COMMENT: The threshold vs “linear no-threshold” (LNT) showdown: Dose rate findings exposed flaws in the LNT model part 2. How a mistake led BEIR I to adopt LNT


Thought your chat group might be interested in this 2009 study with dibenzo[def,p]chrysene (dibenzo[a,l]pyrene), a polycyclic aromatic hydrocarbon (PAH) in 40,000 rainbow trout [see attached 2nd pdf].  We were able to statistically calculate the dose resulting in 1 cancer in 5000 animals, a 50-fold increase in power compared to the 2-acetylaminofluorene ED01 mouse study.

The dose response became sub-linear with dose, and the intercept on the x-axis for 1/106 tumors was more than 1000-fold higher than if the linear extrapolated dose (LED10) had been used.  Surprisingly, the liver DNA adducts remained linear with dose (would not have made for a good biomarker).

I was surprised that this paper did not generate more interest.  Maybe because it (“was only a fish”).  George Bailey spent years documenting the validity (and superiority of fish to mice), with regard to aflatoxin B1 (AFB1)-induced hepatocellular carcinoma (HCC) as a model for human liver cancer.  I repeated the same design –– later with AFB1 –– and that also generated a dose-response that became sub-linear, compared to the LED, but the data were not as dramatically as that for the PAH.

For what its worth,

This entry was posted in Center for Environmental Genetics. Bookmark the permalink.