Необходима регистрация на сайте. Будет опубликован на главной странице после проверки модератором.

Encyclopedia of Cancer, 2015


Cancer without disease; In situ cancer; In situ carcinoma; Occult cancer


Tumor dormancy describes human tumors with three characteristics:

(i) Visible only under a microscope and, therefore, cannot be detected by conventional diagnostic imaging methods and may have an average diameter the size of a pinhead but range from 0.1 to ~2 or 3 mm

(ii) Usually do not expand or spread to other organs

(iii) Usually asymptomatic and harmless but have the potential to resume growth and eventually to be fatal to their host


Virtually all adult humans have dormant cancers, as determined from autopsies of individuals who died of trauma (e.g., auto accidents), but who did not have a diagnosis of cancer during their lifetime. From these autopsies, pathologists report microscopic-sized cancers in different organs, often called carcinoma in situ. In women 40–50 years old, 39 % have dormant in situ carcinomas in their breasts (preneoplastic lesions), but only 1 out of 100 ever develops breast cancer during a normal lifetime (ductal carcinoma in situ). Forty-six percent of men from 60 to 70 have carcinoma in situ of the prostate, but only 1 out of 100 in this age range is ever diagnosed with prostate cancer. In contrast, estimated 98 % of individuals age 50–70 harbor carcinoma in situ of the thyroid gland, but only 1 out of 1,000 develops thyroid cancer. Dormant carcinoma in situ can be found nestled among established capillary blood vessels, but such dormant tumors do not recruit new blood vessels. In other words, most dormant cancers in situ are non-angiogenic.

Blocked angiogenesis

Therefore, one mechanism of tumor dormancy is blocked angiogenesis, i.e., the inability of emerging early tumors to recruit new blood vessels. Cancer arises from a single cell, for example, a liver cell. Normal liver cells rarely divide. A cancerous liver cell, however, can continue to divide without restraint until it has accumulated offspring of up to ~1 million tumor cells. Nevertheless, such a microscopic tumor becomes dormant when its further expansion is arrested by the limits of oxygen diffusion from the nearest open capillary blood vessel. This oxygen diffusion limit is ~180–200 m (about 0.2 mm) for tumor cells and significantly less for normal cells. Virtually every normal cell lives either directly adjacent to a capillary blood vessel or at least not more than two cell widths from a capillary. However, tumor cells can surround a capillary vessel with multiple cells (Fig. 1). Once a microscopic tumor becomes oxygen depleted, tumor cell death (apoptosis) increases to match tumor cell proliferation. This “balance” between proliferating and dying tumor cells is one of the hallmarks of the dormant, microscopic, in situ cancer which virtually all humans harbor (progression).


Fig. 1. Left panel: Blood vessels in a human breast cancer (MCa-IV). Tumors were transplanted into mice, and subsequently the tumors were perfused with fixative so that the microvessels would not be compressed. The white dots show the outer layer of multiple layers of tumor cells surrounding a new capillary blood vessel (lectin binding). Right panel: Endothelial cells are labeled by antibody to CD31 antigen. This section also shows new capillary sprouts in the breast cancer

Maintenance of tumor dormancy by endogenous angiogenesis inhibitors

The majority of new tumors and metastases remain dormant for prolonged periods of time (sometimes for years), in part because of endogenous angiogenesis inhibitors. At this writing, 29 angiogenesis inhibitors have been discovered in the body; none were known before 1980 (anti-angiogenesis). Most of these angiogenesis inhibitors are proteins, such as thrombospondin-1, platelet factor 4, maspin, angiostatin, endostatin, tumstatin, canstatin, interleukin-12, SPARC, and others. Endostatin and tumstatin are under the control of p53. While some inhibitors circulate at low concentrations in the plasma, others are stored in platelets, white blood cells, bone marrow cells, fibroblasts of tissues throughout the body, or in the collagen basement membranes in the stroma underlying most tissue cells. Endothelial cells produce collagen type XVIII basement membrane. Tumor cells themselves also express certain angiogenesis inhibitor proteins directly, such as thrombospondin-1, or express enzymes that mobilize anti-angiogenic peptide fragments from larger proteins. Examples of the latter are angiostatin from plasminogen and endostatin from collagen XVIII. The gene for collagen XVIII is on chromosome 21. Individuals with Down syndrome have three copies of this chromosome (trisomy). As a result, elevated levels of endostatin are found in these individuals. They are protected against abnormal angiogenesis. Down syndrome individuals with diabetes are protected against neovascularization in the retina and also from neovascularization in atherosclerotic plaques. They are also protected against cancer.

