Thông tin tài liệu
DIVISION OF ORTHOPEDIC SURGERY UNIVERSITY OF OTTAWA H K UHTHOFF ANNUAL RESEARCH DAY THURSDAY, APRIL 26, 2012 THE OTTAWA HOSPITAL – GENERAL CAMPUS MAIN AUDITORIUM VISITING PROFESSOR Dr Yoga Raja Rampersaud, MD FRCSC Associate Professor, Department of Surgery, University of Toronto Divisions of Orthopedic Surgery and Neurosurgery, University Health Network Medical Director, Back and Neck Specialty Program, Altum Health Immediate Past President Canadian Spine Society We are pleased to welcome Yoga Raja Rampersaud, MD, FRCSC, as the 2012 H K Uhthoff Visiting Professor South America-born Dr Rampersaud graduated in 1992 with his honours Medical Degree from the University of Western Ontario (UWO) in London, Ontario and completed Orthopedic specialty training at UWO in 1997 Two fellowships followed – in Orthopedic Spine Surgery at UWO in 1997 and in Neurosurgical Spine Surgery in 1998 at University of Tennessee, Memphis Currently an Associate Professor in Surgery at University of Toronto, Dr Rampersaud joined University Health Network’s Division of Orthopedic Surgery and Neurosurgery as a consultant in 1999 Dr Rampersaud is the immediate past president of the Canadian Spine Society His academic interests are in minimally invasive spine surgery, surgical safety and outcomes from a health services perspective RESIDENT RESEARCH REQUIREMENTS THE DIVISION OF ORTHOPEDIC SURGERY UNIVERSITY OF OTTAWA All residents must participate in a minimum of two research projects during their residency Research plan and protocol is presented to the Research Visiting Professor in November Preliminary results are presented to the Division of Orthopedic Surgery Research Committee in early April The final paper is presented at the H.K Uhthoff Research Day in April Papers are chosen for submission to Collins Day in May Two completed manuscripts must be written in style of the Journal of Bone and Joint Surgery and submitted to the Chairman of the Resident Research Committee, one by the end of the PGY-3 year and one by the end of the PGY-4 year RESIDENTS/FELLOWS DIVISION OF ORTHOPEDIC SURGERY 2011 - 2012 PGY-5 PGY-2 Dr Parham Daneshvar Dr Gillian Bayley Dr Luke Gauthier Dr Kelly Hynes Dr Krista Goulding Dr Matthew MacEwan Dr Gregory Hansen Dr Ian MacNiven Dr Maher Khan Dr Travis Marion Dr William Weiss Dr Scott McGuffin PGY-4 PGY-1 Dr Derek Butterwick Christopher Dowding Dr Sasha Carsen Dr Heathcliff D’Sa Dr Michael Creech Dr Adrian Huang Dr William Desloges Dr Brian Le Dr Markian Pahuta Dr Andrew Tice Dr Marc Prud’homme-Foster Dr Kristi Wood Dr Nathan Sacevich Dr PGY-3 FELLOWS Dr Natasha Holder Dr Francesco Blumetti Dr Bradley Meulenkamp Dr Vikram Chatrath Dr Marie-France Rancourt Dr Emmanuel Illical Dr Cai Wadden Dr Milton Parai Dr Geoffrey Wilkin Dr Hani Zamil DIVISION OF ORTHOPEDIC SURGERY RESIDENCY TRAINING COMMITTEE 2011 – 2012 CHAIRMAN Dr Joel Werier MEMBERS Dr Ben Bessette Dr Wade Gofton Dr James Jarvis Dr Ken Kontio Dr Karl-André Lalonde Dr Louis Lawton Dr Allan Liew Dr Peter Thurston RESIDENT REPRESENTATIVES Dr Parham Daneshvar Dr Geoffrey Wilkin DIRECTOR OF RESEARCH Dr Peter Lapner ACKNOWLEDGEMENTS The Division of Orthopedic Surgery greatly acknowledges the support of the H K Uhthoff Research Day by the following companies: Baxter Corporation Bayer Inc Biomet Canada Bristol-Myers Squibb Canada ConMed Linvatec Convatec KCI Medical Canada Inc Kinemedics Medtronic of Canada Ltd Stryker Canada Synthes Canada Tribe Medical Group Wright Medical (Mr Trevor Fisher) PROGRAM 0800 Opening Remarks Dr Joel Werier, Director of the Orthopedic Surgery Residency Training Program, University of Ottawa 0805 Welcome/Introduction of Dr Rampersaud Dr Eugene K Wai, Department of Orthopedic Surgery, University of Ottawa 0810 Comparative Effectiveness of the Surgical Management of Focal Spinal Stenosis Compared to Hip and Knee Osteoarthritis Dr Yoga Raja Rampersaud, Visiting Professor 0830 Discussion SESSION I MODERATOR: Dr Wade Gofton 0840 0848 Quantitative CT and MRI Changes in Arthritic and Prearthritic Hips Andrew Speirs, Graduate Student, University of Ottawa Discussion 0852 Can the Alpha Angle Assessment of Cam Impingement Predict Acetabular Cartilage Delamination? Dr Kelly Hynes, PGY-2 0900 Discussion 0904 0912 0916 0924 A Randomized Controlled