Off-Label Use of Oncology Drugs: The Need for More Data and Then Some

Off-label drug use refers to the prescription of licensed drugs for clinical indications or in a manner different from that approved by the US Food and Drug Administration (FDA) and thus not included in the official FDA-approved labeling for the agents. Off-label drug use takes many forms, and versions of it occur in countries outside of the United States as well. Use of drugs for a clinical indication, in a patient population, through a route of administration, or with a dose not specified in the FDA-approved labeling can all be considered off label.¹ Off-label use does not necessarily mean a lack of evidence demonstrating the efficacy and safety of the used agent²; but the supporting evidence for different off-label indications may vary considerably both in extent and quality.

Although estimates vary, there is general agreement that off-label use of drugs to treat cancer is particularly common.^2–5 For example, in 2005 the National Comprehensive Cancer Network estimated that 50% to 75% of drug or biologic therapy used to treat cancer in the United States was off label.³ The reasons are multifactorial. There is a spectrum of cancer types that may have shared biology and sensitivity to therapeutic agents and yet are different diseases in terms of the drug approval mechanism. FDA approval requires a lengthy and expensive process, even for a single-disease indication. Once a drug is off patent, there is little enthusiasm in pharmaceutical circles to invest the resources to expand the labeled indications for an agent.² Pediatric oncology and rare cancers pose special challenges to the traditional drug-approval paradigm. Given the worrisome prognosis and potentially life-threatening nature of a cancer diagnosis, particularly when recurrent or metastatic, time is critical, and discussions of risks versus benefits differ from many other disease settings.

However, these disease-related factors alone would not lead to the current prevalence of off-label drug use in cancer care unless reimbursement policies supported such off-label treatments. In this regard, oncologic therapies have special status in the United States, which is understandable given the issues outlined above. As part of the Omnibus Budget Reconciliation Act of 1993, Medicare was allowed to pay for off-label use of anticancer drugs and biologics if a “medically accepted indication” was supported by selected drug compendia and the secretary of Health and Human Services agreed.⁶ Given this federal mandate, other payers and the states for the most part have subsequently adopted similar policies. In 2008, the list of approved compendia was revised and expanded.⁷

Various groups, individuals, and the lay press have published on off-label drug use and its reimbursement.^1,2,4,8,9 Certainly, the perceived value of off-label drug prescriptions covers a spectrum. In many clinical settings, reimbursement of off-label drug use clearly provides important and critical access to agents that have a sufficient and often a substantial evidence base supporting efficacy and safety. Indeed, the classification of some drug use as off label may come as a surprise. Take an example from my specialty of head and neck oncology—the use of cisplatin concurrent with radiation therapy for locoregionally advanced head and neck cancer. Although widely used for many years with indications supported by several randomized trials, cisplatin used in this manner is off label.^10–13 In pediatric oncology, off-label drug use is widespread and essential to the application of many standard treatment programs.^2,14 In other settings, however, the supporting evidence base for efficacy or superiority to other drugs may be limited or nonexistent,^15,16 and patients are exposed to potential risks without the associated safeguards, including data and toxicity monitoring responsibilities and institutional review board involvement, which are associated with a formal clinical trial. The accrual to legitimate clinical trials may be undermined.¹⁷ In one self-reported survey of US oncologists, off-label, off-protocol drug access is a common patient request, and the majority of respondents acknowledged discussing such therapy and prescribing it at least once in the last year.¹⁸ The use of newer, targeted biotechnology drugs in patients with advanced cancer generates considerable discussion in this regard. Many patients desire access to this class of agents despite limited or no data regarding efficacy or superiority compared with other agents in a particular setting and the relatively high frequency of adverse events associated with them reported to the FDA.^19,20 The unproven use of these often expensive newer agents is also of concern with regard to their contribution to escalating health care costs.¹⁵ There are, of course, many shades of gray among these examples.

Although specified drug compendia figure prominently in reimbursement decisions and addressing areas of uncertainty with regard to off-label drug use, a systematic review of these compendia noted variability in the consistency and transparency of the policies applied in delineating off-label cancer treatment indications.⁶ The approach to the review, rating, and updating of the evidence was often neither systematic nor comprehensive. Not surprisingly then, in the study by Abernethy et al, there was significant variability among the compendia with regard to their assessment of 14 off-label indications.⁶

Clearly, there are many challenges regarding the off-label use of drugs for cancer treatment. Important gaps in the available evidence make it difficult to fully evaluate many possible off-label indications. Although the FDA provides guidance with regard to initial drug development and approval, there is no coordinated approach to the planning of trials and comparative effectiveness studies to optimally inform decisions with regard to off-label drug use.

