Louis M. Weiner, MD
Vice President, Translational Research
Chairman, Department of Medical Oncology
G. Morris Dorrance Jr Endowed Chair in Medical Science
Fox Chase Cancer Center
Philadelphia, Pennsylvania
Anecdotes add human dimensions and urgency to scientific presentations about cancer. However, nobody wants to read about a little girl who grew up and lived to see her grandchildren marry because she never developed a fatal cancer. We need more of these non-stories. The goal of curing cancer is of course irresistible, but it is even more desirable to prevent cancers from developing in the first place.
Disease prevention can be achieved through numerous interventions. These include the adoption of healthy behaviors, the identification of people at high risk of developing cancer, and through the reduction of harmful environmental exposures. Vaccines that prevent serious or fatal illnesses such as polio, smallpox and influenza have had major impacts on public health. (1) Surprisingly little effort has been expended on developing vaccines that might prevent cancer, despite their potentially huge benefits.
Vaccines can prevent cancers caused by viruses. For example, the discovery of Hepatitis B virus by Blumberg and colleagues led to the creation of Hepatitis B-directed vaccines that prevent hepatitis-B infections and dramatically reduce the incidence of primary liver cancer in vaccinated populations.(2) More recently, a vaccine directed against some forms of human papillomavirus has been approved for use with the reasonable and highly achievable goal of greatly reducing the rate of cervical cancer.(3) We now can attack many different cancers, not only by eliminating infectious causes of cancer-predisposing states, but also by exploiting the many recent advances in understanding how the immune system both fights and nurtures cancer development. However, virus-directed vaccines are only applicable to a few cancers. If vaccines are to prevent most cancers it will be necessary to use them to initiate attacks against developing cancer cells that appear as “self” to the immune system.
For years it has been assumed that the human immune system is either a disinterested bystander or an ineffective weakling with regard to cancer development. However, this is not the case. As early cancers develop, the immune system detects and eliminates cancer cells that appear to be “foreign” to the host. However, since cancers are genetically unstable, cancer cell populations contain cells that are able to evade host-protective immune responses using a variety of mechanisms.(4) The malignancies that eventually emerge are invisible to the host, and may potently and actively suppress the host’s anti-cancer immune response. Hence, established cancers are superbly well adapted to survive in their hosts, having been honed by powerful immune selection mechanisms that have stripped the cancers of their most immune-stimulating properties.(5) This perverse example of “evolution gone wild” has some important implications. The first is that it may prove to be very difficult to induce effective immunity against fully established cancers. However, it is also true that cancers may be most vulnerable to immune attack early in their development, when they are less genetically deranged. At such early stages an additional environmental selection pressure (in this case imposed by a vaccine-initiated immune response directed against one or more tumor targets) can cut off some of the escape paths that cancers may otherwise employ to evade immune destruction. Even if immunization has subtle effects on immune control of cancer progression, the downstream consequences may be amplified, and thus have profound impacts on the ultimate development of human cancers within the ordinary human lifespan. This new understanding of the ways in which cancers effectively defeat the host-protective anti-cancer immune response has led to the development of a large “toolkit” of reagents and ideas that can be employed to target emerging cancers.
Why is this perspective important? Firstly, the idea of targeting emerging cancers focuses on the reversal of cancer-derived immune suppression, and is not restricted to the amplification of an existing immune response or tricking the immune system into “discovering” that the cancer is somehow foreign to the host. Perhaps equally importantly, this new perspective requires initial testing and conceptual validation in a non-traditional population for the development of a new cancer therapy. Ordinary drug development typically commences in people suffering from advanced cancers, and validation of a therapeutic effect in such populations justifies testing in progressively earlier disease states. Tamoxifen is an excellent example of this concept, as it was originally used to treat advanced breast cancer. This hormonal agent then found a role in the adjuvant therapy of the disease, and eventually was used to prevent breast cancer in patients at high risk of developing the disease. However, this linear approach to drug development may not be applicable to cancer vaccines because patients with advanced cancers typically suffer from generally suppressed immune response capabilities and may neither benefit from vaccination (thus failing to provide justification for further testing) nor exhibit toxicities that would be seen in a healthier population with a longer projected lifespan. Moreover, as discussed above, advanced cancers have perfected their immune evasion and immune suppression properties, and may not be suitable predictors of what might happen to arrest or retard the development of a developing cancer. Indeed, there are no current examples of cancer vaccines that effectively treat established cancers, just as vaccines against other illnesses are not used to treat established diseases, but rather to prevent them from occurring.
