Browsing by Author "McKinley, Todd"
Now showing 1 - 10 of 15
Results Per Page
Sort Options
Item A Reproducible Cartilage Impact Model to Generate Post-Traumatic Osteoarthritis in the Rabbit(MyJove Corporation, 2023-11-21) Dilley, Julian; Noori-Dokht, Hessam; Seetharam, Abhijit; Bello, Margaret; Nanavaty, Aaron; Natoli, Roman M.; McKinley, Todd; Bault, Zachary; Wagner, Diane; Sankar, Uma; Anatomy, Cell Biology and Physiology, School of MedicinePost-traumatic osteoarthritis (PTOA) is responsible for 12% of all osteoarthritis cases in the United States. PTOA can be initiated by a single traumatic event, such as a high-impact load acting on articular cartilage, or by joint instability, as occurs with anterior cruciate ligament rupture. There are no effective therapeutics to prevent PTOA currently. Developing a reliable animal model of PTOA is necessary to better understand the mechanisms by which cartilage damage proceeds and to investigate novel treatment strategies to alleviate or prevent the progression of PTOA. This protocol describes an open, drop tower-based rabbit femoral condyle impact model to induce cartilage damage. This model delivered peak loads of 579.1 ± 71.1 N, and peak stresses of 81.9 ± 10.1 MPa with a time-to-peak load of 2.4 ± 0.5 ms. Articular cartilage from impacted medial femoral condyles (MFCs) had higher rates of apoptotic cells (p = 0.0058) and possessed higher Osteoarthritis Research Society International (OARSI) scores of 3.38 ± 1.43 compared to the non-impacted contralateral MFCs (0.56 ± 0.42), and other cartilage surfaces of the impacted knee (p < 0.0001). No differences in OARSI scores were detected among the non-impacted articular surfaces (p > 0.05).Item Award Recognition for "Developing a Minimally Invasive Cell-Based Model to Predict Response to Major Trauma"(2021-08) Gates, Kayla L; Nunge, Rebecca A; Adom, Jamila; Kaplan, Mark; McKinley, ToddBackground: The physiologic response to injury is heavily influenced by the immune system. The complexities of the immunologic response to injury are becoming increasingly understood as researchers have leveraged computational methods that allow temporal and spatial coordination of immune mediator orchestration to be quantified. Recently, early differences in immunologic orchestration have been shown to stratify individual tolerance to injury. Specifically, there are subsets of trauma patients that are either sensitive or tolerant to hemorrhage that demonstrated notably different early immunologic orchestration of mediators from clusters of cytokines that are primarily tissue protective or pro/anti-inflammatory. These differentiating networks of mediators formed and dissipated over the initial 72 hours after injury clearly demonstrating that the immunologic response to injury is an acute dynamic event that has pathomechanistic relevance to outcomes after injury. Additionally, it is distinctly possible that individualized differences in immune response may determine tolerance/sensitivity to injury. The differential immunologic response to trauma represents an opportunity to discover specific factors that may be predictive of a patients’ response to traumatic injury and subsequent hemorrhagic shock. Accordingly, we have embarked on a line of experimentation to explore potential precision approaches based on individual immunologic response to injury. Here we report our initial experimental findings in conceptual model development with the ultimate goal of developing minimally invasive/non-injurious testing that will accurately identify individual tolerance to hemorrhage and injury. In this experiment, an in-vitro cell-based assay was designed to mimic traumatic injury. Specifically, we tested the immunologic response in murine splenocytes to a simulated hypoxic injury, a simulated mechanical injury and a simulated open wounding injury. The development of a reliable cell-based model will allow investigation to determine correspondence and relevance between cell-based responses to non-traumatic injury and in vivo immunologic response to trauma, with the overall goal of developing a reliable test to predict response to traumatic injury in humans. Methods: In-vitro cellular responses of murine splenocytes are reflective of peripheral blood cell responses and were used for pilot experiments. Splenocytes from C57BL/6 (B6) BALB/c or CH3/ HeJ strains mice were used with stimuli that mimic traumatic injury using chemical (hypoxia or sepsis) or mechanical (shear