Novel Vaccines for the Treatment of Gastrointestinal Cancers

Dr. Marshall has written a clear and concise review of the rationale for and preliminary data from studies using therapeutic vaccines in patients with established gastrointestinal (GI) malignancies. He has summarized the recent advances leading to a better understanding of basic immunology. These have had an important influence on the possibility that an effective therapeutic vaccine or vaccines can be developed. Attractive Approach
The idea that a patient's own immune system can be targeted to attack the patient's own malignant cells has long been an attractive one. An immunocompetent surveillance system may prevent the development of malignancy (a different topic) or may allow, with hopefully minimal toxicity, the destruction of established cancer, if an enhanced immune response can be guaranteed. While this approach has been extensively pursued, initial trials were disappointing. There may be several reasons why earlier immunotherapy vaccine trials failed. Dr. Marshall summarizes the difference between tumor-associated antigens that are specific to the patient and the identification of shared selfantigens that are not mutated. The logistics of developing a specific vaccine for each patient may be prohibitive, but one might expect patients to have several similar patterns of mutated genes. If so (for example, ras mutations are common in pancreatic cancer, and limited in number, type of mutation, and site),[1] this would not require a singlepatient vaccine, but rather, allow "customized" treatment to groups of patients. With the alternative strategy, the difficulty is in identifying nonmutated public antigens as well as ensuring that identified antigens are sufficiently immunogenic. Complex Process
It is here that a better understanding of the complex steps required for effective recognition of shared public antigens has led to a renewed sense of optimism in therapeutic vaccines. On the other hand, a better understanding of the role of a variety of molecules and cells in the efficient processing and presentation of antigen fragments has made the development of therapeutic vaccines into a complex process in its own right. Not only must the appropriate antigen or antigen fragment be identified (not a simple task for shared public antigens), but the appropriate vectors to ensure recruitment of the most potent antigen-presenting cells must be found. Dr. Marshall focuses on carcinoembryonic antigen (CEA) vaccines in GI malignancies. The frequent overexpression of CEA by malignant cells makes it an attractive target for a therapeutic vaccine. He describes the reasons for use of viral vectors to deliver antigens such as CEA or MUC-1 and the potential advantages of pox virus over the more commonly used adenoviruses. The use of costimulatory molecules has been explored by his group in several phase I trials. The complexity of this strategy is highlighted in the review. Viral vectors transfecting genes expressing the antigen or antigen fragment and the use of polyvalent costimulatory molecules illustrate the multiple steps being explored in the process of developing an effective therapeutic vaccine. Studies that have been performed to date are primarily phase I trials designed to identify an appropriate schedule and combination of these different factors. While toxicity has been modest, perhaps the most concerning longterm risks have not been explored in depth since the patients chosen for study are appropriately those with far advanced disease. Autoimmune Responses
One of the risks of any vaccine strategy is the development of autoimmunity. Since therapeutic vaccines may, in the end, be most effective in patients who have a minimum disease burden (using them in an adjuvant setting) and since a fairly large percentage of such patients may be expected to be cured of their disease, the risk of an autoimmune response is not insubstantial. For example, a CEA therapeutic vaccine given to patients with stage III colon cancer would be used in a population that, with currently available cytotoxic chemotherapy (an oxaliplatin(Drug information on oxaliplatin) [Eloxatin]-based regimen such as FOLFOX, with fluorouracil(Drug information on fluorouracil) and leucovorin), can be expected to have a 3-year disease-free survival rate in excess of 70%.[2,3] A vaccine would be used to decrease the risk of recurrence among the 30% of patients who will recur within 3 years of surgery. Since at the moment it is not possible to identify which patients have been cured by a combination of surgery and cytotoxic chemotherapy, all patients at risk would be vaccinated. Thus, 70% of patients would be exposed to a vaccine although they would be anticipated to have a normal life expectancy following the completion of therapy. Long-term follow- up of such patients is important. Concluding Thoughts
Lastly, the purpose of phase I trials is to identify the appropriate dose and schedule and composition of the therapeutic vaccine of interest. Response in these patients is a secondary objective. While it is of interest to note occasional responses to vaccine therapy, a better understanding of why they occurred in specific patients would be important. Stable disease is more difficult to interpret in patients who are part of such trials. In summary, therapeutic vaccines are an important area of clinical research. The complexity of these trials is nicely outlined in Dr. Marshall's paper. One might anticipate that vaccines are most likely to be beneficial in patients with a minimal disease burden, such as patients with stage III colon cancer. These studies might initially target very high-risk patients, such as those with stage IIIC disease. If long-term toxicity is minimal, lowerrisk groups could then be treated.