In that context, the article by Mullins et al²¹ that accompanies this editorial is of great interest. Recommendations for clinical trials of off-label drugs for advanced cancer are provided as part of an effectiveness guidance document (EGD), which was “developed to provide a framework to guide the design and implementation of prospective [comparative effectiveness research] clinical trials for FDA-approved oncology drugs used outside currently labeled indications.”²¹ A central objective was to better address gaps in evidence that currently complicate postregulatory decision making in the off-label setting. The EGD was felt to be most applicable to “pretreated patients who are beyond first-line therapy.”²¹

The development methods used for the EGD are transparent and well described. The work was supported by unrestricted contributions from four pharmaceutical companies. The recommendations are based on extensive consultations with experts and stakeholders and a meeting hosted by the Center for Medical Technology Policy. Meeting attendees included representatives of multiple private, public, and selected international entities. The EGD subsequently underwent peer review and a public comment period. The American Society of Clinical Oncology was represented in the proceedings.

The 14 recommendations summarized in Table 1 of the article by Mullins et al²¹ are logically grouped into four categories: trial design and data analysis, patient and site recruitment, comparators, and outcomes. They are all consistent with accepted concepts of clinical trial methodology. Appropriate attention is paid to prespecifying subpopulations of interest; important prognostic, potentially confounding covariates (eg, comorbidity, age, race/ethnicity, socioeconomic status) that may not have been well explored as part of initial drug registration studies; the use of clinically relevant therapeutic comparators; and the selection of clinically meaningful patient-centered outcomes.

The framework provided is an important and thoughtful start to better understanding and defining the indications for, and appropriate use of, off-label drugs in cancer care. To be optimally successful, however, future work will need to address potential challenges to the implementation of the recommendations. Some examples from the EGD for which only limited detail is provided as to how they will be implemented and funded include the following: collaborations between academic and local community practice groups whereby the former may need to provide training and support for the latter; partnerships that “should be pragmatic and not burdened with layers of redundant regulatory or contractural issues”; “practical operating procedures […] that allow easy accessibility to trials and user-friendly data capture” to facilitate clinical trials at community-based sites; a “private-public partnership” as an approach to support coverage with evidence development studies. Recent federal budget debates highlight that available funding for such initiatives, even when meritorious, may be limited.

Developing more and better data to guide decision making with regard to off-label drug use in oncology will take time; even when available, obtaining a consensus on how to best rate, apply, and synthesize these data will have its own set of challenges. Other approaches to addressing and optimizing off-label drug use without compromising appropriate access deserve attention and consideration. In the short run, learning from the results of the systematic review by Abernethy et al⁶ and addressing the methodologic limitations and inconsistencies of drug compendia seems to be an obvious step. The Centers for Medicare and Medicaid Services have previously specified criteria used in determining the additions to or deletions from the list of recognized compendia.⁷ However, optimizing methodologic rigor comprehensively and in an ongoing and timely way will require a substantial infusion of resources and personnel, particularly if there is no consolidation of work across compendia or other significant initiatives to improve efficiency, and such efforts may nonetheless fall short in addressing the methodologic limitations of the data used to support off-label prescribing decisions.⁸ A potentially more feasible approach has been recommended by Gillick,¹⁵ who proposed the application of a two-step process similar to that used for devices and limited to drugs that are both costly and toxic, whereby FDA approval would be followed by a national coverage determination review with regard to reimbursement. Orphan drug programs can assist with the development of labeled indications for rare diseases.^2,22 Modifications of the existing drug approval process to facilitate expedited or fast-tracked approval of labeled indications should help, but with the caveat noted by Mullins et al,²¹ that “the expedited review process […] creates challenges for generating the depth of information desired by postregulatory decision makers […] and may unintentionally encourage off-label use.”²¹ Increased use of adaptive trial designs in the context of an accelerated drug development and approval process is of interest but will require agreement between the FDA and pharmaceutical companies as to their framework.²³ The European Society of Medical Oncology has advocated for a greater role of other stakeholders in addition to the involved pharmaceutical company with regard to label extensions,^{2, 24} and that the process for such label extensions need not be as “as demanding as approving a drug for its first indication, first of all in regard to safety.”2(p1924) High-quality observational research may be a useful source of information for clarifying the indications for drug use as well.

In summary, to ensure the optimal use and availability of off-label drugs in cancer care, more and better data are certainly needed to guide and inform clinical decision making as well as reimbursement policies. The EGD and related recommendations developed by Mullins et al²¹ provide an important step in this regard. But to achieve timely, durable, and comprehensive solutions to the challenges associated with off-label drug use, other approaches and initiatives will also need to be explored and pursued.