Review of NCRP Report No. 136 "Evalation of the Linear-Nonthreshold Dose-Response Model for Ionizing Radiation" (NCRP, 2001-06-04, ISBN 0-929600-69-X)

by Klaus Becker (Boothstr. 27, D-12207 Berlin/Germany, phone/fax x4930-7721284, prof.dr.klaus.becker@t-online.de)

On first glance, this report provides, with 211 pages of well-formulated text, supported by about 600 references and numerous figures, the impression of a comprehensive and balanced up-to-date summary of the present situation, not surprisingly concluding that the LNT hypothesis is basically correct.. The reader will notice a few minor caveats such as the statement in the conclusion (p.210): "The available information does not suffice to define the dose-response curve unambiguously for any neoplasm in the dose range below 0.5 Sv, and it indicates the existence of substantial thresholds for the induction of some types of neoplasms.". It also mentions "osteosarcoma, for which the existing data suggest the existence of effective thresholds" (p. 138). It does, however, not really discuss (or show in a figure as in UNSCEAR 2000) the impressive radium painter data, with 1500 individuals and 45 sarcomas (Rowland 1997) establishing clearly a threshold around 10 Gy (or, according to ICRP, 200 Sv).

In reading the report, the reviewer remembered a pertinent quotation from another book: "Why do we know what we know, and why don’t we know what we don’t know? If the politics of science consists (among other things) in the structure of research priorities, then it is important to understand what gets studied and why, but also what not gets studied and why. One has, in other words, to study the social construction of ignorance. The persistence of controversy is often not a natural consequence of imperfect knowledge, but a political consequence of conflicting interests and structural apathies. Controversy can be engineered; ignorance and uncertainty can be manufactured, maintained, and disseminated." (R. N. Proctor, Cancer Wars – How Politics Shapes What We Know & Don’t Know About Cancer, Basic Books, 1995, ISBN 0-465-00859, p. 8).

Obviously, it is not easy for a body such as NCRP/SC 1-6 to cover reasonably completely this rapidly expanding field of scientific discourse (frequently degenerating into quasi-religious strife) with its swelling tide of publications, and to judge their relative merits, in particular of those outside the U.S. literature and the realm of the English language. Another problem is the necessary cut-off date for publications being considered. On the other hand, certain tendencies to "selective citation" in favour of the currently dominating LNT hypothesis, which enjoys the support of most regulators, administrators, and funding agencies, cannot be ignored in reading NCRP 136. Lacking the time, the administrative and financial support for a detailed analysis comparable to NCRP, only a few examples for this observation may be mentioned.

A man-made large scale "experiment" probably equivalent in importance to the Hiroshima and Nagasaki studies, namely the radiological consequences of the Chernobyl accident, have been treated in Annex J of the UNSCEAR 2000 report on 115 pages, with over 450 references, showing even among relatively highly exposed cohorts no increase in radiation induced genetic or somatic effects, besides the about 30 acute radiation syndrome casualties and the easily treatable thyroid cancers in children, causing another 1-3 deaths during the past 15 y. These data should have been known to the SC, and should have been discussed much more extensively than in a few short scattered remarks.

A highly relevant recent paper by distinguished authors demonstrated that the relative lung cancer risk in humans decreases with increasing external low-LET dose up to about 1 Gy, and exceeds the normal rate at 2 Gy (H.H. Rossi and M. Zaider, Radiogenic lung cancer: The effect of low doses of low-LET radiation, Radiat. Environm. Biophys. 36, 85-88, 1997). These data have not been considered in NCRP 136 (as well as in UNSCEAR 2000) without giving any reason. The highly developed art of "politically correct" data selection is also evident other parts of the report.

For example, the publications of many distinguished scientists working in this field (e.g. M. Tubiana, M. Pollycove, G. Monchaux, R. Mitchel, J. Conrady et al.) are missing in the reference list, while others such as J.H. Lubin are quoted with nine references on the same subject of residential radon epidemiology. This subject is indeed an interesting test case, but NCRP 136 does not adequately discuss the crucial aspect of the high uncertainties in the retrospective smoking determination of the cases, which would probably invalidate most case-control studies. Besides, Fig. 9.4 and 9.5 are obviously not very convincing. The important subject of radon balneology, which is successfully – as demonstrated in randomized medical double-blind studies – applied as a therapy for rheumatic and arthritic diseases for about 75.000 patients just in Germany and Austria annually, is not mentioned at all..

There have been several important conference proceedings, containing a wealth of relevant recent data and references, which should have been available to Subcommittee 1-6 in time before the publication of the report in June 2001, for example

1. "The Effects of Low and Very Low Doses of Ionizing Radiation on Human Health" (539 p., Elsevier Excerpta Medica Internat. Congr. Series 1203, ISBN 0-444-50513-x). Inclusion of some of the contributions, in particular those by M. Tubiana, A. M. Kellerer, K. R. Trott and M. Rosemann, P. Duport, K. Becker, G. Monchaux and J.-P.Morlier, M. Pollycove and L. Feinendegen, R. Masse, and G. Walinder would have definitely influenced the conclusions.
2. "Biological Effects of Low dose Radiation" (Ed. T. Yamada et al, same Series No. 1211, ISBN 0444 50431 1) also contains much information about bystander effects, apoptosis, adaptive response (to which NCRP 136 devotes a total of 3.5 pages), and cancer epidemiology (with important contributions in particular by S. Kondo, H. Tanooka, O. Yamamoto and T. Seyama).

The latter proceedings also contain two well-documented reports about the very comprehensive and careful epidemiological studies in Kerala, one of the world’s highest natural radiation level areas and justly called "a unique natural laboratory", by K. M. Nair et. al., and G. Jaikrishan et al. In areas of 10-40 mGy/y (mostly external gamma) radiation, no detrimental genetic or somatic health effects have been detected in large cohorts in comparison with essentially identical low-level populations living nearby. These results had also been published in Rad. Res., Dec. 1999.   In general, epidemiological studies of populations in high natural, NORM and TENORM situations are underrepresented in the report. This bias in favour of LNT can also be observed in the selection of figures, etc.

Regarding radiation worker epidemiology, studies with nuclear shipyard workers in the U.S. (Matanoski 1993), British radiologists, employees in nuclear energy and research facilities, etc., mostly show a decrease in the cancer incidence, and an increase in life expectancy. This could be explained as adaptive response, or a biopositive response to a moderately increased exposure level. Instead, there is the "politically correct" explanation with the "healthy worker effect". It is difficult to understand, why radiologists in Britain should live much healthier than other medical practitioners. There are many other, more specific points to be criticised in NCRP 136, as pointed out in longer and much more detailed criticism, e.g. by Z. Jaworowski and M. Waligorski with 33 references.

Missing are also some reflections on the economical, social and political consequences of the LNT hypothesis, including the closely related Collective Dose concept . With the widespread habit of misusing this concept to multiply non-demonstrated (and unlikely) "risks" of very small doses with large population numbers, frightening numbers of cancer deaths have frequently been "calculated" with far-reaching consequences, e.g. in cleanup costs. In summary, the report is a valuable compilation for those supporting the LNT hypothesis, but it is easy to imagine an equally profoundly documented report, written by an equally reputable group of experts, which would come to quite different results. "LNT or not LNT, that is the question" remains open to further, hopefully serious and unbiased, studies and analysis.