As tumour cells are genetically unstable, tumour antigens are not ideal targets for anticancer therapy. By contrast, tumour-associated fibroblasts are genetically more stable, which makes them attractive targets. Ralph Reisfeld and colleagues have used a DNA vaccine against fibroblast activation protein (FAP), which is overexpressed in most stromal fibroblasts associated with colon, breast and lung cancer. They have shown that this vaccine can suppress tumour cell growth both on its own and when given with chemotherapy.

When mice were immunized orally with the pFAP DNA vaccine and then challenged with multidrug-resistant colon or breast carcinoma cells, tumour growth was suppressed. To simulate a therapeutic setting, mice were first challenged with colon carcinoma cells and then vaccinated, and the authors observed a decrease in the growth of established lung metastases.

So, what are the effectors of the anti-tumour immune response? After immunization and subsequent tumour challenge, the authors depleted the CD4+ and CD8+ T cells and natural killer cells in the mice using antibodies. Only the depletion of CD8+ T cells decreased the anti-tumour immune response. Furthermore, CD8+ T cells purified from mice that had been vaccinated with pFAP induced the apoptosis of tumour cells transfected with fluorescently tagged pFAP, but not tumour cells that were transfected with fluorescent label alone. This shows that the pFAP vaccine breaks peripheral T-cell tolerance against the self-antigen FAP. In addition, the tumours that did grow in pFAP-vaccinated mice showed a marked infiltration of CD8+ T cells.

Fibroblasts are the primary source of collagen type 1, and the expression of collagen type 1 has recently been shown to correlate inversely with the intra-tumoral uptake of various chemotherapeutic drugs. The authors therefore investigated whether their pFAP vaccine would decrease the amount of intra-tumoral collagen type 1 and increase the uptake of anticancer drugs. Tumour tissue from vaccinated mice did indeed show decreased expression of both FAP and collagen type 1. Vaccinated mice were first challenged with breast carcinoma cells, which are partially sensitive to doxorubicin, and then treated with doxorubicin. Vaccination with pFAP or treatment with doxorubicin suppressed but did not eradicate tumour growth, but vaccination and drug therapy in combination led to complete tumour rejection in 50% of the mice. This experiment was repeated in mice with established metastases, and treatment with pFAP and doxorubicin extended the lifespan of these mice threefold.

As FAP is stably overexpressed in most colon, breast and lung carcinomas, and vaccination increases the intra-tumoral uptake of chemotherapeutic agents, this is an attractive approach to study further.