A comprehensive in silico approach to measure spatio-temporal changes of bone tissue in mouse models of osteoporosis and osteoarthritis.

骨质疏松症(OP)和骨关节炎(OA)等肌肉骨骼疾病是影响数百万人生活质量的主要临床问题，NHS每年花费超过50亿英镑。在OP患者中，患者的骨骼密度较低，与健康患者相比，更容易骨折。患有骨质疏松性骨折的患者失去了活动能力和独立性，死亡风险增加。骨性关节炎会影响关节，关节会变得更加僵硬和疼痛，从而降低患者的活动能力。由于目前还没有治疗骨性关节炎的药物，治疗这种疾病的唯一方法是进行侵入性手术，用植入物取代退变的关节。虽然有治疗OP的有效方法，但它们非常昂贵，而且并不适用于所有患者。因此，有必要开发治疗这些肌肉骨骼疾病的药物。每一种新的治疗方法都需要在两种动物身上进行测试，其中一种通常是老鼠，然后才能在临床试验中测试干预措施。大多数新药疗法在I期临床试验中都失败了，尽管它们在动物试验中被发现很有前途。这可能是由于动物和人类之间的生理学差异，治疗是在动物疾病模型上进行测试的事实，这可能不能完全复制患者的情况，和/或由于我们通常在老鼠身上进行简化的测量。例如，为了评估一种新药降低OP患者骨折风险的能力，在小鼠身上，标准方法建议测量一块骨的两个小部分(如胫骨)的骨特性(密度和形态)，而我们真正感兴趣的是增加骨强度。此外，标准的横断面研究对分配到治疗组或非治疗组的不同动物进行测量，增加了受试者之间的可变性，从而降低了测量的准确性，这可能会掩盖干预的影响。使用标准的方法只能通过机械测试来测量骨强度，这是侵入性的，不能在同一只小鼠身上随着时间的推移而在体内进行，导致大量的动物用于研究并增加测量的可变性。这可以通过使用计算模型来改进。该项目将致力于加强临床前研究中对骨骼和关节特性的评估，将同一小鼠的纵向高分辨率成像、先进的图像处理和计算建模相结合，从活体图像中非侵入性地测量骨强度。这一改进还将大幅减少老鼠在肌肉骨骼研究中的使用。我们将创建并验证计算模型，用于预测使用体内MicroCT扫描的小鼠在每个时间点的骨强度，该扫描允许对小鼠胫骨进行高分辨率扫描。我们将创建一项服务，自动测量胫骨不同部分的骨骼特性，并可供世界各地的其他研究人员使用。最后，我们将修改我们的方案来详细研究小鼠膝关节的骨骼变化，以评估OA的效果。这项工作将改善研究肌肉骨骼疾病和测试新干预措施的临床前研究。此外，我们将通过创建一个带有网络界面的服务来与研究社区分享我们的新方法，该服务将允许采用这些方法，并大幅减少骨和关节研究中使用的小鼠数量。

Musculoskeletal pathologies such as Osteoporosis (OP) and Osteoarthritis (OA) are major clinical problems that impair the quality of life of millions of people and cost over £5 billion to the NHS very year. In OP, patients have less dense bones, which are more likely to fracture compared to those of healthy patients. Patients who suffered of osteoporotic fractures lose mobility and independence and have increased mortality risk. OA affects the joints, which become stiffer and more painful with consequent reduced mobility for the patient. As there are no pharmacological treatments for OA, the only way to treat this disease is to perform an invasive surgery that replaces the degenerated joint with an implant. While there are effective treatments for OP, they are very expensive and they do not work for all the patients. Therefore, there is a need for development of pharmaceutical treatments for these musculoskeletal diseases. Each new treatment needs to be tested in two animal species, one of which is usually a mouse, before testing the intervention in clinical trials. Most new drug treatments fail in phase I clinical trials even though they were found promising during the animal testing. This can be due to the differences in physiology between the animals and the humans, the fact that treatments are tested on animal disease models, which may not completely replicate the condition of patients and/or to the fact that in mice we usually perform simplified measurements. For example, in order to evaluate the ability of a new drug to reduce the risk of fracture in patients with OP, in mice the standard methodology suggests to measure the bone properties (density and morphology) in two small portions of a bone (e.g. tibia), while what we are really interested in increasing of bone strength. Moreover standard cross-sectional studies perform measurements on different animals assigned to treated or non-treated groups, increasing the inter-subject variability, which reduces measurement accuracy, which could hide the effect of interventions. With standard approaches the bone strength can be measured only with with mechanical testing, which is invasive and can not be performed in vivo on the same mouse over time, leading to a large number of animal used in research and increasing measurement variability. This can be improved by using computational models. This project will focus on the enhancement of the assessment of bone and joint properties in preclinical studies, by combining longitudinal high-resolution imaging of the same mice, advanced image processing and computational modelling to non-invasively measure bone strength from the in vivo images. This improvement will also lead to a substantial reduction in the usage of mice in musculoskeletal research. We will create and validate computational models for the prediction of bone strength at each time point in mice scanned with in vivo microCT that allows for high-resolution scans of the mice tibia. We will create a service that measures automatically the bone properties in different portions of the tibia and that can be used worldwide by other researchers. Finally we will adapt our protocols to study in details bone changes in the mouse knee to evaluate the effect of OA. This work will lead to an improvement of preclinical research for studying musculoskeletal diseases and for testing new interventions. Moreover, we will share our novel methodologies with the research community by creating a service with a web-interface that will allow to uptake these methodologies and substantially reduce the number of mice used in bone and joint research.