Summary Basis of Decision for Campral ®

 

A series of drug formulations, either enteric-coated or otherwise, were used in the development. The most significant change was the switch from the initial formulation to the current formulation. With few exceptions, the initial formulation was used in early studies, including three key efficacy studies. The exceptions are the single-dose and repeat-dose bioequivalence studies between the two formulations. The formulation intended for marketing in Canada, was used in the supportive studies.

3.3.1 Human Pharmacology

Once absorbed, acamprosate disassociates into two parts of acetyl-homotaurine and one part of calcium. Unabsorbed acamprosate is eliminated in the feces. Its bioavailability is low at 11%. At steady state, the volume of distribution of acetyl-homotaurine is about 20 L. Acetyl-homotaurine has low protein binding and shows no evidence of biotransformation. Half of the absorbed drug is eliminated via the urine and the other half appears to be via biliary excretion.

The administration of the enteric-coated tablets has shown to have a terminal elimination half-life of 20-32 hours for acetyl-homotaurine, much higher than that seen after intravenous infusion. After repeated dosing of acamprosate 1998 mg three times daily (666 mg, t.i.d), steady-state of acetyl-homotaurine is reached after 5-7 days, resulting in plasma concentrations raging from 379 to 650 μg/L. The pharmacokinetics (PK) of acetyl-homotaurine is time- and dose-dependent, when administered as enteric-coated tablets. The accumulation ratio at steady-state is 2.5.

The PK of acamprosate differs little between men and women. Renal impairment reduces the rate of elimination of acetyl-homotaurine. The reduction is correlated with creatinine clearance. Dose reduction is recommended in patients with mild to moderate renal impairment, and contraindication in patients with severe renal impairment.

The PK of acetyl-homotaurine is unchanged in patients with liver impairment or chronic alcoholism. In patients with mild to moderate liver impairment (Child-Pugh A and B), no dose reduction is required. Acamprosate has not been studied in patients with severe liver impairment (Child-Pugh C). Acamprosate is not recommended in patients with severe liver impairment.

Food reduces the absorption of acamprosate, but the reduction is not clinically significant. The PK of acetyl-homotaurine is unaffected by alcohol, disulfiram, or diazepam. Acetyl-homotaurine has no significant effect on the PK of ethanol, diazepam and nordiazepam, or imipramine and desipramine, when they are co-administered.

Co-administration of acamprosate with naltrexone led to a 33% increase in the C_max and a 25% increase in the AUC of acetyl-homotaurine.

3.3.2 Clinical Efficacy

When Campral^® was submitted to the FDA on December 27, 2001, more than a decade had passed since acamprosate was first marketed in Europe. Most of the Campral^® clinical studies were initiated before July 1, 1991, when the European Community Guidelines on Good Clinical Practice (GCP) came into force.

On May 10, 2002, the FDA convened an Advisory Committee meeting to consult on the acceptability of the studies conducted with older standards. During May 13-24, 2002, a site inspection was conducted for Studies PRAMA and Paille, the two pivotal studies of 1-year duration. Several deficiencies were found. These deficiencies, and other issues identified in the dossier, led to a letter of non-approval on June 27, 2002, in spite of the positive opinions of the Advisory Committee.

In response to the identified deficiencies, mainly regarding the reliability of the data but not the design and conduct of the trials, the sponsor proposed a complete audit of the three pivotal European efficacy trials, re-analysis of the data with the rate of complete abstinence as the primary outcome parameter, and percent days abstinent and time to first drink as the secondary outcome parameters. This proposal was accepted by the FDA, with an additional request to audit the safety data, on March 4, 2003. Campral^® was approved by the FDA on July 28, 2004 with a revised label and post-marketing commitments.

Because of this regulatory history, the review conducted by Health Canada incorporated the revised efficacy outcome measures as the result of the audit. For other European studies not audited, the results from the original clinical trial reports were reviewed.

