Colorectal cancer and antiangiogenic therapy: What can be expected in clinical practice?

Transcription

1 Critical Reviews in Oncology/Hematology 55 (2005) Colorectal cancer and antiangiogenic therapy: What can be expected in clinical practice? Andrea Mancuso, Cora N. Sternberg Department of Medical Oncology, San Camillo and Forlanini Hospitals, Circonvallazione Gianicolense 87, I Rome, Italy Accepted 11 March 2005 Contents 1. Introduction Angiogenesis inhibitors Inhibitors of endothelial cell migration and proliferation Endostatin/Angiostatin Vitaxin Canstatin Vascular endothelial growth factor (VEGF) and blocking strategies Monoclonal antibodies Small molecules that interfere with VEGF binding or receptor signalling AMG NM-3 Isocoumarin Gefitinib (ZD1839, Iressa) and Erlotinib (OSI-774, Tarceva) CI SU PTK Inhibitors of angiogenesis with unclear mechanisms of action Combretastatins Methoxyestradiol Thalidomide and analogues (SelCID-3) Discussion, problems and conclusions Reviewers References Biographies Abstract Angiogenesis is a strongly regulated process, which is dependent upon a complex interplay between inhibitory and stimulatory angiogenic factors. It is essential for tumor growth and metastasis: increased angiogenesis is correlated with poor prognosis in cancer patients. Many novel compounds that potently inhibit formation of neoplastic blood vessels have been recently developed. Major categories of angiogenesis antagonists include protease inhibitors, direct inhibitors of endothelial cell proliferation and migration, angiogenic growth factor suppressors, inhibitors of endothelial-specific integrin/survival signalling, copper chelators, and inhibitors with other specific mechanisms. There is increasing interest in developing angio-suppressive agents for colorectal cancer treatment. Some 20 direct and indirect antiangiogenesis drugs are currently being evaluated in clinical trials in colorectal cancer (CRC). Promising results have been reported. These include an increase Corresponding author. Tel.: ; fax: addresses: mancuso andrea@hotmail.com (A. Mancuso), cstern@mclink.it, csternberg@scamilloforlanini.rm.it (C.N. Sternberg) /$ see front matter 2005 Elsevier Ireland Ltd. All rights reserved. doi: /j.critrevonc

2 68 A. Mancuso, C.N. Sternberg / Critical Reviews in Oncology/Hematology 55 (2005) in overall survival and reduction in the risk of death (Bevacizumab), reversal of cellular resistance (Cetuximab) and activity as second-line therapy in patients who have exhausted other available treatment options (Cetuximab, ABX-EGF, PTK-787, Gefitinib, Erlotinib). This review will outline the mechanisms of action of the principal antiangiogenic drugs, summarize the available data on the use of these new drugs in colorectal cancer, discuss their impact in clinical practice and offer a glimpse into how antiangiogenetic therapy will be integrated into the future care of patients with gastrointestinal cancers Elsevier Ireland Ltd. All rights reserved. Keywords: Colorectal cancer; New regimens; Antiangiogenic therapy 1. Introduction Tumor cells, like many other cell types, require vascularization to receive oxygen and nutrients and to dispose of catabolic products. While it has been known for a long time that new microvessel formation from endothelial cells (angiogenesis) accompanies tumor growth and favors metastatic dissemination, the identity and characteristics of the growth factors involved in this process have been recently defined in the past few years [1,2]. The list of growth factors involved in tumor angiogenesis has grown substantially and includes vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bfgf), Endostatin, platelet-derived growth factor (PDGF), transforming growth factor alpha (TGF-alpha), tumor necrosis factor, and interleukin-8 (IL-8). VEGF induces proliferation of vascular endothelial cells, promotes their survival in newly formed vessels, and increases microvascular permeability. The endogenous antiangiogenic factors Angiostatin and Endostatin, inhibit proliferation and migration of endothelial cells and induce apoptosis (cell death) in endothelial cells. After discovery of these mechanisms, it was only a matter of time before growth factors and receptors involved in angiogenesis would become targets for the development of novel antitumor therapies. Angiogenesis inhibitors can be divided into two main categories of agents according to their mechanism of action [3]: those that act directly on microvascular endothelial cells recruited by the tumor, such as Endostatin, and those that target tumor growth factors (e.g., Bevacizumab, an antibody targeting VEGF) or tyrosine kinase inhibitors (TKIs) (e.g., Gefitinib, Erlotinib) (Table 1). Advanced colorectal cancer (CRC) is a critical health concern, despite improvements during the last years. Overall survival is highly dependent upon the stage of disease at diagnosis. Estimated 5-year survival rates range from 85 to 90% for patients with stage I disease to <5% for patients with stage IV disease. Over 50% of patients with colorectal cancer presenting with metastatic or locally advanced disease experience local recurrence or develop distant metastases after potentially curative surgery. The most important treatment currently available for patients with stage IV disease is systemic chemotherapy. Although there have been recent advances in the field, with randomized trials confirming the activity of Irinotecan, Oxaliplatin and Capecitabine, median survival remains at only months [4]. There is, Table 1 Principal angiogenesis inhibitors in colorectal cancer: cellular targets and stage of clinical development Agent/compound Target Mechanism Clinical trial Angiostatin ATP synthase, Angiomotin, Annexin II Inhibition of endothelial cell proliferation Phase 1 Endostatin Integrin alpha5-beta1 Inhibition of endothelial cell proliferation and migration Phases 1 and 2 Vitaxin (humanized Integrin alpha 5-beta3 Inhibition of endothelial cell proliferation and migration Phases 1 and 2 monoclonal antibody) Canstatin Integrin alpha5-beta3 Inhibition of endothelial cell proliferation and migration Not yet Bevacizumab (humanized VEGF Inhibition of endothelial cell proliferation Phases 2 and 3 monoclonal antibody) Cetuximab (C-225), EGFR, VEGF, bfgf, TGF-alpha Inhibition EGFR signalling and indirectly endothelial cell Phases 2 and 3 ABX-EGF proliferation EMD72000 EGFR, bfgf, TGF-alpha It blocks EGFR and natural ligand binding; it abrogates Phase 1 receptor-mediated downstream signalling AMG706 VEGF, PDGF, c-kit Small molecule inhibitor of multiple kinases Phase 1 NM-3 Isocoumarin VEGF Inhibition of endothelial cell proliferation Phase 1 Gefitinib (ZD1839), Erlotinib VEGF, bfgf, TGF-alpha Inhibition of tyrosine kinase activation Phases 1 and 2 (OSI-774), CI-1033 SU-6668, PTK-787 VEGFR-1 and/or VEGFR-2 Inhibition of receptor phosphorylation Phases 2 and 3 Combretastatins Microtubules Apoptosis of endothelial cells Phase 1 2-Methoxyestradiol Microtubules Apoptosis of endothelial cells Phase 1 Thalidomide and analogues bfgf Inhibition of endothelial cell proliferation Phases 1 and 2 bfgf: basic fibroblast growth factor; VEGF: vascular endothelial growth factor; VEGFR: vascular endothelial growth factor receptor; PDGFR: platelet-derived growth factor receptor.

