Key take-home messages from the lung cancer session, ESMO Signalling Pathways Symposium on targeting the HER/EGFR family

This text presents the key points from the lung cancer session held during the ESMO Signalling Pathways Symposium on targeting the HER/EGFR family of receptors in breast, lung and colorectal cancer (1-2 March 2013, Sitges, Spain). Prof Rolf Stahel, session moderator, summarises the key topics of the interactive discussion. Herewith, we present his talk.

Introduction

Prof Stahel:I will start with a case of a 58-year old woman with an EGFR-mutated adenocarcinoma of the lung who received chemotherapy, which was the standard of care at that time (January 2009), 4 cycles of chemotherapy (platinum combination), and bevacizumab and then was maintained on bevacizumab over one year and the tumour progressed. Now, one knows what the molecular characteristics of the tumour are and institutes treatment with erlotinib; in this case, it could also be gefitinib, and the tumour shrank. The patient became entirely asymptomatic at that time (July 2010).

As time goes on, and this is now awhile, it is January 2012: The patient is again symptomatic and has progression of the disease in the lung. She is treated with chemotherapy again, a combination chemotherapy similar to the one given at the start (carboplatin/pemetrexed) and the disease goes away. This time maintenance with erlotinib is instituted. During all this time the lady had a full life, worked 100%, except for a few days when she came, she was ill and when we did a CT scan. So, upon tumour progression in October 2012, a long time after tumour diagnosis, we institute chemotherapy with a single agent, gemcitabine. The patient becomes asymptomatic, the CT scan shows decrease of the disease. So now we discuss: “Should we maintain her on gemcitabine, gefitinib, afatinib or back to erlotinib?”. It is a 4-year clinical course of the disease.

Current characterisation of lung cancer

Prof Stahel: In our session, Dr Enriqueta Felip had the task of speaking about how we characterise lung cancer these days. For those that do not treat lung cancer, you would be surprised that the lung cancer drug indications were not even distinguished between adenocarcinoma and squamous cell carcinoma before they became available; that bevacizumab was just registered for lung squamous cell carcinoma and pemetrexed was more active in one group than the other.

This is now clear, and today Dr Felip has showed us that the standard is EGFR mutation testing and ALK rearrangement examination in the tumour, if you have access to an ALK agent, and this is not given in all countries yet.

We also discussed targetable oncogenic drivers and actionable mutations; I’m a little confused with this wording, because you can act on a lot of those mutations in a clinical trial but it doesn’t help the patients. So I think we need to find a good way to label those mutations. At least in lung cancer there is a series of rare mutations and then there is a series of rare translocations, which may be, or are already, targets for treatment. The question is of course --in a standard university hospital, “How much should be examined beyond EGFR and ALK and how is it evolving?”.

Again, same as in breast cancer, PD-1 (PDL-1) is going to be on the horizon and it is very important that there seems to be tests evolving that may allow us to predict a response to the antibody. But this is very early and the tests are not standardised at all yet.

So to summarise, we appreciate the histology and all the new things that we need to do. Molecular testing needs a strong work collaboration within the team, which I think is given in most thoracic oncology centres these days.

First-generation EGFR inhibitors: The breakthrough in stratified medicine in lung cancer

Prof Stahel: Dr Benjamin Besse had the task to talk about the first generation of EGFR inhibitors, really a breakthrough. The initial studies on EGFR inhibitors were randomised trials of the EGFR inhibitor given second-line versus placebo. One study was positive with erlotinib, one study was so-called negative with gefitinib; both studies showed a marginal improvement of survival or a significant improvement of survival compared to placebo given as second- or third-line treatment. This was not considered targeted treatment.

The real change came at ESMO 2008, when Dr Tony Mok presented his study that was done in what we now know is an enriched population of patients with EGFR mutations. This study compared a first-line gefitinib treatment with a standard chemotherapy and it showed a superiority in progression-free survival and response with gefitinib treatment. Now we know --and there is no need for another trial as there are 7 trials which demonstrated the better response and highly better progression-free survival with first-line targeted treatment (gefitinib, afatinib, erlotinib) versus chemotherapy.

Dr Besse was a little provocative: the oncology community and the consensus in fact say that the first-line treatment works the best for the patient, so you give the best treatment first and the second best afterwards; he was provocative because he said you can also do it the other way. It’s a point for discussion because the overall survival in all these studies is the same.

I think it is important to show what is being discussed in the lung cancer area. What is important is the current guidelines: in patients with non-squamous cell lung cancer tested for EGFR mutation, if positive, give first the targeted treatment. We will re-discuss this in May [note: EMCTO, Lugano, 9-11 May].

Now all these trials show no survival benefit; how are we ever going to get those drugs approved if there is no survival benefit? The reason no benefit has been shown is that nearly all the patients have a cross-over, and ethically we should never participate in studies with active agents which prevent the patients to cross over to the study arm receiving the active treatment.

Currently, the only data that we have from the past confirming that gefitinib gives nearly a doubling in survival comes from Japan; the survival was doubled by about two-fold after gefitinib was introduced and approved for use. At that time we didn’t know the right way to use the agent so I’m sure the results are quite better now. 

You should not give these agents in the adjuvant setting, they don’t cure lung cancer. An early trial of gefitinib in unselected patients versus placebo showed worse survival in patients with a sensitising mutation. So, the data are not closed, but in no instance should a patient be treated outside a clinical trial in an adjuvant fashion with targeted EGFR.

To summarise, we have to identify the mutations in tumours; there is controversy about how to start, and we shouldn’t use targeted agents in localised disease. But this is not the major message; targeted therapy is a great advance, offering great improvement in survival for patients who have the right genetic characteristics.

New EGFR inhibitors: The cutting-edge frontier

Prof Stahel: Cutting-edge frontier with new inhibitors was the topic that was given to me and I’m just going to discuss how we can improve on what we have now. I want to briefly show that there is a mutation which is present in over 60% of tumours that have an activation mutation, which is a resistance mutation called T790M. Dr Rafael Rosell called it a sub clone. This resistance mutation carries some clinical characteristics with it, like it is a less aggressive tumour, and erlotinib or gefitinib is still active but less active compared to the patients in whom you do not identify the resistance mutation. So the question is: “Is there a more effective treatment for this patient group?”. There are preclinical data from Japan and from a group at MD Anderson that maybe a combination of vascular inhibitor activity may be of benefit and there is an ETOP trial that is being conducted in Spain, Switzerland and Italy --with other countries joining, which is looking in this issue with a lot of secondary biomarker endpoints.

The researchers from the University of Parma (Italy) said the first generation inhibitors were selective but reversible; the irreversible inhibitors have a different chemistry and they should provide a benefit for those patients who have a resistance mutation; these work in vitro, the data are not so convincing in vivo. In the clinical data so far there has been no good large patient cohort so there is not a final conclusion, but it doesn’t look too encouraging.

However, what has not been tested yet is the second generation of irreversible inhibitors in a phase 3 study versus the first generation. This is on-going, for all comers as well as for mutated patients, and we will know the results in 1 or 2 years whether the second generation has more toxicity, but also a potential benefit.

To move from here, there are very interesting data just published, on the mutation selective TKIs. They spare the wild-type, so theoretically some side effects would be spared, and just hit the mutations including the resistance mutation. This is a preclinical finding, we don’t have clinical data yet but this is the way things are developing.

