Normal bone marrow is divided into red and yellow marrow, a distinction made on the grounds of how much fat it contains.
Red marrow is composed of:
- haematopetic cells
- supporting stroma
- reticulum (phagocytes and undifferentiated progenitor cells)
- scattered fat cells
- a rich vascular supply
Conversely, yellow marrow has all the same constituents as red, except that fat cells make up the vast majority, with resulting poor vascularity. Distribution varies with age and from one individual to another, but should be symmetric.
During infancy red marrow occupies the entire ossified skeleton except for epiphyses and aphophyses. Gradually red marrow 'retreats' centrally, such that by adulthood it is essentially confined to the axial skeleton (pelvis, spine, shoulder girdle, skull). Frequently the proximal humeri and neck of femurs have residual red marrow2.
In addition, islands of red marrow may be seen anywhere in the skeleton, typically in a subcortical distribution, often with central yellow marrow giving it a bull's eye appearance on axial imaging. Additionally red marrow is found in subchondral crescents again of the proximal humerus and femur.
- T1 - ALWAYS slightly hyperintense to muscle and disc (due to scatted fat cells)
- T2 - can be difficult to distinguish from yellow marrow as both are somewhat hyperintense.
- STIR - red marrow remains hyperintense, c.f. yellow marrow which is saturated out.
- follows subcutaneous fat on all sequences.
Broadly marrow pathology can be divided into:
- vascular abnormalities
- myelodysplastic syndrome
- polycythaemia rubra vera
- reconversion of yellow to red marrow
- monoclonal gammopathies (plasma cell dyscrasias)
Most of the above conditions (covered individually in the encyclopaedic section) affect the marrow diffusely. The exception is multiple myeloma which has a predilection for focal deposits, and Waldenstrom's macroglobulinemia which causes infarcts.
The MRI appearance is variable:
- normal red marrow appearance (10-25% of all leukaemic patients will have normal appearing marrow)
- abnormal distribution of what appears to be normal red marrow
- abnormal signal from red marrow in a normal distribution
- abnormal signal and distribution
The abnormal signal is due to replacement of the small amounts of fat cells normally found in red marrow, such that T1 signal will decrease to or below the signal from disc or muscle. T2 signal is more variable, but will in general increase when compared to muscle.
Myelofibrosis and mastocytosis incite such prominent sclerosis that the marrow is very dark on both T1 and T2 ; a similar appearance to marrow in haemosiderosis in patients with haemolysis from sickle cell disease and thalassaemia.
The leukaemias typically affect the metaphyses > diaphyses > epiphysese. Changes in the latter indicate a large tumour load, and therefore has prognostic implications.
Yellow-to-red reconversion, obviously, generated red marrow in abnormal distribution. Signal is therefore a very important aspect of correct image interpretation. It occurs in the reverse order to that of red to yellow conversion, and is seen in
Red marrow reconversion can be difficult to differentiate from metastases in the spine. A useful pair of sequences is T1 in- and out-of-phase. If there is focal low signal on T1 in-phase this may be due to either pathology. However the scattered fat cells in red marrow cause marked signal loss on out-of-phase images. There is no such signal loss in metastases.
- regional anatomy
- head and neck
- upper limb
- lower limb
- systems anatomy
- blood vessels
- 1. Musculoskeletal MRI - Kaplan
- 2. Taccone A, Oddone M, Occhi M et-al. MRI "road-map" of normal age-related bone marrow. I. Cranial bone and spine. Pediatr Radiol. 1996;25 (8): 588-95. Pubmed citation
- 3. Taccone A, Oddone M, Dell'Acqua AD et-al. MRI "road-map" of normal age-related bone marrow. II. Thorax, pelvis and extremities. Pediatr Radiol. 1996;25 (8): 596-606. Pubmed citation
- 4. Vande Berg BC, Malghem J, Lecouvet FE et-al. Magnetic resonance imaging of normal bone marrow. Eur Radiol. 1999;8 (8): 1327-34. Pubmed citation