Download this abstract now
You'll also receive future publications in the series. We will not share your information with any other parties.
Medelis: How should a CMO incorporate this data into the protocol?
Dr. Von Hoff: You establish it as an endpoint. So in addition to all the other endpoints that measure anti-tumor activity in a classic phase I trial, you add a line that specifies that one of the secondary endpoints is the time patients are on treatment with a new agent versus the time they were on their “just-prior drug.”
Medelis: You refer to measuring “time on therapy” as opposed to “time to progression?” What’s the distinction?
Dr. Von Hoff:
With “time to progression” you need to be taking regular measurements such as scans. In comparison, “time on therapy” takes other things into consideration.
Medelis: That turns it into a more subjective, general indicator. Is that good?
Dr. Von Hoff: It helps because it takes into consideration clinical judgment, or people’s observational powers, which can be beneficial in addition to scans.
Medelis: And therein lies the art of the matter: the power of observation.
Dr. Von Hoff: What makes a clinical investigator great at his or her trade is the ability to observe keenly. Careful observation is critical at every stage.
Medelis: Are CMOs starting to do this as a matter of course — using time on drug to demonstrate improvement, then using that data to make a case to potential sponsors?
Dr. Von Hoff: I have written about this but have never seen a CMO plot time on a new drug versus the time on a just-prior therapy. This idea of using the patient as their own control is a lost concept in drug development. Dr. Bob Temple at the FDA, an icon in clinical trial design, calls it a lost art. He’s referring to the ability to document changes in the natural history of a patient’s tumor, and how this information can give you a sense of whether the drug will work.
Medelis: Does the approach of using a patient as their own control still work in the case of cyostatic drugs, which affect cells without decreasing tumor size?
Dr. Von Hoff: Even if an agent isn’t shrinking the tumor, it may be keeping it stable for long periods. If you plot the time a patient is on a new drug versus the time they were on the therapy they just progressed on, you
can gauge whether an agent changed the natural history of the patient’s tumor. And, if at the higher doses, you have more people with longer time to progression as compared to lower doses, then you have a trend that your drug is actually doing something. If the
patient had been on the next drug for only one month, then you know it didn’t do anything for them. But if you can beat what the patient just had, you really have a good prognostic sign.
CMOs and others just aren’t giving this information enough value.
Medelis: At what point would you feel validated to invest more deeply in a particular therapy based on information deduced from using the patient as their own control?
Dr. Von Hoff: If you treat 30 patients and 30% stay on a new therapy for a longer time than the just-prior drug they had progressed on, then that would justify a deeper investment. Patients’ tumors grow at an inexorable rate. If you find an agent that can taper that growth, then it is probably doing something and should be pursued.
Medelis: So the phase I shouldn’t just focus on maximum tolerable dose?
Dr. Von Hoff: Exactly. It’s an opportunity to look for therapeutic effect as well. If it’s there, patients benefit and you have a great start.
Medelis: So the payoff for CMOs changing how they see the phase I could be substantial.
Dr. Von Hoff: There’s no question in my mind. If a CMO started comparing time on new drug versus time on just-prior therapy, they would have a better idea as to whether the new agent had promise. All it takes is a more proactive approach to the phase I.
* * *
Download this abstract now
We'll send you a link to download this abstract along with past interviews in our Peer Perspectives in Oncology series. We will also send you an email when we publish future interviews. Please note that we will not share this information with any third parties; we will use it solely to deliver future abstracts in the series.
