Spatial and temporal control of gene therapy using ionizing radiation.

Activation of transcription of the Egr-1 gene by X-rays is regulated by the promoter region of this gene. We linked the radiation-inducible promoter region of the Egr-1 gene to the gene encoding the radiosensitizing and tumoricidal cytokine, tumour necrosis factor-alpha (TNF-alpha) and used a replication-deficient adenovirus to deliver the Egr-TNF construct to human tumours growing in nude mice. Combined treatment with Ad5.Egr-TNF and 5,000 cGy (rad) resulted in increased intratumoral TNF-alpha production and increased tumour control compared with treatment with Ad5.Egr-TNF alone or with radiation alone. The increase in tumour control was achieved without an increase in normal tissue damage when compared to tissue injury from radiation alone. Control of gene transcription by ionizing radiation in vivo represents a novel method of spatial and temporal regulation of gene-based medical treatments.

Enhanced erythropoiesis with concomitant diminished granulopoiesis in preirradiated recipient mice. Evidence for a common stem cell.

Three different methods of measuring the proliferative capacity of transplanted mouse bone marrow were used to study the effects of preirradiation of the recipient. Recipient mice were exposed to 700 R and given graded numbers of syngeneic bone marrow. 7 days were allowed for proliferation of these cells, and then the granulocytic or erythrocytic progeny was measured. The former was determined by the response to endotoxin, and the latter by the incorporation of radioactive iron into newly formed red blood cells. Erythropoiesis, therefore, could be measured independently from granulopoiesis by these techniques. The third method used was the spleen colony method of Till and McCulloch (5). Recipient animals exposed to 150 R preirradiation, 7 days before 700 R and bone marrow transplantation, demonstrated an increase in erythropoiesis with a concommittant decrease in granulopoiesis compared to similar recipients not preirradiated. The spleen colony technique showed that while the number of colonies were the same in both groups, the colonies themselves were significantly larger in the preirradiated animals. Since such colonies are primarily erythropoietic, this finding is consistent with the other methods. The results can be explained by assuming the presence of a hematopoietic stem cell which, in these preirradiated recipients, is directed towards erythropoiesis at the expense of granulopoiesis.

Proliferative capacity of murine hematopoietic stem cells in vitro.

Large numbers of granulocytes can be collected repeatedly from the supernatant medium of long-term cultures of mouse bone marrow cells. A constant relationship was found between the number of adherent hematopoietic stem cells and the lifetime cell production per culture. The data indicate that there is a limit to the proliferative capacity of normal and of irradiated stem cells. A similar limitation was found in the production of marked granulocytes from clonal cultures of "beige" C57 (bg/bgJ) stem cells placed in limiting dilutions into stromal culture layers.

Differing stem cell self-renewal of lectin-separated murine bone marrow fractions.

The success of bone marrow transplantation depends not only on the engraftment of adequate numbers of hematopoietic stem cells but also on their self-renewal capacity, which must be sufficient to provide lifetime hematopoiesis. The lectin peanut agglutinin (PNA), when exposed to bone marrow, causes separation into two distinct fractions. The agglutinated fraction not only is enriched with colony-forming units--spleen but also is devoid of graft-versus-host (GVH) activity when injected into allogeneic lethally irradiated recipients, and therefore, it is considered to be an ideal source for bone marrow transplantation. The absence of GVH activity is presumably due to the separation of mature thymocytes into the nonagglutinated fraction and functionally immature thymocytes into the agglutinated fraction. Although there has been speculation that immature hematopoietic stem cells also selectively bind to PNA, other evidence suggests that the relationship of lectin binding specificity to level of maturity varies in different tissues. This study was performed to assess the self-renewal capacity of lectin-separated bone marrow stem cells. Results indicate that the self-renewal of the agglutinated fraction is significantly lower than that of the unfractionated bone marrow; such self-renewal of the nonagglutinated fraction is higher. This is further evidence for the heterogeneity of the stem cell pool. Stem cell enrichment should not be the goal in bone marrow transplantation; rather, the goal should be utilization of stem cells with the greatest self-renewal potential.

Angiogenesis as a predictor of long-term survival for patients with node-negative breast cancer.

