April 1, 2018 (Vol. 38, No. 7)
Andrew Zupnick Ph.D. Vice President of Oncology Strategy Novella Clinical
Current State of Protocol Design Impacting Small to Mid-Sized Biotech Companies
Over the past century, a greater understanding of how our immune system recognizes and targets foreign invaders for destruction has emerged, resulting in cancer immunotherapy breakthroughs. Today, more than 900 immunotherapy agents are in clinical development, and more than 1,000 in preclinical development, according to the Cancer Research Institute.
Immune-mediated tumor destruction is tantalizing for doctors and patients, and in the last decade, researchers have found ways to point the immune system’s destructive power in the direction of cancerous cells that have previously dodged detection. Clinical use of currently approved immunotherapeutics has demonstrated the power and durability of these therapies in a variety of cancer types, yet much work remains be done before science can realize the potential that exists to modulate the immune system.
During this burst of innovation, several hurdles have emerged that threaten to slow the pace of new cancer immunotherapy development. Some of these hurdles, as outlined by the Society for Immunotherapy of Cancer (SITC), are not necessarily clinical. They include the scarcity of reagents for combination immunotherapy studies and the paucity of funds for the “bench to bedside” transition. The price tag for Keytruda® (pembrolizumab) sits at about $150,000 a year per patient, and unless it is being provided for a combination study by the manufacturer, it can become cost prohibitive to provide the drug to a patient enrolled in a combination trial.
No one can appreciate these specific pain points more than the small to mid-sized biotech companies that are driving innovation in this field, for which both funds and personnel may be scarce. A strategic approach to the design of protocols for endpoint selection and dose escalation can help smaller companies maximize efficiency.
Endpoints and Enrollment
Novel immunotherapeutics have required clinical investigators and trial staff to adapt to a different treatment landscape. Designing an effective protocol is particularly important due to the continuing trend of testing novel immunotherapies in combination with investigational immunotherapies or approved therapies such as chemotherapy and radiotherapy.
One critical piece of the trial design puzzle is choosing the right endpoints. In early-phase studies, primary endpoints tend to focus exclusively on safety and pharmacokinetics. In early-phase evaluations of efficacy, secondary endpoints such as overall response rate, overall survival (OS), and progression-free survival (PFS) are now fairly common. Choosing the right endpoints is of strategic importance, as clues to early efficacy can inform strategy for future trials; identify which indications should be studied further; and in some cases, lead to trial expansion early approval.
In later-phase studies, choosing primary endpoints such as PFS, response rate, and duration of response has been effective in achieving initial approval, with overall survival often included as a co-primary endpoint for immuno-oncology combinations. Other secondary endpoints specific to the pursued indication or the mechanism of action of the drug may also be included (for example, prostate-specific antigen levels in prostate cancer).
Another challenge in protocol design is meeting required patient enrollment numbers. In 2017, there were 469 new trials initiated to test PD-1/L1 checkpoint inhibitors, a specific type of cancer immunotherapeutic, in combination with other drugs. Such trials may require patients of a specific immune status or genotype, potentially complicating recruitment and enrollment, especially if trial participants are drawn from a relatively small population of patients with specific biomarkers. Most current combination trials involving a checkpoint inhibitor exclude patients who have previously received the treatment that is to be evaluated.
It is becoming increasingly difficult to recruit checkpoint-naïve patients as the market continues to expand. Many combination trials also use approved treatments, adding the burden of sourcing, access, and cost in providing these treatments to enrolled patients if a business partnership is not already in place, or if an internal checkpoint inhibitor is not in development.
Dosage and Detecting a Response
Cancer immunotherapy responses can differ from those of classic chemotherapies. For instance, when measuring the response of a tumor to immunotherapy, a “flare effect,” or an increase in tumor size during an immune response, may be seen. Telling the difference between a flare effect and tumor growth related to disease progression is a critical step in determining if a therapy is working or not.
