Zanoni G, Tangl S, van Griensven M, Redl H. Biphasic calcium phosphate ceramics in small bone defects: potential influence of carrier substances and bone marrow on bone regeneration. Clin Oral Implants Res. 2009 Dec; 20(12):1367–74. https://doi.org/10.1111/j.1600-0501.2009.01760.x PMID: 20070742. 20. Yuan H, Fernandes H, Habibovic P, de Boer J, Barradas AM, de Ruiter A, Walsh WR, van Blitterswijk CA, de Bruijn JD. Osteoinductive ceramics as a synthetic alternative to autologous bone grafting. Proc Natl Acad Sci U S A. 2010 Aug 3; 107(31):13614–9. https://doi.org/10.1073/pnas.1003600107 Epub 2010 Jul 19. PMID: 20643969; PMCID: PMC2922269. 21. Cipitria A, Reichert JC, Epari DR, Saifzadeh S, Berner A, Schell H, Mehta M, Schuetz MA, Duda GN, Hutmacher DW. Polycaprolactone scaffold and reduced rhBMP-7 dose for the regeneration of criticalsized defects in sheep tibiae. Biomaterials. 2013 Dec; 34(38):9960–8. https://doi.org/10.1016/j. biomaterials.2013.09.011 Epub 2013 Sep 24. PMID: 24075478. 22. Pobloth AM, Schell H, Petersen A, Beierlein K, Kleber C, Schmidt-Bleek K, Duda GN. Tubular openporous β-tricalcium phosphate polycaprolactone scaffolds as guiding structure for segmental bone defect regeneration in a novel sheep model. J Tissue Eng Regen Med. 2018 Apr; 12(4):897–911. https://doi.org/10.1002/term.2446 Epub 2017 Jun 29. PMID: 28485078. 23. Eastaugh-Waring SJ, Joslin CC, Hardy JR, Cunningham JL. Quantification of fracture healing from radiographs using the maximum callus index. Clin Orthop Relat Res. 2009 Aug; 467(8):1986–91. https://doi.org/10.1007/s11999-009-0775-0 Epub 2009 Mar 13. PMID: 19283438; PMCID: PMC2706350. 24. Fisher JS, Kazam JJ, Fufa D, Bartolotta RJ. Radiologic evaluation of fracture healing. Skeletal Radiol. 2019 Mar; 48(3):349–361. https://doi.org/10.1007/s00256-018-3051-0 Epub 2018 Sep 21. PMID: 30238139. 25. Hammer RR, Hammerby S, Lindholm B. Accuracy of radiologic assessment of tibial shaft fracture union in humans. Clin Orthop Relat Res. 1985 Oct;(199):233–8. PMID: 4042484. 26. Blokhuis TJ, de Bruine JH, Bramer JA, den Boer FC, Bakker FC, Patka P, Haarman HJ, Manoliu RA. The reliability of plain radiography in experimental fracture healing. Skeletal Radiol. 2001 Mar; 30 (3):151–6. https://doi.org/10.1007/s002560000317 PMID: 11357453. 27. Sano H, Uhthoff HK, Backman DS, Yeadon A. Correlation of radiographic measurements with biomechanical test results. Clin Orthop Relat Res. 1999 Nov;(368):271–8. PMID: 10613178. 28. Panjabi MM, Lindsey RW, Walter SD, White AA 3rd. The clinician’s ability to evaluate the strength of healing fractures from plain radiographs. J Orthop Trauma. 1989; 3(1):29–32. https://doi.org/10.1097/ 00005131-198903010-00006 PMID: 2709201. 29. Webb J, Herling G, Gardner T, Kenwright J, Simpson AH. Manual assessment of fracture stiffness. Injury. 1996 Jun; 27(5):319–20. https://doi.org/10.1016/0020-1383(96)00009-5 PMID: 8763283. 