stress) stimuli that might from an open wounding type of injury. Hypoxia was simulated by subjecting cell cultures to hydrogen peroxide. Sepsis was simulated by subjecting cells to lipopolysaccharide (LPS). Some culture conditions included several individual cytokines associated with acute inflammation and external pathogens (interleukin (IL)-6, IL-1β, IL- 33), the damage molecule high mobility group box protein (HMGB)-1, or combinations of LPS and the cytokines. Following treatment, cDNA was prepared and used for qPCR amplification of TNFα, HIF1α, and BAX to assess inflammation, hypoxia, and apoptosis, respectively. Multiplex analysis of IL-21, IL-4, IL-22, IL-5, and IL-10 expression was performed from culture supernatant collected at 24 hours after stimulation. Flow cytometry was performed to assess proliferation of immune cells following treatment. ELISA was conducted to quantify production of the cytokine IL-9 that occurred following splenocyte stimulation. Results: Analysis of C57BL/6 splenocyte viability show that any combination of cytokines or LPS did not impact cell survival, while hydrogen peroxide reduced survival significantly in each treatment group. From the qPCR data, LPS generated a 4x increase in TNFα expression relative to control, while cytokine treatment yielded no expression changes. Treatment with LPS + cytokines closely resembled the LPS treatment group. LPS treatment reduced expression of HIF1α, while hydrogen peroxide increased expression of HIF1α. The addition of cytokines reduced expression of HIF1α in groups that were treated with both hydrogen peroxide and cytokines. ELISA analysis of the proinflammatory cytokine IL-9 indicated increased production of IL-9 following treatment with LPS + cytokines. In the second experiment, the model was applied to three different strains of mice in order to gauge differences in the immune response to the same cellular stress. Multiplex analysis showed no significant changes in IL-4 or IL-21 expression in any of the strains. C3H mice showed no response to LPS, which was expected due to LPS resistance in these strains. In the B6 and BALB mice, IL-10 was induced by LPS treatment. BALB mice also showed increased expression of IL-5 and IL-22 in response to mechanical stress, while the other strains showed no response. IL-10 expression was not induced by mechanical injury in any strain. Flow cytometry analysis was used to assess immune cell response to stimuli. Both B6 and C3H mice showed increased percentages of CD4 and CD8 cells in response to mechanical stimulus, LPS, and LPS + cytokine treatment relative to control. Macrophage levels were more elevated in B6 mice in response to mechanical stimulus, whereas levels decreased in the C3H mice. Discussion: The overall goal of this line of investigation is to develop minimally invasive and non-injurious testing that can be used to determine individual tolerance/sensitivity to trauma and hemorrhage. These pilot studies were used to determine how immune cells can be isolated and stimulated to mimic injury. Splenocytes were used as they encompass a broad cross-section of white blood cells. Clear inter-strain differences were evident between the B6, BALB and C3H mice. Hypoxia stimuli consistently resulted in roughly a 50% loss of cell viability and accordingly may not be a viable strategy. The greatest effects were encountered with LPS +/- addition of stimulating cytokines. We measured changes in five of six cytokines in B6 mice and four of six BALB mice involving reparative cytokines (IL-21 and 22), anti-inflammatory cytokines (IL-10) and in type 2 cytokines (IL-4 and IL-5). Accordingly, these strains and stimulation methods will be expanded to determine effects on production on a broader panel of cytokines. In addition, computational methods will be leveraged on the next experiment to determine in-vitro effects on immunologic mediator orchestration to account for time-dependent mediator networks and spatial networks of mediators. Moving forward, these experiments will be repeated to reproduce our findings and improve our ability to distinguish between varying immune responses. Results will then be paired with studies examining the responses to traumatic injury among these and other strains. The overall goal of this project is to accurately predict the response to an in-vivo injury using an in-vitro non traumatic stimulus. Findings from this project will enable the development of a clinical test that accurately predicts immunologic