FDA Review Safety/Liver - Efficacy - New Cholesterol Drug Alirocumab
Download the PDF here
The central issue regarding this application revolves around the following question: For what population(s), if any, does the LDL-C-lowering benefit of alirocumab exceed its risks to support approval?
Taken together, regardless of how confident we may be in the "LDL hypothesis," we must remember that LDL-C remains a surrogate and not a clinical outcome that reflects how patients feel, function, or survive. This seems particularly relevant now that statins are widely regarded as first-line therapy, owing to their repeated successes in demonstrating benefit on CV outcomes, and now that the conduct of a CV outcomes trial is no longer considered the infeasible endeavor that it was when our advisory committee last discussed issues related to the pre-approval assessment of lipid-altering drugs approximately 25 years ago. I note that high-intensity statins (atorvastatin 40-80 mg; rosuvastatin 20-40 mg) yield LDL-C reductions ranging from -48% to -64%, according to their prescribing information, and some of these doses have proven benefit in CV outcomes trials. Alirocumab has not been tested (with respect to LDL-C lowering) against such statin doses. Even if it had been, however, the Division has expressed concern regarding an indication for monotherapy before benefit has been demonstrated on clinical outcomes in some population (e.g., patients already treated with statins).
One clinical scenario that may seem practical for monotherapy is the use in patients who cannot tolerate statins.
As of the cut-off date of this document, there were no cases of RNA confirmed Hepatitis
Summary of theoretical concerns for alirocumab exposure: Areas of particular theoretical concern for chronic administration of alirocumab were identified based on the very low plasma cholesterol levels attainable with PCSK9 inhibitor therapy, especially with coadministration of statins. These included possible increases in bile acid concentrations in the intestine that could cause tumors, increased risk of HCV infection, modulation of cholesterol-derived hormones, immune suppression (in adults), and similar to statins, risks for increased progression to type 2 diabetes and neurocognitive events. Overall, studies in animals administered alirocumab were reassuring regarding these theoretical concerns.
Theoretical risks have been identified with the PCSK9 inhibitors as a class. The following issues of potential (theoretical) concern have been identified; please refer to Dr. Elmore's review for further information:
Immunosuppression, especially when co-administered with HMG Co-A reductase inhibitors (statins). Immune cells (especially lymphocytes) are critically dependent on adequate membrane cholesterol concentrations. Co-administration of statins, which inhibit intracellular synthesis of cholesterol and are themselves immunomodulatory,18 could theoretically exacerbate the immunosuppressive potential of PCSK9 inhibitors.
Increased susceptibility to hepatitis C virus (HCV) infection. CD81, a critical component of the HCV receptor, is under negative regulation by PCSK9. Therefore, inhibition of PCSK9, by upregulating CD81 expression, might increase the availability of the HCV receptor, thereby increasing susceptibility to HCV infection.
Increased risk of colorectal cancer via increased intestinal bile acid load.
Alirocumab, by increasing the expression of LDL-R, increases hepatic uptake of cholesterol. Given that the primary route of elimination of cholesterol by hepatocytes is conversion to bile acids, treatment with alirocumab may increase the load of bile acids delivered to the intestines, especially in hypercholesterolemic patients. Increased intestinal secondary bile acid load has been shown to increase intestinal cancer risk in rodents.
Hepatitis C virus (HCV) infectivity: A study by Labonte et al. identified regulation of CD81 cell-surface protein expression by PCSK9 as a potential pathway by which PCSK9 inhibitors might cause increased susceptibility to HCV infection and associated liver tumors.16 CD81 is a co-receptor for hepatitis C infection in humans. Labonte showed that expression of PCSK9, especially a modified non-secretable form, reduced CD81 and LDLR levels in immortalized human cells and provided resistance to HCV infection in vitro. Alirocumab may therefore increase CD81 expression resulting in greater infectivity of HCV. The Applicant completed experiments with the more physiologically relevant soluble form of PCSK9. No regulation of CD81 was observed in vitro. The Applicant also evaluated CD81 expression in vivo in PCSK9-/-mice and in hyperlipidemic mice expressing human PCSK9 with reduced LDLR expression (PCSK9hum/hum/LDLR+/-) administered alirocumab; no changes in total CD81 levels were observed. No effects on HCV infectivity or replication kinetics were observed with addition of extracellular PCSK9 in the presence/absence of alirocumab in vitro. Based on conflicting data, the scientific database is currently considered inadequate to conclusively address the potential impact of alirocumab on HCV infectivity. It should be noted that increased cell-surface LDLR has also been postulated to modulate HCV infectivity; statins, like alirocumab, increase cell-surface LDLR.
Impaired liver regeneration: A concern for liver injury and alirocumab treatment comes from a published study conducted with the PCSK9 knockout mouse.17 When compared to littermates, PCSK9-null mice (but not PCSK9+/-mice) were markedly delayed in their ability to regenerate liver tissue following partial hepatectomy. Furthermore, the regenerating liver tissue exhibited necrotic foci. In these foci, the liver architecture was disrupted with swollen hepatocytes undergoing ballooning degeneration. Infiltration of red blood cells and leukocytes was also observed at the border of the necrotic areas. Whether this deficit is likely to be associated only with catastrophic liver injury (e.g., partial hepatectomy) or would also manifest following other liver injury (e.g., acetaminophen toxicity) is unknown. Of particular theoretical concern is the often transient, but sometimes severe liver injury induced by statins, which could theoretically be worsened by pharmacologically-induced loss of PCSK9 analogous to the PCSK9-/-mouse phenotype. However, liver toxicity was not exacerbated in a 3-month combination toxicity study with alirocumab when co-administered with atorvastatin in monkeys at doses that produced modest ALT increases and caused minimal to mild diffuse portal chronic inflammation, periductal chronic inflammation, and biliary ductular proliferation, but in the absence of severe toxicity. It is unknown whether recovery from more serious liver damage would be impacted by PCSK9 inhibitor therapies, including alirocumab.
One possible explanation for the failure of liver to properly regenerate in PCSK9 knockout mice owes to the discovery that HDL-C concentrations regulate bone marrow-derived endothelial progenitor cells18, including the precursors of liver sinusoidal cells. Upon significant liver damage, liver regeneration is dependent upon endothelial progenitor cells to migrate from the bone marrow to the liver, where they are responsible for directing repair of damaged hepatic blood vessels and tissues. This process is required for proper liver repair after partial hepatectomy. 19 It is tempting to speculate that low HDL-C impaired the production and migration of progenitor cells to direct liver regeneration in PCSK9-/-mice. Rats administered alirocumab had low HDL and sinusoidal cell defects were observed in the liver. This phenomenon was not observed in the monkey with alirocumab, even where HDL-C levels were greatly reduced. HDL-C was not reduced in humans administered alirocumab, which may argue that defects in liver regeneration observed in partially hepatectomized PCSK9 knockout mice have limited clinical relevance.
Immune modulation in adult animals: Inhibition of PCSK9 produces profound lowering of circulating cholesterol. The immune system is dependent on cholesterol for proper function. Clonal expansion of rapidly dividing immune cells (e.g., B-cells, T-cells, etc.) and cell-cell signaling are heavily dependent on cholesterol and cholesterol derivatives.20 However, no effects on immune cell populations, T-cell dependent antibody response, natural killer cell activity, or cytotoxic T-cell activity were observed in adult monkeys administered alirocumab or a combination of alirocumab and atorvastatin. No studies designed specifically to challenge the immune system (e.g., introduction of an infectious agent) were conducted, although no imbalances for infections were observed in general toxicity studies in monkeys or rats. Overall, these data indicate that alirocumab is unlikely to affect the immune system in adult patients.
Neurocognitive assessments: Adverse neurocognitive events (e.g., transient confusion and memory loss) have been described, primarily through patient reporting in adults on chronic statin therapy. Cholesterol and other sterols are important for nerve function in both the central and peripheral nervous system. The brain is a cholesterol-rich organ, which depends almost completely on de novo cholesterol biosynthesis for its sterols; peripheral blood lipids are unavailable to the CNS, due to blockade by the blood-brain-barrier. PCSK9 is highly expressed in brain tissues, although its function
there is uncertain.23 Alirocumab is a 150 kDa immunoglobulin, with very low access to the brain. Therefore, it is considered unlikely that alirocumab could directly affect the structure or function of the CNS, but this leaves the possibility of effects on peripheral neurons. No significant alirocumab-related effects on peripheral neurons were observed in toxicity studies of up to 6 months duration in rats and monkeys. The lack of neurological symptoms in animal models is reassuring.
Review of Safety
Within this pool, 1999 (81%) patients were exposed to alirocumab for at least 1 year. The mean duration of exposure was 58 weeks.
Treatment groups within the placebo-controlled (alirocumab versus placebo) and ezetimibe-controlled pools (alirocumab versus ezetimibe) were well matched for demographics and baseline characteristics. In the placebo-controlled pool the mean age was 59 years, 40% were women, 90% were Caucasian, and 30% participated at
U.S. sites. In the ezetimibe-controlled pool the mean age was 62 years, 35% women, 87% Caucasian, and 50% participated at U.S sites. The majority of patients in both the placebo-controlled and ezetimibe-controlled pools had a history of CHD (60 to 70%) - with almost half of patients in both the placebo-controlled pool and ezetimibe-controlled pool reporting a coronary revascularization procedure and approximately a third of patients reporting a history of a myocardial infarction.
On background of maximally tolerated statin therapy in placebo-controlled studies, treatment with alirocumab was associated with a higher percentage of patients reporting events within the SMQ "hepatic disorders" (1.8% placebo, 2.5% alirocumab). These events were primarily associated with abnormal hepatic laboratory values. Evaluation of pre-specified categorical changes in ALT defined as ≥3x ULN (if baseline ALT < ULN) or twice baseline (if baseline ALT ≥ULN) demonstrated a slightly higher percentage of alirocumab-treated patients with this shift in ALT versus either placebo or ezetimibetreated patients, however larger increases in (ALT >5x ULN or >10x ULN) were similar between treatment groups. There was a higher incidence of serious hepatic disorders associated with alirocumab treatment, with the majority associated with elevations of
liver transaminases. Review of these cases suggested alternative etiologies such as hepatitis or concomitant medications as potential causative factors. Of the serious adverse events in which alirocumab was temporarily discontinued due to elevations in ALT, subsequent re-initiation of treatment resulted in negative rechallenge with the exception of 1 case of positive rechallenge(in this reviewer's opinion) with mild elevations in ALT that ultimately did not result in treatment discontinuation. An additional patient experienced an elevation in ALT that resolved with discontinuation of alirocumab, but experienced elevated ALT with rechallenge that led to permanent discontinuation of treatment. There were 3 cases in alirocumab-treated patients that met the biochemical criteria for Hy's Law - however all had alternative etiologies of hepatitis A, cholecystitis, and cholangitis, respectively.
ALT increase and hepatic disorders
The following analyses used the standardized MedDRA query for "hepatic disorders".
A higher proportion of alirocumab-treated patients reported treatment-emergent AEs, SAEs, and discontinuation due to AEs related to hepatic disorders compared to placebo-treated patients. All patients in the placebo-controlled pool were on maximally tolerated background statin therapy.
Within the ezetimibe-control pool, TEAEs did not occur at a higher incidence in the alirocumab-treated group compared to the ezetimibe-treated group. Only 1 patient experienced an SAE (alirocumab-treated) and a slightly higher percentage of alirocumab-treated patients discontinued treatment due to a hepatic disorder.
Selected SAE narratives
- 011717-380-001-003/Alirocumab/ALT increase: 36 year old white man with fatty liver and previous history of ALT increase several years prior to this study. The patient experienced a mild ALT increase 28 days after receiving first alirocumab injection which progressed to a peak ALT of 233 (5.4x ULN) and AST 162 (4.5x ULN), alkaline phosphatase and bilirubin were within normal limits.
Viral serologies were negative. Liver ultrasound confirmed fatty liver infiltration. Alirocumab was temporarily interrupted and liver function tests returned to baseline. Simvastatin and ezetimibe as background therapy were continued throughout this event. After reintroduction of alirocumab, this patient had five additional episodes of increased ALT, no further action regarding the study drug was taken, and the patient completed the study. Since submission of the BLA, the serious event was downgraded to non-serious by the Investigator.
- 011717-250-009-007/Alirocumab/Hepatocellular injury: A 60-year-old female patient (baseline BMI 27.6 kg/m2), with a relevant medical history of coronary artery disease including acute MI, no hepatic disorders or alcohol consumption, experienced hepatic cytolysis on Day 29. ALT and AST increased up to 8.2 ULN (280 IU/L) and 4.9 ULN (168 IU/L), respectively. Bilirubin levels were normal. Alirocumab was temporarily interrupted. The patient recovered without corrective treatment and alirocumab was resumed without recurrence of ALT increase. The patient received ciprofloxacin up to 3 weeks before randomization and the combination of tiliquinol/ tilbriquinol in the month before and in the first week after randomization for the treatment of acute sigmoiditis that started about two months prior to randomization. These drugs are listed to have a risk of liver toxicity and may represent potential alternative cause for the transaminase increases. The ALT values at screening and at baseline were moderately high, 1.8 ULN and 2.6 ULN, respectively as well as an aspartate aminotransferase (AST) value of 1.7ULN at baseline (Day 1).