Trial of a Cemented vs Cementless Femoral Component for Metal-on-Metal Hip Resurfacing: A Bone Mineral Density Study Dr Andrew Tice, PGY-1 Discussion Stress Distributions in the Hip Before and After Corrective FAI Surgery K.C Geoffrey Ng, Graduate Student, Department of Mechanical Engineering, University of Ottawa Discussion 0928 Open Reduction and Internal Fixation of Pilon Fractures: Violating the cm Skin Bridge Rule Dr Geoffrey Wilkin, PGY-3 0936 0940 0948 Discussion Vacuum Assisted Closure (V.A.C TM) Effects on Skeletal Muscle After Compartment Syndrome in an Animal Model Dr Geoffrey Wilkin, PGY-3 Discussion Refreshment Break and Exhibits, Royal Room SESSION II MODERATOR: Dr Karl-André Lalonde 1041 Arthroscopic Acetabular Labral Debridement in Patients Aged >45 Years has Minimal Clinical Benefit Dr Geoffrey Wilkin, PGY-3 1049 Discussion 1053 The Ottawa Experience Using a Modular Neck System for Primary Total Hip Arthroplasty Dr Emmanuel Illical, Clinical Fellow 1061 Discussion 1105 Radiographic Outcomes of Closed Diaphyseal Femur Fractures Treated with the SIGN Nail Dr Sasha Carsen, PGY-4 1113 1117 Discussion 10 Outcome Comparison of Revised Hip Resurfacing with Primary and Revised Total Hip Arthroplasties Dr William Desloges, PGY-4 1125 1129 1137 Discussion 11 Ulnar Placement of the Distal Biceps Tendon During Repair Improves Supination Strength: A Biomechanical Analysis Dr Marc Prud’homme-Foster, PGY-4 Discussion 1141 12 Comparing the Extensor Digitorum Communis Splitting Approach to the Posterolateral Kocher Approach: A Novel Method of Measuring Articular Surface Area Dr William Desloges, PGY-4 1149 Discussion 1153 13 Lesser Tuberosity Osteotomy versus Subscapularis Tendon Peel: Differences in Healing Rates and Fatty Infiltration Dr Milton Parai, Clinical Fellow 1201 Discussion Lunch and Exhibits, Royal Room SESSION III MODERATOR: Dr J Pollock 1315 In Experimental Surgery Is The Use of the Contralateral Limb for Comparison Acceptable? Dr Hans K Uhthoff 1323 Discussion 1327 Systematic approach to Orthopedic Surgical Adverse Events (OrthoSAVES): A Health Services Perspective Dr Yoga Raja Rampersaud, Visiting Professor 1347 Discussion 1357 14 Early High Failure Rate of Large Head Metal-on-Metal Total Hip Replacement Dr Vikram Chatrath, Clinical Fellow 1405 1409 1417 Discussion 15 Percutaneously Assisted Total Hip (PATH) Arthroplasty: Learning Curve and Early Results in a Canadian Centre Dr Cai Wadden, PGY-3 Discussion 1421 16 Does Weight-bearing Status Following Surgery for Hip Fracture Affect Outcomes? Dr Cai Wadden, PGY-3 1429 Discussion 1433 17 Effects of Autogenous Hamstring Tendon Harvesting for Anterior Cruciate Ligament Reconstruction on Hamstring Strength at Deeper Knee Flexion Angles Does Single Versus Double Tendon Sacrifice Affect Hamstring Strength? Dr Parham Daneshvar, PGY-4 1441 Discussion Refreshment Break, Royal Room ANTERIOR CRUCIATE LIGAMENT RECONSTRUCTION ON HAMSTRING STRENGTH AT DEEPER KNEE FLEXION ANGLES DOES SINGLE VERSUS DOUBLE TENDON SACRIFICE AFFECT HAMSTRING STRENGTH? Parham Daneshvar, Krista Goulding, Eugene K Wai, Donald Johnson Purpose The purpose of this study was to evaluate hamstring strength after autogenous hamstring anterior cruciate ligament(ACL) reconstruction with emphasis on deeper knee flexion angles A comparison of hamstring strength between patients undergoing ACL reconstruction using semitendinosus(ST), and those with semitendinosus and Gracilis(STG) tendons was conducted Methods Fifty-Six patients were prospectively followed after undergoing ACL reconstruction surgery Forty-two patients received a ST graft, and Fourteen patients had a STG graft All patients had standard IKDC subjective knee evaluation completed, and had bilateral hamstring strengths tested using isokinetic testing with a Cybex Orthotron machine In addition, a hand held Microfet dynamometer was used to measure hamstring strengths at deeper knee flexion angles at six, 9, 12, and 24 months Results When comparing the surgical hamstring strength compared to the control side at deeper knee flexion angle, the STG and ST group had and average of 54.6% and 73.2% strength respectively at months(p=0.001) While at one year the STG and ST group had 64.3%, and 73.0% of their hamstring strength(p=0.046) Conclusion There is a