Summary Basis of Decision for Avastin ®
Review decision
The Summary Basis of Decision explains why the product was approved for sale in Canada. The document includes regulatory, safety, effectiveness and quality (in terms of chemistry and manufacturing) considerations.
Product type:
Avastin®
Bevacizumab, 25 mg/mL, Solution, Intravenous infusion
Hoffmann-La Roche Ltd.
Submission control no: 089366
Date issued: 2006-05-11
Health Products and Food Branch
Our mission is to help the people of Canada maintain and improve their health.
Health Canada
HPFB's Mandate is to take an integrated approach to the management of the risks and benefits to health related to health products and food by:
- Minimizing health risk factors to Canadians while maximizing the safety provided by the regulatory system for health products and food; and,
- Promoting conditions that enable Canadians to make healthy choices and providing information so that they can make informed decisions about their health.
Health Products and Food Branch
Également disponible en français sous le titre : Sommaire des motifs de décision (SMD), AVASTINMD, Bevacizumab, 25 mg/mL solution, Hoffmann-La Roche Limitée, N° de contrôle de la présentation 089366
Foreword
Health Canada's Summary Basis of Decision (SBD) documents outline the scientific and regulatory considerations that factor into Health Canada regulatory decisions related to drugs and medical devices. SBDs are written in technical language for stakeholders interested in product-specific Health Canada decisions, and are a direct reflection of observations detailed within the evaluation reports. As such, SBDs are intended to complement and not duplicate information provided within the Product Monograph.
Readers are encouraged to consult the 'Reader's Guide to the Summary Basis of Decision - Drugs' to assist with interpretation of terms and acronyms referred to herein. In addition, a brief overview of the drug submission review process is provided in the Fact Sheet entitled 'How Drugs are Reviewed in Canada'. This Fact Sheet describes the factors considered by Health Canada during the review and authorization process of a drug submission. Readers should also consult the 'Summary Basis of Decision Initiative - Frequently Asked Questions' document.
The SBD reflects the information available to Health Canada regulators at the time a decision has been rendered. Subsequent submissions reviewed for additional uses will not be captured under Phase I of the SBD implementation strategy. For up-to-date information on a particular product, readers should refer to the most recent Product Monograph for a product. Health Canada provides information related to post-market warnings or advisories as a result of adverse events (AE).
For further information on a particular product, readers may also access websites of other regulatory jurisdictions. The information received in support of a Canadian drug submission may not be identical to that received by other jurisdictions.
Other Policies and Guidance
Readers should consult the Health Canada website for other drug policies and guidance documents. In particular, readers may wish to refer to the 'Management of Drug Submissions Guidance'.
1 Product and submission information
Brand name:
Manufacturer/sponsor:
Medicinal ingredient:
International non-proprietary Name:
Strength:
Dosage form:
Route of administration:
Drug identification number(DIN):
- 02270994
Therapeutic Classification:
Non-medicinal ingredients:
Submission type and control no:
Date of Submission:
Date of authorization:
2 Notice of decision
On September 9, 2005, Health Canada issued a Notice of Compliance to Hoffmann-La Roche Limited for the drug product Avastin® (bevacizumab), an anti-angiogenic agent for use in combination with fluoropyrimidine-based chemotherapy as first-line treatment for metastatic colorectal cancer. This submission was granted Priority Review due to a promising innovative approach to cancer treatment and possible extension of patient survival time.
Bevacizumab, the medicinal ingredient in Avastin®, is a recombinant humanised monoclonal antibody (IgG1) which binds selectively to, and neutralizes the biologic activity of, human vascular endothelial growth factor (VEGF). Neutralizing the biologic activity of VEGF reduces the vascularization of tumours, inhibiting tumour growth.
The market authorization for Avastin® was based on satisfactory review of quality (chemistry and manufacturing), non-clinical, and clinical data. The safety and efficacy evaluation of Avastin® included a phase III, randomized, double-blind, controlled trial against an active control. Avastin® was administered in combination with irinotecan, 5-fluorouracil and leucovorin (IFL) as first line treatment for metastatic carcinoma of the colon or rectum. The addition of Avastin® to IFL resulted in a statistically significant increase in overall survival of 4.7 months. The data submitted demonstrate that Avastin® can be administered safely when used under the conditions stated in the Product Monograph.
Avastin® is supplied as either a 4 mL or 16 mL sterile solution contained in a single-use glass vial to deliver 100 mg or 400 mg of bevacizumab per vial, respectively. Dosing guidelines are available in the Product Monograph.
Serious side-effects associated with Avastin® include gastrointestinal perforation, haemorrhage, arterial thromboembolism and impaired wound healing. The most frequently observed adverse events across all clinical trials in patients receiving Avastin® with or without chemotherapy were asthenia, diarrhea, hypertension, nausea and pain NOS (Not Otherwise Specified). Detailed conditions for the use of Avastin® are available in the Product Monograph.
Based on the Health Canada review of data on quality, safety, and effectiveness, Health Canada considers that the benefit/risk profile of Avastin® is favourable when administered in combination with fluoropyrimidine-based chemotherapy, under the conditions stated in the Product Monograph, for first-line treatment of patients with metastatic carcinoma of the colon or rectum.
3 Scientific and Regulatory Basis for Decision
3.1 Quality Basis for Decision
3.1.1 Drug Substance (medicinal ingredient)
Description
Bevacizumab, the medicinal ingredient of Avastin®, is a recombinant humanized monoclonal antibody (IgG1) which binds selectively to and neutralizes the biologic activity of human vascular endothelial growth factor (VEGF). Neutralizing the biological activity of VEGF reduces the vascularization of tumours, inhibiting tumour growth.
Manufacturing Process and Process Controls
Bevacizumab is produced by recombinant DNA technology in Chinese hamster ovary cells. The manufacture of bevacizumab is based on a CHO master and working cell bank system, where the master and working cell banks have been thoroughly characterized and tested for adventitious contaminants and endogenous viruses in accordance with the ICH guidelines. Results of these tests confirmed cell line identity and absence of adventitious agents/viral contaminants. Genetic characterization (restriction endonuclease mapping and copy number analysis) also demonstrated stability of the master cell bank ranging from storage to production at the limit of in vitro cell age.
The manufacture of bevacizumab comprises a series of steps which include fermentation, harvest and purification. The purification is performed via a combination of chromatographic and viral inactivation/removal steps. The manufacturing process consistency is ensured through defined production procedures, critical quality tests, in-process limits and bevacizumab certificate of analysis specifications. Microbial control is maintained throughout the manufacturing process by testing for bioburden as well as for bacterial endotoxins. In-process controls performed during manufacture were reviewed and considered acceptable. The specifications for the raw materials used in manufacturing the drug substance are also considered satisfactory.
Characterisation
Detailed characterization studies were performed to provide assurance that bevacizumab is consistently exhibiting the desired characteristic structure. Results from process validation studies also indicate that the methods used during processing are adequately controlling the levels of product and process-related impurities. The impurities that were reported and characterized were found to be within established limits.
Control of Drug Substance
Validation reports were satisfactorily submitted for the analytical procedures used for in-process and release testing of bevacizumab. The drug substance specifications, and analytical methods used for quality control of bevacizumab are considered acceptable.