- 012492-203-405-001/Alirocumab/hepatic enzyme increased: A 67-year-old female patient (baseline BMI 29.6 kg/m2), with a relevant medical history of hypertension and peripheral arterial occlusive disease, experienced hepatic enzyme increased of severe intensity on Day 169 following a viral infection. Liver function tests (LFTs) on Day 169 showed elevated ALT at 129 IU/L (3.8 ULN) with high AST at 133 IU/L (3.91 ULN) and normal total bilirubin; baseline ALT was 54 IU/L (1.59 ULN). On Day 171 ALT values increased to 313 IU/L with high AST at 229 IU/L
(6.74 ULN) and high total bilirubin at 24 μmol/L. On Day 176, ALT values increased to 539 IU/L
(15.85 ULN) with AST at 438 IU/L (12.88 ULN), elevated alkaline phosphatase (ALP) 813 IU/L
(6.61 ULN), and gamma-glutamyl transferase (GGT) 528 IU/L (10.56 ULN). Ultrasonography showed mild nonserious cholesterolosis of the gallbladder. Serology results were negative for hepatitis B, C, and A. No specific corrective treatment was required but paracetamol, ibuprofen, rosuvastatin and ezetimibe were interrupted. The IMP was permanently discontinued, with the last administration on 15-MAY-2013. The patient recovered from the event without sequelae on 06-JUN-2013. On that date, alkaline phosphatase was 275 U/L (2.24 ULN), AST was 23 U/L
(0.68 ULN), and ALT was 31 U/L (0.91 ULN). On 07-AUG-2013, approximately one month after rosuvastatin and ezetimibe were re-started, ALT, AST, and alkaline phosphate were all within normal range.
Discontinuations due to hepatic disorders
A higher percentage of patients treated with alirocumab discontinued treatment due to a treatment-emergent event related to a hepatic disorder. Of the 9 alirocumab-treated patients who discontinued due to a hepatic disorder event, 8 discontinued due to abnormal hepatic laboratory values.
One discontinuation is an example of a positive rechallenge in a 42-year-old female patient (baseline BMI 36.3 kg/m2) with a medical history of obesity experienced ALT
significant evidence of obstruction, and a rise in bilirubin to 2x ULN in the absence of concurrent hepatic infection, hepatotoxic drugs, or injury.
There were 3 incidences of elevated transaminase levels with total bilirubin elevated 2x ULN (1 alirocumab-treated, 2 placebo-treated), which are summarized below. These cases of concomitant increases in liver transaminases and bilirubin do not qualify as Hy's Law cases based on the alternative etiologies of hepatitis A, cholangitis, or acute cholecystitis.
- 001112-528-202-003: A 50-year-old male patient (baseline BMI 30.2 kg/m2), with a history of type 2 diabetes mellitus and daily consumption of alcohol, experienced ALT increase on Day 1 due to hepatitis A (PT: hepatitis A) of mild intensity. IMP was temporarily discontinued. AST reached peak levels on Day 41 at 30.03 x ULN, ALT reached peak levels on Day 50 at 48.7 x ULN, and bilirubin reached peak levels on Day 46 at 7.38 x ULN. The patient was diagnosed with hepatitis A, reported as a treatment-emergent SAE and leading to a temporary drug withdrawal. The patient recovered from the event without sequelae.
- 12492-840-428-001: A 48-year-old male patient had increased ALT (12.36 ULN), AST (17.9 ULN), and total bilirubin (3 ULN) on Day 144. The patient was diagnosed with acute cholecystitis and possible choledocholithiasis reported as a treatment-emergent SAE
- 11717-124-008-007: A 68-year-old female patient, with a history of cholelithiasis and cholecystectomy, had increased total bilirubin (2.78 ULN) on Day 75, and then increased ALT
(8.75 ULN) and AST (6.38 ULN) on Day 136. The patient was diagnosed with a cholangitis reported as a treatment-emergent SAE. Corrective treatment included piperacillin/tazobactam. The IMP (placebo) was temporarily interrupted due to the event of cholangitis. The patient recovered.