Cancer Immune Therapy Edited by G. Stuhler and P. Walden ...

218 10 The Immune System in Cancer: If It Isn't Broken, Can We Fix It?
10.7
Combining the Best of Both Worlds
It is already clear from both in vitro and in vivo tumor model systems [31±35, 38, 39],
as well as from clinical experience of autoimmune disease, that the immune system
can be re-educated to recognize self/near-self antigens to the point of mounting
clearative immune responses against them. In fact, a surprisingly wide variety of different
approaches (Fig. 10.6) have proved effective in animal models of tumor rejection
[4, 5, 12, 40±56]. The literature of tumor immunotherapy in (transplantable) rodent
models (i.e. immunocompetent mice injected with tumor cells derived from
syngeneic mice) is so full of different success stories that it would seem that curing
cancer in mice is little more than a case of choose your gene and use it! Yet immunotherapy
in human oncology remains only very much at the periphery of clinical
utility [57]. This discrepancy highlights the deficiencies of the animal models of tumorigenesis
that have been used in the past and, in reality, such systems may more
faithfully represent models of parasitic infections than growth of human tumors.
This is because many of the tumor cell lines used to establish tumors in animal
models have been passaged for many years in tissue culture and may have lost many
of the characteristics that define true tumors which develop over time in the presence
of an intact host immune system [58]. Nonetheless, despite criticisms of the
models used, it is now possible to rationalize the apparently highly diverse range of
successful approaches into a few common principles. Most successful immunotherapies
for cancer culminate in a common pathway that starts by tumor cell killing
(Figs 10.5B and 10.6).
Tumor cell killing can be achieved in many ways ± directly by cytotoxic gene expression
or indirectly through activation of both specific (T cells) or non-specific [59] (natural
killer, eosinophils, macrophages) immune mechanisms through expression of
co-stimulatory molecules, adjuvants, cytokines, immunogens and so on (Figs 10.5B
and 10.6). However, not all tumor cell death is equal [9, 37, 60±63]. In general, it is
not only important to kill tumor cells, but to make sure that they `suffer' as they die,
i. e. that they show biochemical markers of a `stressful death' [60]. This labels them
as `pathologically abnormal' to the immune system (Fig. 10.4A) ± distinguishing
them from cells that are undergoing `normal' cell death, such as occurs during development
[9, 37, 64]. In particular, cells dying a stressful death express heat shock proteins
[48, 60, 62, 65] and cause the unhindered [60, 66] release of intracellular molecules
that are natural activators of DCs [61, 63, 67, 68] (Fig. 10.5B). By contrast, `normal`
cell death is likely to lead both to induction of tolerance to cellular antigens at
the level of DC function [6] and to the neat packaging and removal of all of these materials
by phagocytes before they can signal immunological danger (see Appendix)
[37, 60, 66, 69±71].
There is also a fine balance between the mechanisms of cell death and the levels of
that death. Even if cells die a stress-free death by apoptosis, if the levels of apoptosis
are too high for phagocytes to clear up the debris, autoimmune pathology results
[72, 73]. It might be possible to exploit the reactions to this form of cell death for
antitumor vaccination [74, 75]: if a good source of tumor antigen is available, then