The retention of 226Ra in human soft tissue and bone; implications for the ICRP 20 alkaline earth model.

Abstract

We have adjusted the ICRP model of alkaline earth metabolism to fit data on the 226Ra content of 87 samples of soft tissue from 17 persons who received radium by injection or ingestion 5 days to 53 yr before measurement. The fitted functions indicate that soft tissue retention rises to 58% of the whole body retention at 18 days after single intake, and then falls steadily reaching 33% at 100 days and 6% at 1000 days. The adjusted model was also found to fit data on bone retention in 5 persons 8–467 days after injection. For both types of tissue, the new functions differ substantially from those originally published. The adjustments improve or do not significantly reduce the model's ability to fit other radium data. The fits to soft tissue and bone data were achieved by changing the quantity B in the ICRP model from its original value for radium {0.06236 (1-RBODY)}−2.410 to {0.1247(1-RBODY)}−2.410 where RBODY is the whole body retention function. This forced new values for five model parameters for radium: beta = 0.6254, omega = 1.436, fC = 5.629, v = 1.521, r = 0.9953. Also changed are the empirical values for the percentages of the body's natural alkaline earth contents which lie in soft tissue: Ca(0.28%), Sr(0.98%), Ba(4.7%), 226Ra(5.5%). For use in dosimetry, effective retention integrals are given for 224,226,228Ra and for 228Th in soft tissue, on bone surfaces, in bone volume, and in the whole body. Integration times of 1 yr, 50 yr and infinity are used. For 226,228Ra, the soft tissue integrals are reduced by factors of 1.5–2.5 below the values originally published; for 224Ra, the reduction factor is about 1.1. The bone volume integrals increase for all isotopes; the percentage increase, about 100%, is greatest for 224Ra. A reanalysis of the autoradiographic data on 226Ra in bone has been carried out using the adjusted functions. The best fit results with dC = 1.5 rather than with the previous value, 10, indicating that the so-called modified version of the ICRP model may be unnecessary. An examination was made of the ability of the ICRP model to predict the outcomes of experiments to which the model functions had not been previously fitted. Results were positive.