We need faster endpoints for targeted cancer drugs, and there is one
I spent more than a decade reviewing marketing submissions at the EMA, and I think we need faster endpoints to measure how targeted cancer drugs can benefit patients. We can create one by combining two commonly used measures.
Objective response rate (ORR) is the proportion of patients whose tumors shrink by a predefined amount for a minimum period. The duration of response (DOR) is the time from treatment response to progression or death. The combination of these two endpoints—sometimes called “durable ORR”—can establish the duration of patients’ response to an experimental drug.
In the right circumstances, durable ORR could support standard approval just as well as progression-free survival (PFS), which the FDA and EMA currently rely on. Durable ORR can demonstrate an agent’s clinical activity even where there is not enough PFS data, giving oncologists additional treatment choices and improving patients’ lives.
What’s wrong with currently used endpoints?
The ideal outcome for cancer patients is to live longer with a reasonable quality of life (QoL). That’s why the gold standard clinical endpoint for cancer trials is overall survival (OS), the time from the start of treatment to death. But proving a clear OS benefit requires:
- Enrolling a large number of patients (difficult for orphan and rare cancers)
- Following patients for years (in the case of some slow-growing tumors)
- Restricting patients to a single treatment during follow-up (which is unethical), and
- Conducting a randomized, controlled trial in which some patients would get a placebo or a therapy with little impact on OS
So since the 1970s, regulators have accepted PFS and ORR as proxies for OS.
For standard approvals, regulators use PFS because it allows smaller, shorter studies with less follow-up. But it is flawed. For example, it does not correlate well with OS, especially for targeted cancer drugs, and it is not meaningful in non-randomized trials. PFS cannot account for the fact that a slow increase in tumor size and growth does not aggravate symptoms in some tumor types. And for first-line treatment of some cancers, PFS requires an extended follow-up.
For accelerated approvals (AAs), regulators use ORR on the rationale that it’s an intermediate endpoint that could reasonably predict clinical benefit.
Why durable ORR is a useful endpoint?
1. It’s clinically relevant for patients. A high response rate that does not last long is meaningless to patients and may be offset by toxicities. But if a patient experiences a complete or partial tumor response that lasts, it’s common sense that their QoL will be better for the same reasons that support PFS use. The patient is alive, without progression of the disease, and probably with a reduction in symptoms.
2. It could work in support of regular approval. Durable ORR could be used to establish efficacy under the following conditions: 1) There is no existing drug that significantly improves PFS/OS; 2) The experimental agent produces a dramatic and durable ORR compared with established treatments, and; 3) There is a high unmet medical need.
3. There is regulatory precedent for using it. The FDA and EMA have granted full approval to several drugs that used ORR as the primary endpoint with DOR as a key secondary endpoint to show efficacy, including two precision medicines for ROS1-positive metastatic non-small cell lung cancer; Xalkori (crizotinib) in 2016, and Rozlytrek (entrectinib) in 2019. The FDA and EMA also approved two chimeric antigen receptor T-cell (CAR T-cell) therapies in 2017 and 2018 based on ORR and remission rates, an endpoint that closely aligns to durable ORR. Kymriah (tisagenlecleucel) was approved for acute lymphoblastic leukemia and Yescarta (axicabtagene ciloleucel) for diffuse large B-cell lymphoma.
Durable ORR could enable physicians to bring better care to more cancer patients more quickly in the right circumstances.