response to trauma and stratify individual tolerance to hemorrhage and injury.Item Developing a Minimally Invasive Cell-Based Model to Predict Response to Major Trauma(2020-07-31) Nunge, Rebecca A; Gates, Kayla L; Adom, Jamila; Kaplan, Mark; McKinley, ToddBackground. Physical trauma results in a systemic inflammatory response. Preliminary research in orthopedic trauma patients suggests that patients with similar demographics and severity of injury vary in their response to traumatic injury. Analysis of the immunological response post-injury showed a sustained pro-inflammatory response with delayed reparative cytokine expression in trauma sensitive patients, while the trauma tolerant patients had an early inflammatory expression with resolution by 72 hours post-injury. Thus, we hypothesize that differential response to non-traumatic injury might serve as a predictive tool for the identification of trauma tolerant and sensitive patients prior to injury. The goal of this research is to test whether immunological changes to inflammatory stimuli can predict tolerance or sensitivity to trauma using an-vitro cell-based assay. Methods. Splenocytes were isolated from naive C57BL/6 mice and subjected to biological trauma in vitro using LPS (100 ng/mL) or hypoxic trauma using hydrogen peroxide (50 µM, 100 µM, and 200 µM) with or without proinflammatory cytokines, IL-1β (1 ng/mL) , IL-6 (200 ng/mL), and IL-33 (150 ng/mL). Inflammation and hypoxia were assessed using IL-6 and HIF-1ɑ expression respectively via qPCR 24 hours post-treatment. Cell death and pro-inflammatory cytokine production using multiplex analysis were used to measure outcomes. Results. Both types of treatments showed increased cell death compared to the control group. qPCR data is pending. Conclusion. With these studies as a core of the experimental approach, this in vitro cell-based assay will be used to assess immunologic response to inflammatory stimuli across the genetic variation of mouse strains. Findings from this project could enable the development of a clinical test that accurately predicts immunologic response to trauma and related-complications based on patients’ sensitivity to pre-traumatic injury.Item Developing a Minimally Invasive Cell-Based Model to Predict Response to Major Trauma(2020-07-31) Nunge, Rebecca A; Gates, Kayla L; Adom, Jamila; McKinley, ToddBackground. Physical trauma results in a systemic inflammatory response. Preliminary research in orthopedic trauma patients suggests that patients with similar demographics and severity of injury vary in their response to traumatic injury. Analysis of the immunological response post-injury showed a sustained pro-inflammatory response with delayed reparative cytokine expression in trauma sensitive patients, while the trauma tolerant patients had an early inflammatory expression with resolution by 72 hours post-injury. Thus, we hypothesize that differential response to non-traumatic injury might serve as a predictive tool for the identification of trauma tolerant and sensitive patients prior to injury. The goal of this research is to test whether immunological changes to inflammatory stimuli can predict tolerance or sensitivity to trauma using an-vitro cell-based assay. Methods. Splenocytes were isolated from naive C57BL/6 mice and subjected to biological trauma in vitro using LPS (100 ng/mL) or hypoxic trauma using hydrogen peroxide (50 µM, 100 µM, and 200 µM) with or without proinflammatory cytokines, IL-1β (1 ng/mL) , IL-6 (200 ng/mL), and IL-33 (150 ng/mL). Inflammation and hypoxia were assessed using IL-6 and HIF-1ɑ expression respectively via qPCR 24 hours post-treatment. Cell death and pro-inflammatory cytokine production using multiplex analysis were used to measure outcomes. Results. Both types of treatments showed increased cell death compared to the control group. qPCR data is pending. Conclusion. With these studies as a core of the experimental approach, this in vitro cell-based assay will be used to assess immunologic response to inflammatory stimuli across the genetic variation of mouse strains. Findings from this project could enable the development of a clinical test that accurately predicts immunologic response to trauma and related-complications based on patients’ sensitivity to pre-traumatic injury.Item Divergence of military and civilian trauma research priorities(BMJ, 2021-06-15) Baer, David; Donaldson, Ross; McKinley, Todd; Guldberg, Robert; Orthopaedic Surgery, School of