Clinical studies plan to enroll a set number of patients, but often analyses are done after a certain number of patients enroll, which is called an interim analysis. At the time of interim analysis, it may be determined that the experimental treatment is too toxic, or less efficacious than planned. The researchers may then need to close the study early. The interim analysis may also determine that the treatment or device is far superior than anticipated.
This would also be justification for early closure. If no unexpected findings are discovered at the time of interim analysis, the study will continue to add patients. Other reasons for early closure of a trial include failure to enroll enough patients in a reasonable time-frame. This could occur if the eligibility criteria are too strict, or if clinicians are reluctant to enroll patients on a trial for any given reason.
Meta-analyses are not clinical studies, but rather an analysis that combines the results of prior Phase III studies that have completed their patient accrual. Statisticians and epidemiologists instead of clinicians or scientists often lead Meta-analyses. Meta-analyses are often performed if the results from studies conflict with other studies and/or many studies show a possible benefit of a drug or treatment, but the results are not statistically significant. Ideally, a meta-analyses will pool the actual individual patient data (as opposed to the reported study outcomes) which allows for a greatly increased number of patients to analyze, and therefore greater statistical power.
When it comes to cancer care, there are certainly controversies about the best approaches to study novel treatments. While Phase III studies and meta-analyses are considered gold standards, Phase II studies can potentially meet accrual goals more quickly (simply by virtue of being a smaller study). A promising drug or treatment may prove to be highly effective based upon Phase II studies, which then poses the ethical question of whether or not a Phase III study is warranted, since a Phase III study necessitates that some patients be randomized to not receive the study drug or intervention.