trend towards weaker hamstring strength in deeper flexion angles after ACL reconstruction using autogenous semitendinosus and gracilis tendons than using semitendinosus alone The use of single tendon ACL reconstruction should be advocated to decrease morbidity of this procedure 18 QUANTIFYING ACETABULAR OVERCOVERAGE ← N Sacevich (first author), A Speirs, P Beaulé ← Division of Orthopedics ← University of Ottawa ← Purpose Femoroacetabular impingement (FAI) involves abnormal contact between the proximal femur and acetabulum during physiologic hip motion and has been implicated in the pathogenesis of early hip osteoarthritis Pincer type impingement can result from generalized acetabular over-coverage An accurate means of quantifying the severity of over-coverage on axial imaging has not been defined Moreover there is a lack of literature defining the relationship between the current radiographic standard used to quantify acetabular coverage on plain x-rays with axial plane imaging The purpose of this study is to further establish normative data for horizontal acetabular sector angles (HASA) as measured on axial imaging, and to evaluate correlations between HASA, center edge angle (CEA), and acetabular depth according to a three-dimensional model Methods 48 hips without pre-existing hip disease underwent pelvic CT scans according to standardized research protocol CEA was measured on CT scout images and HASA measured on axial images according methods previously described 3D measurement of acetabular depth was performed Pearson correlations were calculated for the various measurements Results Moderate correlations were seen for HASA to 3D depth (r 0.42, p 0.009), CEA to 3D depth (r 0.41, p 0.001), and HASA to CEA (r 0.39, p 0.006) The mean HASA was 151.8o(SD 7.3) and mean CEA was 35.0o (SD 5.3) Conclusion Both CEA and HASA may represent accurate means of quantifying global acetabular coverage Defining normative dimensions of the acetabulum are essential for the diagnosis pincer type FAI, and in preoperative planning for accurate deformity correction 19 TIBIAL NAIL DISTAL POSITIONING: A RADIOGRAPHIC STUDY T Marion, S Papp, W Gofton, A Liew Intramedullary fixation is the treatment of choice for diaphyseal tibia fractures The importance of the proper insertion point is well established Intramedullary fixation of distal tibial fractures relies upon the placement of the guidewire distally to achieve and maintain an acceptable reduction once the intramedullary nail has been inserted Inappropriate distal positioning of the guide wire may contribute to malalignment when the nail is inserted, leading to malunion, ankle joint dysfunction, and early arthritic changes The ideal distal position of the guidewire in the distal tibia has yet to be well defined Fifty intact tibial radiographs were selected and evaluated The centre of the medullary canal at the level of the isthmus was established and extended inferiorly through the ankle joint on antero-posterior (AP) and lateral radiographs Transverse lines of reference on the AP and lateral views were established The position in which the isthmic line intersected these reference lines was measured and expressed as a percentage from medial to lateral on the AP view and anterior to posterior on the lateral view The tibial radiographs were manually templated with sized tibial nails The position in which the centre of the implant template intersected the previously established reference lines was measured and once again expressed as a percentage Non Templated AP Radiograph Lateral Radiograph Reference Line Maximal Metaphyseal Width Tibial Articular Width Talar Width Reference Line Maximal Metaphyseal Width Tibial Articular Width Templated AP Radiograph Lateral Radiograph Reference Line Maximal Metaphyseal Width Tibial Articular Width Talar Width Reference Line Maximal Metaphyseal Width Tibial Articular Width Percent From Medial Cortex 59.8 ± 4.1% 63.5 ± 7.6% 60.3 ± 7.6% Percent From Anterior Cortex 55.6 ± 5.1% 60.7 ± 9.0% Percent From Medial Cortex 60.1 ± 3.7% 66.2 ± 6.1% 61.7 ± 7.6% Percent From Anterior Cortex 55.7 ± 4.4% 58.5 ± 8.4% The ideal placement of the guide wire in the distal tibia for intramedullary fixation is not well defined Traditional teaching calls for placement of the guide wire