Vaccination strategies to reduce the risk of leukaemia and melanoma

In the industrialized world, the burden of infectious disease has been vastly reduced by improved standards of hygiene in the home and public places. But numerous studies, reviewed elsewhere,^1-3 suggest that a price may be paid in the rising incidence of diseases such as asthma and autoimmune disorders. According to this hypothesis, the maturing immune system no longer encounters the many microbiological challenges that evolution has led it to ‘expect’, and which contribute to its maturation and pathways of regulation—in other words, to its ‘education’.

Some of the recent reviews of this ‘hygiene hypothesis’^1,3 suggest that the postulates might be extended to other disorders with an immunological element, notably certain cancers. The ability of serious infections to induce regression of cancers has long been recognized. The phenomenon was observed in the 19th century in England, Germany and the USA by, respectively, Campbell De Morgan,⁴ Bruns⁵ and Coley.⁶ Coley induced erysipelas therapeutically in patients with sarcoma, with beneficial results in several cases. Since this was very risky in those pre-antibiotic days, he then experimented with extracts of streptococci and other bacteria and found these ‘Coley toxins’ to be likewise effective.

Not only may infections have an effect on established cancers but it is also possible that, by eliciting or enhancing antitumour immunity, they reduce the risk of tumours subsequently developing. In this context, Greaves observed that acute leukaemia, the commonest malignancy seen in children, occurs particularly in affluent societies.⁷ He therefore postulated that the disease is associated with an abnormal response to common infections of infancy and childhood associated with the altered environment of such societies. According to Greaves' hypothesis, the biological ‘norm’ is for the very young to encounter numerous infections, from the mother around the time of birth and from siblings and other contacts in infancy. Greaves therefore postulated that such infections would modulate the developing immune system, involving the expansion, suppression and elimination of certain T-cell subsets, in line with evolution, and that lifestyles in developed countries, with reduced exposure to such infections, would compromise this evolutionary adaptation of the immune system. The ‘immune proliferative stress’ resulting from an abnormal maturation of the immune system would favour mutations responsible for leukaemia. (It must be emphasized that Greaves' hypothesis refers to factors affecting the maturation of an intrinsically normal immune system. This is quite distinct from congenital or acquired immunosuppressive disorders which are associated with an increased risk of lymphomas rather than of leukaemia.)

Greaves' hypothesis has received strong support from studies in France and the USA.^8,9 Both pointed to a relation between protection against acute lymphoblastic leukaemia and exposure to common infections as a result of daycare attendance, and demonstrated that the degree of such protection was related to the amount of time spent in daycare facilities. The French study also related protection to repeated early common infections, surgical procedures for ear, nose and throat infections and prolonged breastfeeding. The latter, according to Greaves, has multiple immunological effects on the infant and results in the oral transmission of bacteria and viruses that contribute to normal immune maturation.

HYGIENE, LEUKAEMIA AND ALLERGIES—A LINK?

It is noteworthy that the factors protecting against leukaemia are among those that have been shown to protect children against allergies and asthma, both of which are on the rise in affluent nations.¹⁰ Protection against these disorders has been associated with exposure to farmyard muck, bacteria transmitted by the faecal-oral route, time spent in daycare facilities and having older siblings (presumably because they bring infections home from school). Of 53 studies included in a meta-analysis,¹¹ 48 revealed a protective effect of having older siblings on various allergic disorders including eczema (9 of 11 studies), asthma and wheezing (21 of 31) and hayfever (17 of 17). The authors emphasize that the reviewed studies were purely epidemiological in nature and shed no light on the mechanism. While the hygiene hypothesis could explain at least some of the observed findings, other factors (for instance, endocrine and in utero programming) merit consideration.

In addition, there are reports from the 1920s showing that, although patients with asthma or hayfever were often tuberculin positive, injections of small amounts of tuberculin appeared to be curative.¹² More recently tuberculin positivity, indicative of exposure to mycobacterial antigens by BCG vaccination, infection with Mycobacterium tuberculosis or, possibly, sensitization to environmental mycobacteria, was shown to confer protection against atopic disorders in Japan.¹³ This raises the question of whether natural infection by mycobacteria might contribute to immune maturation and protection against childhood leukaemia. Between 1911 and 1959 the UK saw the incidence of this disease rising by 4.5% a year, and this was a time when young children were becoming much less at risk of infection by milk-borne M. bovis, firstly as a result of pasteurization of milk and subsequently by removal of reactor cattle.¹⁴ Clearly, many other socioeconomic changes could, as suggested by Greaves, have led to alterations in immune maturation accounting for the increase in the incidence of leukaemia over that period, but mycobacterial infections as a factor in protection against leukaemia cannot be readily dismissed.