Stability
Based upon real-time and accelerated stability studies, the proposed shelf‑life and storage conditions for bevacizumab were supported and considered to be satisfactory.
3.1.2 Drug Product
Description and Composition
Avastin® (bevacizumab) is a clear to slightly opalescent, colorless to pale brown, preservative-free, sterile liquid solution. The non-medicinal ingredients of Avastin® are trehalose dihydrate, sodium phosphate, polysorbate 20 and water for injection. Avastin® is supplied as single-use glass vials with butyl rubber stopper, containing 25 mg/mL bevacizumab as either 100 mg in 4 mL or 400 mg in 16 mL.
Pharmaceutical Development
During the pharmaceutical development of Avastin®, changes were made to the formulation to increase stability, and to the concentration of bevacizumab per vial to reduce the volume of drug administered per patient. Two vial configurations were also developed: a 100 mg configuration, and a 400 mg configuration that allows dosage delivery with one vial.
The selected formulation retains the potency and protein integrity of bevacizumab during the proposed shelf-life. Compendial excipients used for formulation comply with either the US Pharmacopoeia (USP) or the European Pharmacopoeia (Ph. Eur.) or the British Pharmacopoeia (BP), except for trehalose dehydrate, which complies with an acceptable non-compendial set of specifications.
There were also changes made to the bevacizumab manufacturing process during development. Data pertaining to the physicochemical characteristics and biological activity demonstrated biocomparability between development and commercial batches.
Manufacturing Process and Process Controls
All manufacturing equipment, in-process manufacturing steps and detailed operating parameters are adequately described in the submitted documentation and are found acceptable. Bevacizumab is filled into vials using proper aseptic process techniques, and conventional pharmaceutical equipment and facilities.
Control of Drug Product
Avastin® is tested to verify its appearance, identity, purity, sterility, and potency, as well as the formulation-relevant parameters: protein content, pH, and osmolality. The validation reports submitted for the analytical procedures used for in-process and release testing of Avastin® are satisfactory and in compliance with ICH guidelines.
Analytical testing results, from final batch analyses, were reviewed and considered to be acceptable according to the specifications of the drug product.
Stability
Based on the results of real-time and accelerated stability studies, an expiration period of 24 months is justified for Avastin® 100 mg and 400 mg when stored at the recommended temperature of 2°C to 8°C.
3.1.3 Facilities and Equipment
An On-Site Evaluation (OSE) of facilities involved in the manufacture and testing of bevacizumab has been conducted by the Biologics and Genetic Therapies Directorate, Health Canada . The design, operations and controls of the facilities and equipment that are involved in the production are considered suitable for the activities and products manufactured. All facilities are compliant with Good Manufacturing Practices (GMP).
3.1.4 Adventitious Agents Safety Evaluation
Pre-harvest culture fluid from each lot is tested to ensure freedom of adventitious microorganisms (bioburden, mycoplasma, and viruses). Steps from the purification process designed to remove and inactivate viruses are adequately validated.
Raw materials of animal and recombinant origin used in the manufacturing process have been adequately tested to ensure freedom of adventitious agents. No bovine raw materials are used for the manufacture of bevacizumab. The excipients used in the drug product formulation are not from animal or human origin.
3.1.5 Summary and Conclusion
The New Drug Submission for Avastin® (bevacizumab) is considered to meet the requirements of Division C.08.002 of the Food and Drug Regulations insofar as the Quality (Chemistry and Manufacturing) information is concerned. The Chemistry and Manufacturing information submitted for Avastin® has demonstrated that the drug substance and drug product can be consistently manufactured to meet approved specifications. Proper development and validation studies were conducted, and adequate controls are in place for commercial production.
3.2 Non-Clinical Basis for Decision
3.2.1 Pharmacodynamics
In vitro and in vivo studies were conducted to investigate the pharmacodynamic properties of bevacizumab. The main focus of the pharmacodynamic studies conducted was mainly on the angiostatic effect of the anti-VEGF antibody.
In vitro study results demonstrated that the primary pharmacological effects of the antibody are not acting directly on tumour cells. Rather, the mechanism of action appears to be the growth-limiting effect on vascular endothelial cells.
In vivo tumour growth studies demonstrated that administration of bevacizumab to xenograft models of cancer in mice, in addition to experimental models of tumour metastasis, caused reduction of microvascular tumour growth and inhibition of metastatic disease progression. The effects of bevacizumab were also studied in animals with exposure to cytotoxic agents, radiation therapy, or other angiogenic compounds. Results of these studies show that a combination of anti-angiogenic compounds with anti-tumour therapeutics may result in increased efficacy of treatment.
The studies cited and/or carried out by the sponsor generally support the mechanism of action of the anti-VEGF antibody. In review of the roles of VEGF in other physiologic functions, such as endothelial cell trophicity and survival, or nitric oxide production, this could account for some important adverse effects of bevacizumab.
3.2.2 Pharmacokinetics
Absorption
Absorption of bevacizumab subsequent to a single intraperitoneal (IP) or subcutaneous (SC) administration has been examined in mice, rats, and cynomolgus monkeys. Absorption subsequent to IP administration was complete in mice. SC administration resulted in a slower absorption that was complete in mice and cynomolgus monkeys but was at only 69% in rats.
Distribution
Tissue distribution of 125I-labeleld bevacizumab was examined in male rabbit following intravenous bolus administration. Two and 48 hours after administration bevacizumab was localized (as indicated by precipitable radioactivity) tenfold more in plasma than in tissues. The organs that exhibited the highest levels of radioactivity were kidney, testis, spleen, heart, lung, and thymus. Radioactivity decreased rapidly in the 48 hours after administration but only minimal degradation of the antibody was noted. Bevacizumab was also shown in toxicity studies to distribute into foetal serum and into the amniotic fluid in rabbits.
Metabolism
Following intravenous administration of bevacizumab to mice, rats, rabbits, and cynomolgus monkeys, the antibody concentrations decreased with an initial half-life of approximately 1 day and a terminal half-life of 1-2 weeks. Bevacizumab clearance was slower at higher doses. Nonlinear pharmacokinetic parameters were observed in mice, rats, and rabbits following administration of doses of less than 1 mg/kg. In cynomolgus monkeys, bevacizumab administration was linear over the range of 2-50 mg/kg. The mean clearance was approximately 6 mL/kg/day, the initial half-life was below 1 day and the terminal half-life was 10 days.
Excretion
No specific studies were performed to evaluate excretion. However, a study conducted in rabbits showed minimal excretion; less than 10% of the radioactivity in urine at 2 and 48 hours post dose was trichloracetic acid (TCA)-precipitable.
Study results submitted demonstrate that in all species studied, bevacizumab catabolism, clearance and half-lives were similar to that of other IgGs. Furthermore, the drug interaction studies performed in cynomolgus monkeys revealed no significant potential of bevacizumab for drug interactions.
3.2.3 Toxicology
The animal toxicology assessment was based on repeat-dose toxicity, reproductive and developmental toxicity, local tolerance toxicity, hemolytic potential and blood compatibility, immunogenicity, epiphyseal dysplasia development secondary to inhibition of blood vessel formation, wound healing impairment, and toxicological effects on renal and kidney function.
Single Dose Toxicity
No single dose toxicity studies were conducted for bevacizumab.
Repeat Dose Toxicity
Three repeat-dose toxicity studies of up to 26 weeks in duration were performed in cynomolgus monkeys.
4-Week Recovery (Study 96-181-1751)
The purpose of this study was to determine the toxicity of bevacizumab after 4 weeks of intravenous administration, twice weekly, in young adult male and female cynomolgus monkeys. In males treated with 10 or 50 mg/kg of bevacizumab, a microscopic finding of epiphyseal dysplasia of the distal femur was noted, primarily characterized by thickened growth plate cartilage, clusters of hyperplastic chondrocytes, subchondral bony plate formation and inhibition of vascular invasion of the growth plate. Additionally, degeneration of the cartilage matrix was noted. Severity of epiphyseal dysplasia ranged from slight to moderate in males treated with 10 mg/kg and from moderate to severe in males treated with 50 mg/kg. Epiphyseal dysplasia was not observed in females at any dose level. Epiphyseal dysplasia was present in both male recovery animals after the 4-week recovery period. Additionally, minimal diffuse degeneration and necrosis of the metaepiphyseal bone marrow was present at the recovery necropsy in the high-dose male recovery animal that had severe epiphyseal dysplasia.
13-Week IV Toxicity Study of Bevacizumab in Cynomolgus Monkeys with a 4-Week Recovery (Study 96-182-1751)
The purpose of this study was to determine the toxicity of bevacizumab after 13 weeks of intravenous administration, twice weekly, in cynomolgus monkeys. Ovarian and uterine weights were significantly reduced in females given 10 or 50 mg/kg of bevacizumab twice weekly for 13 weeks. These changes coincided with a reduction in number or absence of corpora lutea in females treated with 10 or 50 mg/kg bevacizumab. These data suggest that the effect of bevacizumab on female reproductive function is at least partially reversible upon cessation of treatment.
Bevacizumab administration caused a dose-dependent increase in the incidence and severity of epiphyseal dysplasia (study 96-181-1751). An additional finding of linear fissuring of the cartilaginous growth plate was occasionally observed. Epiphyseal dysplasia was noted in all males treated with bevacizumab. The severity was dose dependent, ranging from minimal to moderately severe in males treated with 2 mg/kg to slight to moderately severe in males treated with 10 or 50 mg/kg. Minimal to slight epiphyseal dysplasia was present in all females given 10 or 50 mg/kg of bevacizumab. Dysplastic changes were not observed in females treated with 2 mg/kg of bevacizumab.
26-Week IV Toxicity Study of Bevacizumab in Cynomolgus Monkeys with a 12-Week Recovery (Study 97-194-1751)
The purpose of this study was to determine the toxicity of bevacizumab after 26 weeks of intravenous administration, once or twice weekly, in cynomolgus monkeys. Doses of bevacizumab at 10 mg/kg or higher decreased the incidence of menstrual cycles in females; these changes correlated with decreases in endometrial proliferation in these treatment groups. Doses of bevacizumab at 10 mg/kg or higher, regardless of treatment frequency, produced lower uterine weights and reduced endometrial proliferation. In addition, treatment with 10 mg/kg (twice weekly) or 50 mg/kg inhibited follicular maturation at the early Graafian follicle stage in ovaries of female monkeys. These effects were not observed in animals treated with 50 mg/kg following a 12-week recovery, indicating their reversibility. Corpora lutea were absent in females treated with 10 mg/kg (twice weekly) or 50 mg/kg of bevacizumab. Following the 12-week recovery period, only 1 of 2 females in the 50 mg/kg dose group demonstrated this effect, suggesting at least partial reversibility following cessation of bevacizumab treatment.
Mean body weights were decreased by 5% to 10% in males of the 10 mg/kg (twice weekly) group beginning at Week 24 and by 8% to 13% in males of the 50 mg/kg group beginning at Week 14. Epiphyseal dysplasia was also found in males of the 10 mg/kg (twice weekly) and 50 mg/kg treatment groups with an increased severity compared to males treated with 2 or 10 mg/kg once weekly and females in all dose groups.
Bevacizumab administration resulted in a dose-dependent increase in the incidence and severity of epiphyseal dysplasia (study 96-181-1751). Additionally, "cleft formation" synonymous with the linear fissuring (study 96-182-1751) was noted as an occasional, severe manifestation of epiphyseal dysplasia in 1 male in the 10 mg/kg (twice weekly) dose group and 1 male in the 50 mg/kg dose group. Severity ranged from minimal to slight in males treated with 2 mg/kg, from minimal to moderate in males treated with 10 mg/kg (once weekly) and from slight to moderate in males treated with 50 mg/kg. All males treated with 10 mg/kg twice weekly had moderate dysplasia. Epiphyseal dysplasia was not present in animals of the recovery group following 12 weeks without treatment. The decreased incidence of epiphyseal dysplasia in this study, compared with previous studies, may reflect the fact that an effort was made to select animals with closed growth plates by performing radiography prestudy.
Toxicity with Combination Treatment
The safety of bevacizumab in conjunction with two chemotherapeutic regimens irinotecan/5-fluorouracil/leucovorin (IFL) or cisplatin (Platinol-AQ)/paclitaxel (Taxol) has been investigated in the following two studies in cynomolgus monkeys.
Combination Treatment with Bevacizumab and Irinotecan/5-Fluorouracil/Leucovorin (Study 00-376-1756)
The safety of Irinotecan/5-Fluorouracil/Leucovorin (IFL) chemotherapy regimen in combination with bevacizumab was assessed in cynomolgus monkeys. Seven male monkeys were treated intravenously with IFL and 5 male monkeys were treated intravenously with IFL in combination with 10 mg/kg of bevacizumab. All animals were treated once weekly for 2 weeks. Administration of IFL resulted in diarrhoea, body weight loss, decreased white blood cell count and decreased food consumption. A small thymus with corresponding lymphoid depletion was observed in 2 of 7 animals treated with IFL alone and in 2 of 5 animals treated with IFL plus bevacizumab. Treatment related microscopic effects were noted in the sternal bone marrow of most animals in all groups, and consisted mainly of erythroid hyperplasia or myeloid hypoplasia, or both. The co-administration of bevacizumab with IFL did not alter the magnitude of the effects related to treatment with the antineoplastic therapy regimen.
The safety of the Cisplatin and Taxol regimen, or Cisplatin plus paclitaxel regimen, in conjunction with bevacizumab (Study 96-375-1751)
The safety of the cisplatin and Taxol regimen, or cisplatin plus paclitaxel regimen, in conjunction with bevacizumab was assessed in cynomolgus monkeys. Five male monkeys were assigned to each of four treatment groups (20 animals total). Bevacizumab (10 mg/kg) or bevacizumab vehicle was administered twice weekly for 3 weeks. Administration of cisplatin and paclitaxel was associated with vomiting and decreased body weight compared to control animals treated with bevacizumab vehicle and saline. In addition, cisplatin and paclitaxel treatment induced transient decreases in white blood cell and absolute neutrophil counts. Co-administration of bevacizumab had no apparent effect on any of the alterations induced by treatment with cisplatin and paclitaxel.
Genotoxicity
No genotoxicity studies were conducted for bevacizumab.
Carcinogenicity
No carcinogenicity studies were conducted for bevacizumab.
Reproductive and development
As bevacizumab can pass the placental barrier, the following two reproductive/embryofetal toxicity studies were conducted in rabbits.
Embryofetal Development Pilot Developmental Toxicity Study (Study 01-223-1751)
Presumed pregnant rabbits were treated with 10, 30, or 100 mg/kg of bevacizumab administered intravenously during the period of organogenesis. The dosing regimen was selected to maintain an average serum concentration approximately equivalent to the human clinical exposure. In addition, the divided regimen was intended to minimize exposure to anti-bevacizumab antibodies and to antigen−antibody complexes during the period of organogenesis. Neither fetal skeletal nor soft tissue observations were performed. Statistical analysis was not performed because of the small number of animals in each treatment group. Bevacizumab administration induced a decrease in maternal body weight and body weight gain throughout the gestation period and a decrease in food consumption during the post-dose period in the 100 mg/kg dose group. Average fetal body weights were reduced in the 100 mg/kg dose group. Analysis of blood collected at caesarean sectioning indicated that 33%-47% of animals treated with bevacizumab developed antibodies to bevacizumab. Bevacizumab levels were quantifiable despite the presence of antibodies to bevacizumab.
Segment II Developmental Toxicity Study (Study 02-029-1751)
To maximize exposure to bevacizumab during the period of fetal organ development, a standard Study 02-029-1751 was performed in which presumed-pregnant rabbits were treated with intravenous administration of bevacizumab vehicle or bevacizumab at 10, 30, or 100 mg/kg. A dose‑related decrease in maternal body weight gain was observed in the 30 and 100 mg/kg dose groups. In the 100 mg/kg dose group, an increase in the total number of fetal resorptions and the percent dead or resorbed conceptuses per litter was primarily attributed to an increase in the number of late resorptions. Fetal body weights were dose-dependently decreased in the 10, 30, and 100 mg/kg dose. The number of litters with any fetal alteration increased in the 30 and 100 mg/kg groups. In the 100 mg/kg group, the total number of fetuses with any alterations increased, as well as, the percentage of fetuses per litter with any alterations. The number of ossification sites for metacarpals per fetus per litter was reduced in the 10 mg/kg and higher dose groups. The mean number of ossification sites for caudal vertebrae and fore and hind-limb phalanges was also reduced in the 100 mg/kg dose group.
Antibodies to bevacizumab were detected after exposure in maternal serum in 1 animal in the vehicle group, 1 animal in the 10 mg/kg dose group, 4 animals in the 30 mg/kg dose group and 2 animals in the 100 mg/kg dose group. In all, antibody titers to bevacizumab were detected in the maternal serum of 12% (9/73) of the pregnant rabbits in the study. Antibodies were detected in the fetal serum of 13% (9/71) of the animals exposed to bevacizumab and in 5% (1/19) of the animals exposed to bevacizumab vehicle. Antibodies to bevacizumab were detected in the amniotic fluid of 10% (7/73) of the animals exposed to bevacizumab and in 5% (1/19) of the animals exposed to bevacizumab vehicle. The actual occurrence of anti-bevacizumab antibodies may be higher than reported because of the interference of high concentrations of circulating bevacizumab with the anti-drug antibody assay. Based on the transient decrease in body weight noted upon initiation of treatment in the 30 and 100 mg/kg dose groups, the Maternal No Observed Adverse Effect Level (NOAEL) for bevacizumab was 10 mg/kg every third day. However, significant effects of treatment were observed in the fetuses at a dose as low as 10 mg/kg every third day, indicating that bevacizumab can produce developmental toxicity at doses below those that cause minimal maternal toxicity. Doses of 10 mg/kg and higher reduced fetal weights and ossification sites, while doses of 30 mg/kg and higher resulted in multiple malformations in fetuses; thus the developmental NOAEL for bevacizumab is < 10 mg/kg. Pharmacokinetic extrapolation indicated that a 10 mg/kg dose given every third day approximates the average serum concentration at the proposed human clinical dose.
Local Tolerance
No formal studies of this type were performed. However, twice a week and once a week intravenous bolus injections of bevacizumab for up to 26-weeks treatment duration "were well tolerated locally" in cynomolgus monkeys (studies 96-181-1751, 96-182-1751, 97-194-1751).
Other Toxicity Studies
Hemolytic Potential and Blood Compatibility Testing with Bevacizumab (96-221-1751)
The bevacizumab concentration of 5 mg/mL was compatible with human and cynomolgus monkey serum and plasma.
Cross Reactivity of Biotinylated Bevacizumab with Rabbit, Cynomolgus Monkey, and Human Tissues (96-327-1751, 96-326-1751)
Tissue specificity of bevacizumab against a panel of normal rabbit, cynomolgus monkey and human tissues was determined via an immunohistochemical technique using 10 and 400 μg/mL of bevacizumab conjugated with biotin. In the panel of 9 rabbit tissues, 30 cynomolgus tissues and 36 human tissues tested, only non-specific or no staining was observed.
Timing of Antibody Development in the Rabbit (96-407 B-1751)
Rabbits produce antibodies to humanized monoclonal antibodies. The time course of antibody development was thus studied in the rabbit so that further studies conducted could be optimized to avoid anti-bevacizumab antibodies. The data indicate that rabbits are capable of forming antibodies to bevacizumab and that these antibodies develop between 8 and 11 days after the initiation of dosing.
Investigation of Physeal Dysplasia in Rabbits (98-075-1751)
Epiphyseal dysplasia secondary to inhibition of blood vessel formation in long bone growth plates was observed in studies of bevacizumab in cynomolgus monkeys following 4 to 26 weeks of treatment. An investigative study in rabbits was conducted to assess the suitability of the rabbit for further study of epiphyseal dysplasia. Animals that received bevacizumab showed a slight thickening of the growth plate cartilage. In contrast to the effect noted in monkeys, bevacizumab did not inhibit vascular invasion or induce subchondral bony plate formation in rabbits at doses up to 75 mg/kg. The duration of dosing in rabbits is limited by development of antibodies to bevacizumab; this short exposure period may possibly be insufficient for epiphyseal dysplasia to develop.
Effects of Bevacizumab on Female Reproductive Function in Rabbits (97-166A-1751, 97-370-1751A)
Decreases in ovarian weight and the number of corpora lutea were observed in the multiple-dose studies of bevacizumab in cynomolgus monkeys indicating an effect of bevacizumab on female reproductive function. An investigative study in rabbits was performed to further investigate the effect of bevacizumab on luteal function. Rabbits were given four intravenous injections of bevacizumab at 50 mg/kg or bevacizumab vehicle over a 10 day period. To stimulate luteal progesterone release, a single injection of human chorionic gonadotropin (hCG) was given to all animals on the same day as the second bevacizumab injection. The results were consistent with the findings in the cynomolgus monkey studies and indicate that bevacizumab exposure reversibly inhibits the function of the corpus luteum at a dose of 50 mg/kg given every 2 days in the rabbit. In a second study, the dose-response of the effect of bevacizumab on ovarian function was investigated in rabbits. The results of this study are consistent with the inhibitory effect of bevacizumab on female reproductive function observed in the cynomolgus monkey.
Effect of Bevacizumab on Wound Healing in a Linear Incision Model in Rabbits (97-373-A-1751, 97-373-B-1751, 99-009-1751)
Because of the concern that bevacizumab may delay wound healing in patients undergoing biopsy procedures or surgery, the effect of bevacizumab on wound healing was investigated using a linear-incision model in rabbits. Full-thickness linear incisions (2.5 cm), which mimic a surgical incision, were made in the scapular region, closed with sterile sutures and covered with polyurethane membrane for protection. Bevacizumab administration resulted in a dose-dependent and significant decrease in the tensile strength of the wounds in these studies indicating that bevacizumab interferes with wound healing in the rabbit.
Effect of Bevacizumab on Wound Healing in a Circular Wound Model in Rabbits (96-407-1751, 96-407-A-1751)
A second model of wound healing, the circular wound model, was used in rabbits to confirm the results of the studies using the linear incision model. A circular wound mimics an ulcerative lesion and is thus distinct from a linear incision. The results indicate that bevacizumab administered at 50 mg/kg impairs wound healing in the circular wound model in the rabbit. The dose-response of this effect was explored in a subsequent study. The findings indicate a dose-related delay in wound healing followed by complete healing after cessation of dosing. The results of these studies are consistent with the findings of the linear incision model studies indicating that bevacizumab likely delays wound healing. These effects were noted at doses as low as 2 mg/kg given twice weekly for 2 weeks in the rabbits.
Effect of Bevacizumab on Wound Healing in Cynomolgus Monkeys (99-158-1754, 99-159-1754)
Experiments using the linear incision wound model were performed in cynomolgus monkeys. The effects of bevacizumab on wound healing in the monkey were extremely variable. It is surmised that the effect of bevacizumab on wound healing in humans may likewise vary.
Investigation of Thrombosis in a Rabbit Model (01-125-1751)
Occurrences of thrombosis were observed in Phase II clinical trials of bevacizumab. In contrast, no alterations in hematology or coagulation parameters were observed in the repeat-dose studies in cynomolgus monkeys. To investigate the potential for bevacizumab treatment to increase the incidence of thrombosis an acute rabbit model of thrombosis was used. The results of this study indicate that administration of bevacizumab to a normal healthy rabbit does not affect hemostasis or exacerbate thrombosis when a thrombus has been induced by mechanical manipulation.
Selective Renal Effects and Deposition of Bevacizumab in the Kidney (99-537-1751)
A number of patients in Phase II clinical trials were reported to have evidence of proteinuria. In contrast, no alterations in clinical chemistry or urinalysis parameters reflective of impaired renal function were observed in the repeat-dose studies in cynomolgus monkeys. A study was conducted to examine the deposition of bevacizumab in the kidney in rabbits. Examination of the kidneys by light and electron microscopy indicated no histological differences between control and treated animals and no selective deposition of bevacizumab suggesting that bevacizumab does not accumulate in the kidney in rabbits treated acutely with two doses up to 100 mg/kg.
Cisplatin Model of Renal Dysfunction (01-220-1751)
Cisplatin is a chemotherapeutic drug known to be nephrotoxicant; it targets the proximal tubule. This study was conducted in rabbits to evaluate whether a previous insult to the kidney is required for an effect of bevacizumab to be observed. 'As expected with the model', treatment with cisplatin and cisplatin plus bevacizumab resulted in a significant increase in serum blood urea nitrogen (BUN) and creatinine at the end of the treatment period compared with the levels in the control group and the group treated with bevacizumab alone. Microscopic lesions in the kidney were observed in animals treated with cisplatin or the combination of cisplatin and bevacizumab. The results suggest that the administration of 50 mg/kg of bevacizumab does not exacerbate renal injury induced by cisplatin in this rabbit model.
Bovine Serum Albumin Overload Model of Renal Dysfunction (00-038-8095A, 01-193-1751)
A study of protein, bovine serum albumin (BSA), overload in the rabbit was used to complement the cisplatin model to further examine the effect of bevacizumab on renal function in the presence of different types of pre-existing renal damage in rabbits. BSA administration was found to induce mild and variable renal dysfunction as evidenced by changes in serum blood urea nitrogen (BUN) and creatinine. However, it was difficult to make definitive conclusions about these findings because of the small group sizes in this study and an additional study with the BSA model was undertaken. Treatment with BSA, with or without bevacizumab, resulted in an increase in total urinary protein. No significant differences in serum BUN or creatinine arose among treatment groups following bevacizumab administration during Week 8 indicating no initiation or exacerbation of renal injury by bevacizumab treatment. Treatment with BSA induced a relatively mild injury characterized as mild glomerulonephritis. The combined administration of BSA and bevacizumab did not appear to intensify the injury induced by BSA alone. Overall, these data suggest that BSA‑induced renal injury was mild and that 50 mg/kg of bevacizumab did not exacerbate the injury produced by BSA.
3.2.4 Summary and Conclusion
Results obtained from the non-clinical studies support the proposed mechanism of action; bevacizumab administration did suppress tumour blood vessel growth and transformed fast-growing tumours into small slow-growing tumours.
Effects observed in the repeat-dose toxicity studies in cynomolgus monkeys at doses up to 50 mg/kg showed inhibition of blood vessel formation in long bone growth plates resulting in epiphyseal dysplasia. The incidence and severity of epiphyseal dysplasia increased in relation to the dose and duration of treatment. In all studies, evidence of some reversibility was noted upon cessation of treatment. Based on weekly dose, the doses associated with this effect were below (≥0.8-fold) the proposed clinical dose and exposure levels were 1.5 to 2.0-fold below the expected human exposure. It is noted that epiphyseal dysplasia occurred only in actively growing animals with open growth plates. Consequently, it is suggested that epiphyseal dysplasia will potentially not occur in the clinical population because bevacizumab will most likely be administered to adult patients with closed growth plates.
Bevacizumab administration at doses of 10 mg/kg twice weekly or 50 mg/kg once weekly for 26 weeks was associated with a decrease in albumin and albumin/globulin ratio and increase in globulin in male cynomolgus monkeys. These effects were reversible following cessation of exposure.
Decreased ovarian function was observed in cynomolgus monkeys treated with 10 or 50 mg/kg of bevacizumab once or twice weekly for 13 or 26 weeks and was partially reversible upon cessation of treatment. These effects were observed at ≥2-fold over the projected serum concentrations at the proposed clinical dose. An inhibition of ovarian function was also seen in rabbits following administration of 50 mg/kg. Administration of bevacizumab is therefore likely to result in an adverse effect on female fertility.
Administration of bevacizumab to rabbits during organ development resulted in a decrease in maternal and fetal weights, an increase in the number of fetal resorptions and an increase in the incidence of fetal malformations. These incidences, including fetal abnormalities, were observed at doses below those which induced maternal toxicity. The fetal NOAEL was <10 mg/kg. Therefore, administration of bevacizumab during pregnancy could present a risk to the fetus.
Treatment-related delay in wound healing following administration of bevacizumab was observed in the rabbit but not in the cynomolgus monkey. Complete healing was delayed 3 to 5 days. The effect in the rabbit was seen at a dose as low as 0.5 mg/kg which is below the proposed human clinical dose. As effects on wound healing were observed in rabbits at doses below the proposed clinical dose, administration of bevacizumab could adversely impact wound healing in treated patients.
3.3 Clinical basis for decision
3.3.1 Pharmacodynamics
No clinical studies were conducted in humans to evaluate the pharmacodynamics and mechanism of action of bevacizumab. However, the proposed action mechanism has been supported by results obtained in non-clinical studies whereby bevacizumab administration did suppress tumour blood vessel growth and transformed fast-growing tumours into small slow-growing tumours. However, bevacizumab therapy is insufficient for complete tumour eradication. Tumour growth resumes upon cessation of bevacizumab treatment. Despite growing at a slower rate, tumours do appear to continue to grow when suppressing VEGF. As a result, curative treatment for cancer requires administration of cytotoxic agent(s) in addition to bevacizumab therapy.
3.3.2 Pharmacokinetics
Pharmacokinetic parameters were evaluated in two phase I clinical trials. When evaluating the pharmacokinetic parameters of bevacizumab, the two-compartment model was favoured over the one-compartment model given the two-compartment model provides better predictions and consistency with other compounds in the same category.
The results from the phase I study in solid tumours showed that clearance of bevacizumab above 1 mg/kg/day was slow and appeared somewhat proportionally related to dose. Doses lower than 1 mg/kg/day showed higher levels of clearance. The clearance of bevacizumab following administration of four intravenous doses provided at 5 different dosing regimens (0.1, 0.3, 1.0, 3.0, and 10 mg/kg) ranged from 2.75 to 9.29 mL/day/kg. The central volume of distribution of bevacizumab ranged from 37.9 to 48.6 mL/kg. The mean terminal half‑life of bevacizumab ranged approximately from 5 to 15 days.
By pooling the results of eight studies providing pharmacokinetic results and using a population approach, the typical value for clearance was estimated to be 0.207 and 0.262 L/day for female and male subjects, respectively. The corresponding central volume of distribution was estimated to be 2.66 and 3.25 L for female and male subjects, respectively. The initial half-life (t1/2α) was 1.4 days for both sexes, and the terminal (β) half-life estimate was 20 days for the typical female subject and 19 days for the typical male; which is consistent with the values obtained for IgGs.
The ability of bevacizumab to bind to human plasma proteins was not studied. Nor have studies been conducted directly in special populations. In order to address special population concerns, a population pharmacokinetic model was built using eight pharmacokinetic studies. Results from the population pharmacokinetic analysis showed both a clearance and central volume of distribution difference for gender and body weight. However, given the nature of the analysis, it is unclear whether these observed differences are significant.
The means by which bevacizumab is metabolised has not been directly studied. It is assumed that bevacizumab is eliminated through use of similar mechanisms as other IgG antibodies. However, when using a one or two-compartment model to 'fit' the pharmacokinetic data, it is apparent that the fits improve at the higher dose concentrations. This might imply that two simultaneous mechanisms for elimination may be present, such as a saturable enzymatic mechanism for the lower doses and a first order reaction as the principal route for the higher doses.
Studies examining the potential for drug-drug interactions were not performed. However, given current data collected, it is known that bevacizumab does not directly interact with 5-fluorouracil, carboplatin, doxorubicin and paclitaxel.
3.3.3 Clinical Efficacy
A total of 11 clinical studies (1 pivotal and 10 non-pivotal) were submitted for review. All clinical studies submitted for review were presented with appropriate final reports, tables, figures and appendices to allow for a full evaluation. Additional clinical trials of bevacizumab in other tumour types have also been completed and are under analysis by the sponsor. The main clinical studies, which contain the most relevant efficacy and safety data, are AVF2107g, AVF0780g and AVF2192g.
Study AVF2107g
Pivotal study AVF2107g was a large multicentre, randomised, double-blind, active controlled, three-arm, phase III trial to evaluate the efficacy and safety of bevacizumab in combination with IFL chemotherapy or 5-FU/Leucovorin chemotherapy as first-line treatment for previously untreated metastatic colorectal cancer. The primary objective was the efficacy (duration of survival) and safety of the treatment with multiple doses of bevacizumab in combination with IFL chemotherapy (or 5-FU/Leucovorin chemotherapy) as first line treatment for metastatic colorectal cancer. Secondary objectives were to evaluate the efficacy by assessing response rate, time to progression, duration of response and quality of life. The efficacy and safety in the second line therapy were also evaluated.
A total of 923 patients were enrolled within the study. Patients were randomized to one of the following three regimens. First arm: placebo plus bolus-IFL ('Saltz' regimen: 125 mg/m 2 irinotecan, 500 mg/m2 5-FU IV, 20 mg/m2 IV), administered weekly for four weeks followed by two weeks of rest, repeated in multiple cycles. Second arm: bevacizumab 5 mg/kg IV (infusion) once every two weeks plus bolus-IFL as in first arm. Third arm: bevacizumab as in second arm plus bolus 5-FU/LV ('Roswell Park' regimen: 5-FU 500 mg/m2 + LV 20 mg/m2) weekly for six weeks of every eight week cycle. The third arm was initially added as the safety of the second arm treatment (bevacizumab+bolus-IFL) was sufficiently unknown. A preliminary safety analysis carried out by an independent Data Monitoring Committee after the first 100 patients had been enrolled in the second arm determined the safety profile of the treatment in the second arm was acceptable; therefore, enrolment in the third arm was discontinued while it continued in the first two arms until 400 patients per arm had been included. Patients were treated until progression or for a maximum of 96 weeks. Patients in the bevacizumab treatment arms who had a confirmed complete response or unacceptable chemotherapy-related toxicity were allowed to continue on bevacizumab alone. Patients in the first arm (control arm) were not allowed to cross over to the bevacizumab arm.
Results
Of the 923 patients enrolled, 897 (97.2%) patients received at least one dose of the assigned study treatment. Results showed that the response rate was significantly higher in the bevacizumab arm (p = 0.0036) with the majority of responses being partial responses. One of the most statistically significant outcomes of the treatment was the addition of bevacizumab to IFL chemotherapy which demonstrated a significant prolongation of survival with an increase in median duration of survival from 15.6 months in the IFL+placebo arm to 20.3 months in the IFL+bevacizumab arm (p<0.0001). The median duration of progression-free survival was 6.2 months in the IFL+placebo arm and 10.6 months in the IFL+bevacizumab arm (p<0.001).
The average follow-up time for survival was 21 months. A similar proportion of patients in arms 1 and 2 received second-line treatment and the chemotherapy agents used were also similar. Progression-free survival was also significantly increased during the first 20 months of treatment with bevacizumab. Progression-free survival after 20 months was similar in both arms. For the 110 patients randomized to arm 3, the median overall survival was approximately 18 months and median progression-free survival was approximately 9 months.
Discussion
The design of this single pivotal trial along with the randomization and stratification methods used is appropriate and well presented. The results of this trial also showed that bevacizumab, in combination with IFL, gives a prolongation of survival of 4.7 months when compared to IFL alone (20.3 vs. 15.6 months). The percentages of patients who also received second-line therapy during the trial were well-balanced between the treatment arms and unlikely to have affected survival in a manner that may account for the difference in overall survival rate between the control and treatment arms. Addition of bevacizumab to chemotherapy also resulted in a significant improvement in progression-free survival, objective response rate and duration of objective response. The quality of life of the patients in the bevacizumab treatment arm was also not significantly affected when compared to patients receiving chemotherapy alone. These data show benefits for patients with metastatic colorectal cancer treated with bevacizumab + IFL chemotherapy as a first-line therapy. Therefore, it is reasonable to conclude that addition of bevacizumab to IFL chemotherapy results in an increase in patient survival, progression-free survival, and response rate when compared to IFL chemotherapy alone.
