Raltegravir FDA Safety Report
From Jules Levin
These are excerpts from FDA briefing document for hearing on Sept 5 for approval of Raltegravir. Included below are discussions of malignancies, special populations, rash, hepatic events/ALT, resistance, and common adverse events. . I did not include drug interaction report in this because there is so much information. Based on my review the safety analysis finds raltegravir relatively safe and tolerable, here is quote from FDA at end of this report, but you should review this for yourself. Of interest was information on PK variability and food effect:
"In conclusion, raltegravir appeared to be well-tolerated in the Phase 2 and Phase 3 clinical studies with relatively few subjects discontinuing for adverse events. No clinically significant imbalance was observed in mortality rates and ADCs. A higher number of malignancies was observed in raltegravir-treated subjects initially; however, the imbalance appeared to reflect more a paucity of malignancies in control/placebo-treated subjects than an increased rate of malignancies in general or an increase in a specific malignancy. Analyses of rash, hepatic, and CK-related AEs did not detect a specific safety signal associated with raltegravir and additional analyses pertaining to hepatic events are ongoing. Based on review of the available safety data, the benefits of raltegravir in HIV-1 treatment-experienced subjects outweigh the currently identified risks."
Raltegravir was found to readily cross blood-brain and blood-placental barriers.
Summary of Important Clinical Pharmacology Findings
The clinical pharmacology of raltegravir has been characterized in healthy and HIV-1 infected subjects, as well as in vitro studies using human biomaterials. The clinical pharmacology characteristics of raltegravir observed in these studies are summarized in the following sections.
Clinical Pharmacology Conclusions
1. Raltegravir exhibits high pharmacokinetic variability (range of geometric mean C12hr on 400 mg twice daily = 12 to 9151 nM in pivotal studies).
2. The potential sources of variability include: food, pH dependent solubility, UGT1A1 polymorphism, UGT1A1 expression and drug interactions.
3. Defining a clinically significant concentration threshold for potential dose adjustment is challenging because observed raltegravir plasma concentrations span over a 5-log range.
a. Within the concentration range studied, the virologic success rate is similar (77%) for patients with lower C12hr (median C12hr 76nM) compared to those with higher C12hr (median C12hr 1085 nM). This relationship needs careful interpretation in the presence of high within subject variability.
b. It is difficult to define the maximum safe raltegravir concentration because of the size of the current safety database at high exposure levels and the high pharmacokinetic variability
Pharmacokinetics (Absorption, Distribution, Metabolism, Excretion)
After oral administration of single doses of raltegravir in healthy subjects in the fasted state, raltegravir AUC0-Āá and Cmax are dose proportional over the dose range of 100 to 1600 mg. However, the variability is quite large (increasing with increasing dose levels), which implies a large degree of uncertainty in raltegravir exposure levels (See Assessment of Pharmacokinetic Variability section). In treatment na•ve HIV-1 infected patients who received raltegravir 400 mg twice daily monotherapy, raltegravir drug exposures were similar to exposures in healthy subjects.
The apparent terminal t1/2 of raltegravir is approximately 9 hours, with a shorter _-phase half-life (~1 hour) accounting for much of the AUC. The median time to maximum plasma concentration (Tmax) is ~3 hours in the fasted state. Steady state is achieved after two days of dosing at all dose levels.
Raltegravir is approximately 83% bound to human plasma proteins and is minimally distributed into red blood cells (blood-to-plasma partitioning ratio of 0.6). No data are available regarding human central nervous system (CNS) or brain penetration. Raltegravir is a substrate of human P-gp in vitro, which may limit CNS penetration in humans.
The results from a single dose study of 200 mg [14C] raltegravir given to young healthy subjects indicate hepatic clearance via glucuronidation plays a major role in the clearance of raltegravir in humans while renal clearance of unchanged drug is a minor pathway of elimination of raltegravir.
The in vitro metabolism of raltegravir was studied in human hepatic microsomes and hepatocytes. Data indicate glucuronidation of the parent compound to M2 is the major metabolic pathway in humans. Raltegravir is not a substrate of cytochrome P450 enzymes. Correlation and specific chemical inhibition studies in pooled human liver microsomes confirm the glucuronidation of raltegravir is mainly catalyzed by UGT1A1 with a minor contribution from UGT1A9 and 1A3.
UGT1A1 is a polymorphic enzyme. A single-dose, open-label study in healthy subjects with UGT1A1*1/*1 and UGT1A1*28/*28 genotypes is ongoing.
A high-fat meal, on average, resulted in a 19% increase in AUC, 34% decrease in Cmax, 750% increase in C12hr and 7.3 hour delay in Tmax with raltegravir final market image (FMI) formulation. However, the food effect is variable between subjects (See Assessment of Pharmacokinetic Variability section).
Based on the results from the high-fat meal study and the fact that raltegravir was dosed with or without food in Phase 2 and Phase 3 trials, raltegravir can be taken with or without food.
A study to investigate the effects of low, moderate, and high-fat meals on multiple dose pharmacokinetics of raltegravir in healthy volunteers is ongoing.
Assessment of Pharmacokinetic Variability
An attempt was made to understand the factors leading to variability in C12hr. As noted earlier, administration of raltegravir with a high fat meal was found to slow the rate of raltegravir absorption, causing a mean increase in C12hr of 750%. The effect of food on
raltegravir C12hr was variable between subjects. Because raltegravir dosing in pivotal studies was done without regard to food, over the course of the trials (Protocols 018 and 019), day-to-day variability was likely influenced by variability in food intake. In other words, a given patient could have 8 fold higher C12hr on a day when raltegravir was taken with food compared to days when raltegravir was taken without food. In addition to food, there are other determinants of raltegravir pharmacokinetics, such as, UGT1A1 polymorphism and drug interactions.
The Phase 1 drug interaction studies indicated atazanavir/ritonavir increased raltegravir C12hr by 77% and tipranavir/ritonavir decreased raltegravir C12hr by 55%. The mean changes in raltegravir C12hr due to atazanavir/ritonavir and tipranavir/ritonavir were similar between the Phase 1 studies and Protocols 018 and 019. However, because of the high variability in raltegravir concentrations, the range of raltegravir concentrations observed with or without either co-administered drug is similar.
The size of the current safety database at high raltegravir exposure levels and high variability make defining a clinically significant threshold for dose adjustment challenging. The applicant's proposals that raltegravir exposures spanning a 2-fold increase in AUC for safety and a 60% decrease in C12 hr for efficacy are not clinically relevant based on available clinical experience. The cut-off values are under review. Based on the applicant's rationale, a dose adjustment in the presence of atazanavir/ritonavir or tipranavir/ritonavir is not needed. Safety and efficacy data from Protocols 018 and 019 support the administration of raltegravir 400 mg twice daily with either tipranavir/ritonavir or atazanavir/ritonavir, with no dose adjustment. However, doubling of the dose to 800 mg twice daily is proposed in presence of rifampin, phenobarbital or phenytoin. Protocols 018 and 019 prohibited use of phenobarbital, phenytoin, rifabutin, and rifampin. When the protocols were amended, rifabutin (a less potent CYP3A, UGT1A1 inducer) was no longer prohibited. Dose adjustment is under review.
The effects of HIV status, age, gender, weight, and race on raltegravir pharmacokinetics were assessed by evaluation of raltegravir plasma trough concentrations in Phase 2/3 trials. The data indicate age, gender, weight, race and HIV status do not have an impact on raltegravir exposure. No clinically important effect of moderate hepatic insufficiency on the raltegravir pharmacokinetic profile was observed in a study of subjects with Child Pugh scores of 7 to 9. No dosage adjustment is recommended for patients with mild to moderate hepatic insufficiency. No clinically important effect of severe renal insufficiency on the raltegravir pharmacokinetic profile was observed in a study of subjects with 24-hour creatinine clearance of <30 mL/min/1.73 m2. No dosage adjustment is recommended for patients with renal insufficiency.
Resistance Development in Cell Culture
HIV-1 variants resistant to raltegravir were selected by serially passaging the laboratory HIV-1 isolate IIIB in H9 cells in the presence of increasing concentrations of raltegravir. A Q148K substitution in the HIV-1 IN coding region first emerged during selection and was followed sequentially by substitutions E138A, G140A, I208M, S230R, D10F and Y143C. Additional substitutions F181L and D279G were observed in a small number of clones.
The glutamine residue at position 148 is highly conserved among HIV-1 isolates and is located within the central core domain of IN containing the 3 active site amino acid residues D62, D116, and E152.
Phenotypic evaluations of these mutations using a single-cycle HIV-1 infection assay showed that the Q148K substitution conferred 46-fold reduced susceptibility in cell culture to raltegravir. Sequential addition of E138A and G140A substitutions increased overall resistance to 90-fold and 508-fold, respectively. The E138A substitution alone did not reduce susceptibility, while the G140A substitution and the E138A/G140A combination conferred 3-fold and 4-fold reduced susceptibility, respectively.
Thus, it appeared that the Q148K substitution is a primary contributor to resistance to raltegravir, and the E138A and G140A substitutions play a secondary role in augmenting resistance.
Clinical Resistance Analyses
In an as-treated analysis of the Phase 3 studies, paired amino acid sequences of HIV-1 IN from screening and on-treatment samples from 27 evaluable patients experiencing virologic failure on raltegravir were analyzed. A median of 3 (range 1 to 8) amino acid substitutions in HIV-1 IN were detected from the viruses of 26 patients. A total of 48 codons (16.7% of codons in the HIV-1 IN domain) were found to be mutated. Most were mutated once or twice. Seven amino acid changes were observed in 3 or more patients:
• 148 (Q148H/K/R)*
• 155 (N155H)*
• 92 (E92Q)
• 140 (G140A/S)
• 143 (Y143C/H/R)
• 151 (V151I)
• 230 (S230N/R)
These mutations were not observed in patients with virologic response to raltegravir treatment (4 patients from Studies 005 and 018) or with virologic failure to placebo therapy (12 patients from Study 005).
The N155H substitution was the most frequent mutation observed (11 patients [40.7%]) and conferred 13.2-fold resistance to raltegravir in cell culture. N155H was associated with E92Q (5 patients) and/or V151I (3 patients). The addition of E92Q, which by itself conferred 3-fold reduced susceptibility, increased resistance to 64-fold. V151I alone conferred no reduction in susceptibility to raltegravir.
Substitutions of Q148 with basic amino acid residues, arginine (R), histidine (H), or lysine (K) were noted in 7 patients (25.9%) and conferred 24-fold, 46-fold, and 27-fold resistance, respectively. Associated substitutions included E92Q (1 patient), G140A/S (4 patients), V151I (1 patient), and S230N/R (1 patient). Addition of G140A or G140S to Q148 variants substantially increased resistance to 257-fold and 521-fold, respectively. G140A and G140S alone conferred 3-fold and 2-fold reduced susceptibility, respectively.
Viruses from 4 patients harbored the Y143C/H/R substitutions in combination with either E92Q (2 patients) or S230N/R (2 patients). No phenotypic data of these mutations containing Y143C/H/R are currently available.
Protocol 005 (Phase 2 dose-ranging study) yielded a resistance profile of raltegravir similar to that of Protocols 018 and 019. Out of 50 evaluable patients experiencing
virologic failure to raltegravir treatment, key amino acid changes were observed at Q148 (27 patients), N155 (18 patients) and Y143 (2 patients). In addition to key changes at Q148 and N155, E92Q (2 patients), G140A/S (23 patients), V151I (5 patients), and S230N/R (6 patients) substitutions were also observed.
In summary, at least 2 major pathways, the Q148 pathway and the N155 pathway, appear to be involved independently in emergence of raltegravir resistance. Substitution of Q148 with any of the basic amino acids, H, K, or R, and the N155H substitution decreased susceptibility in cell culture to raltegravir 24- to 46-fold and 13-fold respectively. A third pathway is amino acid substitution at Y143 (Y143C/H/R). These substitutions were frequently found with additional amino acid changes.
The list of raltegravir resistance-associated substitutions observed to date includes L74M/R, E92Q, T97A, E138A/K, G140A/S, Y143C/H/R, Q148H/K/R, V151I, N155H, G163R, H183P, Y226C/D/F/H, S230N/R, and D232N.
Clinical Safety Results
A total of 902 HIV-infected subjects received at least one dose of raltegravir during the Phase 2 and Phase 3 studies at the time of the Safety Update Report (SUR, Frozen File date 2/16/07): 758 subjects by initial randomization, 138 subjects by switch from placebo to open label raltegravir after virologic failure, and 6 subjects by switch from placebo to open label raltegravir in the extension phase of Protocol 05. The proposed dose of 400 mg twice daily was received by 41 treatment-naive and 651 treatment-experienced subjects.
In general, in dose-finding treatment-naive Protocol 004 and dose-finding treatment-experienced Protocol 005, no relationship with dose and any adverse event was observed. Safety analyses of common adverse events (AE) and laboratory abnormalities pooled subjects from the Phase 2 and Phase 3 treatment-experienced studies receiving 400 mg raltegravir twice daily or placebo in combination with an optimized background regimen (OBT). The majority of AE analyses were limited to the double-blind treatment period to allow a more direct comparison among treatment arms. For some analyses, clinical AE data from the Safety Update Report (SUR, Frozen File data 2/16/07) was used to capture the most recent AE profile of raltegravir, given the limited duration of exposure in the current ongoing Phase 3 studies. This difference accounts for small discrepancies between Merck results and FDA results.
Clinical AEs were common in study subjects, occurring in >85% of all subjects receiving either 400 mg raltegravir twice daily or placebo. The majority of AEs were mild to moderate in intensity. The most common AEs occurring in > 10% were diarrhea, injection site reactions (due to enfuvirtide use), nausea, and headache, and were observed with similar frequency in each study arm (Appendix 1). Adverse events that occurred at
a higher frequency in raltegravir-treated subjects included: rash (5.3% versus 2.5%) and blood creatine phosphokinase increase (3.7% versus 1.1%).
Eighteen treatment-experienced subjects receiving either 400 mg raltegravir twice daily or placebo discontinued therapy because of adverse events (12, 2.4% versus 6, 2.1%). Overall, these 18 subjects reported 25 AEs as reasons for discontinuation, and 7 were considered at least possibly related to study drug. Raltegravir subjects discontinued due to hepatitis in the setting of bronchopneumonia; recurrent cryptococcal meningitis, hepatomegaly, and lactic acidosis, the latter attributed to concomitant NRTIs; renal failure in the setting of dehydration and concomitant tenofovir use; and flatulence. Placebo subjects discontinued due to lipoatrophy and nausea.
In Protocols 018 and 019, potential AIDS-defining conditions (ADC) identified by the investigator and/or sponsor were reviewed by an external adjudicator who was blinded to treatment assignment. A total of 32 subjects experienced 40 ADCs, 15 "presumptive" and 25 "definitive" diagnoses. The majority of ADCs occurred during the double-blind treatment period (N=34). During the double-blind treatment period, the incidence of ADCs was 4.1% (N=19) in the raltegravir arm and 6.3% (N=15) in the placebo arm (Appendix 2). Notably, the original NDA submission reported more ADCs in the raltegravir arm compared to placebo (3.0% versus 2.5%); however, with longer follow-up from the SUR, more ADCs were reported in the placebo arm than the raltegravir arm.
Further analyses of deaths, neoplasms, rash, hepatic and creatine kinase abnormalities are presented in more detail in the following sections.
A total of 16 deaths have occurred during treatment with study drugs up to the 2/16/07 database lock for the SUR and are summarized in Table 6. All deaths occurred in HIV-positive treatment-experienced adult subjects. Thirteen out of 595 raltegravir-randomized subjects and three out of 282 placebo-randomized subjects died. Adverse events leading to death occurred in the double-blind phase of the study in 11 subjects, in the open-label phase in 2 subjects, and in 1 subject each in the pre-treatment, post-study, and open-label post virologic failure phase.
As noted in Table 6, the majority of deaths were related to infections (N=10), and/or malignancy (N=4). Two deaths were related to cardiac disease and one death was due to suicide. In general, the causes of death were similar to those observed in clinical trials enrolling similar patient populations. No deaths were considered possibly related to raltegravir administration.
An analysis of baseline age, HIV RNA, and CD4+ cell counts was performed to compare the subjects who died to the randomized population (Table 7). Protocols 05, 018, and 019 were evaluated because these protocols enrolled similar populations. All subject deaths occurred in these protocols.
Subjects who died were generally sicker at baseline with higher baseline HIV RNA and lower baseline CD4+ cell counts. In addition, last on study CD4+ cell counts were lower.
An analysis of all-cause mortality in HIV treatment-experienced subjects was performed for the double-blind study period. Patients 3261, 3876, 7005 and 16254, all randomized to raltegravir, are excluded from this analysis because death occurred during open-label post virologic failure treatment.
The subjects enrolled in these trials were highly treatment-experienced patients, and the number of reported deaths does not appear to be in excess of the expected mortality rates in this population. The mortality rates observed in raltegravir clinical trials appear to fall within the range observed in other clinical trials enrolling similar patient populations. A relationship between study drug dose, duration, or other factors and the report of deaths among subjects in the safety population is not apparent.
At the time of database lock for the SUR, an imbalance was noted in rate of malignancies observed in raltegravir-treated subjects as compared to control/placebo-treated subjects. A total of 20 subjects experienced 21 malignant neoplasms through the SUR frozen file date. Twenty malignancies in 19 subjects occurred in raltegravir arms (including one subject who switched from placebo to open-label raltegravir, and two subjects from the expanded access program) and one in the efavirenz arm of Protocol 04 (squamous cell carcinoma of the vocal cord). No placebo-treated subject experienced a malignancy.
Malignancies reported in raltegravir-treated subjects were as follows:
• Squamous cell carcinoma: anogenital (4)
• Anal (1)
• Carcinoma in situ (CIS) (3)
• Lymphoma (4)
• Squamous cell carcinoma: other (4)
• Kaposi's sarcoma (3)
• Hodgkin's disease (2)
• Rectal cancer (1)
• Hepatic neoplasm malignant (1)
• Basal cell carcinoma (1)
Raltegravir-treated subjects with malignancies appear to have more advanced disease at baseline as evidenced by higher baseline HIV RNA (median HIV RNA 90,600 copies/mL versus 56,050 copies/mL in subjects with and without malignancy) and lower baseline CD4+ cell counts (median CD4+ cell count 34 cells/mm3 versus 140 cells/mm3 in subjects with and without malignancy). The malignancy rate for treatment-experienced subjects during the double-blind treatment period was 2.2% (13/595) in the raltegravir arm versus 0% in placebo. Adjusted for 395 patient-exposure years, the rate was 3.3 per 100 patient-exposure years.
Although an imbalance was observed initially in the malignancy rate between raltegravir arms and placebo/control arms, the overall malignancy rate observed in raltegravir-treated subjects was consistent with rates observed in other trials enrolling similar study populations. The identified malignancies are not unexpected in this heavily treatment-experienced HIV population, and no apparent pattern to the types of malignancies was observed. The initial imbalance appeared to reflect more a paucity of malignancies in control/placebo-treated subjects than an increased rate of malignancies in general or an increase in a specific malignancy.
Importantly, a more recent update of malignancies now shows similar rates of malignancies in raltegravir-treated subjects as compared to control/placebo-treated subjects. During the double-blind period 19 patients (2.5%) experienced 21 malignancies in the raltegravir arms and 5 patients (1.5%) experienced 6 malignancies in control/placebo arms. As calculated by the sponsor, the patient-year adjusted incidence rates are 2.32 and 1.92 per 100 patient-years for the raltegravir and control/placebo arms, respectively, resulting in a relative risk of 1.209 with an associated 95% confidence interval of (0.44, 4.14).
In the completed Phase 1 studies, there were 17 reports (5.1%, 17/334) of cutaneous adverse events that included the preferred terms dermatitis, pruritus, rash, rash maculo-papular, rash vesicular, and urticaria. None of these AEs resulted in study drug discontinuation and all were mild in intensity. Two (2) of the seven reports of rash and four of the five reports of pruritus were considered either "possibly" or "probably" drug- related by the investigator. All cases of dermatitis (3 reported) and urticaria (1 reported) were not considered by the investigator to be drug-related. Of the drug-related adverse experience reports of rash, one subject was taking 400 mg of efavirenz alone and the other subject was taking a combination of 400 mg of raltegravir, 500 mg of tipranavir, and 200 mg of ritonavir. Of the drug-related adverse experience reports of pruritus, all 4 subjects were taking raltegravir alone.
