Drugs with antiangiogenic activity began to be approved by the FDA more than three years ago. Since then, they have been introduced into medical practice for the treatment of cancer in the U.S., and in more than 30 other countries (1). This new class of drugs can be administered in combination with other anti-cancer modalities, (including, ionizing radiation, vaccines, and telomerase inhibitors), or with other angiogenesis inhibitors. However, distinctive clinical guidelines are being developed for the administration of angiogenesis inhibitors, because of their fundamental difference from chemotherapy. For example, angiogenesis inhibitors are relatively less toxic than conventional cytotoxic chemotherapy. Furthermore, antiangiogenic therapy appears to be most effective when administered chronically to achieve uninterrupted elevated blood levels over long periods of time (i.e., months to years) (1). [ ]
 These novel properties of angiogenesis inhibitors suggest the possibility that in the future they will also be used: (i) To convert cancer to a chronic manageable disease (2), and (ii) To treat recurrent cancer after “ultra-early” detection, perhaps years before symptoms appear, or before anatomical location is feasible (1). [ ]
 These goals however, require highly specific and sensitive biomarkers that can detect recurrent cancer when it is still microscopic, and before, or at the onset of the switch to the angiogenic phenotype (3). While many laboratories are developing biomarkers for early detection of cancer, the Vascular Biology Program at Children’s Hospital Boston, has had a longstanding interest in developing biomarkers that are based on detection of dysregulated angiogenesis. We have focused on a small set of proteins that regulate angiogenesis, and that would change significantly in the urine or blood when a microscopic tumor was still non-angiogenic, or was in the early stages of switching to the angiogenic phenotype. [ ]
 Matrix metalloproteinases are a class of enzymes that play an important role in tumor angiogenesis (4). Our recent studies show that tumor stage and progression correlate with urinary levels of matrix metalloproteinases in both tumor-bearing animals and in patients with cancer, including brain tumors (5, 6). [ ]
 Studies from our laboratory also reveal that platelets in animals bearing human tumors, either in a subcutaneous, or in an orthotopic location, can selectively accumulate certain angiogenesis regulatory proteins that reach significantly higher concentrations than in plasma (7-9). This change in the "platelet angiogenesis proteome," quantified in platelet lysates by either ELISA or by mass spectrometry, can detect the presence of a non-angiogenic human tumor in the millimeter range, although not its anatomical location. At this microscopic size, such experimental tumors in SCID mice can be anatomically located only by bioluminescence (i.e., by prior infection of tumor cells with luciferase) (10). Non-angiogenic and angiogenic tumor cells were cloned from human tumor cell lines, or from operating room specimens. For a given tumor type, a predictable percentage of non-angiogenic tumors switched to the angiogenic phenotype at a predictable time after inoculation of tumor cells (11). For example, ~95% of human liposarcomas became angiogenic after ~133 days (± 25 days); ~ 5% of human osteosarcomas became angiogenic after ~ 1 year. After the angiogenic switch, tumors grew rapidly, and serum levels of angiogenesis regulatory molecules, as the circulating platelet mass became saturated (12). [ ]
 To begin to evaluate angiogenesis based biomarkers in urine, non-randomized, prospective, open label clinical trials are underway to determine the accuracy of detection of recurrent breast cancer and prostate cancer by elevated urinary metalloproteinases. [ ]
 To evaluate the platelet angiogenesis proteome, similar pilot scale clinical trials have been initiated for the detection of recurrent colon cancer and neuroblastoma. An advantage of evaluating the platelet angiogenesis proteome in colon cancer, is that 50-60% of patients are cured by surgical resection of the primary tumor. In the remaining patients tumors recur at a mean of ~5 years. In this trial we will determine: (i) How soon after surgery the platelet angiogenesis proteome returns to normal in patients most likely to be cured; (ii) How soon it resumes its abnormal pattern in the remaining patients; (iii) How accurately the platelet angiogenesis proteome distinguishes between the two groups; and (iv) How long in advance of symptoms, or conventional imaging, is tumor recurrence detected? A similar clinical trial is being conducted for neuroblastoma. [ ]
 As these angiogenesis-based biomarkers are further developed, it may someday be possible to detect recurrent cancer years before symptoms, or long before anatomical location is feasible. The next step could be to employ relatively non-toxic antiangiogenic therapy guided by angiogenesis-based biomarkers to treat “ultra-early” tumor recurrence. This would be analogous to the treatment of infection guided by an elevated white blood cell count.
 It is not too early to speculate that long-term, non-toxic antiangiogenic therapy, may prevent the angiogenic switch, and thus leave tumors in a harmless dormant state. An animal model that can mimic prolonged microscopic tumor dormancy due to blocked angiogenesis, is the PPAR-alpha null mouse. Implanted tumors remain non-angiogenic and microscopic indefinitely (13). At least one mechanism is that in the absence of PPAR-alpha, thrombospondin-1 (and endostatin) expression are up-regulated in neutrophils. These inflammatory cells then infiltrate tumors and deliver these endogenous angiogenesis inhibitor proteins directly to the tumor bed. When an antibody to thrombospondin-1 is administered to these mice, tumor growth resumes.
 1. Folkman J. Nature Reviews Drug Discovery. (2007) 6:273-286.
 2. Ezzel C. Scientific American (1998) 4:33-34.
 3. Hanahan D, Folkman J. Cell (1996) 86:353-364.
 4. Chan LW, Moses MA, Goley E, Sproull M, Muanza T, Coleman CN, Figg WD, Albert PS, Menard C, Camphausen K. J Clin Oncol (2004) 22:499-506.
 5. Roy R, Wewer UM, Zurakowski D, Pories SE, Moses M. J Biol Chem (2004) 279:51323-51330.
 6. Smith ER, Manfried M, Scott RM, Black PM, Moses MA. Neurosurgery (2007) 60: E1148.
 7. Klement G, Kikuchi L, Kieran M, Almog N, Yip TT, Folkman J. Blood (2004) 104:239a Abstract #839.
 8. Klement G, Cervi D, Yip T, Folkman J, Italiano J. Blood (2006) 108 :426a, abstract 1476.
 9. Italiano J, Richardson JL, Folkman J, Klement G. Blood (2006) 108:120a, abstract 393.
 10. Almog N, Henke V, Flores L, Hlatky L, Kung AL, Wright RD, Berger R, Hutchinson L, Bender E, Achilles E, Folkman J. FASEB (2006) 20: 947-949.
 11. Naumov G, Bender E, Zurakowski D, Kang S, Sampson D, Flynn E, Watnick RS, Straume O, Akslen LA, Folkman J, Almog N. J Natl Cancer Inst (2006) 98:316-325.
 12. Naumov GN, Folkman J. in: “Antiangiogenic Cancer Therapy.” Davis DW, Herbst RS, Abbruzzese JL, eds. (2007) CRC Publishers. Strategies to Prolong the Nonangiogenic Dormant State of Human Cancer, Chapter 1, p 1-16.
 13. Kaipainen A, Kieran MW, Huang S, Butterfield C, Bielenberg D, Mostoslavsky G, Mulligan R, Folkman J, Panigrahy D. PLoS ONE 2:e260.

Second AACR International Conference on Molecular Diagnostics in Cancer Therapeutic Development-- Sep 17-20, 2007; Atlanta, GA