3 A. Mancuso, C.N. Sternberg / Critical Reviews in Oncology/Hematology 55 (2005) therefore, a critical need for more effective and bettertolerated therapies. Recent developments and research have focused upon therapies that are capable of selectively targeting pathways that are critical for tumor growth and development, while sparing normal tissues. Inhibition of signalling pathways can be accomplished through several strategies. The clinical role of angiogenesis inhibitors in advanced/metastatic colorectal cancer will be treated in this review, with an emphasis on three classes of agents: (1) inhibitors of endothelial cell migration and proliferation; (2) inhibitors of vascular endothelial growth factor and related pathways (monoclonal antibodies and small molecules); and (3) inhibitors of angiogenesis with unclear mechanisms of action. 2. Angiogenesis inhibitors 2.1. Inhibitors of endothelial cell migration and proliferation Endostatin/Angiostatin The first antiangiogenesis compounds to be proposed and evaluated for the treatment of patients with colorectal cancer were Endostatin and Angiostatin. Endostatin is a 20-kDa proteolytic fragment of collagen XVIII, which inhibits, with an unclear mechanism of action, proliferation and migration of endothelial cells through induction of apoptosis. Angiostatin, on the other hand, is a circulating inhibitor of angiogenesis derived from plasminogen and is generated by various matrix metallo-proteinases (MMPs) including MMP-2, MMP- 3, MMP-7, MMP9 and MMP-12; it binds subunits of ATP synthase on the cell surface of endothelial cells and inhibits proliferation by a reducing the supply of energy. Preclinical studies on colorectal carcinoma models had shown that Angiostatin and Endostatin effectively inhibit tumor growth and shrink existing tumor blood vessels. Phase I clinical trials demonstrated a favorable toxicity profile but low antitumor activity. In Eder Jr. s study [5] of 15 patients with refractory solid tumors given a min intravenous administration of increasing doses of recombinant human Endostatin (up to 240 mg/m 2 ), no significant toxicities were reported with stable disease in <20% of patients. Results obtained by Thomas et al. [6] in another phase I trial confirmed the safety of the drug and minor activity without significant clinical responses. In spite of this, clinical benefits and promising disease stabilizations were observed, underlining the potential biological role of these compounds. The lack of activity (according the classic criteria) of these compounds could be explained by their relatively short halflife (approximately 1 h) and high in vivo instability with daily serum levels that only briefly reach levels associated with antiangiogenic activity in vitro. To improve upon this serious limitation with the intention of increasing the targeting power, fragments containing the carboxy-terminal end of the molecule (e.g., fragment 4ox with intact disulfide bonds) fully retain the antiangiogenic activity, many hybrid molecules of Angiostatin and Endostatin have been modified. In an attempt to establish a more biologically active clinical dose of these drugs, phase II trials are ongoing Vitaxin Migration of endothelial cells is dependent upon their adhesion to extracellular matrix proteins, such as vitronectin, through a variety of cell adhesion receptors known as integrins. Recent evidence indicates that integrin alpha-v- 3 plays a role in this process [7]. Studies have shown enhanced expression of alpha-v- 3 on newly developing blood vessels in human wound tissue, tumors, diabetic retinopathy, macular degeneration, and rheumatoid arthritis. However, alpha-v- 3 is not generally found on blood vessels in normal tissues. In various animal models, antagonists of alpha-v- 3, such as the alpha-v- 3 -specific antibody, LM609, have been shown to decrease angiogenesis and induce tumor regression or improve arthritic diseases, and this has been associated with the induction of apoptosis within the angiogenic blood vessels [8 10]. Vitaxin (MedImmune, INC) is a humanized version of the LM609 monoclonal antibody that functionally blocks the alpha-v- 3 integrin. This antibody has been shown to target angiogenic blood vessels and cause suppression of tumor growth in various animal models [11,12]. Based upon the selectivity of alpha-v- 3 as a marker of angiogenic blood vessels and the effects of anti-alpha-v- 3 in reversing disease in animal models, clinical phase I/II trials were initiated in colorectal cancer patients to evaluate the safety, pharmacokinetics and efficacy of Vitaxin. Posey et al. [13] in a phase I trial (nine patients) showed that every 3-week schedule of Vitaxin at doses of 200 mg ( mg/kg) could maintain circulating levels of antibody with little or no toxicity. There was no immune response to Vitaxin in any patient. There were no objective antitumor responses but three patients receiving at least two cycles of therapy had prolonged stabilization of disease (85 days). In phase II studies [14], Vitaxin was confirmed to be well tolerated. Altogether, 80% of patients experienced one or more adverse events during therapy: the majority of which were grade 1 (68%) or 2 (32%). No toxicities greater than grade 2 were documented. The most frequent adverse events reported included antibody infusion reactions: fever, chills, nausea, and flushing. Oral premedication with acetaminophen and diphenhydramine before each Vitaxin infusion appeared to prevent fever in subsequent patients. Antibody infusion reactions were noted to decrease in incidence after the first infusion. No clinically significant cardiac, renal, hepatic, or hematological toxicities were observed. In terms of activity (all phase II studies considered), there were 14 evaluable CRC patients. Eight either demonstrated disease stabilization or a partial response. Furthermore, in one patient, treatment resulted in a partial tumor response that was maintained for 22 months. In a second patient, slight tumor shrinkage was noted only after the first cycle of therapy

