In a clinical trial, the endpoint is an event or outcome meant to be an objective measure of whether or not the drug or treatment being studied is in fact effective.
If you have a drug and you run a trial of that drug in patients, you need to have a goal in mind, a goal of what the treatment should achieve, and the endpoint is the goal. Contrary to intuition, there are several endpoints to choose from, and none of them are 'cure rate'. In this entry we'll look at two endpoints.
The US Food & Drug Administration wants to see a 'clinical benefit or improvement' from the drug. The endpoint preferred by the FDA in a cancer treatment clinical trial, and the most reliable measure of efficacy, is called 'overall survival.' Contrary to terminology, 'overall survival' isn't measured by how many people survive their cancer, or by how many people are effectively treated by the drug, but rather by how many people die—of any cause—among the group of people receiving the treatment during a pre-established period of time.
This endpoint is dependent upon time. If your study's final follow-up is 50 months, you can objectively determine how many people who received your drug in your trial are alive 50 months after they were randomized into a group to receive the drug. The presumption is that, the higher the number, the more effective the drug.
Time-dependent endpoints aren't always convenient in the search for cancer treatments. In fact, either cancer has proven especially difficult to treat or drug companies have not shown sufficient innovation, or both; in any event, cancer treatments have forced the FDA to rewrite the rules on endpoints to accommodate drugs that do things other than (plausibly) extend survival times.
While some endpoints rely upon time, others rely upon tumor assessment. One example is objective response rate.
The Objective Response Rate (ORR) measures the drug's overall effectiveness in, for example, shrinking a tumor by some pre-determined size over a pre-determined amount of time. In other words, everybody had a PET-CT scan before the trial started, and three months into the trial everyone is scanned again. How many people had their tumors shrink by, say, at least 20 percent? This is your ORR. It will include patients whose tumors shrunk by twenty percent (partial response) and patients whose tumors are gone (complete response), along with everyone in between.
Sometimes, neither time nor tumor assessment is used as an endpoint; sometimes, a clinical trial endpoint will be an assessment of symptoms. Considered a clinical benefit, symptom assessment generally will not be a primary endpoint because it relies on what the patient says, which throws the term 'objective' out the window. Rather, it is more likely to be a secondary endpoint.
Health professionals are keen on stressing the importance of clinical trials, but this emphasis hasn't proven to be very effective, since in the US less than five percent of (estimated) eligible cancer patients join a clinical trial. It's hardly a secret why: patients regard cancer clinical trials as last-ditch efforts at survival, as the thing to do when you've run out of other, more hopeful things to do.
Health professionals don't really help the matter either, since oncology treatment guidelines don't recommend clinical trials for cancer patients unless they have in fact exhausted at least a couple lines of therapy.
Neither do drug companies help the matter, often by applying a laundry list of restrictions that effectively exclude the majority of patients. You can read that as a scientific step towards instituting the proper set of controls in a study, or you can see it as a veiled form of selection bias in which the company wants to make sure it has as many favorable patients as possible to make their treatment look as good as it possibly can.
Beyond that, the overwhelming majority of experts believe that clinical trials offer patients access to a level of attentive medical care that they won't receive anywhere else. Patients and caregivers interested in clinical trials can search for them using the National Cancer Institute's search tool.
Sources: FDA, NCI, NCCN