In fact, of the ~200 types of cancer, these individuals only develop testicular cancer and a rare but mild form of leukemia. For the other types of cancer, individuals with Down syndrome have <0.1 the expected incidence even though they live into middle age and beyond. In mice genetically engineered to overexpress collagen XVIII so that their endostatin level is increased by ~1.6-fold to mimic individuals with Down syndrome, implanted tumors are poorly neovascularized and grow 300–400 % slower. Conversely, tumors grow threefold more rapidly in mice that lack endostatin. The genes for mental retardation are unrelated to endostatin. Furthermore, a second putative anti-angiogenic gene, called DSCR1 (Down syndrome critical region), has recently been identified on chromosome 21. Although this phenomenon is correlative and not yet proved to be causal, it provides a thought-provoking clinical clue that suggests the question – do individuals with Down syndrome harbor equivalent numbers of microscopic dormant cancers as the rest of the population but a lower incidence of tumors that become angiogenic?


Fig. 2. Angiogenesis in rat sarcoma. In this micrograph, blood vessels grow toward a sarcoma (dark area at right) in rat muscle. This contrasts with the normal grid-like pattern of blood vessels that appears at the upper left. Tumor cells that have begun to surround the capillary vessels are not shown here (Courtesy of L. Heuser and R. Ackland, University of Louisville, USA)

Escape from tumor dormancy by a switch to the angiogenic phenotype

Miles of capillary blood vessels, thinner than a hair, supply every tissue in the body. A pound of fat contains ~1 mile of capillary blood vessels. Vascular endothelial cells which line the inside of these blood vessels normally proliferate only infrequently, to replace lost endothelial cells. The entire endothelial lining is replaced or “turned over” in ~3 years, in contrast to the turnover of intestinal epithelial cells that is measured in days. During physiological angiogenesis, such as reproduction, development, or wound repair, endothelial cells can proliferate rapidly, i.e., with a turnover measured in days or weeks. Physiologic angiogenesis is, however, self-limited. Angiogenesis during ovulation is turned off after a few days and in wounds after ~2 weeks. In contrast, once a tumor has become angiogenic, endothelial cells in the tumor bed proliferate continuously, the beginning of the switch to the angiogenic phenotype. As new capillary blood vessels grow toward the dormant tumor, tumor cells grow around them and the tumor mass now expands rapidly (Fig. 2). Angiogenic tumors become detectable by conventional imaging methods, cause symptoms, and metastasize to other organs. Angiogenic tumors are potentially fatal (progression).

Experimental analysis of the angiogenic switch in human dormant cancers

The escape of human dormant tumors to the angiogenic phenotype has been studied in immunodeficient mice (SCID or nude mice), by cloning single tumor cells from tumor specimens discarded from the operating room or obtained as tumor cell lines from the American Type Culture Collection. When these clones are expanded in vitro, and reimplanted in immunodeficient mice, ~3–5 % form non-angiogenic tumors of ~1 mm diameter and remain dormant with a high proliferation rate of tumor cells balanced by a high apoptotic rate. The microscopic dormant tumors can be visualized in mice if tumors are stably infected either with luciferase or green fluorescent proteins (e.g., green fluorescent protein). For each tumor type, there is a predictable percentage of non-angiogenic tumors that will undergo a spontaneous switch to the angiogenic state (called phenotype), at a predictable time. For example, ~80 % of the non-angiogenic tumors from a given type of human breast cancer become angiogenic at ~4 months. For liposarcoma (a cancer of fat tissue), the angiogenic switch occurs at 4 months in 95 % of non-angiogenic tumors. A human brain tumor (glioblastoma) becomes angiogenic at 8 months in 60 % of tumors. In human bone tumors (osteosarcoma), the angiogenic switch does not occur until after 1 year and in only 5–15 % of mice. After tumors become angiogenic, they escape tumor dormancy and form lethal tumors in 100 % of mice regardless of cancer type. Thus, human tumors studied so far contain a subpopulation of non-angiogenic tumor cells that can form dormant tumors. For some tumor types, the majority of non-angiogenic tumors become angiogenic and escape from dormancy (progression) (e.g., liposarcoma). For other tumor types, only a minority of non-angiogenic tumors become angiogenic and the rest remain non-angiogenic and dormant indefinitely.