4 70 A. Mancuso, C.N. Sternberg / Critical Reviews in Oncology/Hematology 55 (2005) was completed. Considering these findings and the notion that targeting vascular alpha-v- 3 integrin with Vitaxin may provide clinical benefit without significant side effects, other trials are in progress especially on prostate cancer Canstatin Canstatin, a 24-kDa peptide derived from the C-terminal globular non-collagenous (NCI) domain of the alpha2 chain of type IV collagen, induces apoptosis in cultured endothelial cells and inhibits angiogenesis in vitro and in vivo [15]. Its inhibitory molecular pathways are interesting considering that Canstatin-induced apoptosis is associated with phosphatidylinositol 3-kinase/Akt inhibition. The first phase I/II trials on CRC with surrogate markers of activity quantifying microvessel density, tumor blood flow, and tumor metabolism, should be published at the end of 2004 or in Vascular endothelial growth factor (VEGF) and blocking strategies VEGF is a growth factor that is essential for the pathological neoplastic angiogenesis, tumor growth and metastasis. Increased levels of VEGF expression have been found in many human tumors examined to date, including tumors of the lung, gastrointestinal tract, kidney, thyroid, bladder, ovary, and cervix [16]. The tumor-associated stroma may also produce significant amounts of VEGF via tumor-associated induction of the VEGF gene promoter [17]. Colorectal cancer patients show high VEGF levels in % of cases [18,19] with associated poorer outcomes [20 22]. Inhibition of VEGF signalling with several strategies, including monoclonal antibodies to VEGF (Bevacizumab), monoclonal antibodies directed versus epidermal growth factor receptor (EGFR) inhibiting angiogenesis through down-regulation of VEGF, IL-8, bfgf expression (Cetuximab, ABX-EGF), small molecules that interfere with VEGF binding or receptor signalling, ribozymes that degrade VEGF receptor mrna, are currently under clinical development Monoclonal antibodies Bevacizumab. Extensive preclinical studies have demonstrated that treatment with anti-vegf antibodies was effective in suppressing primary tumor growth as well as liver metastasis growth in a murine model of colorectal carcinoma [23]. Based on preclinical data, phase I/II/III programs with Bevacizumab (Avastin, anti-vegf humanized monoclonal antibody) were initiated. Phase I trials showed a good toxicity profile of Bevacizumab, with an increased incidence of thrombosis, bleeding (50% of patients reported transient minor epistaxis), proteinuria and hypertension, which responded to therapy [24 26]. Moreover, bowel perforation and fistula formation have been reported in colorectal cancer patients treated with Bevacizumab (Table 2). The suggested dose of Bevacizumab for phase II trials was 5 mg/kg every 14 days, although higher doses have also been proposed. The clinical activity, noted in phase I trials, was confirmed in phase II concomitant chemotherapy trials. Kabbinavar et al. [27], in a randomized, open-label, phase II multicenter trial evaluated the efficacy, safety, pharmacokinetics and pharmacodynamics of Bevacizumab in combination with 5-FU/LV as first-line chemotherapy in patients with metastatic colorectal cancer. The study enrolled 104 patients between June and November 1998, with patients randomized to treatment with standard 5- FU (500 mg/m 2 )/LV (500 mg/m 2 ) alone or in combination with a high (10 mg/kg) or low (5 mg/kg) dose of Bevacizumab. Compared with 5-FU/LV alone, both combination regimens were associated with higher response rates (17% in the control arm versus 40% in the low-dose Bevacizumab arm and 24% in the high-dose Bevacizumab arm). Combination regimens were also associated with longer median times to disease progression (5.2 months versus 9.0 and 7.2 months, respectively) and longer median survival times (13.8 months versus 21.5 and 16.1 months, respectively). This study and Langmuir s phase I/II experience [26] showed that the addition of Bevacizumab to Fluorouracil plus Leucovorin in first-line therapy, increased response rate, me- Table 2 Bevacizumab: summary toxicity data updated to ASCO 2004 Toxicity Hurwitz et al. s study [28] (813 patients) Kabbinavar et al. s study [29] (200 patients) ECOG 3200 [31] (824 patients) IFL/Beva IFL FL/Beva FL FOLFOX FOLFOX/Beva Hypertension Any Grade (G3/4) Gastrointestinal perforation and fistulae <1 Grade 3/4 bleeding Any thrombotic event Deep thrombophlebitis Pulmonary embolus Proteinuria G The values are given in percentages. I: Irinotecan; F: Fluorouracil; L: Leucovorin; Beva: Bevacizumab; FOLFOX: Fluorouracil/Leucovorin/Oxaliplatin.

