Saturday, March 31, 2012

Massey University and the nuclear test veterans

The supreme court recently ruled that surviving veterans of the United Kingdom's nuclear test series, conducted in Australia and the Pacific during the 1950s, will not be permitted to sue the Ministry of Defence for alleged ill-health caused by the tests.

The story, however, raises a number of interesting issues. Firstly, there is the question of the reliability of unverified personal testimony. Rose Clark, widow of Michael, is reported by The Guardian as stating that her late husband "was so close he could see the bones of the people on the beach beside him. It was like an x-ray."

This is a telling claim because x-rays are actually invisible. As a consequence, x-rays cannot be directly used to see the bones in a human body. In a medical x-ray, the x-rays pass through soft tissue, to be captured on a photographic plate. Human bones partially attenuate the x-rays, hence the bones can be seen as a shadow on the photographic image.

The second issue, not reported by The Guardian, is that the health of the nuclear test veterans has already been rigorously assessed by successive epidemiological studies conducted by what was then the National Radiological Protection Board, now part of the Health Protection Agency:

"Based on this work the HPA conclude that nuclear weapons test participants had, in general, a better life expectancy than members of the general UK population. When compared with the control group, the test participant group had similar overall patterns of mortality and cancer incidence indicating no significant cause for concern.

"The statistical analyses also provided a slight indication that test participation may have caused a very small increased risk of leukaemia but there was not enough evidence to confirm this as a fact and there was evidence to suggest that this finding should be treated with caution."


The third issue is the question of new research. According to The Guardian, "the veterans had contended that they did not have proper knowledge that their illnesses were connected to the atomic tests until medical research was published in 2007."

Now, this new research transpires to be a study, conducted by the Institute of Molecular Biosciences at Massey University, of so-called chromosome abberrations amongst New Zealand veterans who attended the British tests. There are several different types of chromosome abberration, but the primary result obtained by the Massey University team analysed a type of abberration called a translocation, in which parts of different chromosomes are swapped.

The technique involves taking blood samples from the subjects of the study, and counting the number of abberrations. To infer a dose, it is then necessary to calibrate the method by exposing blood samples in vitro to known doses of radiation to obtain a dose-response curve, such as that seen above here.

Everyone's DNA is subject to a continuous barrage of intrinsic damage, so any attempt to infer a radiation dose from DNA damage will need to distinguish radiogenic damage from intrinsic damage. This means that each method of counting chromosome abberrations will have a detection limit associated with it. In other words, each technique will only be able to detect radiation doses above a certain minimum level.

In the case of translocation frequencies, various estimates of the detection limit can be found in the literature. Writing in 1997, A.A.Edwards claimed that "the scoring of translocations...results in reduced sensitivity at low doses so that acute X-ray doses of about 0.3 Gy and chronic doses of about 0.4 Gy are at the limit of measurement...a final limit to these approaches exists because of the higher level of spontaneous translocations...in cells of unirradiated persons."

The unit of dose referred to here is the Gray (Gy). A Gray is a large unit of radiation: most people receive an annual background radiation dose comparable to only about 0.002 Gray. So 0.4 Gray is two hundred times or so the background dose, and that's possibly the detection limit for this technique; if the dose is any lower, it may be impossible to distinguish it from the random level of DNA damage.

But let's have a look and see what the team from Massey University claimed to have found. They published two papers, one in 2007, 'New Zealand Nuclear Test Veterans' Study – a Cytogenetic Analysis', presented to the New Zealand Nuclear Test Veterans’ Association, and a second in 2008, 'Elevated chromosome translocation frequencies in New Zealand nuclear test veterans', published in an academic journal, Cytogenetic and Genome Research 121:79–87.

In their 2007 report they calculate (p31) that there were 37 individuals who received a dose in the range 0 – 0.49 Gy, 6 individuals in the range 0.5 – 0.99 Gy, and 5 individuals who received greater than 1 Gy. The highest dose estimates here are above the range in which the detection limit for this technique might lie. But as a consequence, they're also very large doses; doses so large that, were they to be delivered all at once, could lead to symptoms of radiation sickness!

These dose estimates, we learn, were obtained by comparing the abberrations counted in the veterans with those in "Blood samples from 3 healthy donors (mean age 40.5) [which] were irradiated with 60 Co at a dose rate 0.835Gy/min to different doses (0, 0.2, 0.5, 0.75, 1, 2 Gy)," (p15).

Something strange now happens when we turn to the 2008 paper, published, let us recall, in a peer-reviewed journal. We now find that "Dose estimates ranged from 0 to 0.431 Gy in the veterans (mean = 0.170 Gy)," (p85).

All of a sudden, the mean dose is down to 0.170 Gy, which lies beneath what might well be the detection limit for this technique. And how are these new doses estimates inferred? Well, the blood-samples from the forty-year olds have been discreetly disposed of, and in their stead we find "in vitro exposures of a blood sample of a normal donor of age 60."

There's no explanation of why the blood samples used to calibrate the technique have changed. Which is strange, because if the Massey University team discovered that something was wrong with their initial approach, one would expect them to explain and report this fact, so that the rest of the scientific community could learn from their research.

The Massey University team emphasise the difference between the number of translocations found amongst the veterans, and the number found amongst a control group (whose mean dose was inferred to be 0.037 Gy), but given that both dose estimates are below the detection limit for the technique, it is unclear why this should be considered of dosimetric significance.

Why were samples from several 40-year olds used in the first paper to calibrate the technique, and then a single sample from a single 60-year old used in the second paper? How sensitive are the dose estimates to the choice of calibration sample? Detection limits are not so much as mentioned in either the 2007 paper, or the 2008 paper; why do the team from Massey University not even discuss this issue?

All these questions remain unanswered. Which is disappointing if you're a scientist rather than a lawyer.

No comments: