DR SPIGEL: The idea of trying to use the immune system to fight any cancer is not new. The challenge has been, how do you kind of wake up the immune surveillance to a tumor that’s already established itself in the body? One way that that might be overcome is by interfering with an interaction of a known antigen, PD-L1, with its receptor, PD-1, which when that ligand interacts with that antibody on the T cell, it seems to suppress T-cell activity. So, actually, a number of drugs are in development to target this interaction.
And we first heard about this from Julie Brahmer from Hopkins 2 years ago — actually, at the time of pivotal publication in The New England Journal with nivolumab, an anti-PD-1 therapy. And what we’re starting to see now is the development of other compounds. One of those compounds is named anti-PD-L1. Its more complicated name is MPDL3280A. This is an antibody that, like the nivolumab compound, is designed to target this interaction. But instead of targeting the receptor, PD-1, on the T cell, it’s targeting the ligand, PD-L1, on the tumor cell.
This antibody was first developed in a Phase I study. And the Phase I study was designed to just establish dose and safety, and several dose cohorts were looked at. And then what happened from there were what’s now become so common for us as investigators are these Phase I/II trials, which really have these large expansion cohorts. And what we presented at ASCO this last June were the results of the lung cancer cohort from the expansion phase of the Phase I.
Essentially, the Phase I showed that the drug was safe, that there did not appear to be a maximal tolerated dose, although a dose was chosen for the Phase II development. And in the lung cancer cohort, when you looked at patients who had very refractory disease — so two thirds of the patients had more than 2 lines of prior systemic therapy — looking at those patients in an all-comer fashion and not broken out by PD-L1 expression, there was a response rate of about 20%. And this is somewhat similar to what’s been seen with nivolumab as well, the anti-PD-1 compound, which I’ll talk about in a moment.
Many of these patients had early responses that were sustained. Some patients actually had brief growth in their cancers, something that may be pseudoprogression, before they benefited. And then there were other patients that clearly didn’t benefit. Right away, you knew that they were progressing and not achieving any kind of benefit from this treatment.
DR LOVE: Could I just interject and ask what the current thinking is right now in terms of what’s going on when you see — and I remember Julie had a case actually in the New England Journal paper, where actually it looks like it’s getting worse and then it gets better. Is that an inflammation, or what’s going on?
DR SPIGEL: To me, this is a really fascinating aspect of this therapy and also — if one of these agents or all these agents end up getting approved, a real challenge for how to manage patients in the community.
So the trials have allowed for what’s called pseudoprogression. So that is to say if the patient is clinically stable, doing well, even if there appears to be enlargement of their lesions, they’re allowed to stay on the study, which does complicate how we interpret response rates, the thought being that those patients might actually be benefiting from therapy, that the tumor is being infiltrated by T cells, which is causing this expansion on CT scans.
There have been cases, many anecdotal reports, but certainly documented cases with each of these compounds I’ve mentioned, nivolumab and anti-PD-L1, as well as some other compounds across tumor types of patients who’ve had growth on their scans, whose provider got a biopsy and found that the remaining tumor cell was actually infiltrated by lymphocytes. This has been well established. These are very unusual anecdotal reports of patients with melanoma and lung cancer who’ve had tumors that were thought to be clinically worse/radiographically worse, that actually there were no tumor cells in the biopsy specimens.
DR LOVE: Any way to differentiate clinically, without a biopsy, what’s going on?
DR SPIGEL: I think it’s very challenging and where I struggle really every week in clinic is when I get scans where the patient is feeling well and doing well and they have some lesions that are smaller and some that are bigger, what to do. And I can think of patients — a patient at our center, for example, with melanoma on another trial, who was taken off one of these types of compounds, not one of the ones I mentioned here, because there were new lesions, and ultimately went on to something else, but had biopsies of those lesions and there was no tumor there, and I think a missed opportunity for that patient, perhaps, to have been allowed to stay on therapy and perhaps benefit down the road.
DR LOVE: Can you talk more specifically about what’s been seen, what autoimmune phenomena, if any, have been seen, how it compares between PD-1 and PD-L1 and, let’s say, looking at ipilimumab, the spectrum of these kinds of problems.
DR SPIGEL: Yes. So it’s a little complicated. So with the anti-PD-L1 compound, the safety has been pretty remarkable. There’s been a lot of certainly constitutional low-grade reversible symptoms, certainly things like fatigue, even odd things you wouldn’t expect, like nausea. One interesting side effect that was reported at ASCO was the development of pericardia effusions, not a compilation we’re used to seeing from drugs in lung cancer. And there have been reports of patients who’ve had confirmed tumor cells in the fluid and the cytology specimen from the pericardial effusion that clinically was not recognized before the patient went on study, raising the question of whether immune surveillance identified occult malignancy in the pericardium, as some autopsy studies have suggested in lung cancer, a problem that may be more common than we’ve ever recognized.
DR LOVE: So this is not like an autoimmune phenomena that’s a transudate. This is the tumor?
DR SPIGEL: Unclear, but maybe in these few cases, patients who actually had tumor cells that were unknown in their pericardium that became known because they developed an effusion, perhaps because of the response to therapy. It’s an evolving signal that’s being watched closely on actually all the trials.
DR LOVE: And have these been clinically significant pericardial effusions?
DR SPIGEL: My understanding is, 2 patients had clinically significant effusions, but they were treated and allowed to remain on treatment.
DR LOVE: And what about other autoimmune, like colitis, for example?
DR SPIGEL: The concern so far has been with nivolumab’s development. It’s been also a very well-tolerated, safe drug, but there were 3 patients who developed pneumonitis that led to actually death in 2 of the patients. That was early in the Phase I experience. And as more has been learned about how to manage patients who develop worsening dyspnea, which is a problem for any patients with lung cancer on a trial, is it the drug? Is it the cancer? Is it infection? Is it PE? There are certainly rigid pathways investigators are supposed to follow to deal with patients that develop pneumonitis signs on imaging or symptoms, using steroids, antisuppressive therapy to quickly stop treatment and address those issues for patients before they become more serious.
DR LOVE: And has pneumonitis been seen with anti-PD-L1?
DR SPIGEL: So far it’s not been something that’s been thought to be a problem. And it may have something to do with the fact that targeting PD-L1 may spare pulmonary toxicity where the interaction of PD-L1 and PD-1 may be very distinct from PD-L2 and PD-1, where there may be more of that present in the lungs, creating pulmonary toxicity.
DR LOVE: Maybe you can talk a little bit more about what you presented and what’s been seen in terms of efficacy, but also the concept of combining ipilimumab and one of these anti-PD-L1 or anti-PD-1.
DR SPIGEL: Yes. The efficacy right now is based — whether we’re talking about Julie’s nivolumab, longer-term data set that was recently updated at World Lung or even the ongoing anti-PD-L1 data set that’s been also updated in Sydney. In general, we’re seeing response rates in the heavily pretreated patients across histologies of about anywhere from 15% to 20%. How clinically meaningful is that? Is that really like some older studies where we had single-agent systemic chemotherapy, where you’d see responses in patients who were refractory? I don’t think we know.
I mean, I think it’s remarkable that patients who are heavily pretreated are getting any responses, but more compelling to me are the patients who maybe don’t even get responses but have prolonged disease control on therapies that have very little in the way of serious toxicities. And to me, that’s a big story. Is that going to be enough to improve overall survival compared with docetaxel in a second-line setting? I don't know. And that’s why the randomized trials are going to be so important.
DR LOVE: And, of course, the holy grail of oncology is predictors of response. And one that’s out there on the table is PD-L1. What do we know about the distribution of it, not just in lung cancer, but also in all cancers and, in particular, its predictability and how it’s being integrated into future trials? I was kind of wondering why it’s being used as an entry criteria when we know there are responses in people who are PD-L1-negative.
DR SPIGEL: Yes, you bring up a great point. And it’s why the randomized studies versus docetaxel in second-line lung do not require PD-L1 expression to get on the study, although they are looking at that in a retrospective fashion.
Now, I’m involved — I think Mark is, as well — in Phase II studies looking at selecting patients by PD-L1 expression. Before enrollment to get onto an anti-PD-L1 therapy. So we have trials now at our center in the first-line refractory and even patients with treated brain mets, who have to have PD-L1 expression to get on the study. And that’s based simply on some early data from the Phase I experience that suggest that the patients with PD-L1 expression seem to have a better chance of responding than the patients who don’t express PD-L1.
But to get to your point, there still are patients who are PD-L1-negative, who can respond. So you can’t not study the drug in those patients. And that’s why the pivotal trials include those patients in them.
DR LOVE: What’s been seen also in terms of progression after stopping the drug? How many patients have got out to that point, and do they stay in remission?
DR SPIGEL: So it’s all anecdotal right now. The longest data set we have is with nivolumab. I think Julie, herself, may have patients now going beyond three and a half years. I mean, I’ve heard all different kinds of stories. I have 2 patients on the original nivolumab study who one made it beyond 2 years. Another was right at 2 years but died from an unrelated issue from her therapy or her cancer. So everything else, I think, has been anecdotal. There have been stories of patients on both compounds who have progressed and had a chance to respond again when re-exposed to the therapy, but there are also patients who’ve not responded when they’ve been re-exposed to therapy. It’s unclear how long you need. If these therapies really work, how long you need to be exposed to get benefit and how long that benefit will be sustained when you don’t have the drug in your system.
DR LOVE: What are some of the trials that are going on in this general area that you think are most important to know about?
DR SPIGEL: The big trials I think everyone in here has been probably a part of those. The 2 big trials that have led the pack are — is nivolumab versus docetaxel in a second line-only setting. If you have an ALK rearrangement or an EGFR mutation, you’re allowed to have received a TKI plus first-line platinum chemotherapy. When you progress, you can then go on this to a third-line therapy, but you’re allowed on this trial.
The randomization is one-to-one. It’s not placebo controlled. Docetaxel is every 3 weeks, nivolumab is every 2 weeks, so there’s certainly a bias that’s established there. But overall survival is the endpoint for that trial. There are 2 large trials, one in nonsquamous, one in squamous. The squamous trial finished enrollment Tuesday. The nonsquamous trial is, I’m understanding, about to close to screening any time this month.