5 A. Mancuso, C.N. Sternberg / Critical Reviews in Oncology/Hematology 55 (2005) dian time to disease progression, and median duration of survival. Recently, phase III first-line trials have been presented which have confirmed that the addition of Bevacizumab to IFL (Irinotecan/5-FU/Leucovorin) or FL alone chemotherapy resulted in an improved RR, DFS, and OS [28,29]. In Hurwitz et al. s study, 813 patients with previously untreated metastatic colorectal cancer were randomly assigned to receive Irinotecan, bolus Fluorouracil, and Leucovorin (IFL) plus Bevacizumab (5 mg/kg of body weight every 2 weeks) or IFL plus placebo. The primary end point was overall survival. Secondary end points were progression-free survival, response rate, duration of response, safety, and quality of life. The median duration of overall survival was significantly longer in the group given IFL plus Bevacizumab than in the group given IFL plus placebo (20.3 months versus 15.6 months). This corresponded to a hazard ratio for death of 0.66 (P < 0.00) or a reduction of 34% in the risk of death in the Bevacizumab group. The 1-year survival rate was 74.3% in the group given IFL plus Bevacizumab and 63.4% in the group given IFL plus placebo (P < 0.001). In the subgroup of patients who received second-line treatment with Oxaliplatin, the median duration of overall survival was 25.1 months in the group given IFL plus Bevacizumab and 22.2 months in the group given IFL plus placebo. Likewise, the addition of Bevacizumab to 5-FU/LV in the first-line treatment of metastatic CRC conferred a clinical advantage over 5-FU/LV in another randomized phase 3 study conducted by Kabbinavar et al. [29]. Patients with untreated metastatic CRC (209) who were not suitable for first-line Irinotecan, were randomized to receive Bevacizumab 5 mg/kg intravenously every 2 weeks in combination with bolus 5-FU/LV (5-FU 600 mg/m 2 /week plus LV 500 mg/m 2 /week 6 weeks every 8 weeks) or 5-FU/LV alone. The median survival was 12.9 months on the 5-FU/LV arm versus 16.6 months on the Bevacizumab plus 5-FU/LV arm (P = 0.16). The time to progression was significantly better in the Bevacizumab arm (9.2 months versus 5.5 months; P = ). This study suggests that Bevacizumab/5-FU/LV is also a reasonable first-line option for patients with a suboptimal performance status. Based upon the results of these randomized studies, Bevacizumab has now been approved by the FDA for the first-line treatment of metastatic colorectal cancer in combination with chemotherapy. In phase II/III trials in the salvage setting, the combination Bevacizumab and chemotherapy reports, to date, clashing results. Study TRC-0301 enrolled 350 patients who had received prior therapy; 100 were assessable for response at the ASCO Meeting in The confirmed RR was only 1% and the time to progression was 3.5 months. No survival data were presented [30]. On the contrary, ECOG 3200 study, phase III randomized trial of high-dose Bevacizumab (10 mg/kg, IV, biweekly) in combination with FOLFOX4 versus FOLFOX4 alone in Irinotecan pretreated CRC patients, shows that Bevacizumab in combination with FOLFOX4 is well tolerated and increases survival [31] (Table 3). In conclusion, Bevacizumab plus Fluorouracil-based chemotherapy should be considered a new option for the firstline treatment of metastatic colorectal cancer. In a cumulative analysis by Mass et al. [32], there was a 26% reduction in the daily risk of death with BV/FU/LV compared with 5-FU/LV or IFL alone (hazard ratio 0.742; 95% CI: ). To date, Bevacizumab can be considered also as a salvage option for pretreated CRC but data final analyses must be awaited Cetuximab (C-225). Epidermal growth factor receptor is a member of the ErbB family of receptors. In colorectal cancer, expression or up-regulation of the EGFR gene occurs in 60 80% of cases [33 35]. Moreover, expression of the gene has been associated with poor survival [36,37]. Neoplastic dimerization of the receptor activates the intracellular tyrosine kinase region of the EGFR, resulting in autophosphorylation initiating a cascade of neoplastic progression [38]. Cetuximab (Erbitux, Merck and Imclone Systems) is a chimeric IgG1 monoclonal antibody that binds to EGFR with high specificity and with a higher affinity than either epidermal growth factor or TGF-alpha, thus blocking ligandinduced phosphorylation of the EGFR. Preclinical data has suggested direct inhibition of angiogenesis that was secondary to down-regulation of VEGF, IL-8, and bfgf expression [39]. In clinical practice, Cetuximab has shown activity in the salvage setting in pretreated patients and has the ability to reverse cellular resistance to Irinotecan. In a study by Saltz et al., 57 patients with EGFR-positive colorectal cancer refractory to both Fluorouracil and Irinotecan, achieved an 8.8% partial response (PR) to Cetuximab monotherapy, and 36.8% had stable disease [40]. During the 40th ASCO Annual Meeting, three studies evaluated the role of Cetuximab in patients Table 3 Results from the Eastern Cooperative Oncology Group (ECOG) study E3200 FOLFOX4 + Beva (n = 290) FOLFOX4 (n = 289) P-value Male (%) Median age (range) 62.0 (21 85) 61.4 (25 84) PS (%): 0/1/2 48.4/46.7/ /42.9/5.5 Median OS (months) Sensory neuropathy (%) 14.9 (Grade 3), <1 (grade 4) 8.4 (Grade 3), <1 (grade 4) Beva: Bevacizumab; FOLFOX4: Fluorouracil/Leucovorin/Oxaliplatin; OS: overall survival; PS: performance status.

6 72 A. Mancuso, C.N. Sternberg / Critical Reviews in Oncology/Hematology 55 (2005) who failed all standard therapy [41 43]. Lenz et al. evaluated 350 patients refractory to both Irinotecan and Oxaliplatin. A PR was observed in 12% of EGFR-positive patients, and two major responses were reported in patients who did not express the EGFR. Thirty-four percent of the patients had stable disease for at least 6 weeks, which meant that the disease control rate was 46% [41]. Similar results were reported by Mirtsching et al. in 34 patients and in another protocol in 510 patients. Side effects included difficulty breathing, low blood pressure, acne-like rash, dry skin, tiredness, weakness, fever, constipation, and abdominal pain but safety data demonstrated no unexpected toxicities with less than 3% grade 3/4 hypersensitivity reactions [42,43]. Of most interest is Cetuximab s ability to reverse cellular resistance to Irinotecan, Saltz et al. showed a 17% RR [44] with weekly Cetuximab and Irinotecan in 121 patients with colorectal cancer who had been refractory to Fluorouracil and Irinotecan. These results were confirmed in a randomized phase III trial by Cunningham and published in the New England Journal of Medicine in Cetuximab alone or in combination with Irinotecan was given to 329 Irinotecan-refractory colorectal cancer patients [45]. The rate of response in the combination-therapy group was significantly higher than that in the monotherapy group (22.9% versus 10.8%; P = 0.007). The median time to progression was significantly greater in the combination-therapy group (4.1 months versus 1.5 months; P < 0.001). The median survival time was 8.6 months in the