These observations pose the question of whether vaccination with BCG, an attenuated derivative of M. bovis, affords protection against leukaemia. On initial examination, the published work is confusing and contradictory. Studies in several countries, including the USA, Canada, Australia, Finland and Israel, showed that BCG vaccination protected against leukaemia and other malignancies of childhood and adolescence.¹⁵ On the other hand, a study of New Zealand children vaccinated at the age of 13 years showed no protection,¹⁶ while one conducted by Comstock in Alabama suggested that there might even be a slight increase in risk.¹⁷ On re-evaluation of these reports, it became apparent that protection against leukaemia and, in some studies, other cancers was seen only in regions where BCG was given neonatally and in which it afforded protection against tuberculosis.¹⁵

LEUKAEMIA, MELANOMA AND VACCINATION

It is possible that, in line with the hypothesis of Greaves, BCG vaccination beneficially affects maturation of the immune system in environments where other ‘expected’ stimuli are reduced or absent. A different, but not incompatible, hypothesis was proposed by Rosenthal who, in studies conducted in Chicago, was the first to observe that neonatal BCG vaccination reduced the incidence of leukaemia and other malignancies in childhood.¹⁸ Rosenthal postulated that one important function of T-cell-mediated immunity early in life is to seek and remove embryonic remnants capable of developing into cancers, and that this function could be stimulated by certain infections or vaccines (especially BCG) that induce cellular immune responses rather than antibody production. It is thus noteworthy that in Africa BCG, but not vaccines such as tetanus and diphtheria that primarily induce antibody production, seems to have beneficial impacts on general health beyond the specific infections against which it is aimed.¹⁹

Another cancer affected by previous vaccination is melanoma. A major study conducted in seven European countries and Israel under the auspices of the European Organization for Research for Treatment of Cancer (EORTC) showed that persons who had received BCG or smallpox vaccination, or both, had about half the risk of subsequently developing melanoma than unvaccinated control subjects matched for sex, age and ethnic origin.²⁰ The protection was greatest in younger persons (under 50 years of age). No protection was seen in Dijon, France, and indeed there was a non-significant trend towards an increased risk. The fact that this was the only centre that used the Institut Pasteur strain of BCG raises the possibility that daughter strains of BCG vary in protective efficacy against cancer. This might, for example, explain why Costa Rica, which practises neonatal BCG vaccination, has a high incidence of childhood leukaemia.

In another study, the EORTC group found that certain uncommon and severe infections with high fever (over 38.5°C), including pulmonary tuberculosis and sepsis due to Staphylococcus aureus, were associated with a subsequent reduced risk of melanoma.²¹ Less severe infections associated with high fever, such as influenza, afforded only slight protection, although repeated infections of this type enhanced the protection. Limited published evidence suggests that febrile infections may afford some protection against other cancers.^22,23

CELLULAR BASIS OF PROTECTION AGAINST CANCER

The cellular basis of protection against cancer by natural infection or certain vaccinations is poorly understood, as are the underlying mechanisms of immune regulation. It is, however, noteworthy that infants are born with a weak Th2 polarization which subsequently switches to a Th1 profile.²⁴ This switch is facilitated by environmental factors, principally bacterial infections,^2,25 and an absence of such factors could lead to the observed increase in the incidence of atopic disorders by facilitating a ‘Th2 drift’. Likewise, there have been several reports of an association of cancer with a Th2 drift and in some instances the extent of the drift was related to the extent of the disease or was a predictor of poor outcome.²⁶

In this context, neonatal BCG vaccination induces or enhances Th1 responses.²⁷ Although cause and effect are hard to separate, beneficial effects on cancer induced by immunotherapy with BCG or heat-killed M. vaccae have been associated with shifts towards Th1 cytokine production.^28,29

In considering the impact of environmental factors in diseases associated with immune dysfunction, one must bear in mind that there are several, and probably many, types and patterns of such dysfunction. In all probability, the observed Th1 and Th2 balances and functions are merely superficial aspects of complex immunoregulatory networks and disturbances of the same, with many positive and negative feedback phenomena. Thus, autoimmune diseases are associated with either Th1 or Th2 activity and leprosy reveals a clinicopathological spectrum related to different forms of immune dysfunction.

For this reason, although immune dysregulation may underlie increased risk of allergic disorders and cancer, the relation between the two classes of disease may be complex and even paradoxical. Thus, although some studies indicate that allergy is associated with an increased risk of cancer, others show no association or even apparent protection. When Eriksson et al.³⁰ looked for allergy in a cohort of 6593 patients with cancer they found complex relations between the conditions and called for studies on the exact immunological basis of the allergy and the organ specificity and type of cancer.³⁰ As an example of the complexity of such relations, atopy appeared to be associated with a reduced risk of melanoma: 4.2% of melanoma patients were atopic compared with 10% of controls.³¹ A suggested explanation is that atopic subjects are more prone to febrile infections which, as outlined above, are associated with protection against melanoma. Another is that atopic dermatitis, beginning early in life, leads to a reduction in the number of pigmented naevi in the skin,³² where melanomas commonly originate.

CONCLUSION

Whatever the mechanism, and however complex the immune interactions, there is growing evidence that various environmental factors, specific infectious diseases, regular exposure to low-grade infections and certain vaccination strategies affect the subsequent incidence of acute leukaemia, melanoma and possibly other cancers. As the incidence of these diseases, with their need for burdensome and distressing therapy, is increasing in the affluent world, strategies based on ensuring a more natural maturation of the immune system would be of great benefit. This is especially so in view of increasing evidence that the incidence of other diseases as diverse as asthma, allergy and autoimmune diseases is adversely affected by the same maturation defects. Clearly, nobody would advocate a return to the era of widespread and uncontrolled epidemics of childhood infection. Rather, a simple preventive strategy—namely, universal neonatal BCG vaccination—and encouragement of lifestyles favouring a more natural maturation of the immune system, should receive prompt and serious consideration. Thereby, the empirical observations of Campbell De Morgan published in 1874 could at last be translated into a rational and simple means of reversing the rising tide of cancer in many parts of the world.³³