BACKGROUND: Angiogenesis (the formation of new blood vessels) is necessary for tumor growth and metastasis. PURPOSE: We investigated whether angiogenesis as measured by microvessel count (MVC) predicts clinical outcome in a series of patients with axillary lymph node-negative breast cancer who received no adjuvant therapy and who were followed for a long period of time. Our long-term goal is to identify those patients who may or may not need adjuvant chemotherapy. METHODS: Pathologic archival material and clinical information were analyzed for 167 patients treated with mastectomy from 1941 through 1987; none received adjuvant treatment. The median follow-up time among living patients was 15.4 years (range, 2.6-35.8 years). Ninety-six (58%) patients had a tumor size of 2 cm or less, 52 (31%) had tumors of 2.1-3 cm, and 19 (11%) had tumors of larger than 3 cm. Paraffin-embedded tissue sections were stained for expression of CD34 antigen on microvessel-associated endothelial cells by use of a monoclonal anti-CD34 antibody. Vascularity was defined as the number of microvessels (average of the three highest counts) per high-power microscopic field (400 x magnification) in the area of highest vascular density. A high vascular count was defined as 15 or more microvessels per field. Actuarial survival curves were calculated according to the Kaplan-Meier method and comparisons were made with the logrank test. The Cox proportional hazards model was used for multivariate analysis. All P values were based on two-sided testing. RESULTS: The 20-year disease-free survival (DFS) for the 167 node-negative patients treated with mastectomy and no adjuvant therapy was 74.8% (95% confidence interval [CI] = 64.7%-82.0%). The 20-year DFS was 93.1% (95% CI = 79.9%-97.7%) if the MVC was low versus 68.9% (95% CI = 56.8%-78.0%) if the MVC was high (P = .018). This difference was maintained irrespective of tumor size: for tumor size of 2 cm or less (93.3% [95% CI = 75.3%-98.3%] versus versus 67.8% [95% CI = 50.1%-80.3%]) and for tumor size of larger than 2 cm (92.3% [95% CI = 56.6%-98.9%] versus 70.9% [95% CI = 54.6%-81.6%]). However, the likelihood of a high MVC was greater with large tumors (P = .05). The proportions of tumors with low and high MVC were 33% and 67%, respectively, if the tumor size was 2 cm or less, and 20% and 80%, respectively, if tumor size was larger than 2 cm. There was no significant difference in the 20-year DFS as a function of tumor grade (P = .2). After combining patients with tumors of nuclear grades 2 and 3 compared with those of nuclear grade 1, the 20-year DFS was 93.9% (95% CI = 77.2%-98.4%) for low MVC versus 66.9% (95% CI = 52.2%-78.0%) for high MVC (P = .02). In a multivariate analysis that included the variables tumor size, age, nuclear grade, estrogen receptor status, and MVC, only MVC appeared to be an independent prognostic indicator (P = .04). CONCLUSIONS: Angiogenesis as measured by MVC is a reliable independent prognostic marker of long-term survival in patients with node-negative breast cancer. The prognostic usefulness of this marker is maintained after more than 15 years of follow-up. A low MVC identifies a subgroup of patients with DFS of 92% or more, independent of tumor size or grade.

Improving the therapeutic index in breast cancer treatment: the Richard and Hinda Rosenthal Foundation Award lecture.

Improving the relationship between desired and undesired effects of therapy, the therapeutic index, is a major goal of cancer therapy. Clinical research pertinent to breast cancer treatment attempting to manipulate this therapeutic index is not restricted to studies in patients. Described herein is research in humans, mouse, machine, marrow, and molecule, concerned with either increasing care or decreasing treatment complications. Molecular studies use 125I-tamoxifen as a specific cytotoxic agent for cells which contain estrogen receptors able to bind the agent and transport it into the nucleus where the limited range radiations are cytotoxic. Murine bone marrow stem cells are heterogeneous as regards their self-renewal potential. Chemotherapeutic agents used in the adjuvant treatment of breast cancer have different effects on these cells. L-Phenylalanine mustard is toxic to the most primitive stem cells and produces a permanent stem cell self-renewal deficit when given to mice in an "adjuvant" setting. Cyclophosphamide and 5-fluorouracil act primarily on later stem cells and do not produce such proliferative limitations. Eliminating breast cancer while preserving normal structure and function is the goal of combining radiation therapy and tumor excision. Results with this technique are comparable to those following mastectomy without loss of the breast or chest musculature.

The relationship between nm23, angiogenesis, and the metastatic proclivity of node-negative breast cancer.