endpoint in the ``centre`` of the ankle joint or distal tibia Our findings demonstrate that the ideal endpoint for an intramedullary nail is lateral and central to the centre of the maximum metaphyseal width reference line on AP and lateral view respectively In treating distal third tibial fractures, one cannot rely on the isthmus to guide the intramedullary nail Furthermore positioning of the nail in the ``centre`` on AP may shift the axial position of the talus and contribute to a malreduction, especially in the setting of distal comminution This study supports further investigations to determine the effect of distal tibial nail malposition and its clinical significance 20 NORMAL ANATOMY OF THE DISTAL RADIO-ULNAR JOINT ANATOMY – A BAYESIAN ANALYSIS OF 1,000 WRISTS M.A Pahuta A.A Giachino A.I McIntyre Division of Orthopedic Surgery, The Ottawa Hospital Purpose Three distal radio-ulnar joint (DRUJ) patterns have been described: Type I ulna-neutral with neutral sigmoid notch inclination; Type II ulna-negative with proximally facing sigmoid notch; and Type III ulna-positive with distally facing sigmoid notch Small, biased studies have concluded that the Type III pattern is normal The objective of this study was to collect a representative sample of patients and define normal anatomy For this study, we wished to know “believable” rather than “rejectable” values, and therefore elected to use Bayesian analysis Methods One-thousand consecutiveposteroanterior wrist radiographs, from skeletally mature patients were obtained from the TOH film library Wrists with evidence of previous trauma or malformation were excluded Ulnar variance and sigmoid notch inclination were measured using the “project-a-line” and modified Ekenstamtechniques respectively Bayesian analysis was performed using Markov chain Monte Carlo simulation using Gibbs sampling with OpenBugs Moderately informed prior distributions were developed form the existing data on DRUJ anatomy Posterior distributions and 95% highest density intervals were computed for ulnar variance and sigmoid notch inclination Bayesian ordinal probit regression was used to develop a prediction model for sigmoid notch orientation by ulnar variance Results Based on our analysis, the mean ulnar variance for the population lies between -0.603 and -0.433mm; with a standard deviation of 1.1mm A distally facing sigmoid notch is uncommon, with a predicted prevalence of 20% at neutral ulnar variance, and 60% at 3mm positive ulnar variance Conclusion We have demonstrated that less than Type III DRUJ anatomy is uncommon, and therefore conclude that it is abnormal Furthermore, we have demonstrated a strong association between increasing ulnar variance and a distally facing sigmoid notch This supports a trophic relationship between ulnar variance and DRUJ anatomy A distally facing sigmoid notch is a frequently overlooked cause of radio-ulnar impingement post ulnar shortening procedures As these patients are more likely to have a distally facing sigmoid notch, we suggest that surgeons closely examine sigmoid notch anatomy prior to ulnar shortening, and consider a concurrent sigmoid notch osteoplasty 21 PATELLA TENDON INSERTION FAILURE: EVALUATION OF A RECONSTRUCTIVE TECHNIQUE Zamil, Hani, Dervin, Geoffrey Introduction Rupture of patellar tendon insertion is a devastating complication (< 1%) in primary & revision TKAs & it leads to delayed rehabilitation & inferior outcomes Objective To compare classic primary repair to augmented repair (either double looped wire or free semitendinosus graft) in a cadaveric model regarding load to failure Second, to report early clinical experience in a series of patients Methods – Surgery Five pairs of adult knees were used where patellar tendon was completely detached from its insertion Control gp: Classic primary repair using suture anchors & was augmented by #18 gauge wire double looped through the insertion area below the anchors & proximally in the quadriceps tendon just proximal to the patella Experimental gp: Primary repair using suture anchors was augmented by free semitendinosus graft which was passed through a transverse drill hole in the inferior third of the patella & through a nd drill hole through the tibial tubercle at midpoint of tendon insertion The free ends of