Mayank, I think I'm give wise (ph) not to elaborate your question because of the sensitivity of many levels. But suffice it to say that we are obviously keeping not just the FDA, but other agencies abreast of developments with CRC and some of the other indications as well. Now one thing that is sure that I think needs to be stressed over and over again, when we treat patients, which are not just metastatic patients, for example, in CRC, but also other indications. But these patients are typically third, fourth, fifth line patients. They have been treated with pretty much everything that's available and either haven't responded or failed after they have responded to these other treatments. So these are pretty sick patients. And the kinds of responses that we're seeing, which are in the neighborhood of 20% to 50% depending on the indication, is really something very meaningful for the patient. And of course, from a regulatory perspective, you may say, well, response is not enough, but please explain that to the patient, that the response is not a good thing. So we are diligently pursuing, of course, the next steps. Response is a very important criteria for patient that has exhausted all options particularly to the kind of durable responses we're seeing of the magnitude that I just talked about, they are very meaningful, but we're also diligently pursuing that these responses will translate to longer term benefit to patients enough. Of course, the data that we showed at ASCO-GI with survival curves indicates, and of course, mind you, this is not a randomized trial but the differentiation in this patient population in terms of overall survival is such that we are confident that the responses are going to translate to longer term benefit including survival. And I may add that with CTLA-4s, typically, you do see response rates correlate very well with other benefits. The same is not necessarily true with other higher treatments or other cancer treatments. But with CTLA-4 targeting agents, generally and that we're going to collaborate on this, that it would be impossible to think about a trial where response rates will now translate to survival. Steven, would you like to bring in your experience with that?