combination-therapy group and 6.9 months in the monotherapy group (P = 0.48). Toxic effects were more frequent in the combination-therapy group, but their severity and incidence were similar to those that would be expected with Irinotecan alone. The most frequent grade 3 and 4 events in those who received the combination therapy were diarrhea in 21.2% of patients, weakness in 13.7%, neutropenia in 13.7%, and vomiting in 6.1%. Patients who received Cetuximab alone primarily had fewer adverse effects: 13% developed dyspnea, 10.4% had weakness, and 5.2% reported abdominal pain. The observed tumor-response rate of 22.9% in this setting may be clinically important considering that this response rate is higher than the response rate reported for Oxaliplatin in combination with infusional Fluorouracil and Leucovorin (9.6%) [46]. Nonetheless, the median time to progression and the survival time were similar with the two regimens. International phase II/III studies utilizing Cetuximab in the first-line setting in combination with chemotherapy (FOLFIRI, FOLFOX) are ongoing [47 49]. Results are expected within the next year. Preliminary data suggest that Cetuximab may be safe and effective when combined with FOL- FOX4. In particular, the EXPLORE Study [49] appears to demonstrate an incidence and severity of adverse events comparable to previous reports of Cetuximab with chemotherapy or of FOLFOX4 alone. Similar conclusions can be made for the FOLFIRI/Cetuximab regimen, although only preliminary data are currently available ABX-EGF. ABX-EGF (Amgen, Panitumumab) is a high affinity fully human IgG 2 monoclonal antibody directed against human EGFR. This monoclonal antibody blocks the ligands EGF and TGF-alpha from binding to EGFR avoiding angiogenesis process, inhibiting tumor growth and eliciting both tumor regression and eradication of established tumors in murine xenograft tumor models [50]. The antitumor activity of ABX-EGF appears to be correlated with the number of EGFR molecules on the cell surface and the activity of this signalling pathway. Colorectal tumors expressing 17,000 or more EGFR molecules per cell show significant growth inhibition when treated with ABX-EGF [51]. To date, more than 300 subjects with colorectal cancer have been enrolled in ABX-EGF clinical phase I trials with weekly doses ranging from 0.01 to 5 mg/kg. ABX- EGF has been generally well tolerated when administered as monochemotherapy or in combination with other antiblastic drugs. The most common adverse event attributed to ABX-EGF is a dose-related, reversible, acneiform or maculopapular skin rash, which has been reported in approximately 90% of subjects. Less commonly reported skin effects are vesicular or exfoliative rash, skin erythema and dry skin. Some skinrelated toxicities have been associated with pain or pruritus. At the 2.5 mg/kg dose, rash was observed in 100% of subjects. Diarrhea, asthenia and nausea of any grade appear to be more frequently reported (70 80%) when ABX-EGF is administered with Irinotecan, Fluorouracil and Leucovorin [52]. Phase II studies in colorectal cancer after at least two lines of chemotherapy have shown approximately 10% objective responses and stable disease has been observed in an additional 55% [53,54]. Recently, these results have been confirmed by Hecht et al. [55]. One hundred and forty-eight patients were enrolled in two cohorts, cohort A with EGFR expression of 2+ or 3+ in 10% of tumor cells (n = 105) and cohort B with <10% 2+ or 3+ (n = 43). Of 148 patients evaluated after 8 weeks, 15 (10.1%; 95% CI: %) had confirmed partial responses (12 in cohort A, and 3 in cohort B) and 54 (36.5%) had stable disease (39 in cohort A, 15 in cohort B) by investigator assessment. The overall survival time was 7.9 months, despite a median time to progression that was only of 2.0 months. This analysis suggests that ABX-EGF has moderate single-agent antitumor activity with encouraging survival data in patients failing standard chemotherapy. Further studies are needed to define the precise role of this molecule. Phase II/III studies are ongoing in first-, second- and third-line colorectal cancer patients as monotherapy and in combination with chemotherapy EMD EMD72000 is a genetically engineered humanized monoclonal antibody binding EGFR with high specificity and affinity. EMD72000 competitively blocks natural ligand binding and abrogates receptormediated downstream signalling [56,57]. In human tumor

7 A. Mancuso, C.N. Sternberg / Critical Reviews in Oncology/Hematology 55 (2005) xenograft, antitumor activity of EMD72000 was observed either alone or in combination with Gemcitabine. Vanhoefer et al. reported the results of a phase I dose-escalation study of EMD72000 in patients with advanced solid tumors that express EGFR [58]. Twenty-two patients were treated with this molecule at five different dose levels ( mg/week). Toxicity profile showed a grade 3 NCI headache and fever occurring after the first infusion at 2000 mg/week; no allergic reactions or diarrhea, were observed. Acneiform skin reaction was the most common toxicity, but it was mild (NCI skin toxicity grade 1/2: 64% of patients). Overall response rate (23%) and stable disease (27%) were achieved at all dose levels, and responding patients, without cumulative toxicity, received treatment for up to 18 months. Pharmacokinetic-related study demonstrated a predictable pharmacokinetic profile and pharmacodynamic study on serial skin biopsies revealed that EMD72000 effectively abrogated EGFR-mediated cell signalling (e.g., reduced phosphorylation of EGFR and mitogen-activated protein kinase), with no alteration in total EGFR protein. Ongoing pharmacokinetic/pharmacodynamic studies with EMD72000 (including positron emission tomography computerized tomography (CT) imaging) will confirm safety, activity and mechanism of action defining optimal biologically effective dose and schedule. To date, this drug appears promising considering its toxicity profile better than Cetuximab and ABX-EGF Small molecules that interfere with VEGF binding or receptor signalling AMG706 AMG706 is a potent and selective small molecule inhibitor of multiple kinases, including vascular endothelial growth factor receptor, platelet-derived growth factor receptor and c-kit. Phase I studies have shown mild/moderate adverse events (hypertension, hypercreatinine, hyponatremia) and partial, minor and SD responses across multiple cancer types [59]. Randomized studies in colorectal cancer will start in NM-3 Isocoumarin NM-3, a small molecule Isocoumarin, is a recently discovered angiogenesis inhibitor. In vitro studies have demonstrated that NM-3 specifically inhibits several stages of angiogenesis, including endothelial cell proliferation, migration, tube formation, sprouting, and neovascularization. The mechanisms by which NM-3 exerts its antitumor effects are not completely understood, but data seem to indicate absence of a direct effect on tumor cells with an indirect action delaying tumor growth by inhibition of tumor angiogenesis. Moreover, findings demonstrate that NM-3 is significantly additive