Summary Basis of Decision for Macugen

3.3.1 Human Pharmacology

Studies in Age-related Macular Degeneration (AMD)

DLT = Dose limiting toxicity; AE = Adverse event; BCVA = Best corrected visual acuity; IOP = Intraocular pressure;
PK = Pharmacokinetics; PDT = Photodynamic therapy with verteporfin; S CNV = Subfoveal choroidal neovascularization;
QOL = Quality of life

The clinical program consisted of six studies in Age-related Macular Degeneration (AMD). Five studies (NX 109-01, EOP1000, EOP1001, EOP1003 and EOP1004) were presented in support of the claim for efficacy of pegaptanib sodium in the treatment of patients with exudative AMD. All studies evaluated the safety of pegaptanib sodium. Two studies (EOP1003 and EOP1004) were classified as pivotal and are ongoing with Year 2 extensions; this priority submission only contains data from the first year. Study EOP1006 was designed to further assess the safety of pegaptanib sodium.

All studies were performed in accordance with Good Clinical Practice (GCP) and the Declaration of Helsinki. It was indicated that the studies were designed, conducted, analyzed and reported to a high standard in concordance with current clinical research practices. All studies clearly identified patient populations and employed prospectively-defined standards and accepted measures for the assessment of safety and efficacy.

The development program in humans was conducted entirely in patients; the reason being that the intravitreous mode of administration imposed some risks that are not justifiable in healthy volunteers who derive no benefit.

3.3.2 Pharmacodynamics

Pharmacodynamic studies were not conducted. Based on literature findings on the role of VEGF₁₆₅ or VEGF₁₆₄ on humans and animals, and on the performed non-clinical pharmacology studies, it was postulatedthat pegaptanib sodium will inhibit pathological Choroidal Neovascularization (CNV), without an effect on physiological aspects of VEGF in the eye or in the systemic circulation. Hence, the VEGF₁₆₅-specific antagonist pegaptanib is expected to treat exudative AMD, and this should prevent relative deterioration of visual acuity. The design of the clinical trials included assessments of AMD lesion characteristics before and after treatment with pegaptanib sodium.

3.3.3 Pharmacokinetics

The pharmacokinetics (PK) of pegaptanib in humans was characterized in plasma only and was therefore not fully characterized. The plasma kinetics of pegaptanib were characterized following single and multiple IVT dosing in patients with AMD. For more information on the study designs, see Section 3.3.1Human Pharmacology. The dose of 3 mg/eye, which is 10x the proposed therapeutic dose, was used for the determination of the PK parameters; doses less than 3 mg/eye resulted in plasma levels close to or below the limit of quantification. A validated, Good Laboratory Practices (GLP), dual hybridization method was used to quantify plasma pegaptanib concentrations except for study NX-109-01 where a non-GLP but validated method was used.

Absorption

Results from studies EOP1000, 1001 and 1006 showed similar plasma PK parameters following single or multiple IVT doses of 3 mg/eye of pegaptanib. From these studies, the overall mean C_max was 76 ng/mL (range of individual values 19-136 ng/mL, mean T_max 1-2.5 days) and 77 ng/mL (range of individual values 27-200 ng/mL, mean T_max 1-2.5 days) after the first dose, and the third or fourth dose, respectively. Also, the AUC inf range of individual values was 4-49 µg·h/mL (mean of 25 µg·h/mL) after the first dose and the AUC_tau range of individual values was 13-41µg·h/mL (mean of 25 µg·h/mL) between the second and third dose and, between the third and fourth dose.

These results provide evidence that pegaptanib is unlikely to accumulate in the plasma following repeated dosing (IVT administration every 4 or 6 weeks for 12 or 24 weeks respectively) with a dose of 3 mg/eye. Importantly, results from study EOP1004 provide evidence that there is no plasma accumulation following the administration of the proposed therapeutic dose of 0.3 mg/eye every 6 weeks up to 42 weeks; 100% of the patients (n = 64) with trough samples had non-measurable pegaptanib plasma concentrations (LOQ of 8 ng/mL; samples taken approx. 6 weeks following IVT injection).

Absolute systemic/plasma bioavailability of pegaptanib following IVT administration could not be evaluated since there were no clinical studies with IV administration. Results from animal studies, however, suggest that 70 to 100% of the intravitreously administered dose would be absorbed in the plasma.

