Oncology drug discovery is thriving, but with so many potential cancer treatments in development, recruiting patients for clinical trials represents a significant challenge. Modern cancer drugs are frequently targeted at very specific patient populations, so the ability to select an appropriate subject group for the trial is essential. This adds to the complexity of trial design and recruitment.
While many oncology products are being developed by big pharma companies that have the experience and resources required to run successful Phase I trials, quite a few are not. For instance, a biotech company taking its first steps into the clinic will rarely have the necessary experience in-house, and a contract research organization will be able to assist them. Even big companies are increasingly likely to engage a specialist CRO to speed up recruitment and increase the chances of a successful trial protocol being achieved. They may simply be looking for assistance in recruitment, or a complete drug development and service package might be required.
Oncology therapeutics range from small molecule, orally available products, to biologics that must be delivered parenterally, and include immunotherapies, vaccines, vectors and even adjuvants. Comparative trials also have to be run to gain approval for generics and biosimilars.
Whether the trial is being recruited and run in-house, or managed by a CRO, many factors must be considered. A Phase I cancer trial will often have strict inclusion and exclusion criteria, and the sponsor is likely to have both short timelines and pressure on costs. With so many other drugs in development, competition for patients is a common concern, and the population is demanding to work with.
Patient competition is increased by the fact that it’s not just oncology agents that are tested in this population. Trials need to be run for oncosupportive drugs to treat conditions caused by treatment regiments such as myelosuppression, anaemia and cachexia, and even diagnostics and medical devices involved in therapy.
In such a complex marketplace, trial design is rarely straightforward. The sponsor might have strict requirements for co-morbidities or previous treatment regimens, for example, or the trial might need to include multiple sub-trials. In contrast to most other clinical development programs, cancer Phase I trials are often carried out in patients rather than healthy volunteers, which can significantly increase nursing care requirements. The drugs themselves are often aggressive and fraught with side-effects, and these issues need to be carefully managed.
There are four main models for managing the practicalities of an early stage oncology drug trial, regardless of the type of product that is being tested. Choosing the appropriate model is an important factor in a successful trial, with appropriate trial design and the trial being run in the optimal environment.
Model 1: Healthy Volunteers In A Clinical Pharmacology Unit (CPU)
An early phase clinical pharmacology unit is a dedicated space, usually within a hospital setting, that is specifically used to carry out early phase clinical trials. The SGS CPU in Antwerp, for example, is a 90 bed unit within Stuivenberg Hospital, and is the setting for Phase I trials with mainly healthy volunteers. This means they are less commonplace for oncology trials, bearing in mind these more often involve patients rather than healthy volunteers.
However, not all cancer-related trials involve patients, particularly the type of equivalence studies demanded by the regulators for biosimilars. As an example, a randomized, double blind, three-arm, parallel group, single dose study was carried out in healthy male volunteers on three formulations of a potential biosimilar to a licensed monoclonal antibody. The aim was to compare the pharmacokinetics, safety, tolerability and immunogenicity of the biosimilar with the originator product. As a secondary objective, the plan was to investigate and compare the safety, tolerability and immunogenicity of the two products. Cases such as this are unusual for a Phase I cancer trial, as it was conducted on healthy volunteers rather than patients, and more than 100 subjects were required.
A CPU is the ideal setting for this type of trial, as it makes it simpler to manage and select a homogeneous subject population, and the conditions for the trial will be less variable, as the subjects come in large groups. As all are being seen within the same unit, by the same clinical staff, procedures can be perfectly standardised. The recruitment power of the center is another real advantage – it is easier and faster.
While the subjects clearly do not have the drug target as they are not patients, this is not a problem when studying biomarker profiles and non-target-related side-effects, and efficacy is not in the design scope. It also means that the PK/PK relationship cannot be seen, and there are neither target-mediated clearance nor target-related side-effects to follow.
These types of trials are typically done to meet requests by the regulatory authorities, in situations such as this example, where the study goal was to help demonstrate equivalence of safety for a biosimilar. The results generated by these studies include high quality data and safety data, as well as comparable pK data.
Model 2: Using The CPU With Patients
There are occasions where a CPU is advantageous when running a trial in a specific patient group. In this example, the patient population had brain lesions, but no overnight stays were required, facilitating its management within the CPU.
The Phase I/IIa trial was designed to assess the PK/PD profile and tolerance of a developmental contrast agent in both healthy subjects, and those with metastatic brain lesions. The primary objective of the Phase I part was to evaluate both its clinical and biological safety, and the PK profile in both plasma and urine, following a single dose at ascending dose levels, at first in healthy subjects. Then, the secondary objective in Phase IIa, was to evaluate safety and plasma PK in single ascending doses in patients with metastatic brain lesions. Its PD as a potential contrast agent in MRI scanning, in both healthy volunteers and patients, was also assessed. A total of 12 patients was required, and 24 screened for suitability.