Distant metastases are the major cause of morbidity and mortality in women with breast cancer. The prediction of this metastatic proclivity is essential in determining prognosis and should allow an appropriate choice of therapy. A critical look at the metastatic process and its phenotypic expression offers an opportunity to identify some of the important events in the process that may relate to prognosis, with the goal of identifying those patients with occult metastases and also sparing systemic treatment in those patients whose tumors have not developed the capacity for distant spread. To evaluate the significance of nm23 and angiogenesis in the metastatic cascade, we used archival material from 163 node-negative breast cancer patients who had a median follow-up of 14 years. All patients underwent mastectomy and received no adjuvant chemotherapy or hormone or radiation therapy. Immunohistochemistry was used to detect nm23-H1 expression, whereas angiogenesis was determined by microvessel count (MVC). We found the 15-year disease-free survival (DFS) to be significantly better in patients with high nm23 compared with low nm23 (91% compared with 70%, P = 0.008). Low MVC is associated with excellent (92%) long-term DFS. In those patients with high MVC, high nm23 allows the identification of a subgroup with significantly higher DFS (90% compared with 66%, P = 0.02). Among high nuclear grade tumors, if nm23 is high, the DFS is significantly better (89% compared with 68%, P = 0.03). Thus, nm23 is still associated with excellent survival, even when there is unfavorable angiogenesis or nuclear grade. Multivariate analysis confirms that nm23 and MVC are important prognostic factors. High MVC appears necessary but not sufficient for metastasis to occur, whereas low nm23 may further contribute to metastatic progression. Both nm23 and MVC contribute valuable information in characterizing the malignant phenotype.

Paradoxical effect of radiation on tumor incidence in the rat: implications for radiation therapy.

The high incidence of leukemia in the Fischer rat is reduced by radiation to an incidence below that seen spontaneously. Fractionating the radiation decreased this effect. In contrast, mammary tumors increased with dose until reaching a plateau at the highest doses. Fractionation had little effect. These results are consistent with a hypothesis suggesting that tumor incidence due to radiation is the result of competing processes of tumor induction and cell killing.

Separating favorable from unfavorable prognostic markers in breast cancer: the role of E-cadherin.

Distant metastases are the major cause of morbidity and mortality in women with breast cancer. The ability to predict the metastatic proclivity is essential in choosing the optimal treatment. Tumor size and grade, which are frequently used markers in node-negative breast cancer patients, are inadequate markers for prognosis and individualized treatment design. The steps in metastatic progression include angiogenesis, invasion, and changes in adhesion characteristics. We developed a strategy for choosing biomarkers representing these steps in malignant progression to identify patients with occult metastases who will need chemotherapy and spare those women whose tumors have not developed the capacity to spread. To evaluate the added significance of E-cadherin to that of nm23-H1 and angiogenesis in determining metastatic proclivity, we used archival material from 168 node-negative breast cancer patients who were treated with mastectomy without any adjuvant chemotherapy or hormone therapy. Immunohistochemistry was used to detect E-cadherin and nm23-H1 expression, whereas angiogenesis was determined by microvessel count (MVC) after immunohistochemical staining. The median follow-up is 14 years. We found that E-cadherin is better in identifying the poor prognosis patients. The 14-year disease-free survival (DFS) is 84%, 80%, and 56% in patients with high, intermediate, and low E-cadherin. The worst prognosis group using nm23-H1 and MVC as biomarkers has a 14-year DFS of 62%. In this group, if E-cadherin is low, the 14-year DFS is further decreased to 44%. Nm23-H1 and MVC are better in identifying the good prognosis patients. The long-term DFS is >90% if MVC is low or if nm23-H1 is high. Multivariate analysis shows that E-cadherin, nm23-H1, and MVC are more significant prognostic biomarkers than tumor size or grade. Loss of E-cadherin appears to be a latter step in the metastatic progression compared to angiogenesis and the loss of nm23-H1 expression.

Differential effects of cytotoxic agents on hematopoietic progenitors.