30. Claes L, Laule J, Wenger K, Suger G, Liener U, Kinzl L. The influence of stiffness of the fixator on maturation of callus after segmental transport. J Bone Joint Surg Br. 2000 Jan; 82(1):142–8. PMID: 10697331. 31. Hirasawa Y, Takai S, Kim WC, Takenaka N, Yoshino N, Watanabe Y. Biomechanical monitoring of healing bone based on acoustic emission technology. Clin Orthop Relat Res. 2002 Sep;(402):236–44. https://doi.org/10.1097/00003086-200209000-00023 PMID: 12218489. PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 16 / 18 32. Malizos KN, Papachristos AA, Protopappas VC, Fotiadis DI. Transosseous application of low-intensity ultrasound for the enhancement and monitoring of fracture healing process in a sheep osteotomy model. Bone. 2006 Apr; 38(4):530–9. https://doi.org/10.1016/j.bone.2005.10.012 Epub 2005 Dec 20. PMID: 16361120. 33. Mattei L, Di Puccio F, Marchetti S. In vivo impact testing on a lengthened femur with external fixation: a future option for the non-invasive monitoring of fracture healing? J R Soc Interface. 2018 May; 15 (142):20180068. https://doi.org/10.1098/rsif.2018.0068 PMID: 29743272; PMCID: PMC6000173. 34. Mora-Macı ´as J, Reina-Romo E, Domı ´nguez J. Distraction osteogenesis device to estimate the axial stiffness of the callus in Vivo. Med Eng Phys. 2015 Oct; 37(10):969–78. https://doi.org/10.1016/j. medengphy.2015.07.008 Epub 2015 Aug 28. PMID: 26320818. 35. Wee J, Rahman T, Akins RE, Seliktar R, Levine DG, Richardson DW, Dodge GR, Thabet AM, Holmes L, Mackenzie WG. Using distraction forces to drive an autodistractor during limb lengthening. Med Eng Phys. 2011 Oct; 33(8):1001–7. https://doi.org/10.1016/j.medengphy.2011.04.002 Epub 2011 Apr 30. PMID: 21531604. 36. Branco M, Santos-Rocha R, Aguiar L, Vieira F, Veloso A. Kinematic analysis of gait in the second and third trimesters of pregnancy. J Pregnancy. 2013; 2013:718095. https://doi.org/10.1155/2013/718095 Epub 2013 Jan 31. PMID: 23431450; PMCID: PMC3572696. 37. Wentink EC, Prinsen EC, Rietman JS, Veltink PH. Comparison of muscle activity patterns of transfemoral amputees and control subjects during walking. J Neuroeng Rehabil. 2013 Aug 2; 10:87. https:// doi.org/10.1186/1743-0003-10-87 PMID: 23914785; PMCID: PMC3750514. 38. Bla ´zquez-Carmona P, Mora-Macı ´as J, Morgaz J, Ferna ´ndez-Sarmiento JA, Domı ´nguez J, ReinaRomo E. Mechanobiology of Bone Consolidation During Distraction Osteogenesis: Bone Lengthening Vs. Bone Transport. Ann Biomed Eng. 2021 Apr; 49(4):1209–1221. https://doi.org/10.1007/s10439020-02665-z Epub 2020 Oct 27. 39. Mora-Macı ´as J, Reina-Romo E, Morgaz J, Domı ´nguez J. In Vivo Gait Analysis During Bone Transport. Ann Biomed Eng. 2015 Sep; 43(9):2090–100. https://doi.org/10.1007/s10439-015-1262-2 Epub 2015 Feb 4. PMID: 25650097. 40. Seebeck P, Thompson MS, Parwani A, Taylor WR, Schell H, Duda GN. Gait evaluation: a tool to monitor bone healing? Clin Biomech (Bristol, Avon). 2005 Nov; 20(9):883–91. https://doi.org/10.1016/j. clinbiomech.2005.05.010 PMID: 16009475. 