Distribution

The terminal half-life was estimated to be 10 ± 4 days in humans (study EOP1006). This half-life provides support for a slow transfer from the eye/vitreous to the systemic circulation as observed in animals. Relative to animal data, where mean half-life of 2-5 days was observed following IVT administration, the longer terminal half-life observed in humans may suggest slower passage from the eye to the systemic circulation, lower plasma metabolic activity and/or slower renal elimination. The fact that levels of pegaptanib were close to or below the limit of quantification 4 weeks after IVT administration (3 mg/eye) suggest low residual levels of pegaptanib in the eye/vitreous at the end of the proposed dosing interval (6 weeks) and dose (0.3 mg/eye). Nevertheless, the possibility of accumulation of pegaptanib or one of its metabolites in the eye cannot be excluded. In light of the favourable non-clinical and clinical safety profile of pegaptanib, the potential accumulation of small amounts of pegaptanib and or metabolites does not appear to be problematic.

Information from a pegaptanib tissue distribution study in rabbits, albeit the small number of animals involved (n = 1/sample time collection; 5 sample collection over 42 days), showed that pegaptanib was distributed to the highly perfused organs (liver, kidney, lymph node, spleen, bone marrow).

Metabolism

Pegapatnib is believed to be metabolized by endo- and exonucleases to shortmers and/ or nucleotides. No measurements of pegaptanib metabolites were made in the vitreous, plasma and urine of patients. Metabolism of pegaptanib by nucleases in human is supported by the presence of 2'-fluoro-deoxyuridine, a modified nucleotide of pegaptanib also found in the plasma and urine of pegaptanib-treated rabbits. Also, there was in vitro evidence suggesting that pegaptanib can be metabolized by nucleases in human plasma (study 400014), albeit the metabolic rate in human plasma appears to be much lower relative to other species (dogs and rabbits). Fomivirsen, an oligodeoxynucleotide, was shown to be metabolized by nucleases in human vitreous in vivo. The possibility that pegaptanib (a pegylated modified oligonucleotide) could be metabolized in the human vitreous/eye cannot be excluded.

Elimination

The excretion of pegaptanib and its breakdown products were not evaluated in humans. A study in rabbits, however, indicate that urinary excretion is the predominant route of elimination. The fact that renal clearance affects AUC and C max provides evidence of renal excretion of pegaptanib in humans.

Special Populations

Subjects involved for PK parameters determination (Studies EOP1000, 1001 and 1006) were elderly (mean age range 75.7-78.3; overall range 53-92), mainly female (64% of total subjects) and were afflicted by AMD. Study EOP1006 involved some subjects previously treated with photodynamic therapy (PDT) or had PDT during the study. Study EOP1001 involved subjects scheduled to have PDT treatment in the course of the study. There were a limited number of patients that were below 60 and above 85 years old (EOP1006). Overall, the data did not provide any evidence that the plasma PK parameters were significantly affected by gender, age or PDT treatment.

In rabbits, pegaptanib and 2'-fluoro-deoxyuridine (a nucleotide and metabolite of pegaptanib) were eliminated mainly via the kidney/urine. Results of study EOP1006 showed that renal function may affect the PK parameters of pegaptanib. These changes in the PK of pegaptanib were considered of no clinical consequence as the variation in plasma levels caused by the renal insufficiency were within levels attained with a dose of 3 mg/eye, a dose 10x higher than the proposed therapeutic dose and shown to be safe in animal and human studies. Patients with creatinine clearance less than 20 mL/min have not been studied.

Drug Interaction Studies

Pharmacokinetic drug interaction studies were not conducted. There is no information suggesting that o ligonucleotides are metabolized by the cytochrome P450 (CYP450) system. Following IVT administration, only low concentrations of pegaptanib transit through the liver making the possibility of a classic drug-drug interaction through liver metabolizing enzymes unlikely. As well, even though several antisense oligonucleotides have been specifically designed to inhibit CYP450 enzymes, non-specific control oligonucleotides have been shown to have no effect.

The potential plasma protein interactions/binding with pegaptanib were not investigated. A drug interaction through plasma protein displacement is unlikely given the low systemic exposure following IVT administration. As well, plasma protein binding is not believed to be extensive as results show a lack of effect of 50% plasma on the IC 50 value of pegaptanib in vitro.

3.3.4 Clinical Efficacy

Six studies were submitted to support the use of Macugen for the treatment of exudative AMD. Their design descriptions can be found in Section 3.3.1Human Pharmacology. The two pivotal studies (EOP1003 and EOP1004) were designed to last 2 years. This current submission contains data from only the first year. Both studies had identical designs except that EOP1004 included assessments of the drug plasma concentration measurements and the National Eye Institute Visual Functioning Questionnaire (VFQ 25).

The two pivotal clinical studies were well conducted. One study was conducted internationally including Canada and USA. Of the main efficacy variables used, the last observation carried forward to impute data for missing values, is a potential concern in a condition with declining vision. The concern was greatly mitigated by the low proportion of patients who withdrew from the studies. There was an ethical concern to prohibit treating patients with the currently approved therapy, photodynamic therapy (PDT). Administration of PDT was at the investigators' discretion, making analysis of the concomitant PDT and pegaptanib data flawed (not reported). In addition, the selection of the control group posed difficulties because of the invasive administration procedure, intravitreous injection. Therefore, a sham was selected as control. The problem with sham treatment is that neither the effect of the injection, the drug or the solvent can be segregated in the results for efficacy/adverse events. In addition, the injection alone may result in some type of body reaction that will not be controlled in sham.

In general, the study population reflected the demographic of AMD. The medical and surgical histories reflected the age of the AMD population; the most common vascular conditions were hypertension, cardiac disorders, and muscular-skeletal conditions. Baseline applanation tonometry revealed that patients had normal or medically controlled intraocular pressure of 15.2 (SD 2.9) mm Hg. There was a similar distribution in ocular conditions among the groups. The number of patients that received prior PDT treatment was similarly distributed but the proportion in study 1004 (sham 11%, 0.3 mg 13%, 1mg 14% and 3 mg 14%) was larger than in study 1003 (sham 3%, 0.3 mg 4%, 1 mg 6% and 3 mg 4%). The majority of patients (80%) had a visual acuity 20/30 to 20/100. All lesion sizes were ≤ 12 disc area. The study groups had a similar proportion of patients with the three AMD subtypes. Among the lesions the "classic only" was present in about 25% of the population. Similar proportions of patients had "minimally classic" and "occult only" AMD. The studies were not designed to analyze the efficacy data in each of the AMD subtypes.

In both pivotal studies, the primary efficacy endpoint was the proportion of Responders, defined as patients losing < 15 letters of best-corrected visual acuity (VA) in the study eye from baseline up to 54 weeks, using the Early Treatment Diabetic Retinopathy Study chart at 2 meters. This measurement is considered to be the most meaningful and clinically relevant measure of visual outcome in patients with exudative AMD. Secondary efficacy endpoints were i) the proportion of patients gaining at least 15 letters of vision from baseline to 54 weeks; ii) the proportion of patients gaining at least 0 letters of vision from baseline to 54 weeks and iii) the change in mean VA from baseline to 6, 12 and 54 weeks.

Both pivotal studies demonstrated that Macugen was efficacious in patients suffering from SCNV secondary to AMD (SCNV-AMD) as assessed from results of the primary and secondary variables. The proportion of patients who lost < 15 letters at week 54 in the Macugen 0.3 mg group was statistically significantly higher as compared with the sham group in both studies (Study EOP1003: Macugen 0.3 mg 73% vs. Sham 59%, p-value = 0.0105; Study EOP1004: Macugen 0.3 mg 67% vs. Sham 52%, p-value = 0.0031). Results showed that the proportion of patients losing less than 15 letters at week 54 were similar irrespective of baseline characteristics, including lesion sub-type, prior PDT usage, visual acuity, lesion size, gender, iris pigment and age as assessed by similarity in results (eye bold), not by statistical analyses. The calculated Therapeutic Gain (treated minus sham) in the 0.3 mg dose group was about 15%. It can be concluded that Macugen 0.3 mg treatment over a one-year period reduced the risk of vision loss in patients with SCNV-AMD. However, the benefit over sham was only about 15%.

Secondary variables demonstrated that patients in both Macugen 0.3 mg and sham continued to experience vision loss, but Macugen slows down the disease progression assessed by the number of letters that patients were able to read during the one year study period. Furthermore, the proportion of patients having severe vision loss during the study period, defined as a loss of ≥ 30 letters of VA over 54 weeks, was reduced by more than a half in the Macugen-treated patients (10% of patients experienced severe vision loss) compared with sham (22% of patients).

3.3.5 Clinical Safety

The clinical program consisted of six studies in Age-related Macular Degeneration (AMD) and all evaluated the safety of pegaptanib sodium administration. Two studies (EOP1003 and EOP1004) were classified as pivotal and were designed to last two years. The data for this priority submission was taken from the first year. Study EOP1006 was designed to further assess the safety. The study design of these trials can be found in Section 3.3.1 Human Pharmacology.

The safety population in the two pivotal studies, EOP1003 and EOP1004 consisted of 612 and 578 patients, respectively, who received at least one treatment. Less than 14% of the patients withdrew from the studies during the first year and approximately 90% of patients received 8 to 9 treatments (pooled data from the two studies). Patients were mostly white, a median age of 75 years of age and female, reflecting the general population with AMD. The different groups (Macugen-treated groups and sham-treated groups) were similarly distributed with respect to gender, age, race, performance status, height and weight. The groups were also well-balanced in terms of ophthalmic, medical and surgical history. The most prevalent baseline disorders were hypertension Not Otherwise Specified (NOS, 48-54%), arthritis NOS (19-21%), hypercholesterolemia (18-23%) and drug hypersensitivity (16-23%). There were no imbalances in baseline medication across treatment groups: nearly 85% received medications affecting the sensory organs (mostly anticholinergics), 64-72% received cardiovascular medications (mostly HMGCoA inhibitors), and 52-59% received medications affecting the alimentary tract and metabolism (vitamins and calcium). No patients had baseline electrocardiograms (ECG) with clinically significant findings.

The two studies showed no clinically meaningful imbalances across the groups with respect to the ophthalmic history. Baseline lesions subtypes were classified based on fluorescein angiographic appearance with 38-40% having 0% classic AMD (occult), 34-38% having from 1-49% classic AMD (minimally classic), and 24-27% having ≥ 50% classic AMD (predominantly classic). For the overall baseline ocular history 12-14% of patients across treatment groups had received prior PDT with verteporfin (≥ 8 weeks from first treatment) and 5-6% had received prior systemic treatment for AMD. Subretinal hemorrhages were present at baseline in 71-73% (69% in EOP1003; 75% in EOP1004 of the safety population that included all treatment groups) of all patients.

The rate of patient discontinuation was low. The most frequent reason for discontinuation was "patient request" (5%), adverse events (1%), and death (2%). Death was not unexpected in this age group; none were related to treatment.

Adverse events were reported in nearly all patients (95-100% of patients in study EOP1003 and 99-100% of patients in study EOP1004). The proportion of patients having at least 1 serious adverse event (SAE) was < 20% and only 1-2% of the SAE led to discontinuation.

It should be kept in mind that Macugen was planned to be administered intravitreously (IVT) at 0.3 mg, 1 mg, 3 mg dose per eye in a total volume of 90uL every 6 weeks during Year 1. The planned number of IVT injections was nine. As mentioned previously, 8 to 9 injections were administered to about 90% of the patients. The control group (sham) did not receive IVT injections but were subjected to all the previous procedures and were simulated to receive the IVT injection. They did not know whether they received the injection or not. The adverse events, elicited differentially between the sham group and Macugen groups, could be assessed as being related to the IVT injection or to the drug (pegaptanib sodium). It is difficult to separate injection from drug- related events and this is therefore a deficiency of the study, but accepted because of ethical concerns.

Ocular Adverse Events

The majority of all the adverse events (AEs) were eye disorders (82-88% in study EOP1003 and 92-97% in study EOP1004). Systemic AEs were infrequent, as expected with a locally administered treatment.

Approximately 90% of the patients reported ocular AEs in the study eye; the majority of the events were mild to moderate in severity. With few exceptions, ocular AEs were elicited similarly among all three active treatments (0.3 mg, 1 mg, 3 mg dose per eye) which suggest a relationship to the injection procedure rather than the drug itself.¹ The ocular AEs that appeared to be Macugen dose-dependent were corneal edema, cataract, conjunctival edema, and eye lid edema, but, the incidences in the 0.3 mg group were similar to or lower than in the sham group.

The safety data described below summarize the experience of 295 patients exposed for up to one year (total number of injections = 2478, mean number of injections/patient = 8.4) at the recommended dose of 0.3 mg.

¹ Animal studies, which had IVT injection control groups, demonstrated that repeat intraventricular injection resulted with ocular adverse effects (ocular discharge, conjunctival irritation, vitreal floaters, cataracts, retinal detachment) across treatment groups including controls suggesting that these ocular events were injection-related.

The following ocular AEs were reported as serious in any Macugen group: endophthalmitis (12 cases, 1%), retinal detachment (4 cases, < 1%), retinal hemorrhage (3 cases, < 1%), cataract (3 cases, < 1%), traumatic cataract (3 cases, < 1%), vitreous hemorrhage (2 cases, < 1%), glaucoma NOS (1 case, < 1%), uveitis NOS (1 case, < 1%), and increased intraocular pressure (1 case, < 1%).

The serious events that could be preventable or tightly-controlled are endophthalmitis and traumatic cataracts due to the intravitreous injection procedure. The incidence of endophthalmitis was reduced with improved aseptic conditions, but not eliminated. Adherence to strict aseptic procedures, as well as patient follow-up during the first week after any injection should be emphasized. The signs and symptoms of endophthalmitis should be clearly annotated in the Product Monograph, as well as a description of the procedure before and after IVT injection. To avoid traumatic cataracts, the Product Monograph should instruct doctors to avoid touching the lens with the syringe needle.

The ocular AEs that may be possibly related to undesirable VEGF inhibition in the eye, based on experience with systemic agents that inhibit VEGF, are thromboembolic or bleeding events. Vitreous hemorrhage was the only event that occurred in greater proportion in the Macugen-treated group compared to sham. But the incidence of vitreous hemorrhage did not increase with higher doses of pegaptanib sodium, and did not reoccur in 15 of the 16 patients as would be expected with chronic VEGF inhibition. Fluorescein angiography revealed no retinal vascular abnormalities that are unexpected in the natural history of exudative AMD. There were no notable delays in arterial-venous transit time, abnormalities in choroidal perfusion, or arteriolar occlusions.

The possibility that ophthalmic bleeding could be exacerbated or elicited by co-administered medications such as anticoagulants or platelet aggregation inhibitors was of interest. Study results showed no evidence for an increased risk of vitreous hemorrhage in Macugen-treated patients associated with concomitant anticoagulation medications, platelet inhibitors or non-steroidal anti-inflammatory drugs.

Systemic Adverse Events

Non-ocular AEs related to the systemic inhibition of VEGF were unexpected in the pegaptanib sodium program because of the specificity of pegaptanib sodium for VEGF₁₆₅ and the low or undetectable systemic exposure from intravitreous dosing . In addition, the preclinical assessment of pegaptanib sodium showed no evidence of physiological changes or toxic consequences that would be regarded as reasonably related to VEGF-antagonism, either systemically or in the eye. Preclinical exposures from the intravenous or intravitreous routes achieved exposures considerably higher than exposures achieved in humans administered with 3 mg in one eye. Toxicology and safety pharmacology studies showed no evidence of hypertension, proteinuria, thromboembolic, or bleeding phenomenon.

Systemic AEs were infrequent in the safety studies. The types and incidence of the systemic AEs were not unexpected for an elderly population and most events had similar incidences in all treatment groups. The most common events in the 0.3 mg dose group were hypertension (14/295, 4.7%), nasopharyngitis (19/295, 6.4%), headache (19/295, 6.4%), and bronchitis (16/295, 5.4%). Arthralgia and nausea, and upper respiratory tract infections had an incidence of 4% each.

Systemic thrombotic cardiovascular/pulmonary and hemorrhagic SAEs were few and similar between Macugen groups and the sham group. A review of the safety narrative for these patients indicated that all but one of these cases had strong pre-existing risk factors for thromboembolic cardiac and central nervous system (CNS) disorders; namely hypertension or pre-existing vascular occlusive disease. There is no indication that hypertension/increase blood pressure and myocardial infarction (MI) are related to anti-VEGF effect. Anti-VEGF therapy could potentially inhibit/delay recovery from MI, inhibiting the development of collaterals associated with MI/stroke, inducing or accelerating angina. No time frame for anti-VEGF therapies could be offered if patients recently had an episode of MI/stroke.