Phase I/IIa trials of this nature typically enroll a low number of subjects. As before, the environment is well controlled, and the staff is highly trained, including the clinical teams, lab technicians, and site and database coordinators. A LabPass system ensures on-time capture of high quality data.
There are, of course, drawbacks. Ensuring full enrolment of the entire cohort at a single non-oncology center adds complexity to the recruitment process, and very severely ill patients can be difficult to manage outside a hospital ward. Only two of the 24 potential patients were recruited via the CPU database; eight more were found via a patient organization, and the remainder were referred by an extensive network of radiologists, oncologists and neurologists.
Access to an MRI scanner in a timely fashion was also an issue. This was solved by hiring a mobile MRI van to ensure immediate access to a scanner, thus avoiding waiting time. Although the solution was costly, it did facilitate speed.
This example shows that a CPU can be used in patient cohorts for carefully selected cases. The number of patients, the ability to recruit, the length of treatment, patient care needs and vendor availability must all be considered. If all of these factors can be managed successfully, then it is relatively easy to generate high quality data via the trial.
Model 3: A Satellite Center
The third model involves a satellite center that works as an extension of the CPU, but is part of a hospital rather than owned by a CRO. It can be used to support or augment an existing Phase I unit, or may involve the creation of a new one. Ideally, the CRO’s Standard Operational Procedures (SOPs) will be implemented within the unit to ensure transparency and ease of operation, and the hospital’s experienced clinical trial staff will run the unit to guarantee high quality in its operations.
As an example, an open label, single dose, parallel group Phase I study was run to determine the PK of an anticancer agent in patients with solid tumors, who also had mild, moderate or severe renal impairment or end-stage renal disease. The primary objective was to establish the PK in organ impairment, as kidney damage can affect how the dose behaves in the body; for example, it can increase toxicity if the drug is selective for renal excretion. The study would help inform the future choice of dosage levels. There was a secondary objective of monitoring safety and tolerability. It was a rescue trial for eight patients with endstage renal disease.
The big advantage of this type of setting is that it provides dedicated access to a rare population of patients via the hospital’s dialysis unit, while also offering a CPU-like environment. On the downside, investments are required to take a satellite center on board, form agreements, contracts and infrastructure set-up, to the implementation of SOPs, the availability of the clinical team, and qualifications. It is a semi-flexible model, so while it may have limits in other indications, the advantage is that it may offer other suitable patient populations available within the hospital for running further trials.
Using a satellite center can, in certain cases, aid in the successful management of trials with challenging recruitment requirements as a result of giving direct access to hospital or institutional patient pools. This is particularly important for Phase I studies as these are short, but very busy in operations periods. For example, in order to generate a pK profile, 16 doses with very accurate biomarker measurements must be made. Professional staff dedicated to this type of activity are important, because despite the competence of a nurse on a regular ward, going through the busy workflow chart for a PK trial and meeting all its precise timings will be difficult. Overall, the high quality provided by a CPU can also be provided in a satellite site.
Case Study 4: An External Site
Sometimes, there is no sensible alternative to running a Phase I cancer trial within a third-party facility such as a regular hospital ward. To illustrate this case, the trial was a randomized, multi-center, open label, two-treatment, two-period, two-sequence, single dose crossover bioequivalence study, in a gynecological cancer indication. The trial was to be run in about 50 patients, with the primary objective of comparing the rate and extent of absorption.
This type of trial is challenging in recruitment terms. It is a relatively narrow indication, and there is no therapeutic benefit to the patient, which may lead to a disinclination to give consent. There is also a distinct likelihood that there will be a lack of scientific interest on the part of the investigator as there is nothing new.
Any one facility is unlikely to be able to supply more than a small handful of patients, so the only way to recruit sufficient suitable subjects is to run the trial across multiple sites, with extensive global feasibility searches to identify cooperating sites. In this sense, it is much more like a Phase II or III trial in recruitment terms, with all the complex trial logistics and project management that entails.
The big advantage of this model is, of course, access to a much wider pool of potential subjects across a number of good medical facilities. But accurate feasibility studies are crucial, and as it is likely to involve multiple countries, good regulatory intelligence is needed. Costs will be higher than running it in a single dedicated location. However, a network of external sites is, sometimes, the only feasible way to recruit sufficient patients.
Overall, there is tremendous growth in the oncology therapeutic market, with many indications and product types being investigated. It would be impossible to shoe-horn all trial designs into a single operational model. A flexible, adaptive, and service-orientated attitude is required if the optimal solution is to be identified. While some – notably the satellite center model – require significant investments and strategic thinking, choosing the optimal model can assist in meeting sponsor demands, which in turn can help in facilitating the development of these lifesaving medicines.