To better understand the nature of the evidence obtained from many phase 2 trials and the contributions of phase 3 trials of drugs, vaccines, and medical devices, the FDA has studied 22 recent cases in which promising phase 2 clinical trial results were not confirmed in phase 3 clinical testing. Phase 3 studies did not confirm phase 2 findings of efficacy in 14 cases, safety in one case, and both safety and efficacy in seven cases. These unexpected results occurred even when the phase 2 study was relatively large and even when the phase 2 trials assessed clinical outcomes.
“Some on the libertarian side have said they want to simply toss out the gold standard on efficacy and safety altogether, excoriating the old rules of market engagement,” said columnist John Carroll, writing in Endpoint News. “That toxic argument raises some important questions about each stage of drug development. And there’s a new report out from the FDA that covers just what you can miss if you cut phase 3 out altogether.”
The FDA’s case studies demonstrate that large phase 3 randomized controlled trials can generate critical evidence across all types of products, patients, and diseases, according to the agency. Both safety and efficacy failures occurred even when the phase 2 studies were relatively large, and even when the product was already approved for another condition. In some cases, the phase 3 study revealed that short-term results found in the phase 2 study were not associated with a long-term benefit or that the product had toxicity that was not uncovered in the phase 2 study.
The drugs that showed promise in phase 2 trials but failed to demonstrate efficacy and/or safety in phase 3 studies include the following:
Phase 3 Trials Demonstrating Lack of Efficacy
- Bitopertin (Roche) as an add-on treatment for schizophrenia. The drug failed to improve negative symptoms of the disease.
- Brivanib (Bristol-Myers Squibb) for the treatment of patients with hepatocellular carcinoma. The drug failed to improve overall survival compared with approved treatment and demonstrated unexpected toxicities.
- Capsaicin topical patch (Qutenza, NeurogesX) for the treatment of nerve pain associated with human immunodeficiency virus infection, with previous approval for the treatment of shingles-associated nerve pain. Late-stage studies showed a lack of efficacy compared with control groups.
- Darapladib (GlaxoSmithKline) as an add-on to a statin for the prevention of cardiovascular disease (CVD) complications in patients with a prior heart attack. The drug failed to reduce the risk of heart attack or cardiac death compared with placebo in patients with chronic CVD.
- Dexmecamylamine (Targacept/AstraZeneca) as an add-on treatment for depression. In a phase 3 study, the drug was no more effective than placebo.
- Exhale drug-eluting stent (Bronchus Technologies) for the reduction of shortness of breath in patients with emphysema. The stent showed statistically significant results on measures of lung function and symptoms in phase 2 studies, but failed to improve these parameters in a phase 3 trial.
- HSV-2 vaccine (Novartis) to prevent genital herpes. Despite positive biomarker results in phase 2, the vaccine did not prevent genital herpes in phase 3.
- Glutamic acid decarboxylase vaccine (Diamyd Medical) for the preservation of insulin secretion in patients with recent-onset type-1 diabetes. Despite promising biomarker results in phase 2, the vaccine did not improve pancreatic function or clinical outcomes in a phase 3 study.
- Imiquimod (Aldara 5.0% cream, 3M) for the treatment of molluscum contagiosum lesions in children. Despite efficacy in another viral skin infection and promising phase 2 results, treatment with imiquimod cream was no more effective than placebo in a phase 3 trial.
- Iniparib (Sanofi) as an add-on treatment for patients with triple-negative breast cancer. In a phase 3 trial in which iniparib was added to an established chemotherapy regimen, the treatment did not improve survival.
- Lithium (King’s College London) as an add-on treatment to delay the progression of amyotrophic lateral sclerosis. The treatment did not improve survival, health status, or quality of life in a phase 3 study.
- MAGE-A3 vaccine (GlaxoSmithKline) for the treatment of patients with non–small-cell lung cancer (NSCLC) after surgery. The vaccine showed no clinical benefit compared with placebo in a phase 3 trial.
- NicVAX vaccine (Nabi Pharmaceuticals) for smoking cessation. In phase 3 trials the abstinence rate in the NicVAX group was similar to that in the placebo group.
- Velimogene aliplasmid (Allovectin-7, Vical) for the treatment of patients with metastatic melanoma. In a phase 3 trial, velimogene aliplasmid reduced tumor size in significantly fewer patients with late-stage disease compared with two marketed therapies.
Phase 3 Trial Demonstrating Lack of Safety
- Olanzapine pamoate (Zyprexa Relprevv, Eli Lilly) for long-acting injection treatment of patients with schizophrenia. Although approved as an oral short-acting treatment for these patients, phase 3 studies identified a serious safety risk of the long-acting formulation, requiring safety monitoring.
Phase 3 Trials Demonstrating Lack of Efficacy and Safety
- Aliskiren (Rasilez, Tekturna, Novartis) as add-on treatment for the prevention of congestive heart failure (CHF) complications. In a phase 3 trial, adding aliskiren to standard therapy did not reduce CVD or CHF rehospitalization after discharge and increased the rates of kidney failure and hypotension.
- CoStar drug-eluting stent (Conor Medsystems) to reduce the risk of heart attack in patients with coronary artery disease. In a phase 3 study, patients who received a CoStar stent had worse outcomes than those who received a different stent.
- Figitumumab (Pfizer) as an add-on treatment for patients with advanced NSCLC. Showed promise in Phase 2, but in Phase 3 failed to improve survival. Also, in one combination with another regimen, increased serious adverse events and deaths.
- Recombinant factor VIIa (NovoSeven, Novo Nordisk) to reduce intracerebral bleeding and hematoma size in stroke patients. Although the product is approved for patients with hemophilia, patients with intracerebral bleeding who received recombinant factor VIIa experienced no clinical benefits and had an increased incidence of serious adverse events compared with those given placebo.
- Semagacestat (Eli Lilly) to improve cognitive and functional status in patients with Alzheimer's disease. A phase 3 trial was terminated early because patients who received semagacestat had worsened cognitive and functional status and an increased risk of skin cancer compared with patients given placebo.
- Torcetrapib (Pfizer) for the prevention of cardiovascular events in patients with a history of CVD or type-2 diabetes. In a phase 3 study, torcetrapib increased mortality and cardiac events compared with placebo in patients at high cardiovascular risk.
- V710 vaccine (Valneva/Merck) to prevent Staphylococcus aureus infection. A phase 3 trial of this vaccine was terminated because of lack of efficacy. The vaccine also showed a potential risk for serious adverse events and death.
Sources: FDA Report; January 2017; and Endpoint News; January 20, 2017.