Sequence and rate of bone marrow conversion in the femora of children as seen on MR imaging: are accepted standards accurate?

Abstract
The purpose of this study was to reassess the normal sequence and rate of marrow conversion in the femora of children as depicted on MR imaging. We retrospectively analyzed 81 T1-weighted MR images of the femur for the appearance and distribution of hematopoietic (red) and fatty (yellow) marrow. Eighty-one children 2 days to 15 years old with no known bone marrow abnormalities were divided into four age groups. The signal intensity and homogeneity of the marrow in the proximal epiphysis, proximal metaphysis, diaphysis, distal metaphysis, distal epiphysis, and greater trochanter were compared with the signal intensity and homogeneity of surrounding muscle and fat and graded by two observers. In select cases, region-of-interest measurements of marrow, subcutaneous fat, and muscle were obtained to validate the visual grading system. Conversion of hematopoietic to fatty marrow in the femur followed a well-defined sequence, occurring first in the proximal and distal epiphyses, followed by the diaphysis, distal metaphysis, and then the proximal metaphysis. Although high-signal-intensity fatty marrow could be seen within the femoral diaphysis as early as 3 months of age, fatty marrow with various degrees of heterogeneity was routinely seen in this region by 12 months of age. After 5 years of age, the femoral diaphysis showed homogeneous high signal intensity. These findings are in contrast to previously published data that describe homogeneous red marrow within the femoral diaphysis during the first year of life and homogeneous yellow marrow visualized by 10 years of age. The normal age-related sequence of femoral marrow conversion we saw on MR images conforms to the sequence described in previously published reports, but this transformation, particularly in the diaphysis, occurs significantly earlier in life than has been previously reported. This discrepancy might be explained partially by the sensitivity of signal intensity in the femoral marrow to alterations in window and level settings.