in combination with a broad spectrum of chemotherapeutic agents (especially 5-Fluorouracil) even when given at subtherapeutic doses and in a variety of schedules in human tumor xenograft models. NM-3 has a serum half-life of 3 10 h in preclinical models with a low toxicity profile. Presently, several clinical feasibility phase I and pilot studies are ongoing [60] Gefitinib (ZD1839, Iressa) and Erlotinib (OSI-774, Tarceva) Activation of the epidermal growth factor receptor is a key event in cell proliferation and has been implicated in several of the cellular processes that drive malignancy. These include inhibition of apoptosis, increased metastasis, angiogenesis, and invasion. Gefitinib and Erlotinib are orally active, highly selective, EGFR-TKIs that competitively inhibit adenosine triphosphate binding. Moreover, preclinical in vitro and in vivo studies have shown that these TKIs inhibit VEGF signalling with concomitant antiangiogenetic activity. Substantial antitumor activity against colon carcinoma cell lines and xenografts is well documented [61]. Gefitinib has been extensively investigated in clinical trials, with significant antitumor activity. Four phase I multicenter studies have evaluated ZD1839 monotherapy at doses up to 1000 mg/day in 252 heavily pretreated patients, including 39 with advanced CRC [62 65]. The results of these trials were promising, with ZD1839 showing antitumor activity, predictable pharmacokinetics, and a favorable adverse effect profile. In a phase I trial, Baselga et al. recruited 21 patients with CRC of 88 patients in the trial. He reported durable clinical response to ZD1839 with evidence of antitumor activity over the entire dosage range of mg/day. Overall, 6/21 (28%) CRC patients had stable disease and received ZD1839 for at least 3 months. Evidence of antitumor activity was also shown by >50% reductions in serum concentrations of the tumor marker carcinoembryonic antigen (CEA) in 5/21 patients. One patient received ZD1839 for 6 months, with stable disease and no change in CEA. Overall, ZD1839 was generally well tolerated with manageable and reversible adverse effects at doses up to 600 mg/day and dose-limiting toxicity (diarrhea) observed at 1000 mg/day. A phase I/II trial is under way to evaluate ZD1839 monotherapy in CRC [66]. The first part of the trial was a phase I dose-escalation ( mg/day) study of patients with a variety of solid tumors. From these results, a dose of 750 mg/day was selected for the phase II part of the study, which recruited 27 patients with advanced or metastatic CRC. Among 24 patients evaluable for response, 8 had stable disease (median duration 2.2 months) with 5 of these patients showing evidence of tumor shrinkage. Further evidence of antitumor activity is provided by tumor biopsy data, which showed decreased levels of phosphorylated (p)-egfr, p- AKT, p-erk, and the proliferation marker Ki-67; pharmacodynamic studies show that ZD1839 may also induce apoptosis in colorectal cancer patients via up-regulation of p27 (Kip1) [67]. ZD1839 has also been studied in combination with a number of cytotoxic and other agents such as 5-FU, Irinotecan, Oxaliplatin, Capecitabine, and Celecoxib. Combination therapy of ZD1839 with 5-FU/LV has been investigated as first-line treatment in a phase I trial involving 26 patients

8 74 A. Mancuso, C.N. Sternberg / Critical Reviews in Oncology/Hematology 55 (2005) with metastatic CRC [68]. In conjunction with the Mayo Clinic regimen of 5-FU/LV, ZD1839 was administered at mg/day. This combination treatment was generally well tolerated and produced promising antitumor activity. Objective tumor responses were seen with 1 complete response, 5 partial responses, and 12 patients with stable disease. Pharmacokinetic studies showed that no significant change in exposure of either ZD1839 or 5-FU/LV was observed when given in combination. Preclinical data have also shown that ZD1839 reverses resistance to Irinotecan and enhances its efficacy by improving oral bioavailability. In phase I/II studies, dose-limiting toxicities, especially neutropenia, occurred at an unexpectedly low dose of Irinotecan (250 mg/m 2 every 21 days) in combination with Gefitinib (250 mg daily). The dose reductions required together with the modest activity observed led to the early cessation of these trials. Pharmacokinetic analyses to explore a possible interaction between Gefitinib and Irinotecan are in progress [69,70]. Other clinical trials (phases I and II, respectively) have investigated ZD1839 in combination with 5-FU/LV and Oxaliplatin [71 72]. In the first, treatment involved dose escalation of ZD1839 ( mg/day) and Oxaliplatin (70 85 mg/m 2 ). Sixteen patients were recruited with a variety of advanced refractory epithelial tumors including 12 patients with CRC. Data on overall best tumor response was available for 13 patients, of whom 3 patients with metastatic CRC had confirmed partial responses. An additional 9 patients (of the remaining 10) showed reductions in tumor markers and/or stable disease after four cycles of treatment. Adverse events were usually mild, with the combined ZD1839/Oxaliplatin/5- FU/LV treatment being generally well tolerated. Fisher et al. [72] in a second study reported the efficacy and toxicity data of Gefitinib in association with FOL- FOX4. Fifty-six patients were stratified by prior therapy for metastatic disease; group A had no prior therapy and group B had prior therapy. In group A, 21 of the 27 patients (78%) achieved a partial response (PR): 30% had resection of liver metastases. In group B, 8/22 patients (36%) achieved PR. Grade 3/4 toxicities included neutropenia (53%), diarrhea (49%), nausea (28%) and vomiting (21%). Erlotinib (OSI-774, Tarceva), like ZD1839, is being studied in phase I and II trials [73,74]. The most frequent adverse events in phase I trials were diarrhea and acneiform rash, indicating that these effects may be common to the EGFR- TK-targeted agents as a class. The maximum-tolerated dose of OSI-774 was established at 150 mg/day, which is close to the dose needed to achieve biologically active plasma levels. In a phase Ib dose-escalation trial in combination with chemotherapy (Capecitabine and Oxaliplatin), the recommended dose of Erlotinib for combination phase II trials was 100 mg/mg. There was no evidence of pharmacokinetics interactions [75]. Phase II studies with single-agent Erlotinib showed growth tumor control in CRC patients, with approximately 30% attaining stable disease. Townsley et al. s secondline study of single-agent Erlotinib at a dose of 150 mg PO daily, in 38 patients with metastatic CRC, showed good overall tumor growth control with 39% attaining stable disease. Minor responses were observed. The median number of days until clinical progression was 56 and the median number of days until progression for patients with stable disease was 116 days. The most common adverse events were rash in 34 patients over 51 cycles (48% of cycles), diarrhea in 23 patients over 40 cycles (38%) and lymphopenia in 27 patients over 50 cycles (47%). In conclusion, phase I and II trials have