In order to predict the effect of chemotherapeutic agents on the hematopoietic progenitor compartment, it is necessary to have a hypothesis concerning the dynamics of the cells within the compartment. By viewing this compartment as a continuum of cells with varying self-renewal capacity, one can assess the significance of the stem cell damage incurred following a single drug dose. Vinblastine, 5-fluorouracil, cyclophosphamide, busulfan, and 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) were investigated. The acute toxicity of the drug-exposed marrow was studied by the colony-forming units-spleen (CFU-S) and agar diffusion chamber assay. Busulfan and BCNU were found to kill CFU-S preferentially. By following CFU-S recovery for 14 days post drug, different recovery patterns are noted. Busulfan and BCNU produce prolonged depression in CFU-S, whereas cyclophosphamide and 5-fluorouracil show relatively rapid recovery. If one determines the Rs (a measure of proliferative capacity) after drug, busulfan and, to a lesser extent, BCNU produce a prolonged depression without return to normal even 650 days post drug. No such depression is noted with the other drugs. The data from the recovery curves and Rs support the notion of stem cells being heterogeneous with regard to self-renewal capacity. The agar diffusion chamber and CFU-S acute survival curves would not have predicted which drugs cause significant proliferative damage. Only with the use of CFU-S recovery and ratio of CFU-S can prolonged marrow damage be ascertained.

Genetic radiotherapy overcomes tumor resistance to cytotoxic agents.

We report that radiation enhances gene therapy of a radioresistant tumor by upregulating the induction of a chimeric gene encoding a radiosensitizing protein, tumor necrosis factor alpha (TNF-alpha). We ligated the radiation-inducible CArG elements of the radiation-inducible Egr-1 promoter/enhancer region upstream to the transcriptional start site of the human TNF cDNA (pE425-TNF). This construct was transfected using cationic liposomes into the variant murine fibrosarcoma cell line, P4L. The P4L cell line was both radioresistant (D0 = 188) and resistant to TNF. After a single intratumoral injection of 10 micrograms of pE425-TNF in cationic liposomes and two 20-Gy doses of irradiation, mean tumor volumes were significantly reduced in P4L tumors as compared to those receiving either pE425-TNF in liposomes or radiation alone (P = 0.01). TNf protein in P4L tumors was induced by radiation as high as 29 times control levels and remained detectable for 14 days. Our data indicate that combined gene therapy using liposomes, together with ionizing radiation to locally activate the induction of a radiosensitizing protein, is successful at overcoming resistance to both TNF and radiation.

Tumor necrosis factor alpha (TNF-alpha) gene therapy targeted by ionizing radiation selectively damages tumor vasculature.

Intratumoral injection of an adenoviral vector containing radiation-inducible DNA sequences of the Egr-1 promoter linked to a cDNA encoding tumor necrosis factor (TNF) alpha (Ad.Egr-TNF) enhances the tumoricidal action of ionizing radiation in a human epidermoid carcinoma xenograft (SQ-20B). The dominant histopathological feature in tumor-bearing animals treated with Ad.Egr-TNF and irradiation is extensive intratumoral vascular thrombosis and tumor necrosis. Thrombosis and necrosis are not observed in animals treated with either the viral construct encoding TNF-alpha or radiation and did not occur in irradiated normal tissues adjacent to tumor in animals injected with Ad.Egr-TNF. To determine if the occlusive effects of Ad.Egr-TNF and X-irradiation were specific for tumor vessels, non-tumor-bearing mice were irradiated after receiving i.m. injection of Ad.Egr-TNF at viral titers 20-100 times greater than titers injected intratumorally. No vascular thrombosis was observed in the treated normal tissues. Combined Ad.Egr-TNF and radiation produced occlusion of tumor microvessels without significant normal tissue damage. Taken together, these data suggest that the interaction between radiation inducible TNF-alpha and X-irradiation occurs selectively within the tumor vessels.

Proliferation kinetics of a human breast cancer line in vitro following treatment with 17beta-estradiol and 1-beta-D-arabinofuranosylcytosine.

The effect of 17beta-estradiol on an estrogen receptor-positive human breast cancer cell line (MCF-7) was studied. Low concentrations (10(-9) M) of 17beta-estradiol enhanced the rate of cell proliferation; the overall cell cycle time was shortened; and the proportion of cells in the S phase increased. Higher concentrations (10(-7) M) suppressed proliferation and slightly decreased the proportion of the cells in DNA synthesis. When combined with 1-beta-D-arabinofuranosylcytosine, an S-phase-specific chemotherapeutic agent, 10(-9) M 17beta-estradiol enhanced cell killing. This enhancement was not observed with 10(-7) M 17beta-estradiol. Kinetic changes caused by hormones have profound implications in clinical therapy, since the efficacy of cycle active agents may be altered.