41. Xu D, Zhou H, Quan W, Jiang X, Liang M, Li S, Ugbolue UC, Baker JS, Gusztav F, Ma X, Chen L, Gu Y. A new method proposed for realizing human gait pattern recognition: inspirations for the application of sports and clinical gait analysis. Gait Posture. 2023 Oct 27:S0966–6362(23)01474-1. https://doi.org/ 10.1016/j.gaitpost.2023.10.019 PMID: 37926657 42. Xu D, Quan W, Zhou H, Sun D, Baker JS, Gu Y. Explaining the differences of gait patterns between high and low-mileage runners with machine learning. Sci Rep. 2022 Feb 22; 12(1):2981. https://doi.org/ 10.1038/s41598-022-07054-1 PMID: 35194121 43. Ferber R, McClay Davis I, Williams DS 3rd, Laughton C. A comparison of withinand between-day reliability of discrete 3D lower extremity variables in runners. J Orthop Res. 2002 Nov; 20(6):1139–45. https://doi.org/10.1016/S0736-0266(02)00077-3 PMID: 12472220. 44. Fortin C, Nadeau S, Labelle H. Inter-trial and test-retest reliability of kinematic and kinetic gait parameters among subjects with adolescent idiopathic scoliosis. Eur Spine J. 2008 Feb; 17(2):204–16. https:// doi.org/10.1007/s00586-007-0469-9 Epub 2007 Sep 19. PMID: 17891424; PMCID: PMC2365554. 45. Reichert JC, Saifzadeh S, Wullschleger ME, Epari DR, Schu¨tz MA, Duda GN, Schell H, van Griensven M, Redl H, Hutmacher DW. The challenge of establishing preclinical models for segmental bone defect research. Biomaterials. 2009 Apr; 30(12):2149–63. https://doi.org/10.1016/j.biomaterials.2008.12.050 PMID: 19211141 46. Newman E, Turner AS, Wark JD. The potential of sheep for the study of osteopenia: current status and comparison with other animal models. Bone. 1995 Apr; 16(4 Suppl):277S–284S. https://doi.org/10. 1016/8756-3282(95)00026-a PMID: 7626315 47. den Boer FC, Patka P, Bakker FC, Wippermann BW, van Lingen A, Vink G QM, Boshuizen K, Haarman HJ Th M. New segmental long bone defect model in sheep: quantitative analysis of healing with dual energy x-ray absorptiometry. J Orthop Res. 1999 Sep;; 17(5):654–60. https://doi.org/10.1002/jor. 1100170506 PMID: 10569473 48. Baino A, Novajra G, Vitale-Brovarone C. Bioceramics and Scaffolds: A Winning Combination for Tissue Engineering. Front Bioeng Biotechnol. 2015 Dec 17:3:202. https://doi.org/10.3389/fbioe.2015.00202 PMID: 26734605 49. Miranda P, Pajares A, Saiz E, Tomsia AP, Guiberteau F. Mechanical properties of calcium phosphate scaffolds fabricated by robocasting. J Biomed Mater Res A. 2008 Apr; 85(1):218–27. https://doi.org/10. 1002/jbm.a.31587 PMID: 17688280 PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 17 / 18 50. Miranda P, Saiz E, Gryn K, Tomsia AP. Sintering and robocasting of beta-tricalcium phosphate scaffolds for orthopaedic applications. Acta Biomater. 2006 Jul; 2(4):457–66. https://doi.org/10.1016/j. actbio.2006.02.004 PMID: 16723287 51. Indra A, Hamid I, Farenza J, Handra N, Subardi AA. Manufacturing hydroxyapatite scaffold from snapper scales with green phenolic granules as the space holder material. J Mech Behav Biomed Mater. 