Linear No-Threshold Hypothesis

[JMUCKERHEIDE@delphi.com]

Eric,

To a (rare) radiobiologist:  :-)

You say that Billens table does not have double-strand data. Note that Billens 
Table I reports spontaneous double-strand breaks (the rad data you present is
in Table II): 

---------------------------------------------------------------
Table I  Estimated Sponanteous DNA Degradation Events (Cell/hr)*

Reaction            Single-strand     Double-strand

Depurination             4000              1000
Depyrimidination          200              50
Deamination of cytosine 4000              15
Chain break resulting
 from depurination            ---                  1000
Direct chain break           ---                  4000

* calc from Shapiro, R., 1981, Damage to DNA caused by hydrolysis; In:
_Chromosome Damage and Repair_ (Seeberg and Kleppe, Eds) Plenum pp 3-18
---------------------------------------------------------------- 

[Sorry if the columns don't line up, I'm writing in a proportional font
estimating character count for fixed font output :-)  ] 

If you see Billens [Rad Res 124:242-245 (1990)] reprinted in BELLE Newsletter
[Vol 3, No 1, July 1994], you will also see Sadao Hattori's article "State of
Research and Perspective on Radiation Hormesis in Japan. He presents 20
projects that had been funded in Japan, most in the basic biology
demonstrating fundamental aspects of testing the basis for stimulation in
biological responses to explain the explicit evidence of beeficial effects in
actual populations, especially the Misasa Spa area and other natural
exposures, the Japanese survivors for leukemia, longevity, etc (while RERF may 
be playing with the death cause data to push for the linear model, they can't
very well go around killing people to keep them from living longer! I hope.) 

Note also that your point can go further with the evidence in the biological
sciences that now recognizes proof of the stimulating effects of low-dose
radiation (the protective effect seen when exposure to low doses produces much 
greater survival in a population with later high-dose exposures). This
demonstrates that the biological response in whole, viable, biological
organisms with response/repair mechanisms is non-linear and stimulatory in
response to radiation. (See, eg, Hattori.) 

That is, it is NOT (can not be) the billiard-ball stochastic response that the 
non-biology "physics" models of "statistical probability to the number of
hits" rationales, nor from some evidence in the "neutered" biological
mechanisms/responses of some cell-in-culture and even immune-suppressed
tumorigenic lab animals which are bred to have essentially nothing to
stimulate. 

This of course also applies to some of the chemical response nonsense that is
"demonstrated" to "cause cancer" in such animals (which is especially curious
for those minerals and biological agents which "cause cancer" at low doses
while known to be required nutrients.  

Consider even that potassium is such an essential nutrient, with those of us
subject to the light-headedness and need for orange juice, bananas, or other K 
sources, while the K-40 drives most of our assured (required?) internal body
radiation dose (30-35 mr/yr). 

Combined with explicit data with ranges to thousands of times low natural
exposures, with no adverse effects (eg, high natural and medical and
radium-burden, etc), it would seem incongruous to therefore even remotely
consider a scientific association of low radiation doses (to >>100R for
chronic exposures) to adverse health effects (unless the vested interests of
the regulators and cleanup programs are put on the balance). 

Thanks. I appreciate seeing some actual biology in this discussion.

