Clinical trials can be classified as:
• phase I studies
• phase II studies
• phase III studies.
In addition, some phase III studies are sometimes referred to as phase IV or post-marketing studies.
No study should be started without a protocol that describes in detail:
• the aim of the study
• the patient eligibility criteria
• the screening and follow-up studies
• the treatment
• the criteria to score toxicity and activity.
In addition, rules for informed consent procedures should be specified. Trials of any sort should have approval by a properly constituted ethics committee.
All of these criteria have been specified in guidelines produced by the International Conference for Harmonisation for Good Clinical Practice (ICH-GCP). They are also now embedded in European Union (EU) legislation on the conduct of all trials of new therapeutics.
Phase I studies
Phase I studies are human toxicology studies. Their endpoint is safety, and they usually include 15–30 patients. They are designed to define a feasible dose for further studies. These studies begin at a dose that is expected to be safe in humans. Dose escalation is usually between cohorts, and infrequently in individual patients. It can be:
• according to the Fibonacci method (the dose is escalated in decreasing percentages of the previous dose, i.e. 100%, 66%, 50%, 33%, 25%)
• according to pharmacokinetics (pharmacokinetically guided dose escalation, PGDE), using a method that combines statistics with the experience and expectations regarding side effects (continuous reassessment method)
• variation on these methods.
The aim of the phase I study is to describe the side effects that limit further dose escalation (dose-limiting toxicities, DLTs) and to recommend a dose for further studies with the drug or the new administration method (maximal tolerated dose, MTD).
Phase II studies
In phase II studies, the anti-tumour activity of a new drug or method is the endpoint. There are various statistical designs, including 14–60 patients on average. With the emergence of drugs that create tumour dormancy, rather than cell kill, the endpoint of time to progression becomes important. This is the time from the start of treatment, until the the first evidence of tumour progression. In addition, phase II studies can provide information on side effects related to cumulative drug dose.
Phase III studies have either the time to progression or the survival time as the 1° endpoint. Phase III studies always include randomization against a standard form of therapy, or no treatment when no standard therapy exists. 2° endpoints, such as toxicity, pharmaco-economics, and quality of life, are often included. Phase III trials can involve between 50 and several thousands of patients. The number of patients is dependent on the size of the difference expected/clinically important. Cancer trials have often been criticized in the past for being too small to find realistic differences between therapies. Breast cancer studies involving many thousands of patients have been able to define the long-term benefits of hormone therapy and paved the way for larger-scale trials in other common tumours. In the modern era, many large-scale cancer trials are performed, so that the true level of benefit of a new approach can be proven and to allow for the regulatory approval of new agents.
Most cancer treatments produce unwanted toxicities that interfere with the patient’s quality of life. In many cancers, the benefits of new treatments over existing approaches have been modest. As new cancer treatments are developed, a common problem is the comparison of a novel intensive treatment regimen against a relatively less toxic standard. In such circumstances, if the survival gain from the new treatment is reliably established, but of modest magnitude, then it may be questioned whether the gain is worthwhile for individual patients. In weighing up the risks and benefits of all treatments, it is important to consider many aspects such as:
• the duration of treatment
• the length of hospital stay
• the number of clinic visits
• the short- and long-term toxicities
• the less clinical aspects (perhaps less well appreciated), summarized as QoL.
Assessing health-related quality of life
Several questionnaires for completion by patients have been developed. The European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Study Group has developed a core questionnaire—the EORTC QLQ-C30—to which are added disease-specific modules. A patient who scores high for global health status/QoL is deemed to have high QoL.
Difficulties in quality of life assessment
• Compliance declines, as the patient becomes terminal.
• Compliance may also be poor if the patient feels well.
• Can a surrogate, relative, nurse, or physician fill in the form?
There are important challenges in reporting QoL outcomes in clinical trials. These include the description of compliance, summarizing longitudinal data in a complete, yet clinically meaningful, way, balancing the multiple endpoints under consideration and, perhaps most importantly, relating the findings with regard to QoL to other treatment outcomes such as patient survival and treatment-related toxicity.
Attempts have been made to integrate QoL and survival data into quality-adjusted life years (QALYs). The duration of survival is adjusted according to periods of different levels of QoL before summing, to give the OS time for analysis. The final QALY can then be used for a comparison between treatments, embracing both survival effects and changes in QoL.