Clinical Benefit of Oncology Agents in the Accelerated Approval Era: Broken Promises or Field of Dreams?

Christine Cambareri, PharmD BCOP BCPS
Clinical Pharmacy Specialist in Hematology/Oncology
Hospital of the University of Pennsylvania
Philadelphia, PA

In the past 3 years, the U.S. Food and Drug Administration (FDA) has given 73 oncology-related approvals, 30% of which were under accelerated approval.1 The FDA instituted its Accelerated Approv- al Program (AAP) in 1992 to hasten the approval process and enhance access to lifesaving therapies to treat serious conditions and fill an unmet medical need based on a surrogate end point. A surrogate end point is defined by the FDA as “a marker such as a laboratory measurement, radiographic image, physical sign, or other measures thought to predict a clinical benefit, but … not itself a measure of clinical benefit.”2 The realm of cancer care is most definitely encompassed in the AAP, with more than a million new cases of cancer in 2016 and more than half a million deaths related to cancer in 2016.3

In oncology, overall survival (OS) is the gold-standard end point for clinical efficacy. However, determining the true OS benefit of a drug can take years and can require a large sample size, delaying access to that drug by a patient population in need. A surrogate end point can overcome the barriers of measuring OS.4 With the use of surrogate end points, the work of the AAP has hastened cancer patients’ access to medications and is thought to have aided in improving the 5-year OS rates of all cancer types from 50% in 1980 to approximately 70% in 2008.3

The pillars of drug safety and efficacy date to 1962, with the Kefauver-Harris Amendment of the 1938 Food, Drug and Cosmetic Act. With the awareness that patient needs may accelerate more quickly than the drug approval process, the FDA allowed early- access programs in the 1960s to offer limited patient access to investigational drugs, also known as compassionate-use programs. In the 1980s, the FDA created a fast-track component to its rules to allow for expedited development of agents being studied for serious and life-threatening conditions by, for example, eliminating phase 3 trials from the initial approval phase. This allowed faster medication access to patients in need while the postapproval phase 3 and phase 4 studies were being completed for official FDA approval. This change reduced the average clinical development time from 8.9 to 6.2 years. The enactment of the AAP in 1992 allowed approval based on various surrogate end points, like progression-free survival (PFS) and objective response rates, that were likely to predict OS benefit and shortened the average clinical development time to 4.2 years, with a correlated shortening of new drug application review times from more than 30 months in the 1980s to as little as 9.9 months in 2011.5

Despite these advances, the AAP has also caused a fair share of turmoil. One such example of the controversial use of surrogate end points was the use of bevacizumab for metastatic breast cancer. Bevacizumab in combination with paclitaxel was approved for HER2-negative, treatment-naive metastatic breast cancer in 2008. The accelerated approval was based on a statistically significant 5.9-month increase in the median PFS and increased objective response rates (36.9% vs. 21.2%) for the combination of paclitaxel and bevacizumab over paclitaxel alone in the E2100 study.6 However, the confirmatory phase 4 trials (AVADO and RIBBON-1) and maturation of the E2100 data painted a different picture from what was originally predicted and threatened the regular FDA approval of bevacizumab for metastatic breast cancer.7,8 Ultimately, the results of all three trials did not show an OS benefit. The AVADO trial even demonstrated a smaller benefit in median PFS with bevacizumab 7.5 mg/kg (9 months vs. 8.1 months, p = .045) and 15 mg/kg (10 months vs. 8.1 months, p < .001) than originally proposed. In addition, all three studies showed increased toxicity in the bevacizumab-containing treatment arms, with the most common side effects being hypertension, proteinuria, cerebrovas- cular ischemia, all-grade bleeding, and neutropenia.6-8 In response to these findings, the FDA approval was withdrawn for bevacizum- ab in metastatic breast cancer.9