Molecular mechanisms of angiogenic switching in dormant cancers

If human tumors could be restricted to the non-angiogenic dormant phenotype, or if angiogenic tumors could be reversed to the non-angiogenic dormant phenotype, a novel anticancer therapy could be possible (anti-angiogenesis). Therefore, molecular mechanisms are being studied. For example, transfection of a non-angiogenic human osteosarcoma with the RAS oncogene causes dormant tumors to become angiogenic and escape from dormancy within weeks to 1 month, in contrast to the spontaneous angiogenic switch which can take up to 1 year. Furthermore, after RAS transfection, 100 % of dormant non-angiogenic tumors become angiogenic and grow rapidly, whereas spontaneous escape from dormancy occurred in only 5–15 % of tumors. RAS transfection is followed by a 30 % increased expression of vascular endothelial growth factor (i.e., adrenomedullin) (VEGF, a potent proangiogenic protein) and is accompanied by a 50 % decrease in the expression of thrombospondin-1, a potent angiogenesis inhibitor. Many oncogenes induce increased expression of a proangiogenic proteins and suppression of an anti-angiogenic protein (anti-angiogenesis). This common pattern could lead to a general molecular mechanism of escape from tumor dormancy by activation of the switch to the angiogenic phenotype.

Other forms of tumor dormancy

Immune surveillance as a cause of tumor dormancy

Many experimental models reveal that the immune system can maintain a reduced load of cancer cells based on the manipulation of cytotoxic T lymphocytes (CD8+) which can kill tumor cells expressing specific antigens. Furthermore, escape from tumor dormancy in some models is based on tumor evasion of the immune system. In other experimental systems, growth of mouse lymphoma cells can be suppressed by a T-cell-mediated mechanism. It remains to be determined if therapeutic blockade of angiogenesis, by restricting expansion of a dormant tumor population, will synergize immune suppression.

Hormonal depletion as a cause of tumor dormancy

Certain hormonally dependent tumors, such as prostate cancer, become dormant when the hormone (e.g., androgen) is blocked or depleted. It has been shown that depletion of testosterone from a prostate cancer decreases VEGF expression and thus reduces tumor angiogenesis. However, in clinical practice, this therapy is temporary, and within 1 year or more, androgen-independent prostate cancer cells often emerge.

Dormancy of single metastatic tumor cells

It has been shown in experimental animals that single metastatic tumor cells can exit the circulation at a future metastatic site, for example, the liver or lungs, and survive for long periods near a capillary blood vessel without proliferating (the G0 state).

Future directions

Biomarkers to detect dormant tumors

Many laboratories are developing a variety of molecular biomarkers in the blood to detect the presence of cancer. Some of these may be useful for detecting the presence of microscopic-sized tumors that cannot be located anatomically by conventional imaging methods. This would include non-angiogenic dormant tumors or those just beginning to switch to the angiogenic phenotype. If a biomarker of high sensitivity could be validated in the clinic, then it could eventually be used to guide nontoxic anti-angiogenic therapy, or anti-telomerase therapy, or immunotherapy, to prevent recurrence of cancer years before symptoms or before anatomical location was possible.


Aquirre-Ghiso JA (2007) Models, mechanisms and clinical evidence for cancer dormancy. Nat Rev Cancer 7:834–846

Finn OJ (2006) Human tumor antigens, immunosurveillance and cancer vaccines. Immunol Res 36:73–82 Folkman J (2007) Angiogenesis: an organizing principle for drug discovery? Nat Rev Drug Discov 6:273–286

Holmgren L, O’Reilly MS, Folkman J (1995) Dormancy of micrometastases: balanced proliferation and apoptosis in the presence of angiogenesis suppression. Nat Med 1:149–153

Naumov GN, Folkman J (2007) Strategies to prolong the nonangiogenic dormant state of human cancer. In: Davis DW, Herbst RS, Abbruzzese JL (eds) Antiangiogenic cancer therapy. CRC Press, Boca Raton/London/New York, pp 3–22