Limitation of Weight-Based Pediatric Dosing Guidelines in ^99mTC-DMSA Renal SPECT

Objectives: An important objective of the current North American consensus weight-based dosing guidelines is to provide dose reduction while preserving adequate diagnostic image quality. These guidelines provide for scaling of the administered activity (AA) based on the patient’s weight subject to the constraint of a minimum AA based expert consensus of acceptable image quality. In this work, we investigate these AA based on these guidelines in terms of impact on surrogates for image noise, resolution, and contrast in ^99mTc-DMSA renal SPECT using a population of simulated phantoms.

Methods: The phantom population used was created from segmentations of clinical pediatric patient images scaled based on the Center for Disease Control’s NHANES database. This phantom population realistically models male and female patients having mean gender-specific weights of newborns and 1, 5, 10, and 15-year-olds. For each age, 3 body morphometries corresponding to 90^th (tall and thin), 50^th (reference) and 10^th height percentiles were generated. For each age and height percentile, kidneys having masses equal to the mean kidney mass and 100%±15% of the kidney mass were included. Thus, a total of 2×5×3×3=90 phantoms were simulated. Uptake in the kidneys was estimated using a 47-patient dataset with ages ranging from 1 to 16 years acquired at Boston Children’s Hospital (BCH). Means and standard deviations of the fractional kidney uptake were estimated from these patients. For each phantom, the total activity in the kidneys was equal to this mean kidney uptake fraction times the AA calculated using consensus weight-based pediatric dosing guidelines. Projection images were simulated using an analytic simulation that realistically modeled attenuation, patient scatter, the distant-dependent collimator-detector response and the clinical acquisition protocol at BCH. The projections were reconstructed using filtered backprojection. We calculated three surrogate measures of image quality. In nuclear medicine, noise is a function of counts, so we calculated the mean count density in the kidney region. In SPECT, resolution in the reconstructed image is a function of the distance from the object to the collimator at all the projection views. As a surrogate for resolution, we calculated the average distance from the collimator face to the center of rotation. Finally, image contrast is degraded by scattered photons, so we calculated the scatter to primary ratio (SPR) from projections generated with and without scatter modeling. We investigated the values of these image quality surrogates as functions of weight and height for the phantoms in the study.

Results: The rankings of the three height percentiles at the same weight by values of the 3 image quality surrogates were inconsistent with expectations. The mean count densities were relatively equal except for the newborn, but the SPRs varied substantially across the ages. The average radius of rotation varied substantially. These data suggest that weight-based dosing is unlikely to fully equalize image quality. Thus, although current dosing guidelines may equalize count densities (newborns excepted), there were differences in SPR and average radius of rotation, suggesting there would be differences in image quality. In newborns, the count densities were substantially higher and the scatter fractions were substantially lower than for the other ages, suggesting that image quality would be better for the newborns than for the other ages largely due to the minimum AAs present in the guidelines.

Conclusions: This study suggests that weight-based dosing is suboptimal for equalizing image quality for pediatric patients of different age with different body morphometries. Further, the results indicate that current values of minimum AA in dosing guidelines may result in over-dosing for newborns relative to the other ages. Research Support: This work was supported by NIH under grant number R01-EB013558.