Abstract
Background: The accelerated approval (AA) regulation (21CFR314.510 Subpart H) is granted by the Food and Drug Administration (FDA) when drugs for serious medical illnesses are shown to be an improvement over available therapy. AA provides an option to use surrogate outcomes considered likely to predict clinical benefit. AA was initially developed to hasten access to HIV drugs, then, in 1995, AA was extended to cancer indications. Sponsors receiving AA are required to confirm clinical benefit (termed subpart H trials). Policy makers have several raised concerns: AA is no longer relevant today as the approval bar via this mechanism has been raised too high; many drugs that received AA did not complete subpart H trials; and some drugs approved by AA were subsequently found to be unsafe or ineffective.
Methods: Using publicly available databases, we compared safety, efficacy, clinical development times, and subpart H completion rates for FDA-approved new molecular entities (NMEs) for hematologic and solid tumor cancers from 1995–2008.
Results: 37% of all oncology NMEs received AA versus regular approval (64% during 1995–2003 and 33% during 2004–2008). Twenty oncology NMEs received FDA approval for hematologic malignancies (lymphomas, leukemias, Kaposi’s sarcoma, and myelodysplastic syndromes), accounting for 34% of regular approvals and 53% of AAs for oncology NMEs. Compared to NMEs approved for solid tumors, NMEs approved for hematologic malignancies were more likely to involve Orphan Drug indications (95% vs. 32%); to have shorter development times, defined as the interval between investigational new drug filing and marketing approval, (median 5.6 vs. 7.8 years); and to be approved based on phase II studies (65% vs. 29%). Prior to 2004, development times were similar for solid tumor and hematologic malignancy NMEs. Since 2004, development times have decreased by more than 2 years for hematologic malignancy NMEs, but not for solid tumor NMEs. 50% of NMEs approved for hematologic malignancies versus 71% of NMEs for solid tumor diagnoses are included in first-line cancer regimens in current National Comprehensive Cancer Network guidelines. Drugs approved for solid tumor and hematologic indications have similar safety profiles and efficacy; respectively, 30% and 38% carried black box warnings at initial approval, and 15% and 10% had black box warnings added post-approval. Studies confirming efficacy were completed for 89% of NMEs receiving AA for solid tumor indications versus 30% for NMEs receiving AA for hematologic malignancy indications. Concern that sponsors are not completing subpart H commitments has led the FDA to move from basing AA on final results of single-arm phase II trials to interim results of phase III trials.
Conclusions: AAs for hematologic malignancy indications are less likely to complete Subpart H commitments. In the current era, development times for NMEs are shorter for hematologic malignancy versus solid tumor indications, principally related to the approval based on Phase II versus Phase III studies. Establishing a global policy that AA approval for cancer drugs should be based on interim results of phase III analyses rather than on final analyses of phase II trials may hamper development of novel therapies for hematologic malignancies.
Solide tumor indications Hematologic malignancy indications 1995–2003 (n=21) 2004–2008 (n=10) 1995–2003 (n=11) 2004–2008 (n=9) Drugs receivving AA (%) 29 30 64 33 Orphan drugs (%) 24 40 100 78 Of AAs, Subpart H completion (%) 100 66 27 0 Approval based on phase II trial (%) 33 0 73 78 Median development time (years) 8.4 7.8 8.8 5.2
Bias, benefit, or both: evaluating new glioma therapies