>      A couple of weeks ago, Mark Winslow from EPA requested opinions 
>      and references regarding the linear no-threshold model (LNTM).  
>      There were many responses related to the epidemiological studies 
>      that attempt to demonstrate either an adverse effect from low 
>      doses, no effect from low doses, or even a beneficial effect from 
>      low doses.  There was very little reference to the possible 
>      molecular evidence that shows that the model is at least 
>      conservative, if not inappropriate, for occupational levels of 
>      ionizing radiation exposure.
>      
>      This work below isn't mine, but, as one who formerly defined 
>      myself as a radiobiologist (are we that rare in the HP ranks?), I 
>      find molecular explanations valuable.  The data below are 
>      excerpted from a Commentary Paper by Daniel Billen of Oak Ridge 
>      Associated Universities in Radiation Research, 124 (242-245) 1990.
>      
>      Spontaneous DNA damage events/cell/hour        DNA damage/rad
>      
>                8,000                                   ~ 20
>      
>      
>      Spontaneous DNA damage events/cell/sec        DNA damage/rad
>      
>                  2.2                                   ~ 20
>      
>      
>      If failure to properly repair DNA damage is an initiating event 
>      that leads to cell transformation (and eventually malignancy), 
>      then the rate of damaging events is critical.  For occupational 
>      dose rates which rarely exceed 1 rad/hr, the DNA damage rate is 
>      overwhelmed by the rate of spontaneous damage (and repair).  For 
>      acute radiation dose delivery (such as that experienced by the 
>      survivors of Hiroshima and Nagasaki), the rate of 
>      radiation-induced damage greatly exceeds the spontaneous rate and 
>      probably the capability of repair.  For example, 100 rad (1 Gy) 
>      delivered in a fraction of one second yields ~ 2,000 events 
>      compared to only 2.2 events per second spontaneously.
>      
>      Therefore, while the data from acute dose survivors may be 
>      linear, the data at chronic exposures are likely swamped by 
>      spontaneous events (and the accompanying repair mechanism).  This 
>      doesn't explain any cellular or tissue responses that account for 
>      hormesis, but do indicate the basis for concluding that the risk 
>      from occupational levels of radiation is "lost in the noise."  
>      
>      There are caveats in this argument: 
>           Billen's table does not have data for spontaneous double     
>           strand breaks.  
>      
>           One can't say that a particular event caused by              
>           radiation cannot be the one damage that leads to             
>           transformation.
>      
>           The rates of damage (and repair) are not consistent with     
>           some known serious effects; that is, the tissue damage and   
>           subsequent death of persons exposed to radioactive sources   
>           (e.g. Mexico and Brazil) where the dose rates were not acute   
>           (hence the DNA damage rates are likely less than the         
>           spontaneous rate), yet over a period of many weeks to        
>           months, severe cell death occurred.  Maybe cellular division   
>           delay is a factor.
>      
>      One way to interpret these data is to assume, as Joyce Davis did, 
>      that the dose response curve is a flattened J, with the bottom of 
>      the J immersed in the background of spontaneous DNA damage.  There 
>      are many reasons not to permit the uncontrolled spread of 
>      radioactivity throughout our environment, and there are many 
>      reasons to adopt reasonable occupational and public dose limits 
>      (remember there are likely subsets of the population who are much 
>      more radiosensitive than the average - for example those with the 
>      gene for ataxia telangiectasia).  But we can certainly conclude 
>      that the NRC public dose limits are conservative, and lower values 
>      need not be applied to practices such as residual radioactivity 
>      standards, air emissions, etc.

Note that Joyce Davis' sigmoidal response applied to the radium population
(eg, Robley Evans HPJ 1974; Otto Raabe in HPJ Suppl 1 Vol 44 1983; and others) 
would also make the "linear model" rationale fabricated to justify regulatory
action at low doses "cost" thousands to millions of times more than the $
billions/life-year saved than we are already applying to radiation protection
policies. Also, she ignores the fact that the trough of the "J" is in the 10s
of rem acute, 100s of rem chronic regions so that a rationalization of a
policy to presumes/justify rad protection at low doses can not exist. 

Of course, if one looks at the doses not at a few mrem, or even few rem, range 
it is easy to see that there are consistent statistically significant and
not-significant reductions in health effects when a combination of the
population and the magnitude of the doses, and the natural occurrence of an
endpoint, make it possible (ie, breast cancer in the Canadian flouroscopy
women with 0.66 normal breast cancer at 15 rem exposure, 2.7 standard
deviations below normal, which BEIR ignores to report that the projection from 
high doses to zero is linear Miller, NEJM, 1989; Pollycove, BELLE, 1994).  

The same is true for the 1978-1987 $10M study of 700,000 Nuclear Shipyard
Workers, with a 70,000 worker cohort, the only government study with good
dosimetry and limited confounding effects, which DOE constrained from
publication, (while it sspported Steven Wing's fabricated junk pushed into
JAMA, even trashed by a JAMA editor in the same issue), and since the NSWS is
not published, BEIR and NCRP refuse to consider it in pushing the linear model 
and pushing the fabricated IARC "study" of government workers with junk
science ignoring positive data to project "linear trends" in the 3 groups
(including the high dose group) with negative data. 

Has anybody read the NCRP Report 121 dated Nov 30 that pushes the "Collective
Dose" nonsense at the behest of the regulators?  Pending receipt of a copy,
I'd appreciate any notes and comments from those who have. 

>      Eric Goldin
>      Southern California Edison
>      goldinem@songs.sce.com
>      Standard disclaimers apply

Thanks.

Regards, Jim Muckerheide
jmuckerheide@delphi.com