shown the efficacy and tolerability of Gefitinib and Erlotinib treatment in patients with CRC as monotherapy and in combination with cytotoxic drugs. These clinical results are of particular interest because they build upon the strong evidence from preclinical studies that it would be effective to target the EGFR/VEGF receptors and their pathways in this disease. Future phase II trials are planned or under way to investigate ZD1839 or OSI-774 in CRC either alone or in combination with standard cytotoxic or biological drugs such as Oxaliplatin, 5-FU, Irinotecan, and Capecitabine, and in combination with radiotherapy CI-1033 CI-1033, a pan-erb inhibitor, is an irreversible inhibitor of EGFR-TK that also inhibits VEGF signalling and other EGFR family members such as HER-2 [76]. This EGFR- TK inhibitor has antitumor activity in xenograft models and is currently being investigated in phase I trials with disease stabilization observed in about 11% of treated patients. In addition to the skin rash and diarrhea common to this class of agents, adverse events seen with CI-1033 in this study were grade 3 hypersensitivity and grade 4 thrombocytopenia [76]. Other phase I clinical and pharmacokinetic studies of oral CI in patients with advanced solid tumors find CI-1033 to be generally well tolerated with a favorable pharmacokinetic profile [77] SU-6668 SU-6668 is an oral, small molecule inhibitor of the angiogenic receptor tyrosine kinases vascular endothelial growth factor receptor 2 (FLK-1/KDR), platelet-derived growth factor receptor, and fibroblast growth factor receptor 1 [78]. In preclinical studies on tumor models of HT29 human cells colon carcinoma in athymic mice, SU-6668 was able to greatly reduce vessel count and to inhibit tumor growth, with 60% inhibition at 14 days of treatment [78]. Moreover, SU-6668 augments tumor-suppressive effects of radiotherapy possibly through reducing the survival of tumor endothelium [79]. Xiong et al. at the M.D. Anderson Cancer Center evaluated the biologic effects of SU-6668 (200 or 400 mg/m 2 /day) in seven patients with a variety of solid tumors (including colorectal cancer) using comprehensive measures of pharmacokinetics (PK), functional imaging, and tissue correlative studies. Functional computerized tomography showed that five of six patients had decreased blood flow in tumors in

9 A. Mancuso, C.N. Sternberg / Critical Reviews in Oncology/Hematology 55 (2005) response to treatment, and MRI results indicated significant change of area under the signal intensity versus time curve (AUC) and/or maximum slope (maximum rate of signal intensity change) in two of four patients evaluated with this technique [80]. Phase I studies by Brahmer et al. [81] and Morimoto et al. [82] have shown dose-limiting toxicities including fatigue, pleuritic chest pain, shortness of breath, and pericardial effusions at dose of 400 mg/m 2 BID. Based on these studies and others phase I trials, the recommended phase II dose is 300 mg/m 2 BID with food. Currently, phase II studies are being initiated to evaluate the potential of SU-6668 as an anticancer agent for pretreated colorectal cancer patients PTK-787 PTK-787 is a novel compound targeting inhibition of VEGFR tyrosine kinase with therapeutic potential for the treatment of solid tumors and other diseases where angiogenesis plays an important role. It is a potent and orally active inhibitor of VEGFR tyrosine kinase with higher selectivity for the KDR/FLK-1 receptor tyrosine kinase than the flt-1, flt-4, platelet-derived growth factor (PDGF) receptor tyrosine kinase and c-kit protein tyrosine kinase [83]. Treatment with PTK-787 primarily reduces the number of tumor microvessels accompanied by hemodynamic dilation of the remaining vessels [84]. In phase I studies, 159 patients with a variety of advanced cancers have received PTK-787 as a single agent or in combination with a standard chemotherapy regimen at dose levels ranging from 50 to 2000 mg/day; and several of these patients have received PTK-787 for up to 15 months. The most frequent (>30%) adverse events were nausea (47%), fatigue (39%), vomiting (36%) and dizziness (34%). The 1250 mg continuous once daily dose and schedule of PTK- 787 has been selected as a safe and biologically active dose regimen based upon correlations between pharmacodynamic effects demonstrated on dynamic contrast-enhanced (Gd- DTPA) magnetic resonance imaging (DCE-MRI) of tumor lesions in monotherapy trials [85,86]. In an expanded phase I/II study of PTK-787 in combination with FOLFOX4 as first-line treatment for patients with metastatic colorectal cancer, Steward et al. [87] showed in 35 patients that the pharmacokinetics and toxicity profiles of both PTK-787 and FOLFOX4 were unaffected by co-administration. PTK-787 was well tolerated at doses 1250 mg/day without DLTs. Adverse events at 1250 mg/day included ataxia (grade 3), neutropenia and thrombocytopenia (grades 4 and 3, respectively), two episodes of dizziness (grade 3). CNS adverse events (grade 3/4) were dose limiting in two patients at 2000 mg/day, and expressive dysphasia (grade 3) and intermittent dizziness (grade 3) were dose limiting at 1500 mg/day. Best response data for 28 evaluable patients showed a complete response (4%), 14 partial responses (50%), with 9 patients showing stable disease. In this study, progression-free survival (PFS) was 11 months with an estimated median overall survival of 16.6 months. A parallel first-line study was performed by Schleucher et al. [88]. This study confirmed that the combination of PTK-787 and chemotherapy (FOLFIRI in this trial) was safe, well tolerated and active. The pharmacokinetics of PTK- 787 was unaffected by FOLFIRI but co-administration of 1250 mg/day PTK-787 with FOLFIRI had minimal effects on Irinotecan exposure with serum reductions in of SN38 of some 40%. Efficacy results on only 12 patients included 4 partial responses (33%) and 5 patients with stable disease (42%). In 11 evaluable patients, the median time to progression was 6.7 months. In second and third line, PTK-787 is active in chemotherapy refractory patients. Data from salvage monotherapy studies in colorectal cancer patients who were previously treated with at least one prior chemotherapy reveal a median time to progression of 2.8 months with a 95% confidence interval ranging between 1.9 and 6.2 months, and a median overall survival of 9.2 months with a 95% confidence interval ranging between 5 and 16.9 months [85]. In light of these encouraging results and the need for new treatment options for patients in this setting, phase III randomized controlled combination trials (Confirms 1 and 2) have been developed and it are now closed to accrual. Final results should be published during first 6 months of Inhibitors of angiogenesis with unclear mechanisms of action Combretastatins Combretastatins are small organic molecules found in the bark of the African bush willow, the Combretum caffrum. Combretastatins not only suppress proliferating endothelium, but also specifically targets tumor endothelium. The Combretastatin A-4 prodrug (CA4P) is a