Investigations into Tumor dormancy

Tumor dormancy - a state of tumor development characterized by halted growth - is highly prevalent in both cancer patients and normal populations. The state of tumor dormancy may occur throughout various stages of tumor development. Clinical and experimental observations suggest micro-metastasis in distant organs may be dormant for extended periods, as may minimal residual disease left after surgical removal of the tumor mass or other treatment of primary tumors. For example, often patients believed to have a good prognosis post therapy, nonetheless develop metastases, in some cases many years after apparently successful treatment of their primary cancer. The concept of grossly uneven rates of tumor growth extend to larger tumors as well. Mammographies are revealing new tumor lesions which, based on rates of growth calculated post-detection, should not have been missed on the previous mammogram unless the previous growth rates of these tumors were much faster and they subsequently enter a state of much slower growth or near-dormancy. These punctuated periods of attenuated growth throughout tumor development lead to many questions, such as how best to assign risk of late recurrence, how long to monitor patients, what mechanisms regulate tumor dormancy and which mechanisms enable tumors to advance from the dormant state and renew tumor mass expansion. Despite the high prevalence of dormant tumors, little is known and much remains to be learned about tumor dormancy. Both experimental studies, as well as computational and systems mathematical modeling done at our CCSB are shedding light on the biological and molecular mechanisms responsible. Kinetically, two mechanisms have been proposed. Single-cell dormancy has been attributed to cell cycle arrest, while dormancy of larger tumor masses has been attributed to balanced proliferation and cell death. Cell-intrinsic, population-intrinsic, and host interaction effects are potential causes. Among the latter, the immune system, angiogenesis, and niche 'permissivity' are recognized as potential bottlenecks to tumor progression. These may operate independently. Our group has recently developed a dormancy model in immunodeficient mice, pointing to angiogenesis as a likely controller of exit from dormancy in this case.

Angiogenesis, Dormancy, and the Timeline Paradigm for Tumor Development

Interestingly, the "angiogenic switch" and a "shift in the angiogenic balance" have been used interchangeably to describe the angiogenesis-dependent transition from dormancy to aggressive growth (see e.g. Bouck, Adv Cancer Res. 69: 135-74, 1996). Each, however, really carries a different implication. A "switch" suggests a potentially irreversible transition from a non-angiogenic to an angiogenic state, and has been used often in reference to the acquisition of vasculature by a small, avascular tumor which allows it to grow beyond the oxygen diffusion distance. This has been one major focus of our dormancy studies. Another of our focuses is centering on the angiogenic "balance" - an active dynamic between angiogenic and non-angiogenic states that may shift in either direction depending on ambient conditions. There is reason to believe such a dynamic remains operative even in larger, vascularized tumors. Indeed, the concept of antiangiogenic therapy relies, in essence, on the ability to revert clinical angiogenic tumors back to a non-growing state by re-establishing a limiting angiogenic balance or even an imbalance in favor of regression. This notion raises exciting questions about the control of tumor growth throughout progression. In particular, it points to a need to augment models e.g. the Two-Stage Clonal Expansion (TSCE) Model (see Fig. 1 below), which take various cell-level carcinogenesis events into account, but ultimately equate cancer disease risk to the occurrence of the first cancer cell. As shown in Fig. 2 below, the inevitability of cancer disease assumed by some models after the first cancer cell ("RISK=1" at left) may actually be far less than inevitable ("LOW RISK"), with the difference dictated by whether the tumors can breach subsequent angiogenic, invasive, and other host permissivity barriers to growth.

dormancy timeline paradigmFig. 1. A. The venerable two-stage clonal expansion (TSCE) mathematical carcinogenesis model. Here: initiation is an alteration, such as a mutation, in a normal cell that makes it a stage-1 (that is, pre-malignant) cell; promotion is stochastic proliferation of pre-malignant cells; transformation is another alteration, generating a stage-2 (i.e. malignant) cell from the pool of pre-malignant cells; progression occurs during the time from the production of the first malignant cell to clinically observable cancer. It has often been modeled simply as a fixed lag time. Jagged red arrows indicate possible radiation influences. Single initiation or transformation steps are very brief compared to the promotion and progression steps, which may last many years. More sophisticated models in the literature involve several successive initiation steps. Panel B takes into account dormancy, which gives a more realistic mathematical model that strongly alters risk predictions.

progression bottlenecksFig. 2. In light of the existence of progression-level bottlenecks to tumor growth, the assumption that the risk of cancer disease is certain once a cancer cell is created is in question. The likelihood a tumor will be halted at one of these bottlenecks correspondingly lowers the risk of final cancer disease assessed at the time of the creation of the first cancer cell (shown by the drop in the blue line from "RISK=1" to "LOW RISK").

Researchers at CCSB are extending on these studies experimentally and theoretically. A broad approach is taken that utilizes expertise in molecular and cellular biology to address the underlying mechanisms that leads to tumor dormancy (Nava Almog, Amir Abdollahi), in computational modeling which is helping to dissect the dependencies on cell proliferation, cell crowding, migration and prevalence of tumor stem cells (Heiko Enderling, Lynn Hlatky and Phil Hahnfeldt), in platelets biology which can provide ultra-early biomarkers for presence of dormant tumors (Giannoula Klement and Nava Almog), and in mathematical predictions which can investigate how dynamics in angiogenesis potential affect tumor growth and dormancy (Phil Hahnfeldt, Lynn Hlatky and Ray Sachs). Such emerging research into gene and microRNA signatures of tumor dormancy, modeling of intrinsic cellular dynamics and microenvironment effects, and platelets angiogenic registry, may lead to new approaches for estimating carcinogenesis risk, and for identifying new targets for antitumor therapy based on prolongation of tumor dormancy.

The following articles on the topic of tumor dormancy have been published by researchers who are currently at the CCSB: