Department of Clinical Pharmacology and Cancer Therapeutics, Racliffe Infirmary, Woodstock Road, Oxford OX2 6HE, UK
* Correspondence to: Dr R. Midgley, Cancer Research UK, Churchill Hospital, Headington, Oxford, UK, OX3 7LJ. Email: rachel.midgley{at}cancer.org.uk
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: angiogenesis, bevacizumab, vascular endothelial growth factor
![]() |
Introduction to angiogenesis and VEGF |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The primary stimulus to angiogenesis is oxygen deprivation. When this occurs pro-angiogenic factors are released which activate the formation of new blood vessels through a number of downstream effects, such as proliferation and migration of vascular endothelial cells, survival of immature endothelial cells and induction of vascular permeability in capillaries. Candidate pro-angiogenic factors include vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF) and transforming growth factors alpha and beta (TGF and ß), the most important of these being VEGF [1
, 2
]. There are also anti-angiogenesis factors that have an opposite effect and it is the balance of pro- and anti-angiogenic factors that determines the angiogenesis balance (Figure 1).
|
|
![]() |
History of bevacizumab |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The first phase I trial of Bevacizumab began in 1997 [7]. Twenty-five patients with measurable or assessable solid tumours (sarcoma=8, renal cell carcinoma (RCC)=7, breast cancer=5, lung cancer=2) were enrolled and five dose levels were assessed0.1, 0.3, 1.0, 3,0 and 10 mg/kg. Bevacizumab was delivered intravenously over 90 minutes on days 0, 28, 35 and 42. One patient with RCC achieved a minor response with an approximate 25% reduction in the sum of perpendicular diameters of the pulmonary and lymph node metastases. 48% of the remaining patients achieved disease stabilisation. Bevacizumab showed a linear pharmacokinetic profile and a terminal half life of 21 days. A phase Ib trial then assessed Bevacizumab in combination with a number of standard cytotoxicsdoxorubicin, carboplatin, paclitaxel and 5-fluorouracil (5FU)/leucovorin (LV). Bevacizumab was administered at 3 mg/kg weekly for 8 weeks. This trial demonstrated that adding the anti-VEGF antibody did not significantly increase the known cytotoxic adverse event rates [8
].
Subsequently five parallel phase II trials were commenced, three of single agent Bevacizumab (in hormone-refractory prostate cancer, relapsed metastatic breast cancer and IL2-refractory renal cell cancer) and two in combination with standard first line therapy (in inoperable non small cell lung cancerNSCLC and metastatic colorectal cancer). The most encouraging trials were in renal cancer, NSCLC and colorectal cancer [9, 10
, 11
].
In renal cancer a randomised double blind phase II trial was performed recruiting 116 patients to one of three groupsplacebo, bevacizumab 3 mg/kg or bevacizumab 10 mg/kg, administered every 2 weeks. The time to disease progression was significantly prolonged in the high dose but not the low dose group compared to the placebo group. The percentage of patients being progression-free was 64% in the high dose group compared to 34% and 20% in low dose and placebo groups respectively. Toxicity included increased blood pressure and asymptomatic proteinuria [9].
In the phase II trial in advanced NSCLC ninety-nine chemotherapy-naïve patients were randomised to receive carboplatin and paclitaxel plus or minus bevacizumab (at 7.5 or 15 mg/kg), all agents being administered every three weeks. Bevacizumab was given until disease progression or to a maximum of one year; six cycles of chemotherapy were administered. The trial indicated no significant increase in toxicity compared to chemotherapy alone and, in the 15 mg/kg arm, suggested an increase in time to progression (7.4 versus 4.2 months) and modest increase in survival (17.7 versus 14.9 months) in patients receiving bevacizumab compared to chemotherapy alone [10].
In the first phase II trial in metastatic colorectal cancer, 104 previously untreated patients were randomised to receive 5FU/LV (for 6 weeks of an 8-week cycle) plus bevacizumab, 5 mg/kg (n=35) or 10 mg/kg (n=33) every 2 weeks or placebo (n=36) until disease progression. Interestingly the objective response rate was superior for the 5 mg/kg bevacizumab arm (40% compared to 24% and 17% in the 10 mg/kg bevacizumab dose and the placebo arms respectively). This stands in contrast to the results above in the renal cancer trial where a dose-response relationship was apparent. Whether the colorectal result was due to a subtle randomisation imbalance with more poor prognosis patients receiving the high dose or whether the result was due to a physiological phenomenon, in which the high dose bevacizumab caused excessive regression and destruction of blood vessels and therefore poorer chemotherapy penetration is not known. Again thrombosis was the most significant toxic effect, but hypertension, proteinuria and epistaxis were also observed.
A second phase II trial in CRC was then commenced, principally because the IFL regime (bolus 5FU/LV plus irinotecan) had become the new standard first line treatment for metastatic CRC in the United States. Ninety-two patients with advanced CRC received bevacizumab (10 mg/kg infusion every 2 weeks) in combination with IFL (bolus IV for 4 of 6 weeks). Seventy-two further patients received the same dose of bevacizumab but reduced irinotecan and 5FU starting doses; these doses were then escalated as long as no greater than grade 1 neutropenia or diarrhoea was observed. Overall the complete response rate, partial response rate and stabilisation rate were 5.4%, 38% and 36% respectively, giving a promising overall disease control rate of 79.4% [12].
Based on the results of these two phase II trials and some ongoing concerns about the safety of the IFL regime, a double blind randomised phase III three arm study was commenced. Patients received IFL plus placebo (n=411), IFL plus bevacizumab (n=402) or 5FU/LV plus bevacizumab (n=110). When acceptable safety of the IFL/bevacizumab combination was established the third arm stopped recruitment. The results of the trial are summarised in Table 1. The addition of bevacizumab to IFL chemotherapy produced significant improvements in response rate (44.8 versus 34.8%; P=0.004), time to progression (10.6 versus 6.2 months; P <0.001) and median survival (20.3 versus 15.6 months; P <0.001) [13]. Toxicity is still of some concern with this combination. IFL alone produces grade 3/4 toxicity rates of 30% and 25% for neutropenia and diarrhoea respectively. Adding bevacizumab significantly increased grade 3 hypertension rates (11% versus 2.3%, P <0.01), but interestingly did not impact significantly upon the rates of proteinuria, thrombosis and bleeding. One major concern was six gastrointestinal perforations occurring in patients receiving IFL/bevacizumab and this possibility needs to be highlighted to clinicians who are using the combination and potential symptom profiles outlined to patients to allow early intervention.
|
|
![]() |
Current status of bevacizumab |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
What about patients that are not fit for irinotecan in the first line treatment of advanced colorectal cancer? The companion trial to the phase III trial described above randomised 209 patients to receive Roswell Park 5FU/LV (6 of 8 weeks) plus placebo or bevacizumab 50 mg/kg every 2 weeks [11]. The addition of bevacizumab produced a statistically significant prolongation of progression-free survival and a trend towards improved response rate and survival. However again there was a 2% gastrointestinal perforation rate in the bevacizumab arm and this needs careful consideration when assessing the risks and benefits of the addition of bevacizumab in this population.
In the adjuvant setting of colorectal cancer, bevacizumab is being assessed in combination with capecitabine (1000 mg/m2, 14 of 21 days) and irinotecan (250 mg/m2 every three weeks) in the United Kingdom QUASAR 2 trial.
In advanced breast cancer the first phase III trial (n=462) comparing capecitabine plus or minus bevacizumab as third line chemotherapy (after anthracyclines and taxanes) demonstrated an improved response rate with bevacizumab but no impact upon progression-free or overall survival [14]. It is now being assessed in first line metastatic breast cancer in combination with paclitaxel. Bevacizumab is also being assessed in a phase III trial in NSCLC (with carboplatin and paclitaxel); and in the first line treatment of advanced renal cancer in combination with interferon-
(CALGB 90 206) where the dose of bevacizumab will be 10 mg/kg based on the phase II trial data for this disease.
![]() |
Bevacizumab at ASCO 2004 |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
An exploratory study in inflammatory breast cancer patients assessed response and angiogenesis after treatment with bevacizumab (15 mg/kg). Sixteen patients received a cycle of bevacizumab alone followed by six cycles of neoadjuvant bevacizumab, doxorubicin and docetaxel. Post-mastectomy and radiation, patients were administered eight further cycles of bevacizumab (plus or minus hormone therapy depending on ER status). Tumour biopsies and dynamic contrast-enhanced MRI were obtained prior to treatment, after the single cycle of bevacizumab alone and then after cycles 4 and 7 of combination treatment. Of 10 evaluable patients, nine achieved a partial response and one demonstrated progressive disease. Preliminary data suggested a decrease in vascular permeability (MRI) and endothelial cell proliferation (Cd31/Ki67 expression) even after the single cycle of bevacizumab alone. In responding patients there was a trend towards a decrease in VEGF after bevacizumab alone and in tumour cell proliferation (Ki67) after chemotherapy and bevacizumab. This is an interesting mechanistic study and suggests that bevacizumab warrants further exploration in this setting where treatment is often difficult and inadequate.
Also presented were two phase I trials. The first was a study of neoadjuvant bevacizumab, 5FU and radiation followed by surgery for primary rectal cancer in nine patients with T3 or T4, Nx, Mo tumours. Patients received bevacizumab at a dose of 5 mg/kg (n=6) or 10 mg/kg (n=3) every 2 weeks, with bevacizumab therapy commencing 2 weeks prior to chemoradiation. Surgery was performed 7 weeks after completion of all therapy. All patients to date had completed the course of treatment without DLT and pathological review of the surgical specimens of six of the seven evaluable patients showed only limited microscopic disease (in keeping with the lack of macroscopic disease visualised on follow-up endoscopy prior to surgery). These results demonstrate safety of the combination and significant clinical activity, again suggesting that further investigation would be appropriate.
The second phase I study assessed the combination of bevacizumab and an epidermal growth factor receptor (EGFR) inhibitor erlotinib in patients with advanced squamous cell carcinoma of the head and neck cancer receiving no more than one previous therapy for their disease. The rationale for the combination lies in the observation in pre-clinical studies that resistance to EGFR inhibitors appears related to angiogenesis. A fixed dose of erlotinib (150 mg orally daily) was administered with an escalating dose of bevacizumab (5, 10 and 15 mg/kg every 3 weeks). 10 patients were enrolled. Grade 3 diarrhoea, skin rash and lymphopenia were seen in one patient each. Grade 4 haemorrhage from a tongue tumour was observed in one patient. No dose limiting toxicity was seen. One patient gained a partial response with 7 patients achieving disease stabilisation and a phase II trial is now underway.
The combination was similarly assessed in advanced/recurrent NSCLC. A phase I trial first set a maximum tolerated dose (MTD) of 15 mg/kg bevacizumab every 3 weeks in combination with erlotinib 150 mg once daily orally. Thirty-four patients were treated at the MTD in the phase II portion of the trial. Twenty-four of these patients had received at least two chemotherapeutic regimes. The partial response rate was 22.5% with minor response in a further 5% and stable disease in 35%, giving an overall disease control rate of 62.5%, which compares very well with historical controls treated in this setting. A randomised phase II trial is now planned in a similar cohort of patients.
Another group assessed the same combination in advanced renal cancer who had received a maximum of one previous regimen. Bevacizumab was given at a dose of 10 mg/kg every 2 weeks and erlotinib 150 mg once daily until progression. Fifty-seven patients were treated, 40 of which were evaluable for efficacy (at least 8 weeks of therapy). Progression-free survival at 6 months was 71% with a partial response rate of 25%, a minor response rate of 15% and a disease stabilisation rate of 47%. Grade 3/4 toxicity rates were lowhypertension 11%, diarrhoea 9%, rash 7% and nausea/vomiting 7%. Again the substantial clinical activity means that randomised phase II or even phase III trials should now be initiated.
Bevacizumab was also assessed in phase II trials in a number of new indications, including soft tissue sarcoma, carcinoid tumours and advanced pancreatic cancer. Bevacizumab (15 mg/kg) was combined with doxorubicin (75 mg/m2), both given every 3 weeks, in the treatment of soft tissue sarcomas. The response rate of 14% was no better than historical controls treated with doxorubicin alone; however the authors felt that the disease stabilisation rate of 59% was encouraging and warranted further investigation.
Carcinoid tumour patients were randomly assigned to either bevacizumab (dose not given in abstract) or PEG interferon 2 for the first 18 weeks followed by the combination of the two. Tumour blood flow was assessed by functional CT. The early results demonstrated a rapid decrease in tumour blood flow in patients treated with bevacizumab as opposed to PEG interferon 2
but the data were too immature to draw any conclusions about efficacy.
Updated results from a phase II trial of gemcitabine (1000 mg/m2 weekly for 3 out of 4 weeks)/bevacizumab (10 mg/kg every 2 weeks) in chemotherapy-naïve advanced pancreatic cancer patients were also presented. Forty-two patients were evaluable for response with a median follow-up of 5.7 months. The partial response rate was 21% with a disease stabilisation rate of 45%. Six-month survival was 74%. There were two deaths, one due to a gastrointestinal bleed and one secondary to bowel perforation, so clearly these toxicities are not confined to patients with colorectal cancer. A randomised phase III CALGB study is now under development.
![]() |
Future considerations for bevacizumab and other VEGF inhibitors |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In order to increase our ability to select the most promising anti-angiogenesis drugs overall, and in particular for individual patients a number of surrogate endpoints for effect have been explored. These include treatment-induced depression in plasma VEGF, reduction in tumour blood flow on functional CT or MRI or biopsy proof of a decrease in angiogenesis but to date none of these has been performed in an adequate number of patients or been correlated in any statistically robust way with response to promote routine clinical use. An area of exploration for the future will be studying polymorphisms of the VEGF molecule itself or its receptor and attempting to correlate these with tumour response.
One area of confusion surrounding bevacizumab has been the differential dosing seemingly required for different tumours, with colorectal cancers requiring only 5 mg/kg/ 2 weeks compared to 10 mg/kg or greater for NSCLC and renal cancers. It must be remembered that the dose defined for colorectal cancer was based on a relatively small phase II study (about 35 patients/arm) and was defined according to efficacy. With these numbers a slight swing of two or three patients in either direction (responders being non-responders or vice versa) would result in a completely different conclusion. In the future consideration should be given to dose-confirmation for biological agents being integrated into large randomised phase III trials, perhaps posed in a 2x2 factorial fashion to reduce the numbers of patients required.
One final question has been the conversion of dosing of bevacizumab into a three-weekly schedule to fit in with other chemotherapy regimes. At ASCO this year results from the CRC and NSCLC phase II trials were combined to assess the pharmacokinetic effect of two-weekly (5 mg/kg) versus three-weekly (7.5 mg/kg) dosing on tumour exposure to bevacizumab. The authors concluded that observed and simulated drug exposure were similar in both arms, due to the slow clearance and long elimination half life of 20 days for bevacizumab. This answers any concern about the three weekly schedule and allows QUASAR 2 and other large phase III trials to proceed without further deliberation.
![]() |
Acknowledgements |
---|
Received for publication January 13, 2005. Accepted for publication February 2, 2005.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2. Jain RK. Molecular recognition of vessel maturation. Nat Med 2003; 9: 68593.[CrossRef][ISI][Medline]
3. Ferrara N. VEGF: an update on biological and therapeutic aspects. Curr Opin Biotechnol 2000; 11: 617624.[CrossRef][ISI][Medline]
4. Brown LF, Berse B, Jackman RW et al. Vascular stroma formation in carcinoma in situ, invasive carcinoma and,metastatic carcinoma of the breast. Clin Cancer Res 1999; 5: 10411056.
5. Karayiannakis AJ, Bolanaki H, Syrigos KN et al. Serum vascular endothelial growth factor levels in pancreatic cancer patients correlate with advanced and metastatic disease and poor prognosis. Cancer Lett 2003; 194: 119124.[CrossRef][ISI][Medline]
6. Kim KJ, Li B, Winer J et al. Inhibition of vascular endothelial growth factor-induced angiogenesis suppresses tumour growth in vivo. Nature 1993; 362: 841844.[CrossRef][ISI][Medline]
7. Gordon MS, Margolin K, Talpaz M. Phase I safety and pharmacokinetic study of recombinant human anti-vascular endothelial growth factor in patients with advanced cancer. J Clin Oncol 2001; 19(3): 843850.
8. Margolin K, Gordon MS, Holmgren E. Phase Ib trial of intravenous recombinant humanised monoclonal antibody to vascular endothelial growth factor in combination with chemotherapy in patients with advanced cancer: Pharmacologic and longterm safety data. J Clin Oncol 2001; 19(3) 851856.
9. Yang JC, Haworth L, Sherry RM et al. A randomised trial of bevacizumab, an antivascular endothelial growth factor antibody for metastatic renal cancer. N Engl J Med 2003; 349: 427434.
10. Johnson DH, Fehrenbacher L, Novotny WF et al. Randomised phase II trial comparing bevacizumab plus carboplatin and paclitaxel with carboplatin and paclitaxel alone in previously untreated metastatic or locally advanced non small cell lung cancer. J Clin Oncol 22(11): 21842191.
11. Kabbinavar F, Hurwitz H, Fehrenbacher L et al. Phase II randomised trial comparing Bevacizumab plus 5FU/LV with 5FU/LV alone in patients with metastatic colorectal cancer. J Clin Oncol 2003; 21(1): 6065.
12. Giantonio BJ, Levy D, O'Dwyer PJ et al. Bevacizumab plus IFL as frontline therapy for advanced colorectal cancer: Results from ECOG study 2200. Proc Am Soc Clin Oncol 2003; 22: 1024.
13. Hurwitz H, Fehrenbacher L, Novotny W et al. Bevacizumab plus irinotecan, flurouracil and leucovorin for metastatic colorectal cancer. N Engl J Med 2004; 350(23): 23352342.
14. Miller KD et al. Phase III trial of capecitabine plus bevacizumab versus capecitabine alone in women with metastatic breast cancer previously treated with an anthracycline and a taxane. Breast Cancer Res 76: S37.
15. Assorted authors. Proc Am Soc Clin Oncol 2004. See Table 2.