Cancer Immunotherapy Is More Than a Numbers Game

See allHide authors and affiliations

Science  06 Oct 2006:
Vol. 314, Issue 5796, pp. 68-69
DOI: 10.1126/science.1133893

The concept of manipulating the human immune system to fight cancer has spawned many avenues of potential immunotherapy, from cytokine, drug, and antibody-based treatments to cancer vaccines and gene therapy. Ever since the identification of the first human melanoma molecular marker, or antigen, in the early 1990s (1), tumor immunologists have attempted to boost the numbers of tumor-specific T cells in melanoma patients. The study by Morgan et al. on page 126 in this issue (2) is the first to combine four previously tested cancer treatments—vaccination, genetic engineering, adoptive cell transfer, and cytokine treatment—to tackle human melanoma, or any human cancer for that matter (see the figure). The results are promising, in that two patients who had failed previous therapies showed durable regression of metastatic melanoma. But the success rate of 2 out of 15 patients is disappointing given that all presently available immune weaponry was brought into play. That latter notion raises the question of whether we have now witnessed the maximal potency of cancer immunotherapy. There are many reasons to believe that this is not the case.

An immunotherapeutic regime.

Lymphocytes taken from a cancer patient are engineered to express a T cell receptor that recognizes a melanoma-specific molecule. Once expanded, these antitumor lymphocytes are transferred back into the patient to destroy cancer cells.


Increasing the number of desired tumor-targeting T cells can be achieved in two ways: by vaccination with a tumor-specific antigen or by the adoptive transfer of tumor-specific lymphocytes that have been expanded in cell culture. In the case of tumor-associated autoantigens—those molecules also present on one's normal cells—the breadth and quality of one's T cell repertoire may be limited by self-tolerance. This knowledge has prompted researchers to replenish this repertoire through the introduction of genes encoding tumor-specific T cell receptors into primary T lymphocytes (3, 4). Furthermore, various strategies, such as depletion of host lymphocytes and cytokine therapy, have been explored to overcome suppression of the T cell response by immunoregulatory and homeostatic mechanisms. In recent years, the efficacy of these strategies, and combinations thereof, has been tested extensively in preclinical murine tumor models. These animal studies have demonstrated that even though each of the single strategies may affect tumor growth, successful eradication of preexisting malignant tumors requires combination therapies (5). Because immunotherapy studies in human cancer patients exploiting single strategies have shown very little clinical efficacy, it was important to test the clinical impact of combination therapies. In this respect, the work by Morgan et al. represents a milestone.

As noted by Morgan et al., genetic engineering of antitumor T cells can be improved. Vectors used to transfer transgenes can be optimized to prevent silencing of the transgenes over time. Modification of transgene constructs to circumvent mispairing between T cell receptor chains encoded by transgenes and endogenous genes—resulting in decreased expression of functional receptors—also deserves attention. Furthermore, cells constituting the “helper arm” of the T cell response can be brought into action through the expression of T cell receptors with high affinity for tumor antigens. This is expected to enhance clinical efficacy by supporting long-lasting cytotoxic T cell immunity and by increasing the activity of innate immune effector cells. Previous antimelanoma vaccination studies showed tumor escape through loss of antigen expression (6, 7). Therefore, targeting of other antigens in addition to MART-1 (melanoma antigen recognized by T cells-1) is also expected to increase efficacy.

A striking aspect of the data presented by Morgan et al. is the high number of successfully engineered (MHC-tetramer positive), MART-1-specific cytotoxic T lymphocytes, detected in the circulation of many of the patients. These T cells constitute up to 1% of the cytotoxic T lymphocyte (CD8+ population, even at prolonged times after adoptive treatment. This suggests that the number of tumor-specific cytotoxic T lymphocytes is not the limiting factor. Because there is no overt correlation between systemic cytotoxic T lymphocyte numbers and clinical efficacy, the absence of therapeutic effects in most of the treated patients seems to be related to the failure of these cells to home in on the tumor and/or to exert their effector function in the context of the tumor microenvironment. Previous clinical antimelanoma vaccination studies have similarly shown that induction of high numbers of tumor-specific cytotoxic T lymphocytes does not necessarily result in antitumor efficacy (8, 9). Interestingly, patients in the present clinical study did not have tumor-infiltrating lymphocytes available for the treatment regime, suggesting that their cancers were relatively nonpermissive to lymphocyte infiltration.

These functional limitations are also suggested by the apparent absence of vitiligo in the treated patients, which is surprising because antimelanoma efficacy is often accompanied by autoimmune skin depigmentation (1012). It is conceivable that prolonged in vitro culturing negatively affects the in vivo homing capacity of cytotoxic T lymphocytes. Our knowledge of the requirements for efficient homing by (human) T cells is incomplete, so at present it is difficult to pinpoint how we could improve this further. Morgan et al. also used systemic administration of interleukin-2 (IL-2) to maximize cytotoxic T lymphocyte proliferation and function. However, toxicity limits the use of this cytokine in human subjects, and its impact on T cell immunity has both positive and negative aspects. Other cytokines, such as IL-15, may be more suitable for enhancing the performance of adoptively transferred cytotoxic T lymphocytes (13). Alternative tools for unleashing cytotoxic T lymphocyte function include antibodies that block the immune-regulatory effects of the T cell molecules cytotoxic T lymphocyte antigen-4 and PD1 (14).

The T cell immune response could likely also be improved through better vaccination strategies. In the Morgan et al. study, patients were repeatedly vaccinated with a synthetic peptide representing a minimal MART-1 epitope. But preclinical studies have shown that repeated injection of such minimal peptide epitopes can result in T cell tolerization instead of activation. Minimal epitopes can be exogenously loaded onto all antigen-presenting cells, many of which do not express the desired repertoire of T cell stimulatory signals, the lack of which is known to cause tolerance. This can be circumvented by providing larger antigens that require uptake and processing by dendritic cells, the most potent antigen-presenting cells for T cell activation (15).

The outcome described by Morgan et al. may not be perfect, but modifications of the treatment regime to increase clinical efficacy can readily be envisioned. Notably, the patients enrolled in the present study suffered from end-stage disease, exhibiting progressive metastatic melanoma, and were refractory to prior therapy with IL-2. Therefore, the best opportunity for improving treatment outcome probably lies in immune intervention in patients with less advanced disease. Cancer immunotherapy remains a great hope, and the work by Morgan et al. provides good reason to be encouraged.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
View Abstract

Stay Connected to Science

Navigate This Article