Summary Basis of Decision for Intelence

3.3.1 Clinical Pharmacology

A number of Phase I studies were conducted to assess the pharmacokinetics and the safety of etravirine for the clinical trials. The vast majority of the studies were drug interaction studies, however the studies also included a hepatic impairment study and a QT interval study.

3.3.2 Pharmacokinetics

Absorption

Etravirine pharmacokinetics are dose proportional between 200 mg and 400 mg; however, there is non-linearity upon multiple dosing.

Absorption of etravirine is increased by approximately 2-fold if etravirine is taken with food as compared to on an empty stomach. Subjects who were administered 200 mg etravirine twice daily (BID) within 15 minutes after breakfast and dinner generally reached the maximum plasma concentration at 4 hours after the first dose and after multiple dosing.

Distribution

Etravirine was extensively bound to both human albumin (99.60% at a physiological concentration of 4.3%) and α1-acid glycoprotein (97.66% to 99.02% at physiological concentrations of 0.10% to 0.20%). The fraction of etravirine distributed in blood cells was limited (23%).

Metabolism

The metabolism of etravirine was quantitatively limited; in the plasma unchanged etravirine was more abundant than any metabolite. The most important metabolic pathways involved first methyl hydroxylation and then glucuronidation. No unique metabolites were observed in human plasma compared to the animal plasma tested.

Excretion

Excretion of radiolabelled etravirine was almost complete in 168 hours after dosing and most (93.7%) of the radioactive etravirine was found in the feces; only a minor fraction (1.2%) appeared in the urine. Most (81.2% to 86.4%) of the radioactive dose that appeared in the feces was not modified.

Drug Interaction Studies

Drug interactions were extensively studied covering the integrase inhibitors (raltegravir) and nucleoside reverse transcriptase inhibitors (didanosine and tenofovir); non-nucleoside reverse transcriptase inhibitors (efavirenz and nevirapine); protease inhibitors (darunavir/ritonavir, saquinavir/ritonavir, indinavir, fosamprenavir/ritonavir, atazanavir, atazanavir/ritonavir, lopinavir/ritonavir, lopinavir/saquinavir/ritonavir, tipranavir/ritonavir, ritonavir and saquinavir) and other drugs (atorvastatin, clarithromycin, norethindrone-ethinyl estradiol, methadone, omeprazole, paroxetine, ranitidine, rifabutin, and sildenafil). Of the four classes of antiretrovirals tested, the unboosted protease inhibitors were the most significantly affected by etravirine co-administration, resulting in the recommendation that they not be co-administered in combination. Additionally, with tipranavir, which is boosted by 200 mg ritonavir, there was a substantial decrease in etravirine concentration upon co-administration. Consequently, tipranavir/ritonavir should also not be co-administered with etravirine.

Darunavir/ritonavir and saquinavir/ritonavir both decreased etravirine exposure by about 35%, however, given that etravirine was used safely and efficaciously with darunavir/ritonavir with the Phase 3 studies, both are acceptable for use.

Lopinavir/ritonavir resulted in an increase of etravirine exposure of approximately 85% greater than found in the Phase 3 studies, and given that there is limited safety data, the combination of etravirine and lopinavir/ritonavir should be used with caution.

The combination of nevirapine and etravirine resulted in poor tolerability in healthy female volunteers, including one case of Stevens-Johnson syndrome. Etravirine decreased the exposure of clarithromycin, but increased the exposure of the active metabolite 14-OH-clarithromycin, such that alternatives should be considered for treatment of Mycobacterium avium complex (MAC). As well, etravirine decreased the plasma concentrations of atorvastatin, but the plasma concentrations of the active metabolite were increased, therefore, dose adjustment may be necessary. Sildenafil plasma concentrations were significantly decreased, as well as the plasma concentrations of the active metabolite; concomitant use of PDE-5 inhibitors may require dose adjustment to attain the desired effect.

Hepatic Impairment Study

A pharmacokinetic study was carried out in patients with mild (n=8) and moderate (n=8) hepatic impairment along with age and sex matched controls. There were no statistically significant differences, therefore no dosing adjustments are required in subjects with mild or moderate hepatic impairment. Etravirine has not been studied in subjects with severe hepatic impairment.

QT Interval Study

The results of a well-designed QTc study demonstrated that after 8 days of dosing with etravirine 200 mg BID, 400 mg  once daily (QD), placebo or moxifloxacin 400 mg QD, etravirine did not prolong the QT interval. No subject in any group had an increase in QTcF of ‑60 msec. No subject experienced an interval exceeding the potentially clinically relevant threshold of 500 msec.

3.3.4 Clinical Efficacy

The clinical recommendation for market authorization of Intelence (etravirine) was based on the results of two pivotal Phase 3 studies, known as DUET-1 and DUET-2, and four supportive Phase 2b studies. Analysis of the 24-week data from the DUET studies provided the key efficacy and safety data for Intelence in the target population, treatment-experienced HIV-1 infected adults who have failed prior therapy and have HIV-1 strains resistant to multiple antiretroviral agents including NNRTIs. These studies are ongoing and were identically designed, randomized, double-blind, placebo-controlled, international studies designed to evaluate the long-term efficacy, tolerability, and safety of Intelence compared to placebo as part of antiretroviral therapy (ART) including darunavir/ritonavir (DRV/rtv) and an investigator selected optimized background regimen (OBR), in treatment-experienced HIV-1 infected subjects. A total of 1203 subjects were randomized and treated during the DUET studies; 599 received Intelence and 604 received placebo.

The primary efficacy parameter for the DUET studies was to show the superiority of Intelence compared to placebo, each administered in addition to a background regimen (including DRV/rtv in both treatment arms) in the proportion of subjects with confirmed undetectable plasma viral load values (<50 copies/mL) at Week 24. Various secondary efficacy endpoints were also assessed, including the proportion of subjects with HIV-1 RNA levels <400 copies/mL, the change in log₁₀ viral load, the proportion of subjects with a decrease in viral load of >1.0 log₁₀ copies/mL vs. baseline, and the effects on CD4 cell count.

The benefit of enfuvirtide (ENF) use as a naïve drug in a regimen has been demonstrated in other studies and thus enrolment of subjects with de novo ENF use was limited to 40%. The Breslow-Day (B-D) test was used to determine if a significant interaction existed between the use of ENF in the underlying ART and treatment group (significance level set at 0.2). The result of the B-D test demonstrated that the two treatment groups were highly significantly different, at Week 24, in their response to treatment (p=0.069), indicating that the virologic response rate was influenced by ENF use in the underlying ART. Therefore, the primary efficacy analysis was performed separately for each ENF strata ('de novo ENF use' and 'not de novo ENF use' which included patients re-using or not using ENF).

The results of the primary statistical analysis demonstrated that the difference in virologic response rates (<50 HIV-1 RNA copies/mL according to the TLOVR imputation algorithm) between Intelence and placebo was statistically significant (p<0.0001) in the group of subjects who did not use ENF de novo (56.3% and 33.6% in the Intelence and placebo groups, respectively). As expected, the benefit of Intelence to the treatment regimen was more limited in subjects who were using ENF de novo in the background regimen. A numerical advantage over those receiving placebo was seen, but the magnitude of the effect was small and no statistically significant difference between the treatment groups was found (66.7% and 61.9% for the Intelence and placebo groups, respectively). The secondary efficacy parameters and immunological results supported the results of the primary analysis.

No apparent relationship was observed between the pharmacokinetics of Intelence and virologic response. Subjects that enrolled in the two DUET studies had viruses with NNRTI resistance either at the time of screening or archived. The results demonstrated that Intelence had significant and sustained antiviral efficacy in subjects with NNRTI resistance.

The results of an exploratory Phase 2b trial, TMC125-C227, demonstrated that Intelence -treated subjects exhibited a substantial initial decrease in viral load, but had inferior antiviral responses compared to subjects treated with the control protease inhibitor (PI). This was affected in large part by reduced susceptibility both to Intelence and to the NRTIs in the background regimen. By contrast, since the control subjects were PI- naïve, all of them retained full susceptibility to the PI used. These data indicate that Intelence is not recommended for use in combination with only N[t]RTIs in patients who have virologically failed a first line NRTI and NNRTI-containing regimen. This condition is addressed in the Indications section of the Product Monograph for Intelence.

3.3.5 Clinical Safety

In the pooled safety data from the DUET-1 and DUET-2 studies, 599 subjects received Intelence (etravirine) 200 mg BID. The median exposure for subjects in the Intelence arm and the placebo arm was 30.0 and 29.1 weeks, respectively. The most commonly reported adverse events (AEs) in the Intelence group were diarrhea (15.0% vs. 20.4% in the placebo group), nausea (13.9% vs. 11.1%), rash (individual preferred term, 10.0% vs. 5.5%), injection site reaction (9.7% vs. 10.9%), and headache (9.3% vs. 12.3%). Most AEs were Grade 1 or 2 in severity. Grade 3 or 4 AEs occurred in 24.7% of the subjects in the Intelence group and in 27.2% of the subjects in the placebo group.

Twenty-three subjects died due to AEs starting during the treatment period, 8 (1.3%) subjects in the Intelence group and 15 (2.5%) subjects in the placebo group. All deaths in the Intelence group were considered to be not related or doubtfully related to the study drug. During the treatment period, 35 (5.8%) subjects in the Intelence group and 27 (4.5%) subjects in the placebo group permanently discontinued the investigational medication due to AEs. Rash (any type) led to discontinuation in 2.2% of the Intelence-treated subjects (0% in the placebo group).

Some AEs were considered of special interest, based on their relevance in the target population, their known association with other antiretroviral agents, or their potential importance based on non-clinical or previous clinical data with Intelence. Special attention was given to the following AEs of interest: skin events, neuropsychiatric events, hepatic events, cardiac events, bleeding events, and pancreatic events. Most AEs of interest were Grade 1 or 2 in severity and infrequently led to permanent discontinuation of treatment. The incidence of AEs of interest was comparable between the treatment groups, except for skin events. Rash (any type), combining all rash-related terms reported during treatment and/or follow-up, was reported in 17.0% of subjects in the Intelence group compared to 9.4% in the placebo group. Rash with Intelence treatment mostly emerged during the first weeks of treatment. In the Intelence group, the median time to onset was 12 days and the median duration was 11 days. The incidence of rash (any type) was higher in women (28.3%) than men (15.8%). This gender difference for rash was not seen in the placebo group (8.7% and 9.5% for females and males, respectively). There was no gender difference in severity or incidence of treatment discontinuation due to rash and no relationship with exposure.

Among the 2328 subjects exposed to Intelence in the clinical Phase 1, 2, and 3 studies, there was one case of erythema multiforme reported in a Phase 2b trial (C223; 400 mg bid dose) and no cases of Stevens-Johnson Syndrome (SJS) were reported in any of the Intelence treatment groups (one was reported in the placebo DUET group, most likely due to an allergic reaction to trimethoprim/sulfamethoxazole). Three cases of SJS have been reported in the Early Access Program (EAP) to date. As of January 7, 2008, a total of 5508 subjects have been enrolled in the EAP. With the three cases of SJS assessed by the investigator as at least possibly related to etravirine among a total of 7836 subjects exposed to Intelence, the incidence rate of SJS is 0.038% which is in the category of rare adverse events (<0.1%). In all three cases, treatment with Intelence was stopped and the patients recovered (one died later from end-stage AIDS). The potential for the development of severe skin reactions, including SJS, is identified in the Product Monograph under the Warnings and Precautions and Adverse Drug Reactions Sections. Based on the data to date the incidence of severe rash (Grade 4, including SJS, erythema multiforme, toxic epidermal necrolysis, and exfoliative dermatitis) associated with the use of Intelence appears to be lower compared to that observed in the clinical studies with evavirenz and nevirapine, other members of the NNRTI class.

The evaluation of AEs by age, race, gender and specific disease characteristics showed no significant differences, except for a higher incidence of rash in women.

No apparent relationship was observed between the pharmacokinetics of Intelence and the occurrence of AEs or changes in laboratory or ECG parameters.

Limited long-term safety data (>48 weeks) has been accumulated from the Phase 2b and 3 studies. Of the 1041 subjects treated with Intelence in these studies, 277 were treated for more than 48 weeks; their median treatment duration was 119.6 weeks. In this subgroup, the incidence of AEs was higher in the first 48 weeks of treatment than in the treatment phase after Week 48. The incidence of AEs, including rash, tended to decrease with long-term dosing, except for psychiatric AEs which were similar in incidence in the two treatment phases, and the coronary disorder events which showed a slightly higher incidence in the period after Week 48 (3.2%) than in the first 48 weeks (1.8%). The incidence of myocardial infarction (acute MI and MI grouped) was similar in the group treated up to 48 weeks (1.2%) and in the period after Week 48 (1.4%) of the long-term treatment group. There was no individual AE with an incidence substantially higher in the later treatment phase. No new clinically relevant AEs emerged in the treatment phase after Week 48. Other AEs of interest such as hepatic, pancreatic, and bleeding events tended to be less frequent in the treatment phase after Week 48 compared to the first 48 weeks of treatment. The above data suggest that Intelence is well tolerated, however, the full 48-week data from the Phase 3 DUET studies will need to be assessed when available, particularly with regards to the incidence of coronary disorder events.