These considerations lead to the inescapable conclusion that cancer prevention vaccines should be tested initially with the goal of preventing, but not treating cancer. While the concept is intuitively sensible and attractive, its implementation is associated with a number of formidable challenges. For example, it should be noted that as tamoxifen was tested in progressively less advanced disease settings its modest toxicity profile became a subject of increasing concern.(6) This is an important consideration that is highly relevant to cancer prevention vaccines, since any observed toxicities will occur in people for whom a future cancer is deemed highly possible, but not necessarily inevitable. Vaccines that cause even a low rate of serious toxicities might be considered to be appropriate if they could cause even modest benefit for advanced cancer, but the same vaccines certainly would be considered reckless and unacceptable in the high risk setting. Since vaccines are designed to break immune tolerance to “self”, all effective vaccines will pose the potential risk of autoimmunity, and thus it may be many years before cancer prevention vaccines that target cancer cells will be tested in the general population.
So, in the near term, it seems likely that the best way to develop cancer prevention vaccines will be to carefully match exciting new vaccine concepts with conditions that are characterized by high risks for development of a particular cancer. Such conditions may be based upon genetic analysis (e.g., HNPCC or BRCA1 gene mutations), may be inherited disorders such as familial adenomatous polyposis (FAP) or may be known clinical syndrome that frequently lead to cancer, such as Barrett’s esophagus with high-grade dysplasia. Relatively small studies could prove to be very informative if the proportion of patients destined to develop more advanced neoplastic lesions or invasive cancers is sufficiently high. The field of cancer prevention has struggled to identify biomarkers or surrogates that might indicate that a particular intervention has potential as a cancer prevention agent. Blocking or delaying the development of real cancers in a reasonable time frame could provide an exciting justification for advancing a particular vaccine concept into a larger population of at-risk people. So, while it may be some time before most cancer vaccines can be used to prevent cancer in people who do not have known high risks of cancer development, we now have the ideas, the tools and a blueprint for determining the best path forward.
References
- Sutter RW, Maher C. Mass vaccination campaigns for polio eradication: an essential strategy for success. Curr Top Microbiol Immunol. 2006;304:195-220.
- Blumberg BS. Hepatitis B virus, the vaccine, and the control of primary cancer of the liver. Proc Natl Acad Sci U S A. 1997 Jul 8;94(14):7121-5.
- Schiller JT, Lowy DR. Prospects for cervical cancer prevention by human papillomavirus vaccination. Cancer Res. 2006 Nov 1;66(21):10229-32.
- Pure E, Allison JP, Schreiber RD. Breaking down the barriers to cancer immunotherapy. Nat Immunol. 2005 Dec;6(12):1207-10.
- Smyth MJ, Dunn GP, Schreiber RD. Cancer immunosurveillance and immunoediting: the roles of immunity in suppressing tumor development and shaping tumor immunogenicity. Adv Immunol. 2006;90:1-50.
- Breast International Group (BIG) 1-98 Collaborative Group; Thurlimann B, Keshaviah A, Coates AS, Mouridsen H, Mauriac L, Forbes JF, Paridaens R, Castiglione-Gertsch M, Gelber RD, Rabaglio M, Smith I, Wardley A, Price KN, Goldhirsch A. A comparison of letrozole and tamoxifen in postmenopausal women with early breast cancer. N Engl J Med. 2005 Dec 29;353(26):2747-57.
Possible Entry Points for the Development of Cancer Prevention Agents
Cancer vaccines may not require testing in patients with established malignancies. Arrows depict the direction of drug development. Chemotherapy agents typically are first tested in later stages, and then applied in progressively earlier disease settings. Anti-viral vaccines can be applied to at-risk people because of the low toxicity profiles. As shown, cancer vaccines are perhaps best tested initially in the high-risk setting, and then moved into the average-risk population should they prove to be effective and safe.