derivative of Combretastatin, which is activated by a phosphatase that is selectively amplified in proliferating endothelial cells. Combretastatin A-4 induces apoptosis in human endothelial cells [89]. In tumor-bearing mice, CA4P significantly enhanced the antitumor effects of radiation therapy [90]. Recent results from a dose-finding phase I study of 25 patients with advanced cancer [91] (about 1/4 with colorectal cancer) showed that administration of escalating doses of CA4P was not associated with conventional toxic effects, such as myelotoxicity, stomatitis, or alopecia. Conversely, the toxicity profile of CA4P seemed to be more related to its antiangiogenic properties, including tumor-localized pain (in about 10% of treatment cycles) and acute coronary syndrome events at higher doses ( 60 mg/m 2 ). Over 19 months, a durable stable response was observed in a patient with metastatic colon carcinoma, consistent with preclinical in vitro data showing a high sensitivity of colorectal carcinoma cell lines to CA4P. The cardiovascular safety profile of Combretastatin A-4 phosphate is under intense investigation; preliminary results

10 76 A. Mancuso, C.N. Sternberg / Critical Reviews in Oncology/Hematology 55 (2005) show that CA4P prolongs the QTc interval and a temporal relationship with the CA4P infusion and ECG changes consistent with an acute coronary syndrome are present in about 8% of patients [92] Methoxyestradiol 2-Methoxyestradiol (2-ME) is a physiological metabolite of estrogen, which is excreted in the urine. In contrast to most estrogens, this molecule shows tumor growth-inhibiting effects through several different mechanisms. Besides exhibiting a strong antiangiogenic effect on endothelial cells and tumors, it possesses a tubulin-inhibiting mechanism, causing cells to arrest in the G2/M phase of the cell cycle. In other tumor cells, p53-dependent and p53-independent mechanisms with induction of apoptosis have been shown. Treatment of CRC cells with 2-methoxyestradiol increases expression of p53 and p21 proteins with consequence of induced apoptosis [93]. 2-ME is currently being evaluated in phase I and II clinical trials, under the name Panzem, for the treatment of multiple types of cancer, especially prostate and breast cancers where 2-ME has thus far been well tolerated and appears to demonstrate promising activity. Feasibility and activity data in colorectal cancer are not yet available, but several phase I projects are ongoing based upon its intriguing rationale [93] Thalidomide and analogues (SelCID-3) Thalidomide is an oral drug that has been shown to have clinical potential in a number of conditions in which there are limited therapeutic options, including cancer [94,95]. Clinical efficacy may be associated with a number of diverse properties attributed to thalidomide that include the inhibition of TNF-alpha synthesis [96], co-stimulation of T cells [97] and inhibition of angiogenesis [98]. The effects of Thalidomide and analogues (especially SelCID-3, which are phosphodiesterase inhibitors) on cell viability and cell cycle arrest are pronounced in colorectal and pancreatic tumor lines. Growth arrest is preceded by the early induction of G 2 /M cell cycle arrest, which led to caspase- 3-mediated apoptosis. This is associated with increased expression of pro-apoptotic proteins and decreased expression of antiapoptotic bcl-2. Interestingly, p53 appears not to be involved in the apoptotic process and this is clinically important because many advanced refractory cancers have p53 mutations [99]. Phase I/II clinical trials of Thalidomide monotherapy or in association with Irinotecan in refractory metastatic colorectal cancer have been developed. Dal Lago et al. [100], in a phase I/II trial and pharmacokinetic study showed that Thalidomide mg/day was generally well tolerated, with constipation, somnolence, dizziness and dry mouth as major toxicities. The activity as a salvage regimen seemed minimal without objective responses or stable disease. The reported median survival in this study was 3.6 months, similar to what one might expect from best supportive therapy. Conversely, when every 3 weeks Irinotecan 350 mg/m 2 was combined with Thalidomide 400 mg/m 2 /day responses were demonstrated in a study from the University of Arkansas [101]. Of 10 patients evaluable for response, the authors reported 2 complete responses and 2 partial responses, meaning disease control in 40%. Moreover, in this combination, investigators noted a remarkable absence of grade 3/4 gastrointestinal toxicities, concluding that perhaps the immunomodulatory properties of Thalidomide were able to mitigate the dose-limiting gastrointestinal toxic effects of Irinotecan, particularly diarrhea and nausea (P < ) [102]. Recently, Thalidomide has been studied in an oral combination regimen with Capecitabine in 34 previously treated, refractory metastatic colorectal cancer patients [103]. Treatment was well tolerated without radiographic responses, but 5 patients (17%) achieved a decline in serum CEA of 50% or greater and 13 patients (38%) achieved stable disease. The median progression-free survival was 2.6 months and the median overall survival was 7.1 months. In conclusion, although Thalidomide does not show specific cytotoxic activity, the rate of disease stabilization and overall survival observed in this heavily pretreated cohort of patients suggests that this drug may offer some clinical benefit through cytostatic mechanisms. Other combination phase II studies are under investigation considering its antiangiogenic properties and multiple mechanisms of action. 3. Discussion, problems and conclusions It is increasingly evident that there is an intense interest in targeted therapies for the treatment of colorectal cancer. The addition of the angiogenesis inhibitor Bevacizumab to Irinotecan, 5-Fluorouracil and Leucovorin in the first-line treatment of patients with metastatic CRC conferred a significant improvement in both progression-free and overall survival. However, prudently, must be considered that IFL is currently considered a suboptimal regimen and the Bevacizumab overall survival advantage remains to be confirmed with more effective regimen such as FOLFOX [104]. The addition of the epidermal growth factor receptor inhibitor Cetuximab to Irinotecan as second-line therapy conferred a significant improvement in time to progression by reversing cellular resistance to Irinotecan. The strong association observed in some studies of rash severity with response/survival suggests that rash may serve as a marker of response to Cetuximab treatment and may be used to guide treatment to obtain optimal response. Further studies of the potentially important association between rash and outcome of treatment with EGFR-targeted agents are needed [105]. The availability of other new and promising antiangiogenesis inhibitors such as PTK-787, Gefitinib/Erlotinib and ABX-EGF combined with their optimal toxicity profile as single agents or in combination with Oxaliplatin, has led to investigation of their possible roles in various disease settings.