On behalf of my co-authors I’m going to present the five year follow-up of CA209003, a study of nivolumab in previously treated non-small cell lung cancer. We’re going to focus on the clinical characteristics of the long-term survivors. This is my disclosures.

Prior to the introduction of immunotherapy treatment options were limited for patients with advanced non-small cell lung cancer who progressed after first line combination chemotherapy. The majority of patients with advanced disease died within a year after diagnosis and the five year survival for patients with metastatic non-small cell lung cancer was approximately 4%. Nivolumab, a Programmed Death 1, or PD-1, immune checkpoint inhibitor antibody showed activity in patients with heavily pre-treated advanced non-small cell lung cancer in the phase Ib dose escalation cohort expansion trial. Here we report updated results from this trial based on five years of follow-up.

Just to remind you of the study design, this trial took 129 patients with advanced non-small cell lung cancer who had been heavily pre-treated with at least one, up to five, prior lines of systemic therapy. They were randomised to three different doses of nivolumab given once every two weeks for up to two years. The primary safety, efficacy and three year follow-up has been previously published and this analysis is based on a minimum follow-up of 58 months.

This is the estimated five year survival of 16%. As you can see here, after three years the survivor curve has plateaued out which is similar to what has been seen in the past in other diseases treated with immunotherapy. If you look at tumour PD-L1 status and look at the overall survival curves, about in general 50% of the patients on this trial with non-small cell lung cancer had enough tumour to stain for PD-L1 using the BMS assay. You can see here between patients who did not have PD-L1 staining on their tumours versus those patients with at least 1% staining in their tumours the five year overall survival was very similar: 20-23%. If you look at the thirteen patients who had 50% or greater PD-L1 staining on their tumours the five year overall survival was 43%.

Here’s a table where we’re comparing all the treated patients versus the sixteen five year survivors. You can see here in general there’s really no difference in characteristics – most of them were former or current smokers, very similar rates of different histologies. Interestingly, in the five year survivor group there were two patients with EGFR mutations where typically we don’t expect patients with EGFR mutations to do well with immunotherapy but, as you can see, two of the sixteen did have EGFR mutations. Looking at the rate of adverse events in these treated patients the rate of grade 3 and 4 adverse events were very similar in the two groups as well.

These are the swimmer plots looking at each of the sixteen five year survivors. Just to orient you from left to right the blue bar is basically the time on nivolumab; the yellow bar is the time off of nivolumab. Twelve of the sixteen patients had partial responses as shown by the yellow circles. Nine patients actually completed the maximum treatment of two years on nivolumab, as you can see here. Interestingly, as depicted by the red bar this is progressive disease. Patient number 2 had early progression and then a non-conventional response to nivolumab as depicted by the blue circle. Most of the patients again were treated with the maximum number of cycles with nivolumab but there were four patients who actually stopped therapy early due to adverse events and have not required any further retreatment, even though they didn’t get the full two years of therapy.

You can see here twelve of the sixteen patients received no further therapy after nivolumab and were without evidence of progressive disease at the time of the last follow-up. Of the four patients who received subsequent therapy I want to point out two. Patient number 5 stopped therapy after two years, the disease progressed approximately a year later, they were retreated with nivolumab, had a response that lasted about another year and then their disease progressed. Patient number 15 had progression, stopped therapy and actually had local progression where the right lower lobe nodule was resected and this patient has not required any further therapy after this.

This is a patient that I treated at Johns Hopkins. This gentleman is a 61 year old gentleman who was a former smoker, had squamous cell histology. He originally was diagnosed in 2008 with locally advanced disease, progressed approximately a year later, was treated with chemotherapy as well as on a clinical trial of an HDAC and a demethylating agent, received nivolumab starting in February 2011 and by April of that year had a partial response. He received the two years of nivolumab and as at this time, six years after nivolumab treatment initiation, the patient remains alive and well and in response without evidence of progressive disease. Colleagues of mine at the Bloomberg Institute looked at the tumour and did whole exome sequencing. They found 314 somatic mutations; based on published data in lung cancer 314 would be expected and qualified as high tumour mutation burden.

So in conclusion the CA209003 study has the longest follow-up survival for an immune checkpoint inhibitor in advanced non-small cell lung cancer. The estimated five year overall survival rate with nivolumab was 16%. The long-term survivors had diverse baseline characteristics; the majority of long-term survivors had durable responses to nivolumab although some had stable disease and progressive disease as best overall response. Nivolumab is an established standard of care in patients with previously treated advanced non-small cell lung cancer based on the results of two pivotal phase III trials. Thank you.
Thank you Dr Brahmer. What you’ve heard about just now are results from the very first trial of any anti-PD-1 drug in lung cancer. The results from this trial really were a landmark in the history of immunotherapy for cancer because this showed for the very first time that immunotherapy could be used to treat common cancers, not just melanoma, not just kidney cancer. So this really brought immunotherapy out of the realm of a specialised treatment now into the broader arena of oncology. As Dr Brahmer mentioned, it led to FDA approvals for lung cancer. So this is the longest follow-up to date on patients with lung cancer receiving this therapy and it should be noted that the five year overall survival that you saw today really quadrupled the survival that we would otherwise expect if these same patients had received chemotherapy.

You’re going to see curves such as the one that Dr Brahmer presented with that long tail, we often refer to the tail on the survival curve, and this really speaks to the durability of cancer immunotherapy with checkpoint blockers.

So I’ll take the first question. The patients on this trial, according to the trial design, received a maximum of only two years of therapy but patients are currently receiving standard of care indefinitely. What you showed is that if patients stopped treatment at a certain point 75% of the patients on this trial maintained their responses without any other cancer therapy. How do you think that currently these drugs should be given and is it possible that we’re giving the drugs for too long?

I agree, certainly right now, at least at the time when the phase III trials were actually designed we didn’t know how long we should really keep the immune therapy ongoing. It was really felt that you needed to continue the immune therapy indefinitely. But based on this data as well as other data we can shorten the time where patients require therapy but we really need to be able to identify those patients who develop memory cells to actually continue to combat their tumour long term.

Now the trials that are ongoing are being amended to decrease the time on immunotherapy. In fact, one trial, CHECKMATE-153, will hopefully read out at the end of the year where they’re comparing continuous dosing of nivolumab in patients with lung cancer versus one year of therapy. We’ll see comparatively if there’s any difference in long-term survival. But I think, based on this study and others, we can safely say that they don’t need to be treated indefinitely, at least that’s my personal feeling but, again, we want to be able to better personalise this therapy and be able to tell when is the best time we can stop therapy.

Open it to the floor for questions.

[Audience member] Hi, thank you. Elaine Schattner, contributor to Forbes. The most impressive data looks like that which is for patients who had PD-L1 levels of greater than 50% where you’re calling the overall survival at five years 43%.

Correct.

So was that statistically significant? It’s a very small number of patients, thirteen, and are you looking at that prospectively now?

Yes, we’re looking at this prospectively but due to the small number of patients there was really no statistical differences between the groups. We saw improved survival across the different PD-L1 staining but, again, this was just 50% of the patients with lung cancer on this trial.

[Audience member] Alice Goodman with the ASCO Post. You told us about the characteristics of long-term survivors. What are the take home messages? What characteristics do you think are important and what’s going to be followed?

Right now there’s no characteristic that really stands out to be able to predict who is going to be within the long-term survival category. Certainly the majority of these patients were former or current smokers so we can’t really compare to never smokers. The thought is that the former or current smokers tend to have a high mutational burden and maybe that equates to long-term survival but we just can’t compare. We have to really look at the phase III trials that recruited many more never smokers.

[Audience member] Ted Bosworth, Clinical Oncology News. Last year we heard similar data from melanoma, long-term follow-up of melanoma. The tail seemed to start a little bit earlier but these are completely different diseases. But your tail drops off suddenly after five years, I was wondering if you could just explain that?

I think most of those… only two patients have passed away after five years within the sixteen patients. So, again, it’s very early on and we really need to have long-term survival follow-up. It’s just an exciting time to be able to have long-term survivors to follow-up.

[Audience member] On the basis of these… Charles Bankhead, MedPage Today. On the basis of these results do you think that testing for PD-L1 expression should be required in all patients now before you start the drug?

I think in the second line treatment setting I do not feel that in patients with non-small cell lung cancer that PD-L1 status is required. I do not feel that you need it to be able to start any therapy except for pembrolizumab approval is based on PD-L1 status. That may be different right now in the first line treatment setting where the only drug that’s approved is pembrolizumab based on a high PD-L1 staining result.