Protocol 029 is an ongoing open-label, sequential, 2-period study evaluating the safety, tolerability, and pharmacokinetics of multiple doses of raltegravir administered alone or with multiple does of darunavir and ritonavir. In Period 1, subjects receive 400 mg raltegravir twice daily for four (4) days, immediately followed by Period 2. In Period 2, the same subjects receive 400 mg raltegravir twice daily with 600 mg darunavir and 100 mg ritonavir twice daily for 12 days. At the time of the SUR, four discontinuations due to rash were reported. These four discontinuations were determined by the investigator to be "definitely" related to co-administration of darunavir, ritonavir, and raltegravir. All were Grade 2 (defined as diffuse macular, maculopapular, or morbilliform rash; and/or presence of target lesions), and all occurred during Period 2 after at least nine days of co-administration of darunavir, ritonavir, and raltegravir. One of the subjects who discontinued experienced an SAE. This subject completed Period 1 without complication and initiated Period 2; on Day 12 of darunavir, ritonavir, and raltegravir, the subject developed a diffuse maculopapular rash on the trunk and extremities associated with a temperature of 100.7 F. The subject was discontinued form the study. Skin biopsy showed superficial perivascular chronic inflammation with rare intravascular neutrophils consistent with a delayed hypersensitivity reaction. Rash was observed in healthy volunteer studies when darunavir was administered with other drugs (Source: Team Leader Memorandum NDA 21-897). Given the temporal relationship of rash onset to darunavir initiation, it is more likely that darunavir was the cause of rash.
No cases of Stevens-Johnson syndrome were found in the entire Phase 2 and 3 AE database. One case of erythema multiforme occurred in the efavirenz arm of Protocol 04.
Analyses of rash events were performed for the Phase 2 and Phase 3 studies using AE data at the time of the SUR. To allow more focused analyses, the following preferred terms were selected: exfoliative rash, rash, rash erythematous, rash follicular, rash generalized, rash macular, rash maculo-papular, rash papular, rash pruritic, rash vesicular, and drug eruption.
A total of 87 subjects experienced 91 rash events; none were SAEs. Eight subjects with rash events discontinued from study; seven subjects receiving raltegravir and one subject receiving efavirenz. In all cases, the dates of study discontinuation were greater than 60 days from the onset of the rash event, and the reason for discontinuation was not related to rash.
Four subjects interrupted study therapy due to rash: three subjects receiving raltegravir and one receiving placebo; however, all four subjects resumed study therapy.
The majority of rash events occurred during the double-blind treatment period (N=73); therefore, to allow a more direct comparison among treatment arms, the following analyses of rash events are limited to the double-blind treatment period.
The majority of rash events were mild/moderate in intensity. One subject in the raltegravir arm of Protocol 019 experienced a rash of severe intensity on Day 10 lasting 15 days. The OBT consisted of abacavir, efavirenz, and lamivudine. The rash was assessed by the investigator as probably not study drug related, and study drug and OBT were continued.
A total of 27 rash events were considered to be drug-related by the investigator. Drug-related was defined as definitely, probably, or possibly drug-related. The proportion of subjects with a drug-related rash in the raltegravir arms was 2.4% (18/755) versus 3.1% (10/320) in the placebo/comparator arms. Five of these rash events were considered related to OBT; the rash resolved in 3 subjects with discontinuation of a component of the OBT (fosamprenavir, enfuvirtide, and abacavir).
Twenty-three (23) additional rash events in 20 subjects occurred outside the double-blind treatment period with all subjects receiving raltegravir either in the extension/open-label phase in Protocols 04 and 05, in the interim phase in Protocol 04, or in the OLPVF phase. One subject experienced a second rash in the post-treatment period. None of the rashes were serious in intensity. Eight rash events were considered drug-related by the investigator, including one due to open-label raltegravir occurring 16 days after starting raltegravir with an unchanged OBT.
Six rash events were determined to be OBT-related by the investigator: three due to abacavir (two in the same subject separated by 23 days), one due to amoxicillin, one due to emtricitabine/tenofovir, and one due to delavirdine.
In summary, the majority of rash events in raltegravir-treated subjects were mild to moderate in intensity and no study discontinuations due to rash were reported in the Phase 2 and 3 development program. A clear pattern of rash has not been established and many of the rash events have been confounded by use of concomitant medications associated with rash such as darunavir, abacavir, and delavirdine. All reported rashes in drug-drug interaction Protocol 029, for example, occurred after darunavir was added to raltegravir. In an analysis limited to drug-related rash, no imbalance between the raltegravir and placebo/comparator arms was observed. Therefore, although rash events have occurred during treatment with raltegravir, no consistent pattern is observed and, in general, the events have not led to raltegravir discontinuation.
