Search tips
Search criteria 


Logo of bmjThis ArticleThe BMJ
BMJ. 2007 October 13; 335(7623): 779.
PMCID: PMC2018803

Francis John Caldwell Roe

Dr Francis John Caldwell Roe, who has died aged 82, was for over 50 years active as an experimental pathologist and made major contributions to cancer research and toxicology. He had a wide range of research interests, which included the general toxicology and potential carcinogenicity of foods, food additives and contaminants, drugs, tobacco, pollutants, and industrial chemicals, as well as mechanisms of carcinogenesis, cancer epidemiology, cancer prevention, and the pathology of laboratory animals. This resulted in over 800 publications, including eight books and many leading articles in the BMJ and Lancet. His encyclopaedic knowledge, coupled with his ability to present his findings in an extremely clear and logical way, made his advice highly sought after. He served with distinction for many years on the UK Department of Health and Social Security's committees on carcinogenicity and on toxicity and on the WHO Expert Advisory Panel on Food Safety. He was also a member of numerous other national and international expert committees, as well as of various scientific journal editorial boards, and was a highly valued consultant to a wide range of companies.

Francis Roe was born in London on 16 August 1924 and was educated at St Olave's Grammar School, Orpington, and then at Wadham College, Oxford, and the London Hospital Medical College. He obtained from Oxford his BA in 1945, BM BCh in 1948, MA in 1950, and DM in 1957. He took up a house appointment at The London Hospital in 1948, continuing his pathology experience at the Royal Army Medical College from 1949 to 1951 before returning for 10 years to The London Hospital to be lecturer, then senior lecturer in the Department of Cancer Research. During this period he had a year at the McArdle Memorial Laboratories in Madison as an exchange fellow. In 1961 he moved to the Chester Beatty Research Institute and was there until 1971, during which he obtained his DSc (London) in 1965 and his FRCPath in 1967. He was also associate pathologist and honorary consultant to the Royal Marsden Hospital in this period, but found time to gain considerable experience in general medical practice as a part-time locum. By this time he had already published very widely, and his advice was sought by various industries. In 1971 he joined the Tobacco Research Council as their research coordinator, but in 1973 he decided to set up as an independent consultant in toxicology, experimental pathology, and cancer research. This proved highly successful and his advice continued to be sought over a long period of time by a widening range of food, drug, chemical, and tobacco companies.

Roe's early work was concerned with mechanisms of carcinogenesis and in particular demonstrating the profound effect on tumour incidence rates of the order in which chemicals were applied. These experimental initiation/promotion studies on the skin, many published with his colleague Myer Salaman, underlined the complexities of the cancer process. His writings emphasised this, pointing out that cancer is a group of diseases, each with multiple causes.

When he started his work cure was the main interest for most cancer researchers, but he emphasised the importance of cancer prevention in many of his publications. Towards this aim he investigated the possible carcinogenicity of a large number of agents. These included tobacco smoke (where he conducted much work on its chemical constituents, such as nicotine, 3,4-benzpyrene, and the tobacco-specific nitrosamines), asbestos, talc in mineral oils, the drugs metronidazole and cimetidine, chloroform, and various dietary components (including salt, sugars, lactose, saccharin, aspartame, and vitamin A). Some of his investigations concerned more unusual agents. Thus, in his offices in Wimbledon, he kept species of the euphorbia plant for which he had shown the latex to be a strong cancer promoter.

In the mid-1970s he was involved in thought provoking studies on cancer and ageing. It is well known that the risk of most cancers is much higher in the old than the young, and it was held by many that this was due to decrease in the body's resistance to cancer with advancing age. On the basis of studies in which the age at start of exposure to chemical carcinogens varied, Roe and his colleagues showed that, at least under some conditions, cancer incidence rates depended wholly on the duration of exposure and not at all on the age at the start of treatment. This observation was of profound importance to the understanding of mechanisms of cancer.

Roe's career started when the use of rodents to test new chemicals for carcinogenicity was in its infancy. While the assay's fundamentals have barely changed over several decades, he was well aware of its strengths and limitations and took a keen interest in contentious issues relating to it, in particular the optimum diet to use for the laboratory animals. He planned and executed the huge Biosure study, which showed that restricting the dietary intake of untreated rats to 80% of their usual ad libetum intake dramatically reduced the incidence of cancers of various types. He pointed out that laboratory animals were typically overfed and obese, and that variations in cancer incidence between untreated and chemically treated rodents may arise not because the chemical had any true carcinogenic effect but because it happened to affect the appetite of the animal. He coined the term “pseudocarcinogenicity” to describe the enhancement of tumour risk by a non-genotoxic mechanism (such as endocrine disturbance secondary to overfeeding) in physiologically abnormal animals.