Much debate still surrounds the accurate and appropriate use of surrogate end points, and discussion continues regarding a lack of translation to OS benefit when confirmatory data matures. Despite this concern, the number of trials using surrogate end points has increased. According to the Journal of Clinical Oncology, the proportion of randomized controlled trials of systemic therapy published from 1975 to 2009 using PFS as an end point increased from 0% to 26% in breast, colorectal, and non–small cell lung cancer.7 However, correlation does not necessarily imply causation, or in the case of oncology, surrogacy as described in the example of bevacizumab. It has been postulated that reasons for this discrepancy are multifold, including these: the changes in tumor size needed for progression are too small to affect OS; studies are underpowered to detect an improvement in OS by the same absolute amount as PFS; date of progression is more difficult to measure and capture accurately than date of death; and available studies often allow for crossover, and thus the sequence of administration—not the impact of the new treatment—is being evaluated.10

Although the initial approval under the AAP is based on a surrogate end point, this program places the burden of proof for clinical benefit on the drug companies. Unfortunately, the manufacturers’ diligence in pursuing these phase 4 confirmatory trials is lackluster at best. In a 2009 government accountability report, the agency was criticized for failing to enforce postmarketing study commitments for surrogate approvals.11 As of 2011, postmarketing study commitments for more than 40% of drugs approved through the AAP had not yet been started. Furthermore, according to the Office of Oncology Drug Products, the completion time of these studies has ranged widely from 0.8 years to 12.6 years.5 However, despite these shortcomings, the FDA has not removed the drugs involved from the market.11

Notwithstanding the concerns related to the impact of early access to therapy, shortened development time, and faster review times of agents, the FDA has yet to tighten the reins on the AAP program. In fact, in 2012, a new pathway was sanctioned, referred to as “breakthrough therapy (BT).”12 To be eligible for this des- ignation, drugs must have “an effect on a pharmacodynamic biomarker that does not meet criteria for an acceptable surrogate endpoint, but strongly suggests the po- tential for a clinically meaningful effect on the underlying disease.”12 Drugs approved through BT status must also be eventually approved or rejected under the normal FDA approval standards; however, as observed in confirmatory trials following AAP approval, this confirmation may not be required for several years.5 Drugs that have recently received BT approval include pembrolizumab, nivolumab, daratumum- ab, elotuzumab, and atezolizumab.13

Through use of the BT and AAP path- ways, the FDA runs the risk of allowing products that are clinically ineffective or unsafe, or possibly both, into the market for a period of time before confirmatory data mature. The impact of these pathways was most recently evaluated in the 2016 analysis by Salas-Vega and colleagues, pub- lished in JAMA Oncology, which reviewed the OS, quality of life (QoL), and safety of 53 new FDA and European Medical Agency (EMA) molecular entities with primary oncology indications approved between 2003 and 2013 as evaluated by English, French, and Australian health technology assessments. It was found that 43% of the drugs increased OS by 3 months or longer, 11% by less than 3 months, and 15% by an unknown magnitude. The remaining 30% of cancer drugs did not demonstrate an increase in OS compared with alternative treatments, either because no difference was found or a determination could not be made on the basis of the available evidence at the time. Where OS could be quantified, it was determined that a total mean improvement in OS of 3.43 months was attained related to the treatments that were available in 2003. However, the greatest benefits in OS were seen in breast, renal, and skin cancers, with little or no benefit in OS observed in thyroid, lung, and he- matological cancers. In addition, benefits were concentrated among specific classes of agents, most notably immunologic therapies, which were better at extending OS compared with nonimmunologic drugs (5 months vs. 2.3 months).14

With regard to QoL, it was determined that 42% of the drugs evaluated improved this parameter, 4% reduced it, and 2% were associated with mixed evidence. Notably, 53% did not demonstrate a difference in QoL relative to the best alternative treatments available at the time. However, not all of these opinions were based on empiric evidence with validated QoL instruments. For five drugs (pertuzumab, trametinib, zivaflibercept, sipuleucel-T, and vemurafenib), testimony from patient representatives and clinical experts was used to quantify QoL benefits.14

Regarding the safety assessments associated with these new therapies, it was found that 24 of the 53 drugs (45%) reduced patient safety, as evaluated by the following parameters: incidence of adverse events (AEs), incidence of severe or serious AEs, time to first AE greater than grade 3, treatment discontinuation or dose reduction, overall tolerance and safety profile (not otherwise specified), treatment-related deaths, and input from patient representatives or clinical experts. Of the remaining 29 drugs evaluated, 15% were found to improve safety, 19% had mixed evidence with regard to safety outcomes, and 21% did not demonstrate any difference in safety compared with alternative treatments available at the time of approval.14

The clinical benefit of these agents can be quantified in this way: of the 23 drugs that increased OS by at least 3 months, 65% were found to improve QoL, but 48% reduced patient safety. Salas-Vega and colleagues concluded that gains in OS and QoL often come at the cost of safety.14

The enactment of the AAP has undoubtedly contributed to many advances in cancer care during the last 25 years. However, the magnitude of benefit varies widely and must be considered when one is making treatment decisions, given that one in three newly approved cancer medications has not been associated with an OS benefit.14 Furthermore, in light of the recent BT designation, clinicians must remain steadfastly focused on evidence- based practices, seek clinical benefit data when they are available, and ensure that safety and quality of care are not being sacrificed during expedited approval. Finally, an area not addressed by the FDA or available literature is the importance of education and empowerment of patients with regard to the different ways that agents are approved and brought to market. In order to make well-informed shared decisions with their treatment team, patients should be made aware of whether a given therapy improved overall survival and QoL or reduced safety.


1. U.S. Food and Drug Administration. Hematology/Oncology (Cancer) Approvals and Safety Notifications. Accessed January 22, 2017.

2. U.S. Food and Drug Administration. Accelerated Approval Program. Accessed January 22, 2017.

3. National Cancer Institute. Cancer Stat Facts: Cancer of Any Site. Surveillance, Epidemiology and End Results Program. https://seer.cancer. gov/statfacts/html/all.html. Accessed January 22, 2017.

4. U.S. Food and Drug Administration. Accelerated Approval. https://ww Accessed January 25, 2017.

5. Darrow JJ, Avorn J, Kesselheim AS. New FDA breakthrough-drug category—implications for patients. N Engl J Med. 2014 Jul 3;371(1):89-90.

6. Miller K, Wang M, Gralow J, et al. Paclitaxel plus bevacizumab versus paclitaxel alone for metastatic breast cancer. N Engl J Med. 2007 Dec 27;357(26):2666-76.

7. Miles DW, Chan A, Dirix LY, et al. Phase III study of bevacizumab plus docetaxel compared with placebo plus docetaxel for the first-line treatment of human epidermal growth factor receptor 2-negative metastatic breast cancer. J Clin Oncol. 2010 Jul 10;28(20):3239-47.

8. Robert NJ, Diéras V, Glaspy J, et al. RIBBON-1: randomized, double- blind, placebo-controlled, phase III trial of chemotherapy with or without bevacizumab for first-line treatment of human epidermal growth factor receptor 2-negative, locally recurrent or metastatic breast cancer. J Clin Oncol. 2011;29(10):1252-60.

9. Department of Health and Human Services, U.S. Food and Drug Administration. Proposal to Withdraw Approval for the Breast Cancer Indication for AVASTIN (Bevacizumab). Accessed January 22, 2017.

10. Booth CM, Eisenhauer EA. Progression-free survival: meaningful or simply measurable? J Clin Oncol. 2012 Apr 1;30(10):1030-3.

11. U.S. Government Accountability Office. FDA Needs to Enhance Its Oversight of Drugs Approved on the Basis of Surrogate Endpoints. Originally published on October 26, 2009. Accessed January 22, 2017.

12. U.S. Food and Drug Administration. Breakthrough Therapy. https://www. Accessed January 25, 2017.

13. U.S. Food and Drug Administration. CDER Breakthrough Therapy Designations as of December 31, 2016. HowDrugsareDevelopedandApproved/DrugandBiologicApprovalReports/NDAandBLAApprovalReports/UCM481542.pdf. Accessed January 22, 2017.

14. Salas-Vega S, Iliopoulos O, Mossialos E. Assessment of overall survival, quality of life, and safety benefits associated with new cancer medicines. JAMA Oncol. doi: 10.1001/jamaoncol.2016.4166. [Published online ahead of print December 29, 2016]. Accessed January 22, 2017.