Cancer vaccines: Clinical development challenges and proposed regulatory approaches for patient access to promising treatments


  • Justin Stebbing MD, PhD,

    Corresponding author
    1. Imperial College School of Science, Medicine, and Technology, the Hammersmith Hospitals National Health Service Trust, London, United Kingdom
    • Department of Medical Oncology, the Hammersmith Hospitals NHS Trust, Charing Cross Hospital, First Floor, East Wing, Fulham Palace Road, London W6 8RF, United Kingdom
    Search for more papers by this author
    • Fax: (011) 44-208-8461433

  • Christopher Wood MD,

    1. Department of Urology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    Search for more papers by this author
    • Dr. Wood has acted as a consultant for Antigenics Inc.

  • Michael Atkins MD,

    1. Department of Medicine, Harvard Medical School, Boston, Massachusetts
    Search for more papers by this author
  • Ronald Bukowski MD,

    1. Department of Experimental Therapeutics, Taussig Cancer Center, Cleveland Clinic Foundation, Cleveland, Ohio
    Search for more papers by this author
    • Dr. Bukowski has acted as a consultant for Pfizer, Genentech, Bayer, Antigenics, Novartis, and Wyeth. He also has received research support from Pfizer, Bayer, and Wyeth and honoraria from Bayer, Genentech, Pfizer, and Wyeth.

  • Stephen Litwin MD,

    1. Biologics Consulting Group, Washington, DC
    Search for more papers by this author
    • Dr. Litwin has acted as a paid advisor to Antigenics Inc.

  • Mark Bower MD, PhD,

    1. Department of Human Immunodeficiency Virus, the Chelsea and Westminster Hospital, London, United Kingdom
    2. Department of Oncology, the Chelsea and Westminster Hospital, London, United Kingdom
    Search for more papers by this author
  • Andrew Parsa MD, PhD,

    1. Brain Tumor Research Center, University of California at San Francisco, San Francisco, California
    Search for more papers by this author
  • Hyam Levitsky MD

    1. Department of Oncology, Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University, Baltimore, Maryland
    Search for more papers by this author


The challenges of late-stage development increasingly are becoming clear as the clinical development of therapeutic cancer vaccines continues to progress. Preclinical and clinical research had indicated that cancer vaccines exert optimal benefit in earlier stage disease or in adjuvant/minimal residual disease (MRD) settings. However, clinical trials in these settings can be prohibitively slow from a development perspective because of the better prognosis of the patient population and the current lack of early surrogate markers of efficacy. Therefore, the ‘optimal’ patient population (the patient group in which the greatest benefit is demonstrated) for any given vaccine studied in this setting generally must be identified first through the conduct of a large randomized trial for each indication, then confirmed in a second large randomized trial. On the basis of the current regulatory paradigm, a late-stage development program for a cancer vaccine in the earlier stage disease or adjuvant/MRD setting easily could extend past 10 years. The tight funding environment and constant evolution in medical practice, which can make replication of results from the first trial infeasible over such a long timeline, pose additional challenges. In this report, the authors discuss 3 potential regulatory solutions to better enable the development and commercialization of therapeutic cancer vaccines: a U.S. Food and Drug Administration-proposed cost-recovery program, conditional marketing authorization, and a new development paradigm. All of these solutions aim to balance a complex equation of biologic rationale, weight of the evidence for efficacy and safety, regulatory expectations, and cost and timeline of clinical development. Cancer 2008. © 2008 American Cancer Society.


Compared with many other disease areas, the prognosis for patients with cancer tends to be poorer, with fewer available treatment options. Therefore, the clinical development of a cancer treatment often is more condensed than the typical Phase 1, 2, and 3 development model for other pharmaceutical products. In oncology, an investigational agent usually is evaluated in 1 or more phase 1/2 trials to determine dosing and initial safety and to identify preliminary signals of efficacy, followed by 1 or more controlled, randomized phase 2/3 trials to better define efficacy and safety.

When a late-stage clinical trial fails to meet its primary endpoint, further analyses (either predefined or post hoc) may identify evidence of benefit in a subgroup of patients. According to the conventional regulatory process, a second late-stage study conducted specifically in this patient subgroup is almost always necessary to confirm the benefit observed in the first late-stage trial.

