Purpose or Case Report: To establish the reproducibility of the apparent diffusion coefficient (ADC) measurement in the bone marrow of the clivus and lumbosacral spine.
Methods & Materials: Under operational ethical permission, diffusion-weighted imaging (DWI) was run twice — before and after the morphological sequences — in 6 children having a magnetic resonance (MR) scan of the brain for clinical follow-up. In 4 children DWI was obtained on 2 separate occasions after a 1 to 4-month interval.

Similarly, in 10 children having a spine MR study, the DWI acquisition was repeated. In an additional 4 subjects who had two abdomino-pelvic MR scans with DWI as part of follow-up within a 6-month period, and who had no intervening treatment in the preceding 6 months, scans done on two separate occasions were used to assess reproducibility.

In both cohorts, scans including DWI with b=0 and 1,000 s/mm² for the brain and multiple b values ranging from 0 to 1,000 s/mm² in the spinal cohort were acquired at 1.5T machines; all subjects were at least 6 months off treatment.
A circular region-of-interest was drawn in the clivus (diameter of 4-6-mm) and L3-S1 vertebrae (diameter of 8-10 mm) of the b=1000 s/mm² DWI using in-house software Adept®. Data from the entire volume of interest (2 or 3 contiguous slices) were obtained.
Results: In the brain cohort (8-16 years), 95% confidence intervals of marrow ADC measurements in the clivus ranged from -5.5 to +11%

In the abdominopelvic cohort (3-16 years), the 10 done within the same MR examination showed very similar marrow ADC values of the lumbosacral spine at the beginning and end of the examination. The ADC variability was much greater when scans acquired on different occasions (median of 10 weeks apart) were considered. Bland-Altman plots were constructed using data from all 14 patients and showed 95% confidence intervals of -23.4% to +19.3% for median ADC and -19.5% to +15.8% for mean ADC.
Conclusions: In children, it is feasible to measure ADC from the clivus with 11% variability (Limits of Agreement), whilst ADC quantitation in the lumbar spine can have a measurement variability of up to 25% (Limits of Agreement).
These data will serve as a baseline to assess changes on ADC values of paediatric bone marrow induced by local and systemic oncological treatments, including potential toxicity caused by novel agents within clinical trials, and to assess engraftment in post-bone marrow transplant leukaemic patients in a non-invasive fashion, before changes in blood counts are evident.