2022 Dec: 136:105509. https://doi.org/10.1016/j.jmbbm.2022.105509 PMID: 36240527 52. Bla ´zquez-Carmona P, Sanchez-Raya M, Mora-Macı ´as J, Go ´mez-Gala ´n JA, Domı ´nguez J, ReinaRomo E. Real-Time Wireless Platform for In Vivo Monitoring of Bone Regeneration. Sensors (Basel). 2020 Aug 15; 20(16):4591. https://doi.org/10.3390/s20164591 PMID: 32824259; PMCID: PMC7472372. 53. Gerhardt DMJM, Mors TGT, Hannink G, Van Susante JLC. Resurfacing hip arthroplasty better preserves a normal gait pattern at increasing walking speeds compared to total hip arthroplasty. Acta Orthop. 2019 Jun; 90(3):231–236. https://doi.org/10.1080/17453674.2019.1594096 Epub 2019 Apr 1. PMID: 30931667; PMCID: PMC6534262. 54. King AJ, Wilson AM, Wilshin SD, Lowe J, Haddadi H, Hailes S, Morton AJ. Selfish-herd behaviour of sheep under threat. Curr Biol. 2012 Jul 24; 22(14):R561–2. https://doi.org/10.1016/j.cub.2012.05.008 PMID: 22835787. 55. Van Gelder LMA, Angelini L, Buckley EE, MazzàC. A proposal for a linear calculation of gait asymmetry. Symmetry (MPDI). 2021 Aug; 13(9):1560. https://doi.org/10.3390/sym13091560 56. Schober P, Boer C, Schwarte LA. Correlation Coefficients: appropriate use and interpretation. Anesth Analg. 2018 May; 126(5):1763–1768. https://doi.org/10.1213/ANE.0000000000002864 PMID: 29481436 57. Warmerdam E, Orth M, Pohlemann T, Ganse B. Gait Analysis to Monitor Fracture Healing of the Lower Leg. Bioengineering (Basel). 2023 Feb 15; 10(2):255. https://doi.org/10.3390/bioengineering10020255 PMID: 36829749; PMCID: PMC9952799. 58. Fischer S, Anders A, Nolte I, Schilling N. Compensatory load redistribution in walking and trotting dogs with hind limb lameness. Vet J. 2013 Sep; 197(3):746–52. https://doi.org/10.1016/j.tvjl.2013.04.009 Epub 2013 May 15. PMID: 23683534. 59. Plotnik M, Bartsch RP, Zeev A, Giladi N, Hausdorff JM. Effects of walking speed on asymmetry and bilateral coordination of gait. Gait Posture. 2013 Sep; 38(4):864–9. https://doi.org/10.1016/j.gaitpost. 2013.04.011 Epub 2013 May 13. PMID: 23680424; PMCID: PMC4047486. 60. Larsen P, Laessoe U, Rasmussen S, Graven-Nielsen T, Berre Eriksen C, Elsoe R. Asymmetry in gait pattern following tibial shaft fractures—a prospective one-year follow-up study of 49 patients. Gait Posture. 2017 Jan; 51:47–51. https://doi.org/10.1016/j.gaitpost.2016.09.027 Epub 2016 Sep 28. PMID: 27701034. 61. Patterson KK, Nadkarni NK, Black SE, McIlroy WE. Gait symmetry and velocity differ in their relationship to age. Gait Posture. 2012 Apr; 35(4):590–4. https://doi.org/10.1016/j.gaitpost.2011.11.030 Epub 2012 Jan 31. PMID: 22300728; PMCID: PMC3914537. 62. Jansen H, Fenwick A, Doht S, Frey S, Meffert R. Clinical outcome and changes in gait pattern after pilon fractures. Int Orthop. 2013 Jan; 37(1):51–8. https://doi.org/10.1007/s00264-012-1716-1 Epub 2012 Dec 11. PMID: 23229797; PMCID: PMC3532654. PLOS ONE Gait analysis predicts tissue engineering regeneration PLOS ONE | https://doi.org/10.1371/journal.pone.0296510 December 29, 2023 18 / 18