Rociletinib versus AZD9291 race to market heats up in T790M lung cancer at EORTC
The embargoed press release and abstract for Clovis’s CO–1686 (rociletinib) in advanced lung cancer patients with and without the T790M mutation, originally scheduled for Friday morning in Barcelona, was released last night. The actual presentation is slated for Friday, November 21 during the Plenary session from 11:00 to 13:00 CET.
Thus the ongoing race to market in this segment continues apace, as do the fairly robust and determined discussions on the topic. Without much further ado to read more about our quick insights and reactions to the data, you can sign up or sign in below.
This content is restricted to subscribers
4 Responses to “Rociletinib versus AZD9291 race to market heats up in T790M lung cancer at EORTC”
While the quote about censoring you provide is accurate, it separates from the reality for most oncology drug trials. If a patient is lost to follow-up in an oncology trial, the statistical analysis plan most usually specifies they are marked as progressors at the time of last follow-up instead of being specified as censored. Therefore, the ticks in almost every PFS curve you see will represent a patient who has not progressed as of the corresponding day on therapy on the x-axis.
This does make a difference to the Kaplan-Meier curve shape. A censored patient removes 1 from both the numerator and the denominator. A progressed patient removes 1 only from the numerator.
It is generally folly to try to extrapolate a “new” Kaplan-Meier curve based upon trying to divine where censored patients appear. (Worth noting you know this very well and didn’t attempt this!) This is particularly true for smaller data sets like with these Clovis data. About the only thing you can do is compare the number of censoring ticks to the total size of the database. The larger percentage of patients censored, the less mature the data are.
David, agreed. The deluge of questions that followed this piece from people trying to interpret the ticks always makes my heart sink. There is absolutely NO way that one can read more into the censoring with immature data and which way they will make the curves move. Same with the impact on median PFS – that can shift as the later patients who were enrolled are aggregated into the dataset. We both know that this is simply a snapshot in time and is subject to change…
For all those who are trying to read the tea leaves (ticks), please don’t. Just wait for the mature data.
You correctly pointed out that there are fewer off target effects with CO-1686. There is no rash and diarrhea but there is hyperglycemia. You made it sound like they are rather equally manageable. I contend that for doctors and patients it is clearly easier to deal with hyperglycemia than EGFR rash. While tight glycemic control is necessary for a patient with long life expectancy to prevent such microvascular complications as nephropathy and retinopathy, it’s no necessary in a patient with life expectancy of less than 2 years. In a lung cancer patient with increased blood glucose my goal is to prevent diabetic ketoacidosis whose risk increases with blood glucose over 500. As long as the sugar is below that, there is no problem. EGFR rash, on the other hand is much more difficult to manage. The current treatment is Doxycycline and Elidel cream. The latter is not usually covered by insurance, is expensive and the rash rarely resolves completely. Worse, the rash is unsightly and contributes to the cancer stigma. There is no effective treatment for EGFR paronychia. Overall, the only thing that makes rash better is time. And that’s what I tell patients to encourage them to continue taking Tarceva.
Thanks for your detailed perspective, Dr Kreditor. I’m not sure that they are equally manageable (apologies if I gave that impression!), more different types of events requiring individual approaches. Certainly, the Academic experts we spoke to at AACR and ASCO earlier this year were not bothered by either. Interestingly, similar discussions with oncologists in the private setting definitely tended more towards preferring to deal with rash than hyperglycemia so that’s where that perspective came from.
The good news is that the incidence of hyperglycemia reported earlier this year with rociletinib seem much lower with the 650 and 500 mg doses than 750mg. There were more at the higher dose, for example, causing initial consternation. If the current trend continues at the lower dose then most likely the concern will abate over time.
Comments are closed.