Natural Killer Cells and Cancer: Immune Surveillance Revisited
February 1, 2001
Natural Killer Cells and Cancer: Immune Surveillance Revisited
ABSTRACT & COMMENTARY
Synopsis: The presence of immune surveillance against cancer has been difficult to establish in humans. In this longitudinal analysis of a relatively large cohort of individuals for whom a baseline level of lymphocyte (NK cell)-mediated natural cytotoxicity, it was found after 11 years that a lower baseline level of activity was associated with a greater risk for cancer. This report may provide the strongest evidence for the presence of immune surveillance in humans.
Source: Imai K, et al. Lancet 2000;356:1795-1799.
The concept of immune surveillance, originally proposed by Burnet over three decades ago, remains controversial.1 Although there is an increase in cancer in those with profound immune deficiency, such as in patients with AIDS or those receiving immunosuppressive medications after organ transplantation, the tumors most commonly encountered are those that theoretically reflect immune dysregulation rather than immune deficiency (e.g., lymphoma, leukemia, and Kaposi’s sarcoma).2 However, there have been published reports that would suggest that individuals with diminished immune function, particularly natural cytotoxicity, are at increased risk for development of malignancy.3,4 In the current report, a large cohort (n = 3500) of people from a single Japanese community were followed for the development of cancer over an 11-year period.
In 1986, 8552 residents (95% of those ³ 40 years of age) of a single Japanese town (Saitama) were sent an epidemiological questionnaire and, of these, 3625 self-selected individuals (1371 men and 2254 women) agreed to provide peripheral blood samples. The questionnaire topics were those primarily of lifestyle and included dietary habits, cigarette smoking, alcohol ingestion, body weight, and height. Peripheral blood was analyzed for a variety of biochemical and immunological parameters, including natural killer cell activity by standard chromium (51Cr) release assay using K562 leukemia cells as targets. Eleven years later a follow-up survey on cancer incidence and death from all causes was undertaken.
After excluding early cancers (those which occurred within 2 years of the baseline assessment), there were 154 cases of cancer among the 3500 participants. Individuals were analyzed in tertiles based upon the initial cytotoxicity response (low, medium, and high responders). The age-adjusted relative risk of cancer (all sites) was 0.72 (95% confidence interval [CI] 0.45-1.16) for men with high cytotoxic activity, and 0.62 (0.38-1.03) for men with medium cytotoxic activity, taking the risk of those with low cytotoxic activity as reference. For women with high cytotoxic activity, the relative risk was 0.52 (0.28-0.95), and for women with medium cytotoxic activity the relative risk was 0.56 (0.31-1.01). For both sexes with high and medium cytotoxic activity risks were 0.63 (0.43-0.92) and 0.59 (0.40-0.87).
The results were interpreted as supportive of the concept that NK cells mediate resistance to cancer development, as those with low activity were shown to have increased cancer development over this 11-year span.
COMMENT by William B. Ershler, MD
Many immunologists have laid to rest the concept of immune surveillance, primarily because its very existence has been difficult to establish in human beings. For one thing, the tumors that develop in profoundly immune deficient people (and animals, for that matter) have been those more reflective of immune dysregulation than immune deficiency. If one were to implicate a failure of immune surveillance as an explanation of the common cancers in the general population, then one would expect a similar pattern of malignancy (e.g., lung, colon, breast, and prostate cancers) in patients with AIDS or after organ transplantation. Or should we? Perhaps those common cancers are associated with a more subtle decline in immune (or natural killer) function and because of their longer preclinical stage (estimated to be a decade or more for most epithelial cancers), their development is just not apparent in patients with AIDS or transplant recipients, many of whom will not live long enough with their illness to develop these cancers.
The Japanese cohort is a remarkable contribution in this regard. A careful analysis of natural cytotoxicity was performed in a relatively large group of adults, some of whom later went on to develop cancer (primarily stomach, lung, and colon). Those with lower baseline cytotoxicity, particularly women, were at significantly greater risk for cancer development. This analysis is perhaps the strongest evidence to date that a decline in lymphocyte function is associated with the occurrence of the common malignancies encountered in the general population. Inasmuch as NK cell function is influenced by lifestyle characteristics (such as chronic stress, dietary habits, and cigarette smoking), positive adjustment of these factors may possibly reduce cancer incidence by immunological mechanisms. This, of course, is theoretically appealing but totally lacking in scientific documentation.
References
1. Burnet FM. Progr Exp Tumor Res 1970;13:1-27.
2. Penn I. Transplant Res 1977;9:1121-1127.
3. Herberman RB, et al. Science 1981;214:24-30.
4. Hersey P, et al. Br J Cancer 1979;40:113-122.