Medical oncologist Dr Michael Thompson is the medical director for the Early Phase Cancer Research Program at the Aurora Research Institute, with a research focus on clinical trial patient participation and precision medicine. In part one of our conversation with Dr Thompson, he discusses innovative clinical trial designs that aim to break through bureaucracy in running early-phase clinical trials, as well as designing clinical trials in newly diagnosed blood cancer patients that may ease the pressure in recruitment.
Reynald Castañeda: Let’s start with a broad question: what are the challenges facing oncologists in running clinical trials?
Dr Michael Thompson: In adults with cancer, only about 3–5% of patients are put on clinical trials, and there have been many publications noting the many barriers to recruitment. For example, there are geographic barriers where there is limited access to clinical trial sites in certain parts of the US. There are trials that only open in large academic medical centres or in cancer centres designated by the National Cancer Institute (NCI), which can hurt accrual. Yet, while barriers to recruitment have been written about for decades, not a lot has been done about it.
There are two main types of clinical trials in cancer. There are NCI trials that have been chronically underfunded and so oncologist participation is more in a voluntary basis. These trials do not really bring in revenue to the clinical trial sites.
Then there are industry-sponsored studies that do bring in revenue but are more difficult to get started. There’s increasing bureaucracy in opening trials, and there are multiple levels of contracting. It also involves Confidential Disclosure Agreement (CDA) evaluations and feasibility questionnaires, which in total creates barriers to opening studies and even a barrier to patient accrual.
RC: On bureaucracy, can you expand on that and have you seen ways to break through this?
MT: NCI trials use centralised institutional review boards (IRB) which helps with initiating trials, but in the same trials there are also local IRBs in certain institutions that seem completely unnecessary. With industry-sponsored trials, there are a bunch of steps involved to start a trial. First, investigators receive the protocol title or the concept of the trial. Then, investigators must go through CDAs that involve lawyers – and when lawyers are involved nothing’s fast.
Investigators must go through CDAs that involve lawyers – and when lawyers are involved nothing’s fast.
How well do you really know your competitors?
Access the most comprehensive Company Profiles on the market, powered by GlobalData. Save hours of research. Gain competitive edge.
Company Profile – free sampleThank you!
Your download email will arrive shortly
Not ready to buy yet? Download a free sample
We are confident about the unique quality of our Company Profiles. However, we want you to make the most beneficial decision for your business, so we offer a free sample that you can download by submitting the below form
By GlobalData
Once that’s done, the investigator receives the protocol, followed by a feasibility questionnaire, which can be straightforward. It would ask if you have patients that might be adequate for the study or if you’ve participated in such a trial before. But these questionnaires can also be very long – up to 20 pages – and involve detailed questions on things the investigator won’t easily know and would require a lot of work to dig up, which can delay trial initiation.
There are ways of getting around these delays. We are one of the first sites in the US to participate in the Novartis-driven Signature program, which is a modular Phase II cancer trial design for precision medicines. This approach worked well: it has master agreement and budgets are non-negotiable. The idea was that each arm would investigate a single treatment, the arm wouldn’t start unless there is a patient ready, and that the trial would open within two weeks.
There is also the ASCO TAPUR (Targeted Agent and Profiling Utilization Registry) study, which has a pragmatic research design. It studies FDA-approved drugs to find out if they can work in a different disease. If you find this genetic alteration in a different type of tumour, what’s the efficacy of that approved drug in this cancer?
A third example is the NCI-driven NCI-MATCH (Molecular Analysis for Therapy Choice) study. It is designed to figure out if available treatments can be tailored for cancer alterations. Opening one of these arms requires the use of a clinical trial management system (CTMS) to record your research, your billing, and how to build chemotherapy treatment around it.
There is an appetite for these types of trials that have a master protocol to streamline things.
Even if I think NCI-MATCH was kind of hard to initially open, it had been successful so there is an appetite for these types of trials that have a master protocol to streamline things. You can open the trial at one site, but the patient pool can be nationwide as they can be treated at different locations.