Across the three pivotal efficacy studies, Pelc II, PRAMA, and Paille, a total of 623 patients were treated with acamprosate (251 with 1332 mg daily, 236 with 1998 mg daily, and 136 with a weight-adjusted dose) and 375 were treated with placebo. All three studies were conducted in Europe (France, Germany, and Belgium). Enrolled patients were adult outpatients with alcohol dependence, aged 18-65, who had at least a 12-month history of alcohol dependence, ranging from a minimum of 12 months (Pelc II and Paille) to 2-3 years (PRAMA). All patients underwent medically-supervised detoxification and remained abstinent from alcohol for at least five days (minimum of five days in Pelc II, minimum of 14 consecutive days and maximum of 28 days in PRAMA, and 7-30 days in Paille) before receiving study medication.

Patients in the Pelc II and Paille studies were randomly assigned to treatment with either acamprosate 1332 mg/day, acamprosate 1998 mg daily, or placebo. Patients in the PRAMA study were randomly assigned to treatment with acamprosate (1332 mg daily for patients ≤60 kg, or 1998 mg daily for patients >60 kg), or placebo.

Treatment durations for the pivotal efficacy studies were 90 days for Pelc II, 360 days for PRAMA, or 48 weeks (336 days) for Paille, respectively. Patients in these studies were to receive psychotherapy or other psychosocial therapy at the discretion of the investigator.

The drug exposure in these three studies was low, resulting from a low retention rate in all three studies. The table below summarizes number and percentage of those who remained at the end of study.

Clinical Efficacy

Study	Statistic	ACAMP 1332 mg/day	ACAMP 1998 mg/day	Placebo
Pelc II (90 days)	n / N (%)	18/63 (29%)	19/63 (30%)	19/62 (31%)
PRAMA (360 days)	n / N (%)	-	14/136 (10%)	12/136 (9%)
Paille (336 days)	n / N (%)	48/188 (26%)	49/173 (28%)	40/177 (23%)

The dose selection process was not satisfactory by today's standards; yet the current dosing regimen (maximum dose 1998 mg daily, dosing based partially on body weight, dose titration according to tolerance, considerations for renal function, old age, etc.) seems acceptable.

Results for the primary efficacy outcome parameter - the rate of complete abstinence - are shown below.

Clinical Efficacy

Study	Statistic	ACAMP 1332 mg/day	ACAMP 1998 mg/day	Placebo	P-value1
Pelc II	n / N (%)	26/63 (41%)	24/63 (38%)	8/62 (13%)	0.001**
(90 days)
PRAMA	n / N (%)	-	38/136 (28%)	17/136 (13%)	0.002**
(360 days)
Paille	n / N (%)	26/188 (14%)	28/173 (16%)	16/177 (9%)	0.044*
(336 days)
Data Source: Tables 8.7.1.5.1, 8.7.1.5.2, and 8.7.1.5.3 [Module 5, Vol. 150, 5.3.5.3.1, Pgs. 168-170]

* Significant at the 0.050 level; ** significant at the 0.010 level.
¹   P value is from the comparison of ACAMP 1998 vs. Placebo (Pelc II and Paille) or ACAMP vs. Placebo (PRAMA) based on a chi-square test of complete abstinence vs. non-abstinent or missing.
Note: Results from the ACAMP group in PRAMA are presented in the ACAMP 1998 mg/day column.

It is noted that the best efficacy outcome was not consistently observed in the higher dose (i.e. 1998 mg tid in Pelc II). Of the two studies of 1-year duration, the study with a flexible dosing regimen (PRAMA) had an overall better efficacy outcome. Due to these observations, the dosage should be 1332 mg daily for patients ≤60 kg, or 1998 mg daily for patients >60 kg, dose titration should be permitted based on tolerability, and dosing higher than 1998 mg daily is not recommended as there is no convincing evidence to indicate a higher dose results in better efficacy.

Efficacy Interactions

Some trends have been identified, but they have not been systematically investigated. With respect to gender, female sex seemed to be a predictor of a more promising outcome. With respect to age, there were some data to suggest that younger patients (age 16-39) tended to have improved efficacy. Body weight did not seem to be a predictor of efficacy outcome. The suggestion that a shorter history of alcohol dependence was associated with a better treatment outlook was not strongly substantiated. Subanalyses of efficacy interactions were conducted for polysubstance use, antidepressants, anxiolytics, hypnotics, sedatives, and analgesics. No lead was identified.

Dose Response

The sponsor attempted to present evidence of dose-related response in the pivotal studies. Given the regulatory history of acamprosate, and its unique pharmacokinetics (enteric coating, low bioavailability, flip-flop pharmacokinetics) and the complex psychoneuropharmacology in the treatment of alcohol dependence, it probably matters little if a clear dose-response could be established. This argues in favour of individualized, flexible dosing in the course of acamprosate treatment, taking into account specific patient pathophysiologic conditions, drug tolerability, etc.

The efficacy results for acamprosate indicate that a one-dose-fits-all conclusion is not acceptable for this drug. Treatment should be initiated according to body weight, with permission to titrate to the optimal dose level. Dose adjustments should be allowed according to renal clearance, overall health conditions, etc.

3.3.2  Clinical Safety

To facilitate the analysis and presentation of the safety data, the sponsor categorized all of the clinical studies into five groups. Group I included all of the randomized, double-blind, placebo-controlled studies at least 26 weeks but less than 52 weeks in duration. The only US study, US 96.1, which cannot be presented to support efficacy, is included in this group for safety analysis. Studies grouped from II through IV, and Other, included studies for clinical pharmacology, initial efficacy exploratory studies, post-marketing open-label studies, etc. The clinical safety profile for acamprosate has been mainly derived from the Group I studies and it is used for labelling. Safety issues identified in other studies are corroborated with those in Group I.

The Intent-to-Treat (ITT) safety population included a total 12,336 subjects: Group I, 4246 (acamprosate 2280, placebo 1966); Group II, 522; Group III, 924 (acamprosate 480, placebo 444); Group IV, 3773; and Others, 663. With the exception of Group I and III studies, the safety data from other Groups were not placebo controlled. According to the duration of the studies, Group I studies were further divided into short-term studies (≤26 weeks) and long-term studies (>26 weeks). In Group I studies, the total exposure at ≥26 weeks was 701 on acamprosate vs. 525 on placebo. The exposure at ≥52 weeks was 123 on acamprosate vs. 65 on placebo. Except Study US 96.1, all other studies were conducted in Europe. The predominant ethnic group was Caucasians. The age range of the studies was 18-65, with few exceptions. Women consist of a quarter to a fifth of the total study population.

3.3.3 Clinical Safety

To facilitate the analysis and presentation of the safety data, the sponsor categorized all of the clinical studies into five groups. Group I included all of the randomized, double-blind, placebo-controlled studies at least 26 weeks but less than 52 weeks in duration. The only US study, US 96.1, which cannot be presented to support efficacy, is included in this group for safety analysis. Studies grouped from II through IV, and Other, included studies for clinical pharmacology, initial efficacy exploratory studies, post-marketing open-label studies, etc. The clinical safety profile for acamprosate has been mainly derived from the Group I studies and it is used for labelling. Safety issues identified in other studies are corroborated with those in Group I.

The Intent-to-Treat (ITT) safety population included a total 12,336 subjects: Group I, 4246 (acamprosate 2280, placebo 1966); Group II, 522; Group III, 924 (acamprosate 480, placebo 444); Group IV, 3773; and Others, 663. With the exception of Group I and III studies, the safety data from other Groups were not placebo controlled. According to the duration of the studies, Group I studies were further divided into short-term studies (≤26 weeks) and long-term studies (>26 weeks). In Group I studies, the total exposure at ≥26 weeks was 701 on acamprosate vs. 525 on placebo. The exposure at ≥52 weeks was 123 on acamprosate vs. 65 on placebo. Except Study US 96.1, all other studies were conducted in Europe. The predominant ethnic group was Caucasians. The age range of the studies was 18-65, with few exceptions. Women consist of a quarter to a fifth of the total study population.

Adverse Events

Based on data from spontaneous reports, the most frequent adverse events (AEs) were gastrointestinal (GI) upsets, especially diarrhea and flatulence. The incidences were higher in the pooled group on acamprosate, based on Body System/Preferred Terms. Based on the short-term data gathered from work-sheet review, AEs of diarrhea, nausea, and insomnia were most frequent. The most frequent AE in the cardiovascular system was palpitation, followed by syncope. In the long-term data gathered from work-sheet review, diarrhea, palpitation, and syncope remained to be the most frequent events in the cardiovascular system. GI events, specifically diarrhea, were the predominant reasons for treatment discontinuation. Of note, AEs of a suicidal nature and AEs of the nervous system, including psychiatric disorders, caused nominally more discontinuation in those on acamprosate, without a dominant AE. The cause and clinical significance of this observation are uncertain at this time. Pruritus and rash also resulted in a nominally higher frequency of discontinuation.

There were a total of 62 reports of death in the clinical programme of acamprosate, 45 of them treatment-emergent. The combined incidence of death was 0.44% (34/7718) on acamprosate vs. 0.45% (11/2406) excluding healthy volunteers. In Group I studies, the incidence of death was 0.55% (12/2189) on acamprosate vs. 0.46% (9/1962) on placebo. There were 17 reports of non-treatment-emergent death, 9 from those on acamprosate, 6 on placebo, and 2 on other treatments. Of all reports of death, there were 12 reports of suicide. Of these 12, 8 were on acamprosate (one non-treatment emergent) and 4 were on placebo (two non-treatment emergent). The incidence of death due to suicide was 0.09% (7/7718) on acamprosate vs. 0.08% (2/2406) on placebo.

The sponsor conducted a further analysis of AEs of a suicidal nature, including completed suicides, suicide attempt, intentional overdose, suicidal ideation, and suicidal tendency. The results showed an increased frequency of these events in those on acamprosate. The combined incidence was 1.78% in those on acamprosate (36/2019) vs. 0.59% in those on placebo (10/1706). The risk ratio was nearly 3-fold higher in those on acamprosate. The incidence of non-treatment-emergent AEs of a suicidal nature was 0.30% in those on acamprosate (6/2019) vs. 0.12% in those on placebo (2/1706).

Alcohol consumption/relapse were reported as serious adverse events (SAEs), ranging from 2-3% in short-term studies and 6-7% in long-term studies.

In addition, the combined SAEs of the cardiovascular system had a nominally higher frequency in those on acamprosate than on placebo in Group I short-term (12/1346, 0.89% vs. 4/1218, 0.33%) and long-term studies (12/126, 9.52% vs. 4/84, 4.76%). No specific event stood out.

Effects on Electrocardiogram

The safety monitoring of electrocardiogram has not identified any consistent signal. Multiple-dose integrated Group II clinical pharmacology studies did not exclude the possibility of a <5 msec increase in the QT interval with high doses of acamprosate, however this fell within the range of natural variability.

Drug Interactions

Analysis of the combined safety data has not found any drug interactions with anxiolytics, hypnotics, sedatives, and non-opioid analgesics. The AEs observed were consistent with the safety profile of individual medications. There is no information available on the concomitant administration of acamprosate with diuretics.

Safety Issues with Pregnancy and Lactation

No human data are available to exclude a teratogenic effect or an effect on child development with confidence. Given the animal data on teratogenicity, acamprosate is not recommended during pregnancy and in women of child-bearing potential not using a reliable method of contraception. The sponsor has undertaken a post-marketing animal study to further evaluate the effects of alcohol and acamprosate on the fetus in pregnant animals.

Acamprosate is secreted and concentrated in milk in nursing animals. No human safety data are available. Alternative therapeutic options are available. Acamprosate is not recommended in nursing women until safety issues of acamprosate in nursing women with alcohol dependence receives further clarification.

Overdose

The symptoms and signs of an overdose of acamprosate are severe gastrointestinal adverse reactions. The highest reported overdose has been 56 grams, resulting only in symptoms of gastrointestinal adverse reactions.

