Naiyer Rizvi, MD: One of the challenges is knowing how long to continue the immunotherapy. Some of the trials were stopped at 2 years. Some of them are continuing indefinitely. If you have a great response to immune checkpoint blockade, what are you doing? Are you continuing it indefinitely, are you stopping it, or what's your plan?

Jacob Sands, MD: This is a tough one. Generally, at 2 years, if patients have done well on therapy, and you have 2 years of ongoing disease control, in most patients we are stopping at 2 years and then monitoring them. We'll see increasing numbers of these, but in what we've seen so far, it seems like generally patients continue to do well. And there are people who I'm monitoring off therapy at this point, for many months. So it seems like there's variability across the country in how this is being done, and it's a good area for further study.

Naiyer Rizvi, MD: I think another point to make along that line is that patients who are on immune checkpoint blockade can get isolated areas of progression, which can be managed locally. And these patients still can do really well with continuing on the immunotherapy alone. The most common site being nodal disease, which we'll frequently radiate or resect. Josh, your thoughts on duration of treatment?

Joshua Bauml, MD: I used to feel the same way, Jacob. CheckMate-153 really changed my mind about this. This is a study, of course, that looked at patients who were on nivolumab, who had completed 1 year and randomized them to either continuing the nivolumab in perpetuity or stopping. Now, granted that's 1 year and not 2 years, but the hazard ratio for the progression-free survival strongly favored continuation of the nivolumab indefinitely. The hazard ratio was something like 0.4, and that really scared me. And most importantly we saw that difference in outcomes regardless of whether the patient was in a CR [complete response], a PR [partial response], or stable disease. Patients weren't able to regain the response when they were re-exposed. So if I have a patient who is tolerating treatment and has a response, I tend to continue until toxicity or progression. And I think that this is a different disease, so when we look at the melanoma literature, when you stop at 2 years, 90% of patients remained in response at 1-year follow-up. I know we're seeing some data from the lung cancer study showing some retained response, but I am concerned about patients losing their response when we stop treatment. So I do tend to continue indefinitely, and I recognize this is an area where research is required.

Leora Horn, MD, MSc: Yes. I think I use CheckMate-153 to say 1 year wasn't enough but 2 years is. We have a lot of patients at our institution who are on CheckMate-003 and KEYNOTE-001, and I have some patients who have been off their checkpoint inhibitor for 8 years now. I don't even know, should I scan them every 6 months, which is what I keep doing, or just go to an annual CT [computed tomography] scan? But we have so many patients now who have been off their checkpoint inhibitor for years, and so I stop pretty much everyone after 2 years. Now, if I go and radiate someone at 18 months, I might keep them going 2 years past that 18 months because that was progression in my mind. But if they've had 2 years of therapy and they're doing well, I set them free. I scan them a little bit more often at the beginning, but eventually they get out to that every-6-months scan.

Jacob Sands, MD: Yes, and I guess the question then becomes, if you're not stopping at 2 years, Josh, is there a point at which you are? Like, would you treat out for 8 years, or are there some of these people who might possibly be getting cured of their incurable disease? If that's happening, this is a small subset of patients, but is there a point at which you are maybe not benefitting them by ongoing treatment?

Joshua Bauml, MD: I think quite possibly. The problem is that from my perspective, I don't think that that's a time cutoff. I don’t think it’s: 1 year is not OK, 2 years is. I think it's: there are people for whom you can stop treatment, and I at this time do not have any way of knowing who that person is, and because I don't, I guess I would keep treating. It reaches the point of ridiculousness, though, you're right, if we're getting at 8 years, 10 years, are we really treating at that point? I haven't had to deal with that yet, and I’m hoping that by the time I do, I'll have a better answer. So if one of you guys could do a trial and answer the question, I'd really appreciate it.

Naiyer Rizvi, MD: There are some data I think from Memorial Sloan Kettering Cancer Center recently looking at plasma DNA on patients who are out long term on I/O [immunotherapy], and if you're negative, you didn't recur, whereas if you're positive, you're more likely to recur. It doesn't necessarily tell you whether you could or should stop or not, but I think if you're clearly negative, you may have a greater comfort zone around stopping, but that comes down to which is the best assay to use for that assessment and a more complicated discussion.

Jacob Sands, MD: I like studies like that. I think that's good, and certainly if it was positive at that time, that would make you more uncomfortable with stopping. I think that's fair.

Transcript Edited for Clarity