Study AVF0780g
AVF0780g was a phase II, active controlled, open-labelled, randomized, clinical trial to evaluate the efficacy, safety and pharmacokinetics of bevacizumab in combination with 5-FU/LV as compared to 5-FU/LV alone in patients with metastatic colorectal cancer.
The primary objectives were time to progression, response and safety. Secondary objectives included survival, duration or response, quality of life and pharmacokinetic analyses.
A total of 104 patients were randomized within the study. All patients received the Roswell Park regimen (FA 500 mg/m2 and bolus 5-FU 500 mg/m2) given weekly for 6 weeks for every 8 week cycle. In addition, each patient was randomized to receive either placebo, 5 mg/kg or 10 mg/kg of bevacizumab administered every two weeks for a maximum of 24 doses (48 weeks). Unless showing early progression, chemotherapy was administered up to six cycles. At the end of the treatment period complete responders, partial responders and patients with stable disease receiving bevacizumab in the two treatment arms were permitted additional bevacizumab treatment in the extension study AVF0778g, if progression occurred within six months. Patients who were randomized to the placebo arm were allowed to cross over to the 10 mg/kg dose of bevacizumab in the extension study after progression. Administration of treatment continued for the remaining 48 weeks treatment period or following a minimum of four bevacizumab doses given evidence of disease progression. The ECOG Tumour Response Criteria were used to evaluate responses and determine disease progression.
Results
Among the 104 patients enrolled, 102 patients received at least one dose of the assigned study drug. Review of study results demonstrated that time-to-progression was notably consistent with exposure to study treatment. Patients receiving bevacizumab 5 mg/kg showed a significantly longer progression‑free survival than placebo and 10 mg/kg treatment arms. When comparing the placebo arm against both the 5 mg/kg and 10 mg/kg treatment arms, the overall survival estimates indicated a similar and favourable trend for the two treatment arms, even though patients in the placebo arm were allowed to cross over to the 10 mg/kg bevacizumab treatment arm. No differences were noted in the quality of life scores among the treatment groups.
Discussion
Although the number of patients in this trial was insufficient to assess all end points, data from the study did show that bevacizumab administered at 5 mg/kg every two weeks in combination with fluoropyrimidine chemotherapy resulted in a remarkable increase in time-to-progression when compared to chemotherapy alone. The response rate was also higher in the treatment arm. Therefore, these data support the results obtained in pivotal trial AVF2107g previously presented.
Study AVF2192g
Study AVF2192g was a phase II, double-blind, randomized, active controlled trial to evaluate the efficacy and safety of bevacizumab in combination with 5-FU/Leucovorin chemotherapy in patients with metastatic colorectal cancer who were not optimal candidates for first line Irinotecan (CPT-11). The primary efficacy endpoint for this study was duration of overall survival. Secondary endpoints were progression-free survival, response rate, duration of response, quality of life and safety.
A total of 214 patients were currently randomized within the study at the time of review. All patients received the Roswell Park regimen (FA 500 mg/m2 and bolus 5-FU 500 mg/m2) weekly for 6 weeks of every 8-week cycle. In addition, each patient was randomized to receive either placebo (arm 1) or bevacizumab 5 mg/kg (arm 2) every 2 weeks. Patients were treated until progression or for a maximum of 96 weeks.
Results
Among the 214 patients enrolled, 204 patients received at least one dose of the assigned study drug. The hazard ratio of death among subjects in arm 2 (bevacizumab 5 mg/kg + 5-FU/FA) compared with subjects in arm 1 (placebo+5-FU/FA) was not statistically significant. Median survival was 12.9 months in arm 1 and 16.6 months in arm 2. Median progression-free survival was 5.52 months in arm 1 and 9.17 months in arm 2 (p=0.0002). The objective response rate was 15.2% in arm 1 and 26% in arm 2 (p=0.552; all were partial responses). The treatment effect was greater in patients with a baseline albumin level of < 3.5 g/dL.
Discussion
Notably, the chemotherapy regimen used in this study was different than the chemotherapy used in arm 2 of pivotal trial AVF2107g. The subjects enrolled in this study were generally older and sicker than subjects enrolled in the two other first-line colorectal cancer trials. The study did not demonstrate a significant improvement in the primary endpoint of overall survival. However, it did demonstrate a significant improvement in the secondary endpoints of progression-free survival, objective response rate, and duration of objective response.
Conclusion
Overall, the data obtained in the pivotal trial indicates that bevacizumab (Avastin®), in combination with IFL chemotherapy is an effective first-line therapy in metastatic colorectal cancer [statistically significant prolongation of survival of 4.7 months compared to IFL alone (survival: 20.3 vs. 15.6 months, and progression-free survival: 10.6 vs. 6.2 months). The results presented are in support of a beneficial effect of the bevacizumab + Irinotecan + 5-FU/Leucovorin combination in patients with metastatic colorectal cancer.
The data obtained in study AVF2192g, suggest that bevacizumab might be also effective in metastatic colorectal patients in combination with 5-FU/Leucovorin alone (without Irinotecan). Whether or not bevacizumab can indeed be combined with other "fluoropyrimidine-based" chemotherapies, such as 5‑FU/Leucovorin alone, or in other formulations, or using other routes of administration to achieve an efficacy comparable to that of IFL as administered in the pivotal trial remains to be determined. Further data of Avastin ® in combination with intravenous 5-FU/leucovorin suggest superior objective response rates, progression free survival, and longer survival compared with 5-FU/leucovorin chemotherapy alone.
Bevacizumab can only decrease the rate of tumour growth and thus, it is insufficient for complete tumour eradication. This is a combination therapy where bevacizumab is added to chemotherapy; therefore the choice and regimen of chemotherapy remain critical to the prognosis of the disease and the patient's quality of life.
Certain facts also remain outstanding in connection with the efficacy of this drug. The median prolongation of survival and progression-free survival under the proposed treatment is only approximately 4.5 months (survival 20.3 vs. 15.6 months; progression-free survival 10.6 vs. 6.2 months). Bevacizumab can only decrease the rate of tumour growth and thus, it is insufficient for complete tumour eradication. Also, this is a combination therapy where bevacizumab is added to chemotherapy; therefore the choice and regimen of chemotherapy remain critical to the prognosis of the disease and the patient's quality of life. Last but not least, in the pivotal study the main reason for treatment discontinuation was disease progression.
Avastin® inhibits the binding of VEGF to its receptors on the surface of endothelial cells. Neutralising the biologic activity of VEGF reduces the vascularisation of tumours, thereby inhibiting tumour growth. The efficacy data are available essentially for bevacizumab in combination with fluoropyrimidine-based chemotherapy. Consideration should be given to current standard of care guidelines for colorectal cancer.
