Fat plays an important role in MSK research, both in terms of fat suppression as well as in quantification of the fat content. The main reason to suppress fat in MSK research is because it hampers the quantitative measurement of relaxation times and diffusion coefficients and to avoid image distortions, while quantitative fat imaging can facilitate comparative studies across groups, help in the understanding of pathophysiology, and determine the efficacy of therapies and interventions by capturing temporal changes. This talk will explore what the advantages and disadvantages are of both approaches and highlight recent research advances.
Introduction – where is the fat in MSK research and how much is there?
This talk will focus on fat in MSK research; and will cover research areas in bone marrow and skeletal muscle fat suppression and quantification. While water has a single resonance peak, the fat spectrum has several peaks, and around 30% of the fat protons resonate at frequencies different from that of the major methylene (–CH2–) resonance [1-3]: in addition many of the peaks exhibit scalar coupling. These factors make both suppression and quantification of fat protons complicated. In skeletal muscle, fat exists in two compartments, intra-myocellular lipids (IMCL) and extra-myocellular lipids (EMCL). IMCL is located in droplets in the myofiber close to the mitochondria and is metabolically active, but concentrations are typically very low. EMCL is located between the myo-fibers and has much higher concentrations. As a result, the fat content assessed by quantitative imaging approaches and the efficacy of fat suppression techniques are both dominated by EMCL. A third fat compartment surrounding skeletal muscles is subcutaneous fat. In healthy skeletal muscle, proton density fat fractions (PDFF) are in the range of 2-5%, and depend mainly on age [4] [5], while in severe diseases muscle can be completely replaced by fat. The other major tissue containing fat is bone marrow, accounting typically for up to 4–5 % of the human body weight. It mainly consists of adipocytes, hematopoietic stem cells, and mesenchymal stem cells. Bone marrow is one of the few tissues in the human body in which both water and fat can be present in equal amounts [5].Fat suppression – when do we need it and how do we do it?
The main reason to suppress fat in MSK research is because it hampers the quantitative measurement of relaxation times and diffusion coefficients. Fat has very different MR relaxation, diffusion and magnetization transfer properties compared to water, and tissue water parameters can be overshadowed by their fat counterparts. Two prime examples are the assessment of edema via the T2 relaxation time in both bone marrow and muscle, and diffusion weighted imaging (DWI) in muscle. An increase in the T2 relaxation time can be an indication of intra and/or extracellular edema or even ‘disease activity’ in the muscle tissue [6, 7]. In muscle, the T2 relaxation of muscle water is around 30 ms at 3T, whereas the T2 of fat is much longer (>80 ms) and the increase in T2 relaxation time of inflamed/edematous tissue rarely exceeds 10 ms [6]. Since in many muscle diseases, particularly in chronic stages, muscle tissue is progressively replaced by fat, assessment of the T2 relaxation time without accounting for the presence of fat will lead to an overestimation of the T2 of the muscle water and possible wrong conclusions about the presence of edema. A similar rationale can be followed for DWI; as the diffusion of fat is two orders of magnitude lower than water, incomplete fat suppression leads to an underestimation of mean diffusivity [8]. Moreover, as an EPI readout is often used to speed up acquisition, a rim of fat surrounding the tissue of interest leads to extra problems when it is not properly suppressed due its associated image distortions and chemical shift displacement artefact [8]. Also for other MR techniques like magnetization transfer, MR elastography and ultra-short echo time fat infiltration in muscle diseases is a challenge and needs to be dealt with either using fat suppression or during post-processing. An overview of different fat suppression methods can be found elsewhere [7], but it is important to realize that it is very challenging to suppress completely. For instance, while frequency selective methods such as chemical shift (spectral) selective fat saturation (CHESS) or hybrid methods like spectral adiabatic inversion recovery (SPAIR) can achieve robust suppression of the main methylene (CH2) and methyl (CH3) peaks, the olefinic peak at 5.3 ppm with a resonance frequency close to the water signal is not suppressed [8, 9]. Therefore, in the assessment of T2 relaxation times in muscle, research is now moving towards acquisition of non fat-suppressed images and decomposing the signal decay into a sum of fat and water contributions using complex models based or extended phase graph (EPG) analysis [6, 10]. Alternatively, approaches have been shown where water and fat are decomposed prior to relaxation time mapping in muscle [11, 12] and vertebral bone marrow [13].Fat quantification - why do we need it and how do we do it?
It is also useful to quantify the fat signal, with applications in a wide range of musculoskeletal diseases, orthopedic injuries, aging-associated conditions like sarcopenia and osteoporosis, differentiating malignant from benign vertebral fractures and the metabolic syndrome as well as in the study of normal physiology. Quantitative fat imaging can facilitate comparative studies across groups, help in the understanding of pathophysiology, and determine the efficacy of therapies and interventions by capturing temporal changes. Overall, the main techniques used for fat quantification in a wide range of tissues are chemical-shift based approaches, and reviews related to MSK research have been published recently [14, 15]. Sequence parameters and post-processing need to be tailored to the tissue of interest. For instance, bone marrow can enclose trabecular bone, which results in significantly broadened linewidths for all its chemical species, which requires the inclusion of T2* decay effects in the analysis of bone marrow chemical shift encoding-based methods [15].In summary - Fat-Suppression & Quantification: Friend or Foe?
Removal of the fat component is essential for accurate assessment of several key parameters in quantitative MRI, but due to the complexity of the fat spectrum sophisticated methods are needed to perform complete fat suppression – and as such it can be considered a foe. Decomposing the signal into fat and water components prior to or in the process of relaxation time fitting might well be the way forward in that area. Fat quantification on the other hand is definitely a friend, with increasing impact in a wide range of applications. It remains a challenge to standardize quantitative imaging methods across vendors, sites and post-processing methods.