the sutures were sewn next to the repaired tendon & a Bioscrew was inserted to fix the two ends within the transverse tibial tunnel Results – load to failure ( n=5 specimens) Control Wire and Semi tendinosus and suture suture anchor(N) anchor (N) Mean ± SD Mean ± SD Difference control – Recessed tendon (N) 469.3 ±212.4 465.71.1 P = 0.015 935±178.4 Clinical Series Five patients were treated: (2 for intraop disruption with (1 revision TKR, primary TKR hemophilia), primary case augmented repair, for associated patella baja with arthroplasty An augmented repair as described and post op – passive ROM to 90 degrees allowed with quads setting and electrical stimulation and PWB with extension splint for weeks At final F/U range 2-8 years, there were no rerupture or reoperations; cases – no extensor lag and cases had a 10 degree extensor lag Discussion and Conclusion The goal for this injury would be to ideally allow full range of motion to prevent risk of arthrofibrosis which would compromise long term result and perhaps require nd procedure for manipulation Reinforcement with semitendinosus autograft is sufficiently strong for passive mobilization and clinically produced satisfactory long term results in small clinical series 22 STIMULATION OF MACROPHAGES BY CHROMIUM (III) IONS INDUCES AN INCREASE OF TRAP EXPRESSION IN VITRO Stephen J Baskeya, Eric A Lehouxa, and Isabelle Catelasabc a Department of Mechanical Engineering; b Department of Surgery; c Department of Biochemistry, Microbiology and Immunology; University of Ottawa, Ontario, Canada Purpose The long-term effects of metal ions from metal-on-metal bearings remain a cause for concern, and little is known about their effects on osteoclastogenesis, a process commonly demonstrated by the expression of tartrate-resistant acid phosphatase (TRAP) Receptor activator of nuclear factor kappaB ligand (RANKL) is an essential mediator of wear-particleinduced osteoclastogenesis Therefore, the purpose of the present study was to analyze the effects of macrophage stimulation by trivalent chromium ions (Cr3+) on the expression of TRAP, in the presence or absence of RANKL Methods RAW 264.7 macrophages were stimulated with 50 ppm of Cr3+ and/or 20 ng/ml of RANKL for days Unstimulated macrophages served as negative controls TRAP expression was visualized through cell staining using a commercial kit Separately, TRAP activity in cell lysates was assayed spectrophotometrically, and normalized on nuclei number determined fluorometrically Results Cell staining indicated increased TRAP expression in macrophages stimulated with Cr3+, RANKL, and Cr3+ and RANKL together, relative to unstimulated (negative) controls The enzymatic assay of cell lysates also showed increased TRAP activity in macrophages stimulated with Cr 3+ (about fold), RANKL (about 2.5 fold), and Cr3+ and RANKL together (about fold) Conclusion Results showed that macrophage stimulation by Cr 3+ ions induced an increase of TRAP expression, which was enhanced by the presence of RANKL This increase suggests osteoclastogenesis activation However, macrophage differentiation into mature osteoclasts remains to be tested by the presence of multinucleated cells capable of inducing bone resorption Overall, the results suggest that Cr3+ ions may activate a transcription factor functioning downstream of RANKL 23 LEVEL AND UPSLOPE WALKING AFTER TKA: BIOMECHANICAL IMPLICATIONS FOR THE NONOPERATED KNEE Sarah Reynolds, Graduate Student, University of Ottawa Total knee arthroplasty (TKA) has proven to be a highly successful method of treatment of severe knee osteoarthritis (OA), with high rates of patient satisfaction, reduced pain, and improved knee function1 Despite this success, clinical studies show that patients not achieve normal knee kinetics and kinematics2 Recently, a medial pivot (MP) design has been proposed to better mimic normal knee mechanics This design is anticipated to reduce polyethylene wear and allow for anteriorposterior translation and rotation3 However, few studies have inquired into whether the theoretical advantages of the MP knee translate to improved clinical results It is important to understand how TKA influences gait patterns, as an abnormal gait can accelerate damage to the implants and predispose patients to further lower