|Reviewer: Ryan P. Smith, MD|
|The Abramson Cancer Center of the University of Pennsylvania|
Authors: Sosman JA, Weeraratna AT, Sondak VK
This article serves to update the reader on the current state of vaccines for the treatment of melanoma. In doing so, it highlights some important studies, points out some inherent flaws in nonrandomized studies, and gives direction to future investigations. The article illustrates many points that are important to emphasize.
For more than 40 years, investigators have attempted to make cancer vaccines efficacious. Perhaps in no disease is this more prevalent than in melanoma. The reasons for this are many. First and foremost, melanoma usually has an aggressive course, with clinical remissions rare, with virtually no proven benefit to any nonsurgical treatments. Therefore, patients could benefit greatly from a novel and effective therapy. However, in order for a cancer vaccine to be efficacious, the immune response must be achieved. The immune mechanism requires that peptide (protein) fragments called antigens that are derived from within cancer cells are bound to the major histocompatibility complex (MHC) so that it can be presented to the T-cells. Also, there must be other cell-surface ("costimulatory") molecules on the antigen presenting cells (APC) as well as the T lymphocytes. In melanoma, functional T cells have been isolated and shown to recognize melanoma antigens (the peptide fragments). These T cells can be activated by the tumor stimuli and hence cause an immune response against the tumor cells.
The scientific community's knowledge of these facets of the immune response has greatly improved over the past decades. The development of an effective cancer vaccine, though, still awaits development. Our immune system is known to respond to melanoma and other cancer antigens, so the problem is not the antigen presentation or the presence of T-lymphocytes. Rather, the problem likely lies in the immunosuppression that is secondary to the tumor.
In the same issue of the JCO, Berd, et al. (JCO 22(3), 2004, 403) reports on 214 melanoma patients with advanced (with lymph node involvement) disease. The vaccine used was derived from modified tumor cells from each individual patient. With the vaccine tested, patients had a 5 year relapse free survival of 33% and an overall survival of 44%. Though these results are better than would normally be expected, it is impossible to say for sure whether this represents an advance, since the only comparison is a historical control group. This, of course, is open to numerous biases.
What are some of these biases-which could surface in this and any other non-controlled vaccine study? First, the vaccine was tested by placing a small amount under the skin (much like a test for tuberculosis). In patients with a skin reaction to the vaccine, there was an improved relapse free survival and overall survival. However, correlating this skin reaction to actual immune response against tumor cells is an intermediate end point. The significance of this artificial endpoint is unknown. It has been argued that the correlation between increased skin reaction and improved outcome is simply due to the fact that responding patients are inherently healthier, with better immune systems, and would therefore have improved outcomes, regardless of the efficacy of the vaccine.
Another bias that could explain the better results in patients receiving the vaccine is the fact that the expression of specific antigens on a tumor cell indicates that it is better differentiated and less aggressive. This has been demonstrated in melanoma and in many other tumors-that increased antigen expression equals well differentiated tumors with better prognoses.
Though these biases should be remembered, it is also very possible that the tumor vaccine was, in fact, efficacious. There has been one randomized trial of a melanoma vaccine. This study showed no improvement in overall survival or relapse free survival as a whole, though certain subsets of patients with specific types of major histocompatibility complexes had improved outcomes.
Future studies should also include intermediate end points, at least equal to the skin test type shown in Dr. Berd's study. Also, consideration should be given to enroll patients in randomized trials, to answer the question of the vaccine's efficacy. Given the heterogeneity of melanoma and peoples' immune systems, the question will not be answered in a definitive sense until a randomized trial is done. The major problem thus far has been the cost of producing these vaccines. As they are individually curtailed to each patient, these are expensive vaccines to produce. Therefore, the cost of proceeding with a randomized trial has thus far been too high to complete a study. A couple of trials have been designed and started, but not completed because of these issues. These trials must find a way to be completed, so that the efforts of vaccine development up until this point can be built upon, with improved outcomes as the hopeful result.