When reading slides from experimental studies he was keenly aware of the importance of consistent diagnostic criteria. Later in his career he also became involved in studies into the accuracy of diagnosis of human cancer. Based on autopsy studies in Hungary, a country where postmortem examinations are routinely conducted in patients who die in hospital, he found that a large proportion of some cancers diagnosed at necropsy are not detected clinically and that a similarly large proportion of cancers diagnosed in life are not confirmed at necropsy.

Roe did not have statistical qualifications but was conscious of many important ways in which inadequate use of statistics could produce misleading results, and was one of the first to be keenly aware that analysis of tumour incidences without adjustment for survival differences can lead to erroneous conclusions as to whether a chemical is deemed carcinogenic or not. He worked closely with statisticians such as Malcolm Pike, Richard Peto, and, during his period of independent consultancy, Peter Lee. In the late 1970s Roe and Lee collaborated in ideas for the development of a computer system for pathology data which would allow relevant data to be collected and statistically analysed in an unbiased way. With the major contribution of the statistician John Fry, this ROELEE system has continued development to this day.

While he had extremely high standards and his papers were of an excellent quality, his writing was never dry. He had a mischievous sense of humour, and his papers and reports could be very amusing to read. In 1990 after a spell in a well-known London hospital involving considerable time on a toilet seat which proved more uncomfortable than any pain stemming from his scars and tubes, he published a letter in the Lancet entitled “Flat seats for convex bottoms,” advocating a superior design.

Roe had a long and distinguished career and one that I am sure he found most fulfilling and rewarding. In recognition of 50 years of scientific endeavour, a fitting tribute was paid to him in 2002, when a special issue of Food and Chemical Toxicology was published in his honour. He had been closely associated with this scientific journal since its inception in 1964. Leading scientists from the United Kingdom, other European countries, and the United States who knew him well, submitted papers to this special issue as a tribute to a man who had engendered so much respect in the pathology, cancer research, and toxicology communities.

One charity that was particularly close to him was Marie Curie Cancer Care. His enormous contribution over many years as a trustee and committee member of the charity's cancer research institute was rewarded on becoming a life vice president of Marie Curie Cancer Care in 1996. He was extremely supportive of the high grade molecular-biology research undertaken in the research institute's laboratory in Oxted.

Roe was an extremely kind man with time for everyone, and he was always willing to help young pathologists and other scientists to develop their careers. This kindness, as well as his exceptional breadth of knowledge in pathology and toxicology, his critical and balanced judgement, and his considerable abilities to interpret toxicological data and evaluate its significance to humans, is among the reasons why he will be sorely missed.

In his spare time he was an extremely skilful portrait sculptor. He became honorary treasurer of the Medical Art Society in October 1972 when the society was at a low ebb. He circulated a questionnaire to the members, one of whom, Baron ver Heyden de Lancey, a philanthropic colourful character qualified in medicine, dentistry, and the law, responded positively by putting the society on a sound financial and organisational footing. With Roe's encouragement he supported a link between the society and the Royal Society of Medicine, and also provided in perpetuity funds for art prizes. In recognition of his generosity the Medical Art Society asked the Baron to accept the title of patron of the society. Both he and the Baroness sat for Roe for portrait heads which were cast in bronze and are on display at the Royal Society of Medicine. Roe's study of Professor Cuthbert Dukes, the famous histopathologist, is on permanent display at the Royal College of Pathologists. Thus Roe was instrumental in regenerating the Medical Art Society and remained president until 1979.

Francis Roe died on 8 August 2007, following complications of pneumonia.

Roe married, in 1948, Brenda Beckett, who survives him with their two sons, two daughters, and seven grandchildren.


Former independent consultant in toxicology, experimental pathology, and cancer research (b 1924; q Oxford/The London 1948; MA, DM, DSc, FRCPath), died from complications of pneumonia on 8 August 2007.

Articles from The BMJ are provided here courtesy of BMJ Publishing Group