MedicineItem Effects of Thrombopoietin (TPO) on Longitudinal Mouse Hind Limb Crush Injury Model(Office of the Vice Chancellor for Research, 2014-04) Rothchild, Greg; Lipking, Kelsey; McKinley, Todd; Kacena, Melissa A.; Sandusky, George E.Approximately 645 people suffer from blunt force trauma injury to the femur every day. The recovery time of such injury can last anywhere from 3-6 months. Thrombopoietin (TPO) was used as a growth factor to induce bone and muscle healing. In this study, nine separate mouse groups (10 mice per group) were used: Crush PBS, Crush TPO, Surgery PBS, and Surgery TPO at day 3 and day 17, and controls with no surgery/crush/treatment. Skeletal muscle was harvested from the following sites: experimental impact, experimental adjacent, and normal contralateral skeletal muscle as a control. The muscles were fixed, processed, sectioned, and stained with H&E and Massons Trichrome stains. The slides were reviewed for skeletal muscle injury, muscle necrosis, inflammation, muscle repair, and regeneration. In addition, F4/80, an immunostain for macrophages was performed. On microscopic examination at day 3 the most common histologic changes seen were sporadic muscle fiber vacuolation, focal necrosis of varying sizes, muscle contraction bands, and infiltration of macrophages. On day 17, the skeletal muscle injury was generally healed. The main histologic lesions seen were variable sizes of muscle fibers, early fibroplasia, fat infiltration, some macrophages, satellite cells, and neovascularization. Comparing the TPO treated mice versus the PBS control group, the lesions at both time points were less in the TPO treated mice.Item Effects of Thrombopoietin (TPO) on Longitudinal Mouse Hind Limb Crush Injury Model(Office of the Vice Chancellor for Research, 2015-04-17) Rothchild, Greg; Lipking, Kelsey; McKinley, Todd; Kacena, Melissa A.; Sandusky, George E.Approximately 645 people suffer from blunt force trauma injury to the femur every day. The recovery time of such injury can last anywhere from 3-6 months. Thrombopoietin (TPO) was used as a growth factor to induce bone and muscle healing. In this study we utilized 9 separate mouse model groups (10 mice per group) were used: Crush PBS, Crush TPO, Surgery PBS, and Surgery TPO at day 3 and day 17, and controls with no surgery/crush/ treatment. Crush models were introduced to hind limb crush injury by a mechanical-gravity driven Einhorn device. Skeletal muscle was harvested from the following sites: experimental impact, experimental adjacent, and normal contralateral skeletal muscle as a control. The muscles were fixed, processed, sectioned, and stained with H&E and Masson’s Trichrome stains. The slides were reviewed for skeletal muscle injury, muscle necrosis, inflammation, muscle repair, and regeneration. In addition, F4/80, an immunostain for macrophages was performed. On microscopic examination at day 3 the most common histologic changes seen were sporadic muscle fiber vacuolation, focal necrosis of varying sizes, muscle contraction bands, and infiltration of macrophages. On day 17, the skeletal muscle injury was generally healed. The main histologic lesions seen were variable sizes of muscle fibers, early fibroplasia, fat infiltration, some macrophages( less than day 3) , satellite cells, and neovascularization. A follow-up immunostain (CD206 specific for M2 double labeled with F4/80) was performed to characterize the macrophages in and around the lesions at day 3. M2 macrophages were seen around the periphery of the lesion and none in the middle of the lesion. There were very minimal differences in M2 numbers between the PBS and TPO treated groups at day 3. In conclusion, comparing the TPO treated mice versus the PBS control group with F4/80 immunostain showed the lesions at both time points were less in the TPO treated mice.Item Electroacupuncture Promotes Central Nervous System-Dependent Release of Mesenchymal Stem Cells(Wiley, 2017-05) Salazar, Tatiana E.; Richardson, Matthew R.; Beli, Eleni; Ripsch, Matthew S.; George, John; Kim, Youngsook; Duan, Yaqian; Moldovan, Leni; Yan, Yuanqing; Bhatwadekar, Ashay; Jadhav, Vaishnavi; Smith, Jared A.; McGorray, Susan; Bertone, Alicia L.; Traktuev, Dmitri O.; March, Keith L.; Colon-Perez, Luis M.; Avin, Keith; Sims, Emily; Mund, Julie A.; Case, Jamie; Deng, Shaolin; Kim, Min Su; McDavitt, Bruce; Boulton, Michael E.; Thinschmidt, Jeffrey; Calzi, Sergio Li; Fitz, Stephanie D.; Fuchs, Robyn K.; Warden, Stuart