In March 2004, the U.S. Food and Drug Administration (FDA) issued its major report on the Critical Path Initiative, which was designed to modernize the clinical development process by obtaining fundamentally better answers regarding how the safety and effectiveness of new products can be demonstrated in faster time frames, with more certainty, and at lower costs. A key element of the initiative includes use of an adaptive clinical trial design in which early evaluation of incoming trial data allows for refinement in patient selection. For trials in which an apparent benefit is identified for a subgroup of patients but not in the overall study population, adaptive clinical trial design potentially can eliminate the need for a second trial to confirm subset analysis findings. However, adaptive trial design is only feasible in cases in which early determinants of efficacy, such as tumor response or biomarkers, are available.

In patients who have little to no measurable tumor burden, the use of tumor response as a surrogate measure of potential efficacy is not possible. The advent of molecular biomarkers of minimal residual disease (MRD) (eg, prostate-specific antigen in prostate cancer or polymerase chain reaction-based quantification of tumor DNA in serum) provides a potential index of treatment-associated change in tumor burden. To our knowledge to date, however, a correlation of such parameters with accepted clinical endpoints such as survival has not been validated. In addition, although there remains great hope and effort, years of research have yet to identify quantitative indices of vaccine-induced immune activation that correlate with clinical responses. Accordingly, currently, there appears to be no near-term solution(s) that would allow key advantages of adaptive trial design to be realized in the clinical setting best suited for cancer vaccines.


Extensive preclinical studies evaluating a variety of vaccine approaches across more than 15 animal models of human cancers consistently demonstrate that therapeutic vaccines benefit the host.1–55 These approaches include tumor-derived heat shock protein-peptide complexes, tumor cells modified to secrete cytokines, tumor cells modified to express the immune costimulatory protein B7, and tumor cells mixed with bacillus Calmette–Guerin, as well as a select number of strategies that target defined tumor-associated antigens, such as lymphoma-derived immunoglobulin (idiotype).56–63

Two key points emerge from these studies. First, therapeutic vaccination against cancer results in benefit to the host, as measured by complete tumor rejection, prolonged stabilization of tumor growth, and/or improved survival time. Second, when it has been examined, greater efficacy has been observed in the minimal disease setting compared with the efficacy observed in the setting of more advanced disease.1, 20, 25, 27, 29, 30, 32, 34, 38, 41, 48, 53, 64

In clinical studies, cancer vaccines generally are safe and well tolerated. Several late-stage, randomized clinical trials of cancer vaccines indicate treatment activity in subsets of patients.65–69 This activity, as predicted by the preclinical body of evidence, typically is observed in disease settings in which tumor burden is lower—ie, earlier stage disease or with adjuvant treatment (eg, in conjunction with surgery, such that there is no evidence of disease or residual disease is minimal). Summarizing decades' worth of clinical trials of cancer vaccines conducted across multiple tumor types, Choudhury et al. recently concluded that those patients who have limited disease or patients in the adjuvant settings have benefited most from this kind of targeted therapy approach.70 Examples are discussed below.


OncoVAX (Intracel Resources) was evaluated in the adjuvant setting in a phase 3 clinical trial that involved patients with stage II or III colon cancer comparing vaccination versus observation after surgical resection. The study indicated that, with a 5.8-year median follow-up, there was a statistically significant benefit associated with vaccine for both recurrence-free survival and overall survival in stage II patients (n = 157), but not in stage III patients (n = 84). For stage II vaccine and control groups, the 5-year recurrence-free survival rate was 79% versus 62%, respectively (P = .009), and the overall survival rate was 82.5% versus 72.7%, respectively (P = .010).69 A phase 3 confirmatory trial in stage II colon cancer is planned that will enroll an estimated 600 patients and will take at least 5 years to complete.


An exploratory subset analysis of a phase 3 trial evaluating the Sipuleucel-T (APC8015; Dendreon) cancer vaccine that involves 127 patients with asymptomatic, androgen-independent, metastatic prostate cancer indicated that a subset of patients with Gleason scores ≤7 (n = 75) appeared to derive the most benefit from vaccine therapy. In this better prognosis subset, the median time to disease progression was 16 weeks for patients who received vaccine versus 9 weeks for patients who received placebo (P = .002; hazards ratio [HR], 2.2).71 Final 3-year follow-up data demonstrated a median survival benefit of 21%, or 4.5 months, and a 3-fold improvement in survival at 36 months compared with placebo, regardless of Gleason score (P = .010; HR, 1.7).65 A biologics licensing application for Sipuleucel-T has been reviewed, and the FDA will consider this further with updated overall survival data.