There have been a total of 16 reports of overdose from post-marketing data thus far, resulting in deaths, serious adverse reactions, or no apparent symptoms. In the three cases of death, causes of death by drugs other than acamprosate were likely. In the six cases with serious adverse reactions, several symptoms were attributable to acamprosate, such as gastrointestinal adverse reactions, dizziness, somnolence, etc., however alternative explanations were possible. In the five cases with no apparent symptoms, the total amount of overdose ranged from 10-56 g. No hypercalcemia was reported in any of the overdose cases.

Although acamprosate itself has a high therapeutic index, in overdosed alcoholics, it is a frequent observation that multiple drugs are involved. Thus, if acamprosate overdose is present, co-overdose with other drugs should be suspected.

The cumulative animal and clinical data have not identified any potential of tolerance or dependence for acamprosate, nor a risk for abuse.

Post-marketing Safety

The global exposure during the period from August 1, 2002-July 31, 2005 was 232,600 patient-years. The safety analysis showed that the safety profile of acamprosate was consistent with that observed in the clinical studies. The data on overdose have been discussed above.

3.3.4 Bioequivalence

The formulation of Campral used in the clinical studies and the proposed commercial formulation of Campral are considered by Health Canada to be significantly different. To support this significant formulation change, the sponsor has provided the following studies comparing the clinical formulation to the proposed commercial formulation:

AD993H - randomized, open-label two period crossover study comparing the pharmacokinetic parameters of the clinical and proposed commercial formulation of Campral when given as a single dose (2 x 333 mg)

AD1044H - randomized, open-label two period crossover study comparing the pharmacokinetic parameters of the clinical and proposed commercial formulation of Campral when given as multiple doses (2 x 333 mg t.i.d. for eight days)

Normally a significant change in the formulation would require a single dose study conducted under fasting conditions. Additional studies may be required depending on the characteristics of the drug. However, the design of the study AD993H was flawed, specifically the study was not sufficiently powered (N=12) to demonstrate bioequivalence between the two products containing a drug which potentially has high inter-subject variability. Consequently, for this submission, study AD1044H was reviewed in depth instead of study AD993H.

The analytical method used in the analysis of subject samples from study AD1044H was flawed. In an appropriately designed analytical protocol, in order to confirm the accuracy of the data, each of the quality control samples should be analyzed in duplicate on each run. In study AD1044H, the majority of the quality control samples were analyzed only once per concentration on each analytical run. On four of the analytical runs, quality control samples that were analyzed only once per concentration were outside of the acceptance criteria. As the four analytical runs were not repeated, the data generated from these runs for subjects 2, 3, 10, and 15 after administration of the commercial formulation cannot be confirmed as accurate. Further, 24 subject samples from 4 subjects had acetylhomotaurine concentrations that were above the upper limit of quantitation (ULOQ) of the analytical method. These subject samples were not reanalyzed and the data generated for several samples could not be confirmed as accurate.

Due to the above reasons, all data from subjects 2, 3, 10, and 15 were removed from the dataset, in addition to the samples from subjects 4, 5, and 13 which were above the ULOQ.  With these subjects and samples removed from the dataset, the appropriate pharmacokinetic parameters could not be adequately defined. Therefore, Prempharm Inc. has failed to establish bioequivalence between the Campral 333 mg modified release tablet formulation used in the clinical studies and the Campral 333 mg modified release proposed commercial tablet formulation.

The were, however, a number of more recent open-label studies that used the Campral 333 mg modified release proposed commercial tablet formulation. The studies were the New European Alcoholism Treatment (NEAT) studies from Belgium, Switzerland, Austria, UK, and Portugal. Studies Integral, Capriso, Micado, and Ares provided further support. The clinical outcomes were overall consistent with the outcomes of the European pivotal studies (Pelc II, Prama, and Paille) with regard to safety and efficacy. The batches used in the NEAT studies are considered pharmaceutically equivalent to the proposed market product. Consequently, it is considered that these studies can be used to bridge the formulation used in the three pivotal studies (the initial formulation) and the proposed commercial formulation to provide clinical equivalence between those formulations.