Potentiation of the antitumor effect of ionizing radiation by brief concomitant exposures to angiostatin.

Angiostatin, a proteolytic fragment of plasminogen, inhibits the growth of primary and metastatic tumors by suppressing angiogenesis. When used in combination with ionizing radiation (IR), angiostatin demonstrates potent antitumor synergism, largely caused by inhibition of the tumor microvasculature. We report here the temporal interaction of angiostatin and IR in Lewis lung carcinoma (LLC) tumors growing in the hind limbs of syngeneic mice. Tumors with an initial mean volume of 510 +/- 151 mm3 were treated with IR alone (20 Gy x 2 doses on days 0 and 1), angiostatin alone (25 mg/kg/day divided twice daily) on days 0 through 13, or a combination of the two as follows: (a) IR plus angiostatin (days 0 through 13); (b) IR plus angiostatin (days 0 and 1); and (c) IR followed by angiostatin beginning on the day after IR completion and given daily thereafter (days 2 through 13). By day 14, tumors in untreated control mice had grown to 6110 +/- 582 mm3, whereas in mice treated with: (a) IR alone, tumors had grown to 2854 +/- 338 mm3 (P < 0.05 compared with untreated controls); and (b) angiostatin alone, tumors had grown to 3666 +/- 453 mm3 (P < 0.05 compared with untreated controls). In combined-treatment groups, in mice treated with: (a) IR plus longer-course angiostatin, tumors reached 2022 +/- 282 mm3 (P = 0.036 compared with IR alone); (b) IR followed by angiostatin, tumors reached 2677 +/- 469 mm3 (P > 0.05 compared with IR alone); and (c) IR plus short-course angiostatin, tumors reached 1032 +/- 78 mm3 (P < 0.001 compared with IR alone). These findings demonstrate that the efficacy of experimental radiation therapy is potentiated by brief concomitant exposure of the tumor vasculature to angiostatin.

NM-3, an isocoumarin, increases the antitumor effects of radiotherapy without toxicity.

We examined the effects of a new antiangiogenic isocoumarin, NM-3, as a radiation modifier in vitro and in vivo. The present studies demonstrate that NM-3 is cytotoxic to human umbilical vein endothelial cells (HUVECs) but not to Lewis lung carcinoma (LLC) cells nor Seg-1, esophageal adenocarcinoma cells, in clonogenic survival assays. When HUVEC cultures are treated with NM-3 combined with ionizing radiation (IR), additive cytotoxicity is observed. In addition, the combination of NM-3 and IR inhibits HUVEC migration to a greater extent than either treatment alone. The effects of treatment with NM-3 and IR were also evaluated in tumor model systems. C57BL/6 female mice bearing LLC tumors were given injections for 4 consecutive days with NM-3 (25 mg/kg/day) and treated with IR (20 Gy) for 2 consecutive days. Combined treatment with NM-3 and IR significantly reduced mean tumor volume compared with either treatment alone. An increase in local tumor control was also observed in LLC tumors in mice receiving NM-3/IR therapy. When athymic nude mice bearing Seg-1 tumor xenografts were treated with NM-3 (100 mg/kg/day for 4 days) and 20 Gy (four 5 Gy fractions), significant tumor regression was observed after combined treatment (NM-3 and IR) compared with IR alone. Importantly, no increase in systemic or local tissue toxicity was observed after combined treatment (NM-3 and IR) when compared with IR alone. The bioavailability and nontoxic profile of NM-3 suggests that the efficacy of this agent should be tested in clinical radiotherapy.

Aging, progression, and phenotype in breast cancer.

The malignant potential of cancer is dynamic, changing throughout the natural history of a tumor. For breast cancer, this is especially important because the clinical presentation has been altered by the increasing use of screening mammography. The varied outcomes of similarly staged patients is most consistent with breast cancer not being a homogeneous disease, but rather a spectrum of disease states that have varying capacities for growth and metastasis. Evolutionary pressures are at play in both tumor development and during the clinically apparent portion of the life of a tumor and are responsible for this spectrum of tumor heterogeneity. Required of tumors is the development of critical phenotypic attributes: growth, invasion, metastagenicity, and angiogenesis. The combination and permutation of genetic changes that result in the acquisition of these characteristics may vary, but they must result in some expression of each of these phenotypes. The expression of these attributes will differ as tumors evolve to become more adept at each of these characteristics. Recognizing tumor heterogeneity emphasizes the need to determine an individual tumor's place in the evolutionary spectrum. This may be accomplished using clinical features such as size, nuclear grade, and patient age, as well as by examining markers of angiogenesis, metastatic capacity, and proliferation. Identification of the extent of tumor progression with regard to these major tumor phenotypes should allow individual therapy to be fashioned for each patient.