J.; McKinley, Todd; Shekhar, Anantha; Febo, Marcelo; Johnson, Phillip L.; Chang, Lung Ji; Gao, Zhanguo; Kolonin, Mikhail G.; Lai, Song; Ma, Jinfeng; Dong, Xinzhong; White, Fletcher A.; Xie, Huisheng; Yoder, Mervin C.; Grant, Maria B.; Ophthalmology, School of MedicineElectroacupuncture (EA) performed in rats and humans using limb acupuncture sites, LI-4 and LI-11, and GV-14 and GV-20 (humans) and Bai-hui (rats) increased functional connectivity between the anterior hypothalamus and the amygdala and mobilized mesenchymal stem cells (MSCs) into the systemic circulation. In human subjects, the source of the MSC was found to be primarily adipose tissue, whereas in rodents the tissue sources were considered more heterogeneous. Pharmacological disinhibition of rat hypothalamus enhanced sympathetic nervous system (SNS) activation and similarly resulted in a release of MSC into the circulation. EA-mediated SNS activation was further supported by browning of white adipose tissue in rats. EA treatment of rats undergoing partial rupture of the Achilles tendon resulted in reduced mechanical hyperalgesia, increased serum interleukin-10 levels and tendon remodeling, effects blocked in propranolol-treated rodents. To distinguish the afferent role of the peripheral nervous system, phosphoinositide-interacting regulator of transient receptor potential channels (Pirt)-GCaMP3 (genetically encoded calcium sensor) mice were treated with EA acupuncture points, ST-36 and LIV-3, and GV-14 and Bai-hui and resulted in a rapid activation of primary sensory neurons. EA activated sensory ganglia and SNS centers to mediate the release of MSC that can enhance tissue repair, increase anti-inflammatory cytokine production and provide pronounced analgesic relief.Item Epidemiology of Fracture Nonunion in 18 Human Bones(JAMA, 2016-11) Zura, Robert; Xiong, Ze; Einhorn, Thomas; Watson, J. Tracy; Ostrum, Robert F.; Prayson, Michael J.; Della Rocca, Gregory J.; Mehta, Samir; McKinley, Todd; Wang, Zhe; Steen, R. Grant; Department of Orthopaedic Surgery, School of MedicineImportance Failure of bone fracture healing occurs in 5% to 10% of all patients. Nonunion risk is associated with the severity of injury and with the surgical treatment technique, yet progression to nonunion is not fully explained by these risk factors. Objective To test a hypothesis that fracture characteristics and patient-related risk factors assessable by the clinician at patient presentation can indicate the probability of fracture nonunion. Design, Setting, and Participants An inception cohort study in a large payer database of patients with fracture in the United States was conducted using patient-level health claims for medical and drug expenses compiled for approximately 90.1 million patients in calendar year 2011.The final database collated demographic descriptors, treatment procedures as per Current Procedural Terminology codes; comorbidities as per International Classification of Diseases, Ninth Revision codes; and drug prescriptions as per National Drug Code Directory codes. Logistic regression was used to calculate odds ratios (ORs) for variables associated with nonunion. Data analysis was performed from January 1, 2011, to December 31, 2012, Exposures Continuous enrollment in the database was required for 12 months after fracture to allow sufficient time to capture a nonunion diagnosis. Results The final analysis of 309 330 fractures in 18 bones included 178 952 women (57.9%); mean (SD) age was 44.48 (13.68) years. The nonunion rate was 4.9%. Elevated nonunion risk was associated with severe fracture (eg, open fracture, multiple fractures), high body mass index, smoking, and alcoholism. Women experienced more fractures, but men were more prone to nonunion. The nonunion rate also varied with fracture location: scaphoid, tibia plus fibula, and femur were most likely to be nonunion. The ORs for nonunion fractures were significantly increased for risk factors, including number of fractures (OR, 2.65; 95% CI, 2.34-2.99), use of nonsteroidal anti-inflammatory drugs plus opioids (OR, 1.84; 95% CI, 1.73-1.95), operative treatment (OR, 1.78; 95% CI, 1.69-1.86), open fracture (OR, 1.66; 95% CI, 1.55-1.77), anticoagulant use (OR, 1.58; 95% CI, 1.51-1.66), osteoarthritis with rheumatoid arthritis (OR, 1.58; 95% CI, 1.38-1.82), anticonvulsant use with benzodiazepines (OR, 1.49; 95% CI, 1.36-1.62), opioid use (OR, 1.43; 95% CI, 1.34-1.52), diabetes (OR, 1.40; 95% CI, 1.21-1.61), high-energy injury (OR, 1.38; 95% CI, 