Benchmarks in Clinical Trial Design Require Closer Examination by Oncologists

Maurie Markman, MD

Development, interpretation, and reporting of results from clinical trials in peer-reviewed literature form the basis of innovation in modern cancer medicine. In the hierarchy of such trials, the randomized phase 3 study, in which an experimental strategy is compared with a control arm, is supreme.

In most oncology settings the control arm is considered the standard-of-care option for the patient population at the time of trial initiation. Depending on the overall intent of the trial protocol, selection of the control arm may be determined by the study investigators, a sponsor (eg, pharmaceutical or device manufacturer) or a regulatory agency (eg, FDA).

In recent years there has been discussion in oncology literature regarding multiple aspects of randomized trials in cancer medicine. Topics often center on appropriate study end points required to achieve approval for commercial sale and the adequacy of subpopulation representation. A particularly vexing issue for some has been the role of so-called surrogate end points, such as progression-free survival (PFS) in obtaining regulatory approval vs other objective measures that may independently define clinical benefit.

However, less discussion surrounds the initial selection of appropriate study control arms. It is critical to appreciate that the results of most randomized trials in the oncology sphere are not available for at least several years following study initiation. As a result, it is possible that the control arm of a particular study will no longer be considered the standard of care when results become available to the oncology community. This outcome may occur if a presentation of study results during the interval following study activation reveal an alternative experimental strategy to be meaningfully superior to the standard-of-care regimen.

This issue once might have been far more theoretical than a reality because major changes to standards of care occurred infrequently. Consider, for example, primary therapy in advanced epithelial ovarian cancer, where use of a platinum agent (ie, cisplatin or carboplatin) plus a taxane (generally paclitaxel) was the undisputed standard of care for decades. Then, data became available regarding the favorable effect of bevacizumab (Avastin) on clinical outcomes.1 These findings resulted in regulatory approval adding this third antineoplastic agent to the dual cytotoxic drug strategy.¹

With regulatory approval of the 3-drug combination in the primary chemotherapeutic treatment of advanced ovarian cancer, the question is whether this regimen should be the required standard-of-care control arm in future studies in this clinical setting, or simply an option to be considered. If a phase 3 trial conducted for regulatory approval of a novel agent reveals an improvement in PFS compared with a control arm not containing bevacizumab, how should the results of this study be interpreted by the FDA, practicing oncologists, and patients with ovarian cancer requiring therapy in this clinical setting?

This thought experiment is not unique to ovarian cancer. In a recent commentary on randomized trials designed to examine new agents in the management of metastatic castration-resistant prostate cancer, Van Wambeke et al highlighted several clinical trials with clinically questionable control arms.² In one case the control arm was stated to be physician’s choice but was limited to 1 of 2 antihormonal agents. These agents were selected based on the regimen the individual’s cancer had progressed on before entry into the study.² In addition, a substantial percentage of patients eligible for enrollment had not received drugs known to be active in the clinical setting being examined.

This experience highlights a serious concern in trial development: protocols are not directly comparing novel strategies with relevant standard of cares. Further, trials are enrolling patients who have either not received the standard of care or have progressed following standard of care the setting being examined.

Unfortunately, the impressively rapid establishment of multiple pharmaceutical agents (with differing mechanisms of activity) approved by regulatory agencies in several clinical settings make it increasingly difficult to ensure that the trial-based populations objectively and appropriately represent a population for demonstrating a new drug’s effectiveness.

One strategy for circumventing this issue is to only permit patients entry into a trial if they have previously received all approved agents/drug classes in the clinical setting being examined. It is reasonable to suggest this is approach when a patient has previously received and progressed, failed to respond, or responded and then quickly progressed following standard-of-care therapy. For example, in ovarian cancer, this could include a patient with documented disease progression 3 months following initiation of primary platinum-based therapy.

However, what if a patient with ovarian cancer treated with a first-line platinum-based regimen initially responds to therapy but has disease recurrence 6, 12, or 24 months following completion of the treatment program? Do they f it the definition of prior exposure to a platinum agent? The FDA has considered such patients to fall into 2 different regulatory categories, defined as follows:

• Platinum resistant: patients with no response, initial progression, or response followed by disease progression within 6 months of completing primary chemotherapy.
• Platinum sensitive: patients with response and progression more than 6 months after completion of primary chemotherapy.

What about patients who have been treated with the standard-of-care cytotoxic chemotherapy regimen plus bevacizumab, followed by singleagent maintenance bevacizumab? Would it be fair to conclude that documented disease recurrence 12 months after completion of maintenance bevacizumab implies resistance to this antiangiogenic agent? Would a clinical trial examining a novel agent compared with a standard-of-care regimen not need to include bevacizumab?

Such a conclusion would be inappropriate because data from randomized trials have revealed that the administration of bevacizumab (plus chemotherapy) in certain ovarian cancer populations that previously received bevacizumab (plus standard-of-care cytotoxic regimen) can experience a superior outcome compared with chemotherapy alone.³ Therefore, if bevacizumab is not included in the control arm with the standard-of-care cytotoxic regimen, how does one know whether an outcome favoring the experimental agent is better in both efficacy and toxicity?

There are no easy answers to the questions posed in this commentary. These are factors to acknowledge and that should encourage open discussions between patients and oncologists about what data from randomized trials and regulatory approvals realistically provide in helping to select optimal therapy.

References

1. Markman M. Pharmaceutical management of ovarian cancer: current status. Drugs. 2019;79(11):1231-1239. doi:10.1007/s40265-01901158-1
2. Van Wambeke S, Vera-Badillo FE, Gyawali B. Controlling the control arm in metastatic castration-resistant prostate cancer trials: best standard of care or the minimum standard of care? J Clin Oncol. 2022;40(14):1518-1521. doi:10.1200/JCO.21.02304
3. Pignata S, Lorusso D, Joly F, et al; MITO16b/MANGO–OV2/ENGOTov17 Investigators. Carboplatin-based doublet plus bevacizumab beyond progression versus carboplatin-based doublet alone in patients with platinum-sensitive ovarian cancer: a randomized phase 3 trial. Lancet Oncol. 2021;22(2):267-276. doi:10.1016/S14702045(20)30637-9