(190z) Effects of Immune Modulation on Melanoma Progression

Adeyinka Lesi, Richard M. White and David S. Rumschitzki

Cancer is a disease whose patient outcomes are difficult to predict. Some patients do not respond to treatment, some respond well and the rest appear to recover only to suffer a recurrence of the disease after a few years. In melanoma, for example, the disease reappears in patients 10 years or more after treatment at a 7% rate. This recurrence after a long period of dormancy (during which the disease is medically undetectable) is a feature in many cancer including breast cancer and metastatic melanoma. The causes of dormancy and recurrence are poorly understood though the fact that cancers evolve over time is an important factor. Melanomas can accumulate ~70,000 mutations, many of which increase the tumors ability to grow by, for instance, inducing angiogenesis or suppressing the immune system. In fact, recent development in treatments that activate immunity, so-called checkpoint inhibitors, show promise in treating otherwise intractable cancers such as metastatic melanoma. However, our current research neither focuses on situations where immunity is strong enough to completely prevent nor weak enough to be helpless to prevent tumor growth, since in these cases the patient outcomes are clear without a model. We instead investigate the case in which immunity balances tumor growth, and the patient can persist with the disease for a significant length of time. We describe a mathematical model, compare it to data and use it to gain insight into a potential mechanism for tumor dormancy and recurrence that is connected to the situation where these processes are in balance. We then seek to gain insights into dormancy and recurrence by conducting animal experiments where immunity and growth forces are balanced and interpreting the results using our mathematical model.

The overall effect of the immune system can best be studied inside a living host. The zebrafish animal model provides a platform to study the interactions of immunity and cancer. Zebrafish stripes are composed of melanocytes that are susceptible to human-like melanoma. A stripeless zebrafish (called casper) has been used along with a virulent fluorescently-labeled zebrafish melanoma cell line (ZMEL) to study how the disease progresses and to track it over time. We have developed a mathematical model that complements and rationalizes the data.

We posit a population balance model that tracks the tumor size distribution of an ensemble of tumors given the population average growth (mitosis) rates and shrinkage (cell death) rates as well as metastasis generation rates at any given tumor size. Since larger tumors have undergone more divisions than smaller ones, they have had the chance to accumulate a few more mutations and thus one can consider the rate parameters to be continuous functions of tumor size. Because tumor growth and shrinkage events occur randomly, a simple mathematical transformation shows that the tumor size distribution has the potential to expand is a diffusive manner. Most interestingly, for sizes at which growth and shrinkage rates are near equal, this diffusion becomes the primary mechanism by which the size distribution changes. The model is meant for large groups of patients, but in the context of an individual patient, diffusive behavior makes possible otherwise unexpected tumor behavior. Since we expect growth rates to depend on size, diffusion in tumor size space can make it possible for tumors to slowly pass from a size where growth and shrinkage are equal to a size where growth in dominant. This means the tumor may grow slowly or not at all for a very long time until it suddenly starts to grow very quickly. This behavior may correspond to the dormancy and recurrence experienced by cancer patients. The model presents a mechanistic approach to tumor dormancy that may be a useful new guide for research.

Previously, we measured melanoma progression after transplantation of a fixed number of ZMEL cells into zebrafish whose immune systems had been suppressed by irradiation and into zebrafish with no immune suppression. We compared these data to our model and found good agreement for optimal parameters. For our experimental conditions, the full effect of normal zebrafish immunity was able to kill all cancer cells in roughly 5 days after the cancer cells induce an immune response that begins 4-5 days after cell transplantation. Thus the immunity, once activated, is far stronger than mitosis. Our theory, in contrast, requires growth and immunity to balance in order to generate the most interesting situation of dormancy and recurrence. We would like to access such conditions experimentally.

Immune ablation via irradiation is inadequate to achieve the sort of immune modulation required. At any radiation dose that is nonlethal, fish immunity eventually returns, and a small dosage of irradiation means the fish immune system recovers quickly. In either case, the immune parameter is not time-independent. Moreover, repeated application of irradiation throughout the experiment is unworkable because irradiation would kill the tumor cells along with the immune cells and the effects of irradiation are cumulative so that even repeated exposure to small individual doses would eventually kill the fish. An alternative is immune suppression by chemical means. Dexamethasone, a steroid known to have immune suppressive effects, offers a tool for tuned immune modulation. Constant treatment with specific doses of dexamethasone can result in maintaining a constant level of immunity for long times.

We conduct experiments with several different dexamethasone dosage levels to allow the measurement of tumor growth at various levels of immune suppression. We then use the model to obtain parameters for the immunity as a function of dexamethasone dosage. Immunity should decrease with increasing steroid dose. This set of parameters will allow the identification of a dosage that achieves immunity parameters that balance those of growth (which should be unaffected by dexamethasone). Finally, we conduct experiments at the balance dosage to see if dormancy and recurrence are achieved. We will report the relationship between dexamethasone dosage and immunity and identify any dormancy/recurrent behavior observed. These experiments are ongoing and we shall report the latest results available at the time of the conference.