Mortality and cancer in Porton Down subjects

BMJ 2009; 338 doi: https://doi.org/10.1136/bmj.b358 (Published 25 March 2009) Cite this as: BMJ 2009;338:b358
  1. Malcolm R Sim, director, Monash Centre for Occupational and Environmental Health
  1. 1School of Public Health and Preventive Medicine, Monash University, Melbourne, VIC 3004, Australia
  1. malcolm.sim{at}med.monash.edu.au

    Risk is not increased, but some health questions remain

    The experimental studies of the effects of chemical warfare agents conducted on thousands of British military personnel over several decades at the Porton Down facility in the United Kingdom have been an ongoing source of controversy regarding the scientific, ethical, and moral environment in which they took place.1 Concerns about effects on health increased following a coroner’s finding in 2004 of unlawful killing regarding the death of a Porton Down subject in 1953 after cutaneous administration of a chemical warfare nerve agent in a non-therapeutic experiment. This finding understandably raised important questions about the longer term health of the 18 000 British veterans who took part in the Porton Down experiments.2 To investigate this, two linked studies have assessed whether the risks of cancer or mortality are higher in veterans who took part in tests compared with those who did not.3 4

    Established in 1916 in response to the use of chemical warfare agents against British troops in the first world war, the Porton Down facility was later expanded because of the threat posed by Germany during the second world war; the use of chemical weapons by Iraq; and the later attacks by terrorists against civilians, such as the release of sarin in a subway in Tokyo in 1995. In response to increasing public concern, the UK Ministry of Defence commissioned an epidemiological study to investigate a range of health outcomes in the military personnel who took part in the Porton Down chemical weapon experiments.

    More than half of the Porton Down veterans were exposed to known or probable human carcinogens, most commonly dermal exposure to sulphur mustard,5 so Carpenter and colleagues’ study (doi:10.1136/bmj.b655) on cancer morbidity is of considerable scientific and public interest. The authors found no significant differences in the overall incidence of cancer nor in the incidence of most cancer types in Porton Down veterans compared with non-veterans.3 Although a few subgroups had some isolated excesses of specific cancer types—such as excess of cancer of the trachea, bronchus, and lung in Porton Down veterans exposed to Lewisite—the lack of smoking data and the large number of analyses make it difficult to draw any firm conclusions about the causal nature of these few findings.

    Venables and colleagues’ study of mortality (doi:10.1136/bmj.b613) reported a small excess of overall mortality in the Porton Down veteran group.4 However, this overall excess was mainly the result of excesses in non-cancer causes of death, such as deaths from infectious and parasitic diseases, circulatory diseases, genitourinary causes, and external causes. Some of these excess deaths are probably related to the longer period of service of Porton Down veterans compared with that of the comparison veteran group in the study, rather than to their involvement in the Porton Down experiments. Importantly, the authors found no increase in cancer mortality, consistent with the main finding in the cancer morbidity paper.3

    The authors acknowledge that their studies have some limitations. Because UK cancer data have been available only since 1971, cancer rates in the early years of the Porton Down programme could not be studied. Also, it was not possible to adjust their findings for important confounders, such as cigarette smoking. One of the major strengths of these studies was the painstaking and time consuming attention to detail in the assessment of exposure, which is not easy when extracting data from records that are several decades old.6

    The overall findings are consistent with other studies of veterans with short term or intermittent exposure to chemical weapons. These include the study of US navy veterans from the second world war, who participated in mustard gas chamber tests, which found no excess in any cause specific mortality.7 Also, the mortality study of US army Gulf war veterans who were potentially exposed to nerve agents during the March 1991 chemical weapons demolition at Khamisiyah in Iraq, reported no significant findings, apart from a borderline excess of deaths from brain tumours.8

    The findings of the two studies should provide some reassurance to Porton Down veterans and their supporters that they have no excess risk of cancer or major causes of death. Although human carcinogens such as sulphur mustard and benzene were used in the Porton Down experiments, it should be remembered that the level and duration of exposure are important determinants of future cancer risk. In this study, the veterans took part in the experiments for only one to four weeks, for an average of two days a week. Therefore, the cumulative doses received were probably small compared with industrial exposure in manufacturing facilities, and—as the studies found—too low to cause a measurable excess of cancer.

    Although these findings are reassuring for cancer and mortality outcomes, we need to take a more global view of the health of Porton Down subjects. This research group has previously published results for some other outcomes, which—although they have some methodological limitations—showed excess reporting of symptoms and poorer quality of life in a subgroup of Porton Down veterans.9 The finding of excess symptoms has some similarities with reported effects in many other veteran groups after deployment to areas of high personal threat.10 Although the nature of the threat to Porton Down veterans was different from that experienced by other veterans, non-fatal and non-cancer health outcomes in Porton Down veterans remain an important area of future research and public interest.


    Cite this as: BMJ 2009;338:b358



    View Abstract

    Sign in

    Log in through your institution