To help clarify such matters, the Response Evaluation Criteria in Solid Tumours (RECIST) working group issued a consensus guideline. The original guideline, which appeared in 2000, was called, appropriately enough, RECIST. It was succeeded in 2009 by a more refined guideline, RECIST 1.1. Other response criteria include immune-related response criteria (irRC), which appeared in 2009; immune-related RECIST (irRECIST), which appeared in 2013; and immune RECIST (iRECIST), which appeared in 2017.
irRECIST and iRECIST have been developed to accurately determine how a tumor is responding to immunotherapy. In contrast to irRC, irRECIST measures tumors in one dimension along its longest diameter. This helps improve reproducibility and inter-rater reliability.
Analysis of data from a Phase Ib trial with pembrolizumab in advanced melanoma patients suggests that using irRC to measure disease progression underestimated the benefit of treatment in approximately 15% of patients. Using irRECIST can more accurately determine response to treatment and thereby prevent early cessation of treatment.
Measuring tumor volume through cross-sectional, two-dimensional imaging can assist in making a clinical assessment of response to treatment. Computerized tomography (CT) is often an accurate way to measure treatment response; however, it can often be difficult to define tumor boundaries. Magnetic resonance imaging (MRI) can accurately determine three-dimensional tumor volume and has an acceptable rate of inter-rater reliability, though it is used less commonly than CT.
Frequently, first-in-human Phase I trials abide by a classic 3+3 trial design for determining the maximum tolerated dose. Variations on the 3+3 design have emerged more recently, so that the number of patients treated at suboptimal doses is limited.
One alternative is an accelerated titration design, where one patient is enrolled per dose, starting at the lowest dose. Dose escalations proceed until the patient experiences a Grade 2 toxicity, at which point the trial reverts to 3+3 rules.
Another design, pharmacologically guided dose escalation, uses plasma drug concentrations to direct dose escalation. Using preclinical data, a target plasma concentration is determined based on area under the curve (AUC) for drug concentration as a function of time. Drug concentrations are monitored in each patient, in real time, and once AUC is reached or if dose-limiting toxicities (DLTs) occur, the trial reverts to 3+3 rules.
A “rolling six” design is often used in pediatric Phase I trials when prior information about the dose range is available from adult studies. In this trial design, two to six patients are simultaneously assigned a dose level on the basis of DLT data, specifically, the number of currently enrolled patients who experience a DLT.
Lastly, continual reassessment models use statistical analysis to predict safety as a function of dose escalation. The probability of a DLT is determined as new patients enter the study, and the occurrence of a DLT stops the trial.
As discussed here, some of these pitfalls can be avoided through meticulous consideration of protocol design including endpoint choice, design, and methods used for detecting a response. With limited budget and personnel, small to mid-sized biotech companies can benefit from careful consideration of these factors. As the field continues to evolve, so will the potential barriers that stand in the way of approval of new cancer immunotherapeutics
Pitting Natural Killer Cells against Cancer
Immunotherapy research that leverages live cells to attack tumors is on the upswing, especially on the heels of FDA approvals of CAR-T cell therapy and wide use of checkpoint inhibitors. Yet, for patients who are immuno-depressed from a first-line treatment, T-cell-based therapy could face challenges, including obtaining functional T cells to modify or stimulate.
For applications like these, natural killer (NK) cell therapy is the next frontier. Varied immunotherapies have the potential to leverage NK cells’ ability to bind to and kill cells; approaches in development include bringing NK cells and cancer cells together through an antibody-based approach to cause direct cytotoxicity, or induction of antibody-dependent cell cytotoxicity.
To study these treatment options requires sufficient, functional NK cells and an in vivo environment that supports their survival, says Azusa Tanaka, Ph.D., product manager, Taconic Biosciences.
“Humanized immune system animal models can support delivery of the human cytokines essential for NK cell survival, most notably human interleukin-15 (hIL-15),” notes Dr. Tanaka. Using a model that transgenically expresses human IL-15, investigators can engraft human donor-derived NK cells onto the model and use it for target validation studies.
Taconic’s super-immunodeficient hIL-15 NOG model is playing that role, helping researchers determine if a drug of interest will bind to NK cells in vivo, according to Dr. Tanaka, who notes that “the hIL-15 NOG provides an environment in which to test if the drug can traffic to the right place at the right time, then recruit NK cells to attach to and kill cancer cells.”