3.3.4 Clinical Safety
In summary, over 1,100 patients with metastatic carcinoma of colon or rectum were enrolled in the clinical studies of which greater than 550 patients received bevacizumab in combination with IFL chemotherapy. Serious adverse events included arterial thromboembolism, gastrointestinal perforations, haemorrhage, and impaired wound healing. Gastrointestinal perforations were seen in seven patients in the pivotal trial, of which two were serious and one was fatal, and in two patients in one of the other trials. The most frequently observed adverse events in patients receiving bevacizumab with or without chemotherapy were hypertension, asthenia, diarrhoea, nausea and pain. Hypertension could be medically controlled and seemed to be reversible. The occurrences of hypertension and proteinuria were likely bevacizumab dose‑dependent. Proteinuria was not associated with renal impairment. An increased incidence of diarrhoea could be due to increased exposure to the Irinotecan metabolite SN38.
Some of the commonly occurring adverse events in patients treated with Avastin®, and which are potentially related to bevacizumab include the following conditions.
Gastrointestinal perforations
Gastrointestinal perforations ranged in severity from free air seen on the plain abdominal X-ray to colonic perforation with abdominal abscess and fatal outcome. The common feature among these cases was intra-abdominal inflammation associated with various other medical conditions. Thus, patients receiving concomitant anti-inflammatory therapy may be at a higher risk.
Delayed wound healing
Post-operative bleeding or wound healing complications were observed in up to 20% of patients treated with Avastin® who underwent major surgery while receiving treatment.
Hypertension
Hypertension of any grade occurred in up to 32% of the patients treated with Avastin® and it was generally controllable with medication. At week 24 of treatment, the mean change in blood pressure from baseline in those patients was diastolic 4.1-5.4 mm Hg and systolic 5.5-8.4 mm Hg. Patients under treatment who develop hypertension that cannot be controlled with standard medication should discontinue Avastin® therapy.
Proteinuria
Proteinuria was reported in up to 38% of patients receiving Avastin®. It was most often asymptomatic (grade 1), bevacizumab dose-dependent, and it generally didn't impair renal function.
Haemorrhage
There was no significant difference between the incidence of grade 3-4 bleeding events in Avastin®-treated patients (3%-5%) and controls (2.5%-2.9%). These events were mostly minor muco-cutaneous haemorrhage (mostly Grade 1 epistaxis, in up to 34% of the patients; resolved without medical intervention) and tumour-associated haemorrhage (1%-3%).
Thromboembolism
While the incidence of general thromboembolic events was similar in treated patients and controls, the incidence of arterial thromboembolic events was higher in patients receiving Avastin® (3%-10%) compared to controls (1.3%-4.8%). The incidence of fatalities due to arterial events was 0.8%. The incidence of venous thromboembolic events was similar in patients receiving Avastin® (95-16.6%) compared to controls (13.5%-15.2%). Avastin® therapy should be discontinued in case of thromboembolic events.
Decreased neutrophil count and white blood cell count
Decreased neutrophil count and white blood cell count may be associated with Avastin® therapy. Grade 3-4 leukopenia occurred at an incidence of ≥5% in patients ≥65 years compared to patients <65 years. These laboratory findings may be correlated to the increased concentration of Irinotecan active metabolite SN38.
Other Considerations
The risk of central nervous system (CNS) haemorrhage could not be evaluated in Avastin®-treated patients as such patients were not enrolled in the clinical trials.
Avastin® should not be used in paediatric patients given that no data are available for this age group. Avastin® should not be used in pregnant women given that bevacizumab is expected to inhibit foetal growth (experimentally-proven teratogenic effects in rabbits) or women who breast-feed since bevacizumab may be excreted in human milk.
Notwithstanding all the preceding, the toxicity profile of bevacizumab remains acceptable, in general. The adverse effects induced by bevacizumab in conjunction with chemotherapy are clinically important and should be actively monitored.
3.4 Benefit/Risk Assessment and Recommendation
3.4.1 Benefit/Risk Assessment
The data submitted by the sponsor support the indication of Avastin® as a first-line anti-angiogenic therapy in combination with chemotherapy in metastatic colorectal cancer. However, the adverse effects of this combined therapy should be closely monitored and the healthcare provider should decide on the merits of stopping Avastin® treatment in a patient when a serious adverse event develops.
Pivotal study AVF2107g is the first trial of an anti-angiogenic drug that demonstrates efficacy. The significant improvement in overall survival of 4.7 months was achieved with an acceptable level of toxicity when taking into account the medical condition of the patients and the poor prognosis with chemotherapy only. Also, in study AVF2192g, the addition of Avastin® to 5-FU/Leucovorin therapy resulted in the prolongation of median survival by 3.7 months and increases of both the objective response rate and the duration of response in the Avastin® arm were noted in this study. Avastin®, per se, appears to have added little toxicity to the relatively toxic background of chemotherapy. However, healthcare providers need to be vigilant for hypertension (which can be controlled), proteinuria, and gastrointestinal perforation (one fatal case), which are the main adverse reactions of concern.
The clinical benefit achieved by adding Avastin® to IFL (Irinotecan+5-FU/Leucovorin) and 5-FU/leucovorin chemotherapy in patients with metastatic colorectal cancer could weigh against the relative toxicity of this therapy. However, healthcare providers and patients should be advised and alert as to the potential adverse events so they can effectively anticipate, prevent and find appropriate therapy should such events occur.
While there are several other analyses in this package, their value in a Priority Review Assessment is to confirm that the survivals, duration of these survivals and the response rates found are not an artefact but are consistent. It is also encouraging in this context that, for studies of this type, the Phase III pivotal trial was of a sufficient magnitude to allow for confidence in the data generated. It is to be noted that the majority of responses obtained are partial in both arms but with a difference of 10% between the bevacizumab and the placebo arm, in favour of bevacizumab.
Hence, from the perspective of clinical significance there is no question that the addition of bevacizumab to a standard chemotherapeutic regimen of Irinotecan , 5-FU and Leucovorin provides a significant advantage without significant disadvantages.
3.4.2 Recommendation
Based on the Health Canada review of data on quality, safety and efficacy, Health Canada considers that the benefit/risk profile of Avastin® (bevacizumab) in combination with intravenous fluoropyrimidine-based chemotherapy IFL (Irinotecan+5-FU/Leucovorin) is favourable for first-line treatment of patients with metastatic carcinoma of the colon or rectum. The New Drug Submission complies with the requirements of sections C.08.002 and C.08.005.1 and therefore Health Canada has granted a Notice of Compliance pursuant to section C.08.004 of the Food and Drug Regulations.
4 Submission Milestones
Submission Milestones: Avastin®
| Submission Milestone | Date |
|---|---|
| Pre-submission meeting: | 2003-12-16 |
| Request for priority status | |
| Filed: | 2003-12-22 |
| Approved: | 2004-01-08 |
| Submission filed: | 2004-01-27 |
| Screening 1 | |
| Screening Deficiency Notice issued: | 2004-02-24 |
| Response filed: | 2004-03-16 |
| Screening Acceptance Letter issued: | 2004-03-29 |
| Review 1 | |
| On-Site Evaluation: | 2004-11-15 - 2004-11-20 |
| Quality Evaluation complete: | 2005-06-21 |
| Clinical Evaluation complete: | 2005-09-07 |
| Labelling Review complete: | 2005-08-29 |
| NOC issued by Director General: | 2005-09-09 |
Related Drug Products
| Product name | DIN | Company name | Active ingredient(s) & strength |
|---|---|---|---|
| AVASTIN | 02270994 | HOFFMANN-LA ROCHE LIMITED | BEVACIZUMAB 25 MG / ML |