In police circles, entrapment is frowned upon as a means of securing an arrest. But, as reported by Akira Takashima and colleagues in the January issue of Nature Biotechnology, it could be a legitimate way of creating cancer vaccines without having to resort to costly ex vivo approaches.

A promising way of generating cancer vaccines uses dendritic cells — specialized antigen-presenting cells. But to generate an effective vaccine, they must be collected from the patient and then subjected to time-consuming and costly 'customizing' procedures — in which they are expanded and loaded with tumour antigens and/or cytokine-encoding genes — before they can be re-administered to the patient. Could this lengthy ex vivo process be replaced by a simpler procedure?

Langerhans cells (LCs) — dendritic cells that reside in skin — generally stay put unless they are induced to mature. Maturation involves acquiring receptors for chemokines such as macrophage inflammatory protein 3β (MIP-3β), allowing them to move along a chemokine gradient from the epidermis to draining lymph nodes. This response can be triggered by haptens — small molecules that are not antigenic unless they are associated with a larger molecule such as a protein. But what if a 'decoy' lymph node could be produced that diverts LCs away from their real destination to a site where they can easily be pulsed with antigen? To this aim, the authors produced ethylene–vinyl–acetate (EVA) rods that released MIP-3β, and implanted them just under the abdominal skin of mice. Application of fluorescein isothiocyanate (FITC) — a hapten that also doubles up as a fluorescent signal for tracking LC migration — revealed that MIP-3β-expressing rods 'trapped' LCs by attracting them to the rods. By contrast, after 24 hours, a significant number of LCs had migrated from the epidermis to the draining lymph nodes in mice with either no implanted rods or rods expressing a control protein.

So, LCs can be trapped in one location, but can they also be loaded with tumour antigens? To investigate this, rods expressing ovalbumin were implanted with those expressing MIP-3β, and were, again, treated with a hapten. T cells harvested from the spleens of these mice five days after hapten treatment had a strong cytotoxic T-lymphocyte response against an ovalbumin-transduced cell line, and this was comparable to what could be achieved with a standard dendritic-cell vaccine.

The therapeutic and preventative efficiencies of the vaccine were also tested by inoculating mice with ovalbumin-expressing tumour cells either one day before or five days after administration of the vaccine, respectively. The vaccine had a 50–60% therapeutic efficacy, but provided mice with almost full protection when administered before tumour inoculation.

But will this vaccine strategy also work with tumour antigens isolated from a patient's tumour — an essential prerequisite for its successful translation into the clinic? The authors incorporated crude extracts from the S1509a fibrosarcoma cell line into the EVA rods and co-implanted them in mice with rods expressing MIP-3β. Following hapten treatment, the mice had cytotoxic T-lymphocyte activities that were able to lyse S1509a targets.

This anticancer vaccine strategy therefore seems to be as effective as traditional DC vaccines, but has the advantage of being in situ. Let's hope that these new technologies will soon provide real benefit to patients in the clinic.