Advertisement
Commentary Free access | 10.1172/JCI18359
Department of Medical Oncology, Dana-Farber Cancer Institute and Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts, USA
Address correspondence to: Glenn Dranoff, Dana-Farber Cancer Institute, Dana 510E, 44 Binney Street, Boston, Massachusetts 02115, USA. Phone: (617) 632-5051; Fax: (617) 632-5167; E-mail: glenn_dranoff@dfci.harvard.edu.
Find articles by Dranoff, G. in: JCI | PubMed | Google Scholar
Published April 15, 2003 - More info
The most important achievement of cancer immunology thus far may be the development of robust techniques for the identification of tumor antigens (1, 2). This work underlies our current understanding that cancer patients frequently generate specific cellular and humoral antitumor responses. Moreover, the expression of transformation-associated stress genes commonly provokes innate immune reactions (3). Together, these findings unveil a previously unsuspected breadth of immune recognition in tumor bearing hosts.
The characterization of cancer cell antigenicity has fueled efforts to delineate protective immune-effector mechanisms. The task is complicated by the dual role that immunity plays during cancer progression. Recent studies disclosed a marked increase in the incidence of spontaneous and chemically-induced tumors in immunodeficient mice compared to littermate controls (4). Since immunocompetent animals efficiently reject tumor transplants from immunodeficient hosts, the experiments support the idea that the immune system functions as an extrinsic tumor suppressor. Consistent with this concept, clinical-pathologic investigations established a strong association between the presence of dense intratumoral T cell infiltrates and favorable clinical outcomes in patients with malignant melanoma or carcinomas of the colon, kidney, and ovary (5, 6).
Other work indicates, though, that tumors may subvert the immune system to facilitate disease progression (7). Unresolved inflammation, whether due to infection, autoimmunity, or environmental agents, markedly increases the risk of cancer. Dysregulated cytokine production promotes cell proliferation and attenuates apoptosis. Phagocyte-derived reactive oxygen species damage DNA. Tumor cell invasion and metastasis exploit the normal cues for leukocyte migration. Collectively, these studies illustrate diverse ways in which the immune system sculpts the hallmarks of cancer (8).
The dual role of immunity in tumor suppression and progression underscores the possible benefits and risks of cancer immunotherapy. In this issue of the JCI, Forni and colleagues provide important insights into the differing requirements for vaccine-induced elimination of incipient and established tumors (9).
The investigators examined the immune response to Her-2/neu–positive breast carcinomas originating in transgenic mice. Although the tumors express rat Her-2/neu, which differs from the murine homolog by 6%, the model system is intriguing, since Her-2/neu contributes to the pathogenesis of human breast cancer and is a target for clinically efficacious monoclonal antibodies (10). Whereas previous work showed that prophylactic vaccination with tumor cells or DNA encoding Her-2/neu prevented spontaneous tumor formation in transgenic mice (11, 12), the current study explores the mechanisms of tumor rejection. Breast cancer cell lines, established from spontaneous tumors, were implanted into syngeneic mice rendered deficient, through gene targeting or antibody depletion, in various immune components. While the ectopic injection of transplantable tumors does not fully recapitulate spontaneous carcinogenesis, the analysis of tumor rejection in a large panel of immunodeficient mice constitutes a major strength of the investigation.
Prophylactic immunization with DNA encoding Her-2/neu efficiently stimulated the rejection of subsequent tumor challenges in wild-type mice. CD4+ T cells were essential for immune priming but not tumor rejection. CD8+ T cells partially contributed to tumor destruction, likely through the concerted actions of perforin and IFN-γ, but neutrophils were absolutely required for rejection. In contrast, antibodies, Fc receptors, macrophages, CD1d-restricted NK T cells, perforin, and interferon (the latter two individually) were dispensable. Whereas neutrophils are typically associated with acute inflammatory reactions, a role in cancer defense was previously suggested by experiments using cytokine-secreting tumor cells (13). Activated neutrophils may lyse cancer cells directly and compromise the tumor vasculature.
Therapeutic vaccination against established tumors proved far more stringent than prevention, and nearly all of the immune components tested contributed to tumor rejection (Figure 1). Successful treatment required a coordinated response involving CD4+ and CD8+ T cells, antibodies, Fc receptors, CD1d-restricted NK T cells, macrophages, neutrophils, perforin, and IFN-γ. Although the basis for this rigorous requirement remains to be clarified, it is tempting to speculate that the rapid growth kinetics of established tumors demand a robust reaction. A large number of tumor cells may also include variants with acquired resistance to some modes of immune attack.
Coordinated tumor immunity. Forni and colleagues (9) show that immune-mediated rejection of established Her-2 positive breast tumors requires CD4+ and CD8+ T cells, CD1d-restricted NK T cells, neutrophils, macrophages, antibodies (Ab’s), Fc receptors, IFN-γ, and perforin.
The substantive hurdles to rejecting established tumors suggest that immunotherapies will need to stimulate broad and sustained host responses. Unfortunately, these same requirements also imply that tumors may accomplish immune escape by devising strategies that undermine diverse immune mechanisms. The combination of immunotherapies and other treatments that target resistance pathways may thus be worth exploring.
The study also provides insight into why prophylactic immunization, in which there is considerable redundancy of immune effectors, may be more easily achieved than therapeutic vaccination. Indeed, preventive immunization against pathogen-related tumors already demonstrates great promise for reducing the burden of hepatocellular and cervical carcinoma (14, 15). Whereas autoreactivity, as an unintended consequence of immunity to aberrantly expressed self-antigens in tumors, may ultimately limit cancer vaccination, the rules underlying the discrimination of tumor and self are still poorly understood. Carefully conducted vaccine trials in patients at high risk for cancer may shed light on this crucial issue.
This work was supported by NIH grants CA74886, CA39542, and CA66996, the Cancer Research Institute, and the Leukemia and Lymphoma Society.
See the related article beginning on page 1161.
Conflict of interest: The author has declared that no conflict of interest exists.