Clinical progression of breast cancer malignant behavior: what to expect and when to expect it.

PURPOSE: Seemingly localized breast cancer is a heterogeneous mix of truly localized cancers and cancers with occult metastases. Our purpose is to determine the parameters of metastatic proclivity for the different clinical presentations of operable breast cancer and to present quantitative prognostic information useful to both doctors and patients. PATIENTS AND METHODS: A series of regionally treated breast cancer patients was analyzed to determine the likelihood and time of the appearance of clinical metastases for different clinical subgroups. Patients operated on at the University of Chicago from 1927 to 1987 for clinically regionally localized breast cancer, who received no systemic therapy as a part of their initial treatment, were included. Overall survival and distant disease-free survival in this mature series are analyzed. RESULTS: Metastagenicity, the metastatic proclivity of a tumor, increases with both tumor size and nodal involvement. This is also true for virulence, which is the rate at which these metastases appear. Each clinical group has a cured population, even those with extensive nodal involvement. A table provides a tool for determining the proportion of risk expended in each clinical group as a function of the distant disease-free survival. Whereas the likelihood of metastasis increases with tumor size and nodal involvement, the time to their appearance decreases. CONCLUSIONS: Breast cancer metastagenicity and virulence are heterogeneous even within clinically similar groups of operable breast cancer patients. Tumor progression is correlated with increasing tumor size and nodal involvement. Markers are needed to identify individual tumor virulence and metastagenicity.

Functional organization of the hematopoietic stem cell compartment: implications for cancer and its therapy.

Recent discoveries indicate that hematopoietic stem cells have limits on their proliferative capacity and are unable to divide indefinitely. There is great heterogeneity within the compartment as to the extent of this proliferative limitation. At any given time it appears that hematopoiesis is maintained by the progeny of only a few stem cells. When these are exhausted the progeny from other stem cells take their place. The observations of proliferative limitation, heterogeneity, and clonal succession must be incorporated into any model of stem cell organization. These new discoveries and the models incorporating them have important clinical implications. They may explain the inability of normal tissues to develop drug resistance and they also offer a mechanism by which cell renewal systems decrease the development of malignancies. In the selection of chemotherapeutic agents not only the effectiveness of the drug upon the tumor must be considered, but also how specific agents affect the stem cell compartment. These data have important implications in the use of bone marrow transplantation for both malignant and nonmalignant disease.

Natural history of node-negative breast cancer: a study of 826 patients with long-term follow-up.

PURPOSE: We were interested in examining the long-term outcome of patients with node-negative breast cancer to address the following questions: (1) Is node-negative breast cancer a disease that is curable by local modalities? (2) Are there predictors of disseminated disease in node-negative breast cancer? (3) Are there subgroups of tumors that have different times to recurrence? METHODS: From 1927 to 1984, 826 women with node-negative breast cancer were treated at the University of Chicago. Patients underwent either a radical or extended radical mastectomy (83%) or a modified radical mastectomy (13%). RESULTS: Follow-up evaluation ranged from 9 to 523 months (43.6 years); the mean follow-up period of survivors is 162 months (13.5 years). On multivariate analysis, the strongest predictor of outcome and time to relapse was pathologic tumor size. Patients with tumors less than 2 cm had a 20-year disease-free survival (DFS) rate of 79% and a median time to recurrence of 48 months as compared with patients with tumors greater than 2 cm, who had a survival rate of 64% (P < .001) and a median time to recurrence of 37 months (P = .01). CONCLUSION: With extended follow-up evaluation, node-negative breast cancer is a curable disease. Size is the strongest predictor of dissemination and rate of relapse. These data suggest that given the natural history of node-negative breast cancer, analysis of clinical trials with short follow-up periods can be misleading, since it may identify those patients whose tumors have a greater virulence but not necessarily a greater likelihood to metastasize.