ecules that are continuously secreted and degraded by the gastric acids; although it is about 95% water, it substantially slows the diffusion of acid relative to water alone. It has been hypothesized that the continual movement of water and new mucin molecules away from the epithelial surface also carries acid away from the epithelium by convection (Livingston and Engel 1995). The interplay of these factors has been mathematically modeled (Engel and others 1984), but many of the physiological parameters of the model have not been measured.
In view of the importance of the diffusion mechanism in estimating the dose to the stomach, the committee found it useful to formulate a model within which it could investigate this mechanism. The model consisted of a spherical representation of a stomach with a volume of 250 mL. A mucus layer 50 µm thick was assumed. This was followed by a layer of surface cells 50 µm thick. The stem cells were considered to be distributed throughout a layer of tissue 200 µm thick. Below this layer, diffusion into capillaries was assumed to remove radon and reduce the concentration to zero. The concentration of radon in the contents of the stomach, assumed to be well mixed, was taken to decline exponentially with a half-time of 20 min. The model and its results are discussed in detail in appendix B. When the above parameters were used with a radon diffusion coefficient for the gastric wall of 5 × 10-6 cm2 s-1, the time-integrated concentration of radon at the depth of the cells at risk (200 µm) was found to be 30% of the time-integrated radon concentration in the contents of the stomach. The time-integrated concentration was found to be insensitive to the value assumed for the diffusion coefficient and to depend somewhat on the depth to which radon was assumed to diffuse.
Although the diffusion model of the stomach does not permit definitive conclusions, it does suggest that both radon concentration and its time integral vary over a rather limited range for a wide range in the diffusion coefficient. If the mucus layer is a barrier to radon diffusion, concentration in the wall could be substantially reduced. The chemical composition of the layer (95% water, degraded mucin, and soluble polymeric mucin secreted by the mucosa) does not suggest a strong diffusion barrier to inert substances like radon. The concentration of radon reached in the wall is controlled by the blood flow through the gastric mucosa, and the depth of microvasculature may be of considerable importance. The influence of the microvasculature of the small intestine on absorption of gases has been investigated (Bond and others 1977), but little information is available on the stomach. Further studies clearly are needed to determine the influence of the mucus layer and the capillary structures on the concentration of radon in the stomach wall. It should be noted that because the PBPK model cannot fully adhere to the microvasculature which removes most radon directly to the blood before it can diffuse near stem cells, the model is a conservative model.
The calculations of dose and risk reported below assume that the time-integrated concentration of radon at the depth of the stem cells is 30% of that in