1.27-1.49), anticonvulsant use (OR, 1.37; 95% CI, 1.31-1.43), osteoporosis (OR, 1.24; 95% CI, 1.14-1.34), male gender (OR, 1.21; 95% CI, 1.16-1.25), insulin use (OR, 1.21; 95% CI, 1.10-1.31), smoking (OR, 1.20; 95% CI, 1.14-1.26), benzodiazepine use (OR, 1.20; 95% CI, 1.10-1.31), obesity (OR, 1.19; 95% CI, 1.12-1.25), antibiotic use (OR, 1.17; 95% CI, 1.13-1.21), osteoporosis medication use (OR, 1.17; 95% CI, 1.08-1.26), vitamin D deficiency (OR, 1.14; 95% CI, 1.05-1.22), diuretic use (OR, 1.13; 95% CI, 1.07-1.18), and renal insufficiency (OR, 1.11; 95% CI, 1.04-1.17) (multivariate P < .001 for all). Conclusions and Relevance The probability of fracture nonunion can be based on patient-specific risk factors at presentation. Risk of nonunion is a function of fracture severity, fracture location, disease comorbidity, and medication use.Item Large-magnitude Pelvic and Retroperitoneal Tissue Damage Predicts Organ Failure(Springer, 2016-06) Gaski, Greg; Frantz, Travis; Steenburg, Scott; Bell, Teresa; McKinley, Todd; Department of Orthopaedic Surgery, School of MedicineBACKGROUND: Pelvic and retroperitoneal trauma is a major cause of morbidity and mortality in multiply injured patients. The Injury Severity Score (ISS) has been criticized for underrepresenting and inaccurately defining mechanical injury. The influence of pelvic injury volume on organ dysfunction and multiple organ failure (MOF) has not been described. Through the use of CT, this investigation sought to precisely define volumes of mechanical tissue damage by anatomic region and examine its impact on organ failure. QUESTIONS/PURPOSES: (1) Do patients with MOF have a greater volume of pelvic and retroperitoneal tissue damage when compared with those without MOF? (2) In patients who sustained pelvic trauma, does the magnitude of pelvic injury differ in patients with MOF? (3) Does the magnitude of organ dysfunction correlate with pelvic tissue damage volume? METHODS: Seventy-four multiply injured patients aged 18 to 65 years with an ISS ≥ 18 admitted to the intensive care unit for a minimum of 6 days with complete admission CT scans were analyzed. Each identifiable injury in the head/neck, chest, abdomen, and pelvis underwent volumetric determination using CT to generate regional tissue damage volume scores. Primary outcomes were the development of MOF as measured by the Denver MOF score and the degree of organ dysfunction by utilization of the Sequential Organ Failure Assessment (SOFA) score. Mean pelvic and retroperitoneal tissue damage volumes were compared in patients who developed MOF and those who did not develop MOF using Student's t-test. Among patients who sustained pelvic injuries, we compared mean volume of tissue damaged in patients who developed MOF and those who did not. We assessed whether there was a correlation between organ dysfunction, as measured by the SOFA score as a continuous variable, and the volume of pelvic and retroperitoneal tissue damage using the Pearson product-moment correlation coefficient. RESULTS: The average volume of tissue damage was greater in patients with MOF when compared with those without (MOF: 685.667 ± 1081.344; non-MOF: 195.511 ± 381.436; mean difference 490.156 cc [95% confidence interval {CI}, 50.076-930.237 cc], p = 0.030). Among patients who sustained pelvic injuries, those with MOF had higher average tissue damage volumes than those without MOF (MOF: 1322.000 ± 1197.050; non-MOF: 382.750 ± 465.005; mean difference 939.250 [95% CI, 229.267-1649.233], p = 0.013). Organ dysfunction (SOFA score) correlated with higher volumes of pelvic tissue damage (r = 0.570, p < 0.001). CONCLUSIONS: This investigation demonstrated that greater degrees of pelvic and retroperitoneal tissue damage calculated from injury CT scans in multiply injured patients is associated with more severe organ dysfunction and an increased risk of developing MOF. Early identification of polytrauma patients at risk of MOF allows clinicians to implement appropriate resuscitative strategies early in the disease course. Improved stratification of injury severity and a patient's anticipated clinical course may aid in the planning and execution of staged orthopaedic interventions. Future avenues of study should incorporate the ischemic/hypoperfusion component of pelvic injury in conjunction with the mechanical component presented here for improved stratification of multiply injured patients at higher risk of MOF. LEVEL OF EVIDENCE: Level III, prognostic study.