Now, I am a part of a team looking at a trial that could potentially have 50 arms that would be ideal for rare cancers. Building this new trial will require a tonne of work and likely sunk costs. So, one could argue a way to optimise this logistical work is to loop in the best of these approaches. For example, first find the patient that would match a specific drug, then build the trial arm around this. And with the CTMS, it would be easy to separate what different costs are involved.
But then again, these approaches will work on specific diseases only. If it were a breast cancer trial with 5,000 patients there is no initial rush in recruitment as the cancer is not as aggressive. The master protocol type approach is ideal in early-phase trials with much fewer patients. It is hard to initiate separate smaller trials that only accrue a handful of participants.
RC: Is there any difference in recruiting for blood or solid tumour trials?
MT: When we have multiple myeloma trials in our site, they tend to be one of the top accruing trials, but these are the patients I usually see so I might be a bit biased. But in blood cancers, historically, there are not many good therapies compared to breast and lung cancers where there are many options. In short, a high unmet need can be one of the catalysts in helping with recruitment.
One potential difference between blood and solid tumour trials is that initiation of dosing may be different in newly diagnosed patients. In solid tumour cancers, while there are emergency cases where the patient needs to be treated fast, the pace is bit slower. There can be a delay in dosing to wait for biopsies and analysis.
In contrast, one of the perceptions with blood cancer is that the patient needs to be treated early, which can be an obstacle for trial recruitment. For a patient to be recruited, samples must be collected and analysed. All this information is needed to be reviewed by the research chief if the patient meets the inclusion criteria. And all these steps can take time.
But there is increasing experience showing certain blood cancers might not have to be treated as quickly.
But there is increasing experience showing certain blood cancers might not have to be treated as quickly, allowing some time for consideration of being recruited in trials. In acute myeloid leukaemia (AML), there is a study by the Leukemia & Lymphoma Society stating that, for precision medicine, you could wait a week before treatment – you don’t have to treat them immediately. In diffuse large B cell lymphoma (DLBCL), there is work by the Mayo Clinic showing that these patients can also wait before initial treatment.
Chronic lymphocytic leukaemia (CLL) typically develop slowly so this can be counted as having a more delayed treatment timeline unless the patient is accelerating quickly. On the other end of the spectrum, there is acute promyelocytic leukaemia (APL) where the patient needs to be treated fast – within hours – because of the high mortality rate.
I am on the NCI Lymphoma Steering Committee and we’re increasingly trying to say you can initially treat patients with what the oncologist thinks the patient needs. For instance, in DLBCL, the patient can be given the chemotherapy cocktail R-CHOP for cycle 1 and the investigational drug from cycle 2. This allows some time to enrol patients or consider enrolling patients to a clinical trial. This time is helpful if it’s a biomarker-driven trial or if a bone marrow biopsy is needed, where there can be a wait for these results for recruitment.
RC: How about in the relapsed or refractory setting?
MT: In patients who have relapsed or are refractory cancer patients, certain patient profiles have a higher weighting in decisions with recruitment – such as heart disease or kidney disease risk – so there’s a lot more individualisation. If it is a drug with a new mechanism, it can draw patient interest for recruitment.
A challenge in trials recruiting relapsed/refractory patients is that the trial design should quickly reflect the standard of care. When Johnson & Johnson and AbbVie’s Imbruvica (ibrutinib) was first investigated in CLL there was a high unmet need, which helped with recruitment. After its approval, following trials must be redesigned to reflect Imbruvica’s efficacy benchmark.
The standard in some cancer indications moves so fast that there is the need to predict what will be the new standard in a year to 18 months and design the trial accordingly. You need to know the treatment landscape, where the unmet crowds are, and what the efficacy results are likely to be in yet-to-be-approved drugs. Some of these predictions might be wrong but at least it helps with planning for the future.
The interview has been edited for length and clarity.