Large tumors (circles) appear more than a year after mice turn on an oncogene, long after T cells are disabled.

Cancer cells outwit the immune system long before they debut as visible tumors, say Willimsky et al. (page 1687).

Despite a wealth of studies, researchers still don't know when the immune system is alerted to the presence of cancer cells or when its response is subverted. Part of the problem stems from the fact that T cell responses in most animal models are tracked only after tumors form.

To get an earlier start, Willimsky et al. designed a mouse model in which a gene silencer was inserted near an oncogene. When a random mutation disabled the silencer, the mice expressed the oncogene and eventually developed cancer. These mice could be identified by the presence of antioncogene antibodies, which appeared before tumors formed. The presence of antibodies indicated an activated immune system, allowing the authors to study T cell responses as the earliest possible time during cancer development.

The researchers now find that the T cells fail months before tumors appear. Antitumor antibodies often appeared by six months of age, at which point the T cells were already disabled. Tumors became visible about nine months later. As the T cells were unable to keep the cancer cells at bay, the lag until tumor formation might instead stem from the requisite time needed to accumulate additional, proliferation-inducing mutations. The growing tumors did not lose their ability to elicit a T cell response, as injecting them into precancerous mice provoked tumor rejection.

As tumors grew, the defect in cancer-specific T cells spread to all cytotoxic T cells. Proponents of past tumor models attributed this generalized apathy to a threshold level of tumor-derived, T cell-inhibiting molecules that increase as a function of tumor size. The authors ruled out this idea by engineering mice that expressed the oncogene during embryogenesis. T cells in these mice treated the oncogene as “self” and thus allowed tumors to grow unchecked. These tumors did not, however, hamper the ability of other T cells to recognize foreign antigens.

Others have blamed the generalized T cell defects on the suppressive cytokine TGF-β and/or a population of immature myeloid cells that inhibit T cells in vitro. Indeed, mice that turned on the cancer gene had high TGF-β levels and growing numbers of immature myeloid cells at the time they had developed antitumor antibodies. How oncogene-induced antibody responses might lead to TGF-β production and myeloid cell growth remains to be seen.