Yes. Michael, thanks for the question. Given the Celgene’s non-inferiority data versus TNF therapy, particularly in RA patients, particularly looking at a venous thromboembolism, malignancy, and then major adverse cardiac events, you asked a question on the read-through to rux itself, and then I guess, potentially to topical rux. We've been with rux on the market since 2011. So we have many, many thousand years of patient exposure, including with our partner, Novartis, as well as long-term follow-up on our clinical trial programs. So let me just talk a little bit about the clinical trial programs. If you look at the COMFORT data in MF, now with five years of follow-up on those studies, there's been no signal for any of those events that are worrying in that particular exposure. In polycythemia vera, the response study, that is a prothrombotic disease, also now have five years of published follow-up, and we've looked across the board at thromboembolic events, cardiac events and malignancies there. And in fact, on the treated arms, the rates are lower in both the primary treated arm, the crossover arm versus the best available therapy arms there. So in keeping as well with our market experience that we're not seeing a signal for any of those events as we asked to on a yearly basis by regulatory authorities. And just remind you, rux cream has no warnings or black box for any of these. Topical rux itself, as we have in two published papers now, one in AD and one in vitiligo, showing that the bioavailability of the cream is about 4% to 7% of that applied, on average about 5%. So – and those are two published papers, one in AD and one in vitiligo. And thus the effective oral exposure there is very small and not at pharmacologically relevant concentrations. So given the parent compound itself not having an issue, rux cream having that sort of bioavailability, and now our safety from those clinical programs, we don't expect any read-through there at the moment.