Early oncology trials have changed for the better over the last few years thanks to novel investigational agents, innovations in trial design, and changes to regulatory practices. Among other improvements, these changes have helped to perfect the way study designers plan early phase dosing.
Dosing strategies in Phase I trials
When an investigational agent is administered to a patient for the first time, dosing is based on preclinical and toxicological data. Initial Phase I trials provide the first information on a drug’s safety profile, as well as foundation data for investigators and potential subjects for further investigation. Investors, regulatory authorities, institutional review boards, and ethics committees also use this data to make key regulatory and financial decisions. Although safety is the primary goal for first-in-human trials, ideally these trials can also produce key pharmacokinetic/pharmacodynamic (PK/PD) and mechanistic data.
Single ascending dose
Single ascending dose studies are first done when there are questions about an agent’s tolerability. They involve a single agent, and dose ranging is done with patients or healthy volunteers. However, for ethical reasons, it is not always possible to test drugs of known cytotoxicity on healthy people, and so their participation is less common in oncology research.
If preclinical or theoretical data suggest toxicity, testing is done in single subjects with staggering between and conservative dosing, with observation and testing for a period of time to confirm safety. If no adverse effects are observed and PK data conform with predicted safe values, a new group of participants is given a higher dose.
Single ascending dose studies may have a crossover design to help control for interpatient variability and a placebo to control for environmental conditions. A single-blind approach allows trial staff and investigators to make informed decisions regarding dose escalation.
Multiple ascending dose
Multiple ascending dose design is similar to single ascending dose design, and such trials may follow an initial single ascending dose study. A multiple ascending dose trial involves a single agent, and dose ranging is done in patients with disease who have limited treatment options. An advantage of multiple ascending dose design is that it limits the number of patients who may receive doses predicted to be ineffective. Multiple ascending dose studies can also be aligned with standard-of-care treatment cycles to anticipate combinations in later studies.
Patients are given multiple doses of the investigational agent over a period wherein dose-limiting toxicities, or DLTs, are expected to occur. Following a DLT observation period, dosing is continued until stopped for tolerability issues or progressive disease. Multiple ascending dose studies may provide first efficacy data when performed in patients with the targeted indication.
Combination studies with an SOC regimen backbone can be used to simplify trial design. In these studies, the investigational drug is given in combination with other agents and requires dose ranging based on tolerability and PK/PD. Preclinical evidence that supports a biological or pharmacological combination rationale and exploration of PK/PD interactions between drug partners is necessary for Phase I trials of combination therapies. These preclinical models are used to assess additive or synergistic effects of therapy combinations.
A common — and effective — strategy for combination studies is to include an arm comprised of SOC with placebo and an arm with both SOC and the study drug to allow all patients to receive standard existing treatment. The combination is done in a stepwise fashion, keeping SOC at the approved dose and escalating doses of the investigational agent. Study designers should take care to consider the mechanism and overall order of combination; one compound may be necessary for the other to elicit maximum effect.
Establishing a starting dose
But how are the exact doses determined for these early trials?
First-in-human dosing can rely on the minimum anticipated biological effect level, which is based on existing data from in vivo animal models and in vitro human cell culture. It represents the smallest amount of drug necessary to see a pharmacological effect on a patient. The European Medicines Agency tends to encourage using this as the starting dose.
Another standard favored by the EMA is the no-observed adverse effect level, or NOAEL, which is the greatest amount of agent that will not cause detectable adverse changes to morphology, functional capacity, growth, development or lifespan. However, NOAEL and similar designations might not be feasible to determine dosing for highly cancer-targeted compounds with low toxicity.
According to the Food and Drug Administration, the starting dose for safety testing should have a pharmacological effect and be reasonably safe to use. “Safe to use” has traditionally meant one-tenth the lethal dose in mice or one-sixth of the highest nonseverely toxic dose in a model more sensitive than a rodent — marmosets, for example.
What everything comes down to
What is the most important part of all aspects of trial design, no matter the phase? Begin with the end in mind. To plan optimal dosing for your product, you’ll need to tailor everything to meet its unique needs and challenges.