Analyses of hepatic events were performed for the Phase 2 and Phase 3 studies, limited to the double-blind treatment period. The following preferred terms were combined to define "hepatic event": abdominal pain upper, ascites, gastric varices, haematemesis, oesophageal varices haemorrhage, varices oesophageal, cytolytic hepatitis, hepatic
function abnormal, hepatic pain, hepatic steatosis, hepatitis, hepatitis acute, hepatitis toxic, hepatomegaly, hepatosplenomegaly, hepatotoxicity, hyperbilirubinemia, jaundice, liver tenderness, portal hypertension, portal hypertensive gastropathy, ALT increased, AST increased, blood alkaline phosphatase increased, blood bilirubin increased, blood unconjugated bilirubin increased, GGT increased, spleen palpable, hepatic neoplasm malignant.
A total of 129 subjects experienced 189 hepatic events. There was no apparent dose-response relationship; therefore, the raltegravir dose groups are combined.
A higher rate of laboratory-related hepatic events was reported in the raltegravir arm; however, the remainder of the hepatic AEs were balanced between the two groups.
Seven hepatic AEs occurring in five subjects were reported as SAEs, all occurred in the Phase 3 studies: one in the placebo arm (hepatitis toxic in the setting of tipranavir therapy) and four in the raltegravir arm (two subjects with hepatitis in the setting of pneumonia, one subject with history of hepatomegaly incidentally discovered to have portal hypertension and esophageal varices, and one subject with hepatocellular carcinoma attributed to hepatitis B). The subject diagnosed with hepatocellular carcinoma died.
Liver enzyme data submitted at the time of NDA submission were examined for hepatic abnormalities. Table 11 shows the rates of AST, ALT, alkaline phosphatase and bilirubin abnormalities from the raltegravir and placebo arms of the Phase 2 and Phase 3 studies.
Overall, the rates of liver enzyme elevations were similar between the raltegravir and placebo arms. A higher rate of Grade 3/4 total bilirubin was observed in the raltegravir arm. The majority of subjects with elevated total bilirubin levels had elevated indirect bilirubin (85.7%, 24/28), and all of these subjects were receiving atazanavir as part of the OBT. Additional analyses are ongoing to identify and define potential Hy's Law cases.
Increased Creatine Kinase:
An analysis was performed for elevated creatine kinase (CK) and associated musculoskeletal AEs. This analysis used data submitted at the time of NDA submission from Phase 2 and 3 studies, limited to the double-blind treatment period. A total of 63 subjects experienced Grade 2 - Grade 4 CK elevations, displayed in Table 12. For the purpose of this analysis, all raltegravir doses were combined as no dose-response relationship was observed for elevated CK levels.
Overall, there was a small increase in the rates of CK elevations in the raltegravir arms compared to the placebo arms.
The AE database for Protocols 2 and 3 was examined for potential AEs associated with elevated CK, including: arthralgia, myalgia, myositis, blood creatine phosphokinase increased, rhabdomyolysis, musculoskeletal pain, muscle fatigue, muscle strain. The following table reports the rates of elevated CK values, defined as Grade 1 or higher, and potential CK-related AEs. Of note, no AEs were associated with Grade 1 CK levels.
The association between CK elevations and the clinical AEs of myalgia, myositis, and arthralgia were balanced between the two groups. There were no reported SAEs or study discontinuations due to elevated CK levels. The "blood CPK increased" preferred term occurred at a higher rate in the raltegravir arms; however, this preferred term
categorization was determined by the investigator and the laboratory data provides a more accurate reflection of the CK data.
Overall, there was a modest increase in Grade 2 - 4 CK elevations in the raltegravir arm; however, association with clinical symptoms was balanced between the two groups. In addition, no SAEs or study discontinuations were associated with elevated CK levels.
In conclusion, raltegravir appeared to be well-tolerated in the Phase 2 and Phase 3 clinical studies with relatively few subjects discontinuing for adverse events. No clinically significant imbalance was observed in mortality rates and ADCs. A higher number of malignancies was observed in raltegravir-treated subjects initially; however, the imbalance appeared to reflect more a paucity of malignancies in control/placebo-treated subjects than an increased rate of malignancies in general or an increase in a specific malignancy. Analyses of rash, hepatic, and CK-related AEs did not detect a specific safety signal associated with raltegravir and additional analyses pertaining to hepatic events are ongoing. Based on review of the available safety data, the benefits of raltegravir in HIV-1 treatment-experienced subjects outweigh the currently identified risks.
Appendix 1: Most Common AEs by MedDRA Preferred Terms Reported in ≥
2% of Subjects Without Regard to Causality (Protocols 05, 018, 019 400 mg Twice Daily Raltegravir and Placebo arms), Frozen File Date 2/16/07