Biomedical Engineering Reference
In-Depth Information
As more is learned in radiation biology, greater confidence can be placed in the
assessment of risk estimates for exposure of persons to ionizing radiation, partic-
ularly at low doses. Understanding the basic molecular mechanisms of radiation
damage in cells will greatly facilitate the task. We turn next to the subject of dose-
response relationships, which underlie radiation-protection regulations in use to-
day.
13.13
Dose-Response Relationships
Biological effects of radiation can be quantitatively described in terms of dose-
response relationships, that is, the incidence or severity of a given effect, expressed
as a function of dose. These relationships are conveniently represented by plotting
a dose-response curve, such as that shown in Fig. 13.11. The ordinate gives the
observed degree of some biological effect under consideration (e.g., the incidence
of certain cancers in animals per 100,000 population per year) at the dose level
given by the abscissa. The circles show data points with error bars that represent a
specified confidence level (e.g., 90%). At zero dose, one typically has a natural, or
spontaneous, level of incidence, which is known from a large population of unex-
posed individuals. Often the numbers of individuals exposed at higher dose levels
are relatively small, and so the error bars there are large. As a result, although the
trend of increasing incidence with dose may be clearly evident, there is no unique
dose-response curve that describes the data. In the figure, a solid straight line, con-
sistent with the observations, has been drawn at high doses. The line is constructed
in such a way that it intersects the ordinate at the level of natural incidence when
a linear extension (dashed curve A) to zero dose is made. In this case, we say that
a linear dose-response curve, extrapolated down to zero dose, is used to represent
the effect.
Curves with other shapes can usually be drawn through biological dose-effect
data. An example of this kind of response is found for leukemia in the atomic-bomb
survivors, shown by the inset in Fig. 13.5. Also, extrapolations to low doses can be
made in a number of ways. Sometimes there are theoretical reasons for assuming
a particular dose dependence, particularly at low doses. The dashed curve B in
Fig. 13.11 shows a nonlinear dependence. Both curves A and B imply that there
is always some increased incidence of the effect due to radiation, no matter how
small the dose. In contrast, the extrapolation shown by the curve C implies that
there is a threshold of about 0.75 Gy for inducing the effect.
For many endpoints of carcinogenesis, mutagenesis, and other effects, dose-
response functions at low doses and low dose rates can be analyzed in the following
way, contrasting high- and low-LET radiations. With low doses of high-LET radia-
tion, the effect is presumably due to individual charged-particle tracks: their spatial
density is small, and there is a negligible overlap of different tracks. Since the den-
sity of tracks is proportional to the dose, the incidence
E
(
D
)
(above controls) should
also be proportional to the dose
D
at low doses. This general behavior of the dose
