Thanks for having me. Can you see my screen? No, just kidding. We've done this enough, so you should see it.

I'm going to talk about a case of driver mutation lung cancer. It's nice to meet you all. I'm Vinay Prasad. I'm a professor here who attends at San Francisco General Hospital and works in epidemiology.

All right, let's get started. I have a bunch of learning objectives.

I don't know how many we're going to get through, but we'll see what we can do.

One question I always have, and people talk about, is driver mutation lung cancer better

than non-driver mutation lung cancer.

Is single agent pembrylizumab ideal for someone who doesn't have driver mutations but has a PDL1

expression of over 50%, let's say 55%.

Why is disease control rate and overall survival actually unreliable in single arm

studies?

Are waterfall plots visually accurate?

How much time does surrogate approval save?

How often are control arms in randomized trial suboptimal?

How does censoring undo randomization? And what happens to alpha error when you give people second

chances? All right. So maybe we'll unpack these as we go through. I'll keep an eye on the time.

So this is a case of like so often presents a 69-year-old female who has stage four non-small

cell lung cancer, patient presented with the right side at three centimeter lung mass.

I think some hyler nodes were involved.

And we have liver metts and adrenal meds at presentation and associated a fusion, which would

automatically make the patient stage four.

Brain MRI is negative.

The patient presents to the hospital.

...