The intuitive appeal of eliciting an effective immune response against cancers has long been recognized but, until quite recently, rarely fulfilled. At the end of the 19th century, William Coley, a New York sarcoma surgeon, noted some tumour regressions in cancer patients infected with streptococci, by provoking an immune response.1 But such responses proved hard to replicate and his treatments quickly fell out of favour. The idea, however, remained potent. Occasional instances of spontaneous tumour regression2,3 or prolonged dormancy suggested some form of ‘host restraint’, and a number of clinical successes kept the field of immunotherapy alive, despite the many failed attempts at treatment.
As the science of immunology developed, together with an emerging understanding of the innate and adaptive immune system, so interest in its function in modulating tumours was rekindled. The use of intravesical Bacillus Calmette–Guérin (BCG) to stimulate regression of superficial bladder cancers was the first true immunotherapy to be adopted, in the 1960s. The mechanism of spontaneous regression was investigated and, with a viral aetiology for cancer suspected, antibodies were thought to play a key role, as were newly discovered cytokines and other signalling peptides.2
The invention of hybridoma technology in 19754 overcame previous difficulties in making large quantities of specific immunoglobulin, allowing antibodies directly targeting tumour antigens to be tested for the first time. The results were initially disappointing, owing to the short half-life and poor recruitment of human immune effector mechanisms by murine antibodies; however, with the development of molecules including human constant regions more successes were seen, such as the targeting of clusters of differentiation (CD) 52 antigens on lymphoma by the Campath-1H antibody.5
Despite theoretical reservations about the depletion of normal B cells and poor effector capacity in patients with advanced lymphoma, rituximab proved an effective treatment for B cell malignancies, in 1997 becoming the first monoclonal antibody to be licensed for use in humans. This chimeric monoclonal antibody, targeting CD20, has several mechanisms of action once attached to the target cell, activating antibody-dependent cytotoxicity and complement-dependent cytotoxicity, as well as driving apoptosis.6 Its impact in improving survival from B cell lymphomas has been so marked that it has been classified by the WHO as an essential medicine.7,8 Subsequently, an ever-increasing number of monoclonal antibodies targeting cancer cells themselves have been licensed. They either evoke effector mechanisms, in a similar manner to that of rituximab, or act by inhibiting aspecific stimulatory pathways, such as the use of trastuzumab to block activation of the epidermal growth factor (EGF) pathway in human epidermal growth factor receptor 2 (HER2)-positive breast cancer.9
In the 1990s, through the work of groups at the Ludwig Institute for Cancer Research, in Brussels, and others, it became apparent that malignant cells do evoke responses by CD8-positive cytotoxic T cells and CD4-positive T helper cells through presentation of self ...