extremity joint damage This type of analysis should be focused on the operated and non-operated limbs of patients, as several studies 5,6 have reported a significant correlation between the strength of the non-operated leg and functional performance following TKA In addition, some TKA studies5,7 have reported that patients experience weakening of the non-operated limb following unilateral TKA, as well as exhibiting differences in knee adduction angle and adduction moment between the operated and non-operated limbs after surgery, possibly a result of pre-operative compensation patterns This type of asymmetrical gait mechanics can predispose the non-operated knee to mechanical overload and contribute to OA progression7 Previous gait studies for TKA patients have been limited to level surfaces; however, we propose to examine both level and inclined gait as this is more realistic of the environment many patients experience on a daily basis It is also assumed that inclined walking is a more demanding task that may provide insight into how TKA patients compare to their healthy counterparts during more effortful* tasks The aim of this study was to explore differences in gait patterns of patients with a medial pivot knee prosthetic and that of healthy controls, to understand the extent to which both the operated and nonoperated limbs compare to a healthy limb during both level and inclined walking tasks Seven patients (age: 62 ± 7, BMI: 31.1 ± 2.7 kg/m 2) were recruited for post-operative analysis one year after receiving unilateral total knee arthroplasty with a medial pivot knee prosthetic Normative data was obtained from seven healthy control subjects (age: 57± 12, BMI: 26.8 ± 2.9 kg/m 2) that were recruited from the local community Subjects were recruited if they were between 50-75 years of age, had a BMI below 35 kg/m2, and if they were not suffering from any other lower limb joint disorders Approval was obtained from the hospital and university’s research ethics board, and informed consent was provided by all subjects Three-dimensional (3D) kinematics and kinetics of the lower limbs were measured during inclined walking using a ten-camera motion analysis system (Vicon MX, Oxford Metrics, UK) Subjects were outfitted with 45 reflective markers, placed according to a modified Helen Hayes model Two force plates were embedded within a 4m steel ramp, inclined at a 12.5% slope, during dynamic trials Level walking trials were performed along a flat walkway with forces plates embedded within the floor Knee kinetics were obtained with inverse dynamics During both level and inclined walking, the operated limb of TKA patients exhibited kinetic and kinematic values within the ‘normative range’ of control subjects Generally, this would be indicative of a successful surgery, allowing patients to experience similar gait patterns as their healthy counterparts However, further analysis of gait patterns at the non-operated limb indicate that certain functional deficiencies remain Compared to both the operated limb and control participants, the non-operated limb of TKA patients exhibited decreased hip power generation during early stance for level walking, as well as a greater peak knee adduction moment during both level and inclined walking During inclined walking, TKA patients exhibited greater hip flexion moment, knee extension moment, and knee power absorption during early stance, followed by increased knee power generation in late stance These gait patterns are consistent with findings from previous studies that have noted increased loading at the non-operated knee of TKA patients following surgery With implications for OA progression at other joints of the lower limb, future research and rehabilitation should focus on both the operated and non-operated limb to measure functional outcomes and success following TKA References Hatfield GL et al 2011 J Arthroplasty, 26(2), 309-318 Yoshida Y et al 2008 Clin Biomech, 23(3), 320-328 Bae DK et al 2010 J Arthroplasty, 26(5), 693-698 Milner CE 2009 J Orthop Sci, 14, 114-120 Farquhar, S & Snyder-Mackler, L 2010 Clin Orthop Relat Res, 468:37–44 Zeni Jr., J.A & Snyder-Mackler, L 2010 Phys Ther, 90:4354 Alnahdi, Zeni, and Snyder-Mackler 2011 J Ortho Res, 29, 647-652
Ngày đăng: 20/10/2022, 01:12
Xem thêm:
