Adaptive trial designs are being used increasingly in early-phase clinical trials and cover many different aspects of study designs, such as the number of participants, the amount of compound dosed or dosing intervals, safety elements, and the number of cohorts. The bases for all these adaptive elements have one thing in common: they allow alterations in the design within preset limits without amending the protocol. The preset limits are very important aspects for the approval of adaptive designs by regulatory agencies and ethics committees. In many cases, changes to the study design do not require further review of the protocol. However, depending on the risk, regulators or ethics committees may ask to be notified of, or to fully review, the changes. In the rare case that a full review is required, the protocol and informed consent form do not need to be rewritten, thus saving time.

Tackling the element of uncertainty

Applying adaptive trial designs to first-in-human (FIH) studies of monoclonal antibodies (mAbs) and the use of biomarkers to look at efficacy and safety is a great example. When moving from preclinical testing to human trials, there is always an element of uncertainty since preclinical in-vitro and in-vivo testing only provide good approximations of what will happen in humans. However, the uncertainty is even greater when dealing with mAbs since the targets are often different in humans and there is a lot of variability with subcutaneous dosing.

There are many cases of preclinical testing for biologics where the product was not acting on its target in humans as predicted. In fact, the same can be said for small molecules. That is why, for FIH studies, a safety factor is applied to reduce the initial dose by a tenth or greater from the no observed adverse effect level (NOAEL) or the minimal active biological effect level (MABEL). Furthermore, with mAbs, small increases in doses have often resulted in large downstream effects leading to adverse events. When the FIH studies start, basing the escalation decision on biomarkers of the downstream effect of the mAb can give a much more reliable marker of the biological effect of the treatment; hopefully before adverse events are seen.

Many mAbs are intended for subcutaneous dosing for efficacy and safety reasons, but also because subcutaneous dosing allows patients to administer the treatment themselves, without going to an infusion centre. The downside to subcutaneous dosing is the pharmacokinetic (PK) variability it introduces, often with non-linear kinetics. This causes difficulty in predicting the efficacy and safety of the next dose escalation.

Ensuring safety in clinical trials

A study run by Altasciences at its clinic in Montreal dealt with this uncertainty by adding a biomarker to the activity of the mAb. The antibody targeted an enzyme upstream of the biomarker and its action resulted in increases in the levels of the biomarker, allowing it to be a good pharmacodynamics (PD) marker. The study called for extensive measurements of the mAb in the blood and the biomarker. The design was aided by the fact that the biomarker’s action is within the blood. In many cases, when the compound acts in local tissues, biomarkers in the blood still provide reliable reflection of tissue levels. The protocol was written so that the dose escalations could be adjusted based on either the PK or PD response, and details the information the safety review committee would analyse to make those decisions.

The stopping criteria per subject and for each cohort was based on the usual set of specific symptoms and clinical labs, and also included a cut-off for the biomarker, as preclinical data suggested an 80% change was sufficient for efficacy. As often seen with subcutaneous dosing, there was large variability in the bioavailability, the exposure didn’t increase linearly and the Tmax shifted with each dose increment.

In addition, the biomarker showed that the PDs were not correlated with the PK in a linear manner. Finally, the PK/PD relationship was very different from what was seen in preclinical analysis, with a much lower PD effect than expected. By having the PK and PD measures, the safety committee was able to adjust the dosing to ensure the safety of the study participants and establish proof of concept through the biomarker levels.

There are many types of adaptive designs being used in clinical research these days. Using biomarkers in addition to PK analysis offers a great way to allow adaptation to rising dose studies to ensure the safety of patients, and to ensure a potentially efficacious dose is reached without causing delays or amending the protocols.