The prospects for cancer gene therapy

Gene therapy is a flexible technology with which to look for new ways of inhibiting cancer. However, the marginal success achieved has made it clear that direct engineering of cancer cells is more complex than had been supposed. The main barriers are raised by the difficulty of securing gene delivery into cancer cells in vivo and the selective advantages of those against which it is ineffective. These drawbacks do not arise when an immunological approach is adopted. Genes coding for tumor-associated peptides are used to engineer professional antigen presenting cells (APC). Alternatively APC pulsed with tumor antigens are engineered to overexpress costimulatory molecules or release cytokines. A more conservative approach is to engineer whole tumor cells with costimulatory and MHC molecules. Tumor cells can also be engineered to secrete cytokines and chemokines. The sustained presence of these factors in the tumor microenvironment recruits and activates distinct repertoires of APC and skews the antitumor response towards Th1 or Th2 reactivity. Engineered tumor cells are quickly rejected while mice acquire an immune memory against subsequent challenges, even when the tumor involved is poorly immunogenic. They also cure mice bearing incipient tumors and small metastases. This efficacy, however, vanishes as the tumor progresses. Even the best-induced specific immunity, therefore, is of no avail against advanced tumors. By contrast, the experimental data endorse the rational expectation that cancer vaccines will soon be both an established treatment of minimal disease after conventional management and a way of securing preventive antitumor vaccination.