Subset analysis from a phase 3 trial of the Vitespen cancer vaccine (Antigenics) also indicated a significant improvement in recurrence-free survival among a group of patients with a better prognosis. The trial, which involved 728 patients who had nonmetastatic renal cell carcinoma (RCC), compared nephrectomy versus nephrectomy plus Vitespen. In the per-protocol population of 604 patients (nonmetastatic RCC without residual disease at baseline after surgery), there was a 43% improvement in recurrence-free survival associated with vaccine compared with observation among a subgroup of 361 patients with earlier stage disease: stage I, high-grade; stage II, high-grade; or stage III (T1, T2 and T3a), low grade (P = .018 [nominal, 2-sided]; HR, 0.567).72 This group of patients with a better prognosis correlates with “intermediate risk” patients, a definition currently being used by the Eastern Cooperative Oncology Group in a large, ongoing RCC adjuvant study.73 Data from that trial continue to be collected and analyzed to gain additional efficacy information.


Although the majority of clinical research indicates greater benefit of cancer vaccines among patients with earlier stage disease, a subset analysis of a phase 3 trial of the aTL cancer vaccine (LipoNova) in patients with nonmetastatic RCC indicated a significant reduction in tumor progression for patients with T3 tumors but not for patients with T2 tumors.74 The 5-year progression-free survival rate was 81.3% for patients with T2 tumors who received vaccine (n = 119) versus 74.6% for similar patients in the control arm (n = 145; P = .216). For patients with T3 tumors, the 5-year progression-free survival rate was 67.5% for those in the vaccine arm (n = 58) compared with 49.7% for those in the control arm (n = 57; P = .039). The apparent lesser degree of vaccine benefit among patients with T2 tumors most likely was influenced by the 5-year timeframe of observation, which may not have been long enough to capture the vaccine's effect among this group of patients with earlier stage disease. An updated report with additional overall survival data is expected, and a second phase 3 study is planned.


Compared with traditional cancer drugs, therapeutic cancer vaccines are biologic agents designed to activate the immune system to target tumor cells with high specificity, and they generally are regarded as safe and well tolerated. However, an important barrier that limits vaccine efficacy in the setting of established cancer (as opposed to prophylactic vaccination) is the development of immune tolerance to cancer antigens that arises during tumor progression.75 This contributes to the well documented observation that therapeutic cancer vaccines are far more efficacious in settings of lower tumor burden, such as early-stage or adjuvant/MRD settings, findings that are supported by a large body of preclinical and clinical data.

Traditional cancer agents typically are evaluated in settings of high tumor burden, allowing for quicker data collection and the use of surrogate endpoints, such as tumor response. In contrast, evaluating therapeutic cancer vaccines in ‘better prognosis’ patients with lower tumor burden is challenging because of the long timelines required to measure effects on endpoints that are meaningful in this disease setting (eg, overall and recurrence-free survival).

Identification of the ‘optimal’ patient population for treatment also is difficult in lower tumor burden settings and entails large randomized trials for each vaccine in any given indication. Although preclinical studies indicate that a vaccine has greatest benefit in early disease, the vaccine's effect still must be evaluated in humans across a range of early disease stages to identify exactly which stages or subcategories of disease are most responsive to the treatment in each indication and for each vaccine being tested. In addition, any compelling benefit observed in a subgroup of patients—even if it is statistically significant in predefined or post hoc analyses—must be confirmed subsequently in a second randomized trial involving this subgroup.


To facilitate the clinical development of cancer vaccines in the disease setting in which they are likely to have the most benefit—the ‘better prognosis’ patient setting (earlier stage of disease or adjuvant/MRD)—3 solutions are summarized here for consideration separately or in tandem: an FDA-proposed cost-recovery program, conditional marketing authorization, and a new development paradigm.

Cost recovery

In December 2006, the FDA issued a draft of a proposed rule to amend the regulation concerning charging patients for investigational new drugs. If the proposed rule becomes effective as it is written currently, then it will permit charging for a broader range of investigational uses than presently allowed, which provides a potential mechanism to partially fund the long and expensive, late-stage trials of cancer vaccines. The proposed rule strictly defines criteria that must be met to qualify for cost recovery and further stipulates that the price charged for the drug under this allowance generally would be only for the direct costs of manufacture and supply of the product. However, given the extraordinary expense of large and long trials currently needed to demonstrate efficacy in patients with earlier stage disease, the applicability of cost recovery in this clinical setting is questionable. In addition, the success of such an initiative depends on working with the Centers for Medicare and Medicaid Services and private payers to ensure adequate patient reimbursement.

Conditional marketing authorization

Another model that has been proposed for consideration is similar to that recently adopted in Europe and grants conditional marketing authorizations (CMAs) before full marketing approval for treatments that provide preliminary scientific evidence to demonstrate a positive risk-benefit assessment. In life-threatening or orphan disease settings, CMAs allow patient access to treatments while confirmatory data are gathered in a postmarketing environment. Approval under this scenario requires that the sponsor complete or initiate the studies to provide the necessary additional data, and authorization is subject to annual renewal or revocation.

Under the conditions of a CMA, it is important for regulatory authorities to consider the types of trials that can be executed ethically and reasonably to provide the confirmatory evidence needed. CMAs provide an opportunity for greater access to investigational treatments and more comprehensive cost recovery compared with the FDA cost-recovery program. Given the significant costs associated with caring and treating cancer patients who have advanced disease, prevention of recurrence through the use of cancer vaccines could translate into significant patient benefit as well as significant cost savings to the health care system. However, as with all adjuvant studies, benefit will need to be balanced with the finding that a proportion of patients with earlier stage disease never will have a recurrence after initial intervention and ultimately may receive unnecessary treatment.

New development paradigm

A new development paradigm for cancer vaccines recently was proposed by the Cancer Vaccine Clinical Trial Working Group (CVCTWG), a group of > 50 experts from academia, regulatory bodies, and the biotech/pharmaceutical industry from America and Europe that was organized by the International Society for Biological Therapy of Cancer and the Cancer Vaccine Consortium.76 The authors proposed a clinical development model in which therapeutic cancer vaccines are investigated in 2 general types of clinical studies: proof-of-principle trials and efficacy trials. Designed to account for biologic features of cancer vaccines, the CVCTWG's proposed paradigm supports a more flexible, prompt, and focused process of clinical development with early and informed decision making through prospectively defined decision points (ie, ‘go’ or ‘no go’), the use of biologic endpoints and adjusted clinical endpoints, and the early use of randomized trials and adaptive design components when applicable.

The CVCTWG must be commended for their contributions toward developing a more feasible regulatory pathway for therapeutic cancer vaccines, which represents an important step in highlighting the need for regulatory accommodation to enable accelerated and efficient development of cancer vaccines. However, it remains unclear how the proposed development paradigm could be applied effectively in the near term toward the acceleration of cancer vaccine development in the better prognosis patient setting. The lack of currently established, early markers of efficacy and the issue of ‘late events’ (the long timeline required for a meaningful proportion of patients to experience disease recurrence) remain impediments to the accelerated decision making inherent in the CVCTWG's model.

Because therapeutic cancer vaccines appear to be associated with far fewer toxicities and greater specificity of action compared with traditional cancer treatments, this newer category of treatments is a strong candidate for applying a novel development and regulatory path. Discussions at a recent cancer vaccine development and licensure workshop, which was sponsored jointly by the National Cancer Institute and FDA, indicated that there exists broad consensus, based on decades' worth of preclinical and clinical data, that the best biologic setting for a therapeutic cancer vaccines is in patients with earlier disease, which makes findings from clinical trials in these patients especially compelling. Given the current lack of novel approaches to accelerating phase 3 clinical trials in this disease setting, consideration should be given to the type and amount of clinical efficacy data needed to support regulatory approval for cancer vaccines, especially given the strong indications of favorable safety and high specificity as well as the apparent biologic rationale for the application of therapeutic cancer vaccines in the setting of patients with earlier stage disease.

Changes in regulatory convention are needed to ensure the continued advancement of this promising field. Innovation and dialogue in the regulatory process for cancer vaccine development will be critical for the sustained investment of time, effort, and funds—without which progress in this novel and important area may never come close to realizing its potential.