建立适用于人体红骨髓内脂肪酸成分分析的高分辨磁共振定域谱新方法

Fatty acids of the red bone marrow (RBM) play a significant role in affecting energy metabolism, bone mass and hematopoietic function. Thus, investigating the fatty acids composition of RBM can not only help us to understand the pathological mechanism of related diseases, but may also provide extra biomarkers for early diagnosis and evaluating the stage of diseases. Magnetic resonance spectroscopy (MRS) as the only noninvasive technique has been widely used to explore the metabolites of in vivo, like brain and liver. Since RBM always co-existes with honeycomb-like trabeculae, the local B0 is extremely inhomogeneous and the signal linewidths obtained by conventional 1H-MRS is extra broad due to the magnetic susceptibility effect. Thus, some small resonances, which not only allow us to infer the molecular structures but may also be used as potential diagnostic markers at molecular level, from the long-chain fatty acids are overlapped typically by the main fatty acids peak at 1.3 ppm and the water peak at 4.8 ppm. Intermolecular double quantum coherence (iDQC) signals originates from the dipolar couplings between distant spins. Theoretically, only the magnetic inhomogeneous field in this range of the distance between interacting protons will broaden the line-width originate. In this project, based on our previous studies on iDQC methods, a novel MRS method with corresponding quantitative correction method via iDQC technique is proposed to explore the fatty acids composition in RBM of human. This project will focus on the following work: (1) Study the iDQC signal behaviors of different spins of fatty acids through theoretical deducing and computer simulation; analyze the field inhomogeneity, susceptibility, J-coupling and anisotropy effects on the each kind of spins quantitatively. (2) Through introducing spin-lock technique to ameliorate the fast transverse relaxation time and three ways of achieving the localization purpose, a new iDQC-MRS pulse sequence will be proposed and implemented on a Siemens 3.0 T Trio scanner to test its feasibility to enhance the spectrum resolution. Then signal characteristic and effect factors for different spins, especially for unsaturated fatty acids related resonances, will be explored on a series of lipids-water emulsion phantoms using the new pulse sequence. In addition, a post processing and corresponding semi-quantitative/quantitative method will be developed and a fatty acids composition index will be established. (3) Finally, the new method will be applied to study the fatty acids compositions of vertebral bone marrow between healthy and patients (osteoporosis and leukemia patients) in vivo to investigate whether there is difference. This project is expected to find new biomarkers for diagnostic and prognostic purpose regarding diseases status.

红骨髓内脂肪酸成分与骨量和造血功能相关，研究其成分有助于对相关疾病的理解，也可能为诊断提供新的生物标志物(biomarker)。磁共振波谱（MRS）可以无损检测代谢物，然而对于成年人的红骨髓，由于骨小梁的存在，局部磁场极度不均匀，常规MRS谱峰展宽，一些可能与疾病相关的谱峰被掩盖。研究发现分子间二量子相干(iDQC)技术对不均匀磁场不敏感。本课题拟在我们较好的研究基础上, 建立基于iDQC的MRS新技术, 并将其应用于探索红骨髓内脂肪酸成分：（1）理论推导和仿真脂肪酸上不同自旋的iDQC信号演化。模拟研究磁化率不均匀，J耦合效应，各向异性的影响；（2）在人体成像仪上将自旋锁定技术和空间定位模块与iDQC技术结合，设计新的脉冲序列。在体外模型上研究不同自旋的信号，计算定量指标（index）和校正因子；（3）对健康志愿者和骨质疏松以及白血病患者的脊椎红骨髓进行扫描，分析不同脂肪酸含量。

虽然红骨髓脂肪已被证实在骨骼形成和能量代谢中发挥着重要作用，但对人体内红骨髓脂肪成分的认识仍然较差。磁共振波谱(MRS)是一种获取红骨髓脂肪的非侵入性方法。然而，传统1D MRS对红骨髓脂肪的分辨率较低，特别是对随骨小梁一起存在的红骨髓脂肪的分辨率较低。因此，即使非常小心地匀场,脂肪酸峰与主脂肪峰重叠在1.3 ppm。在本研究中，体素选择性二维MRS-iDQC方法被证明比传统1D MRS在股骨远端和胫骨近端识别更多的脂肪酸峰。体外的探索性实验结束后，我们按照计划，建立了新的脉冲序列方法，并且在初步的探索了人体红骨髓脂肪酸成分，获得了理想结果。 在13名健康志愿者参加的这项研究中，我们在临床 3.0 T全身MRI扫描仪上进行了iDQC定位的MRS扫描。为了减少系统误差，所有受试者均在两周内进行两次扫描。我们发现股骨远端和胫骨近端红骨髓脂肪的组成表现出强烈的个体间差异。根据先前的研究，我们以2.2 ppm 的α-亚甲基脂峰作为内标，然后用α-亚甲基脂肪峰划分出各交叉峰的体积比，进行校正分析。通过计算每个脂肪酸峰的变异系数(CV)来评估该方法的定量重现性。右股骨远端的红骨髓脂肪的组成不同于右胫骨近端的成分。我们发现，股骨远端红骨髓脂肪更不饱和(乙烯基，P<0.01;这种高不饱和是由多不饱和脂肪酸引起的(r=0.67，双烯丙基，P<0.01)。本研究以相对年轻的健康受试者为研究对象，并没有观察到左右腿和女性与男性之间红骨髓的脂肪酸组成的显著差异。我们的研究证明了二维定域iDQC谱能较好地描述红骨髓的脂肪酸的组成。所提出的方法可能是一个潜在的研究在骨小梁存在情况下红骨髓脂肪酸组成的工具，而脂肪酸交叉峰可能为红骨髓相关疾病的研究和诊断提供新的生物标志物。