Even after decades of research and everincreasing R&D spending, the overall success rate of oncology programmes remains low. At an industry level, over $50bn is spent annually on failed oncology trials, leading to a 95% attrition rate. The expected cost to develop a new drug can be anywhere up to $2bn, including the expenses associated with drugs that fail to reach the market.
Research has shown that trials using biomarkers for patient selection had almost twice the probability of success compared to trials that do not use biomarkers (10.3% vs 5.5%). Imaging biomarkers are an essential part of oncology trials, tracking the efficacy of the new treatment and comparing it to the existing gold standard therapies. The insights gleaned from imaging biomarkers steer the course of oncology clinical trials, informing decision-making and endpoints. While RECIST/RECIST 1.1 is the most widely used imaging criteria, it has certain limitations. In recent years researchers have debated whether RECIST is still “the sharpest tool in the oncology clinical trials toolbox”.
In this article we look at some of the limitations of conventional oncology imaging biomarkers (e.g. RECIST 1.1), comparing them with the advantages gained by adopting newer methods such as tumour growth rate (TGR) modelling.
Limitations of Conventional Imaging Biomarkers
Require Large Patient Cohort
Oncology studies using RECIST 1.1 require a comparative arm to derive robust statistical conclusions. More patients must be enrolled which can make clinical trials more expensive. For rarer forms of cancer, larger patient populations are not available, making developing treatments more challenging.
Inefficient Endpoints in Non-curative Trials
Statistically proving equivalence (or superiority) of the treatment arm compared to the control arm in trials with variability in end-point assessment requires large numbers of patients. This means that conventional imaging biomarkers are not fully suitable for non-curative trials that may need a longer follow up time. For metastatic solid tumours there is a need for newer biomarkers to capture longer survival and correlate with developing biomarkers.
Do Not Effectively Capture Whole Body Tumour Burden
A number of studies have shown that the drugs designed to treat primary tumour may not be as effective in treating distant metastasis.6 There are different escape mechanisms for tumours and heterogeneity between primary tumours and metastasies. Conventional imaging biomarkers do not always capture the heterogeneity and whole-body tumour burden accurately. As more specialised and targeted therapies are developed to treat one portion of a tumour over another, the RECIST protocols cannot efficiently capture tumour heterogeneity.
Lack of Tailored, Patient Specific Treatments / Patient Selection for Trial Participants
Biomarkers should be able to inform patient decisions about their treatment. More effective imaging biomarkers could be used to select patients for whom the trial treatment is most likely to prove effective.
Lack of Flexibility in Trials
Patients participating in oncology treatment studies need to have their tumours measured at timed intervals with little room for flexibility. However, there are many reasons why a patient may miss an appointment for a scan. Patients dealing with a serious illness, its associated financial and psychological burdens, and treatment side effects require more flexibility than RECIST-based biomarker studies offer.
Tumour Growth Rate (TGR) Modelling
TGR modelling provides growth/decay rate imaging biomarkers that can predict patient survival and drug responses potentially from a limited number of patients.