HER2/neu in non-small cell lung cancer

Prof Stahel:Dr Solange Peters talked about HER2. It’s been known for a long time that HER2 is present in lung cancer, there were trials done in the United States and in Europe which couldn’t show a benefit. Obviously not, because the trials were not done in selected patients. Also Dr Peters clearly discussed that HER2 amplification may not occur at the same level in gastric and breast cancers as in lung cancer. It is not known, there are just no data and no criteria yet. So the lung cancer community certainly is interested to review this area using all the new agents that are currently available. How frequent is HER2 amplification? One study shows that it occurs in 6% of patients that don’t have another identified mutation. And what is interesting, in lung cancer it appears that those patients without an activating mutation in exon 20 are not the same patients who have a gene amplification. The proportion of patients that are immunohistochemistry positive or FISH positive is quite large so there maybe two populations for clinical trials. 

There’s not much data regarding HER2 in lung cancer and it’s a large undertaking that France, Switzerland and Spain have tried to collect their patients retrospectively, putting clinical data together. This will soon be available in JCO to read in detail.

But it appears clear that afatinib has some activity in HER2 lung cancer and there is a case report out on trastuzumab with chemotherapy. So the conclusion here is that trastuzumab is currently being tested as a single agent and in combination in certain clinical studies, pertuzumab is being tested and afatinib hopefully will be tested, so we will hopefully have, in a certain time, some data. The question is: “If it is a rare disease and everybody does a different kind of clinical study, how long is it going to take to make any advancement?”. Again this shows that if we do not band together we are going to lose.

Mechanisms of primary/secondary resistance to EGFR inhibitors

Prof Stahel:There is resistance; as in all other diseases patients, or their tumours become resistant to treatment. How to deal with this is certainly a topic. I hear everywhere that re-biopsy will lead to eventual treatment. I just want to say that I am sceptical: if you have a patient with a disease seen in 2% of all patients and then resistance mechanisms develop in 10 different ones, how rare are those patient subgroups that you have to identify to determine a specific treatment?

A big issue is that RECIST is useless for the patient care and we need to find out how we’re going to deal with this. So RECIST response is the same when an EGFR mutated tumour shrinks back to practically nothing and then it grows 2 millimetres: this is a progression. But the community has learned that in no way should we stop treatment at that time, because you should only stop the treatment when the patient really becomes symptomatic or if it’s localised disease like in GIST, you provide local treatment (radiotherapy or surgery). This has been demonstrated in cases and case series, so a big issue in lung cancer is how to deal with this.

What we have also learned in the past is that if the treatment doesn’t work, with RECIST progression, you stop the chemotherapy. However, 25% of patients who stopped EGFR TKI had a rapid progression, some needed to be hospitalised, and some died. So how to stop, how to overlap treatment is an issue that is under research.

A case that Dr Silvia Novello brought up involved a patient in whom EGFR treatment was stopped and after 21 days massive disease progression was seen; treatment with TKI was resumed and a large improvement was seen by day 42. So again, the old rules of chemotherapy do not apply anymore and we don’t have good definition of how to deal with this issue. Very strange things happen in lung cancer. Small groups of patients progress with small-cell lung cancer and the small-cell lung cancer is EGFR mutation positive. 

Another issue is EMT (epithelial to mesenchymal transition) has to be associated with AXL and with other things, so several investigations involving MET are going on. And every month at least there is another mechanism, preclinical or in a model, and we have a hard time to grapple with it.

What is the current understanding of targeting? This is an interesting thing: afatinib is a second generation inhibitor and it was shown preclinically that when combined with cetuximab, it provided activity in T790 mutated tumours, but the combination with erlotinib did not. So this has been taken from the mouse and into the clinical setting; it’s toxic but this combination is a very effective treatment.

So we need more information, but what to do with the information is a different story. And yes, we want to be clever and molecularly driven but we have to consider all the information together to make an advance.

Building on EGFR inhibition: Synthetic lethality strategies

Prof Stahel: Dr Raphael Rosell and his group are always at the forefront. What he says here is that chemotherapy is not good for lung cancer. The idea is that chemotherapy may remove a lot of tumour but the tumour that is left may get more aggressive. He provided an hypothesis about that and discussed how this can be looked at. So he provided an update on synthetic lethality strategies. He talked about an erlotinib/bevacizumab trial, a completed trial of olaparib/gefitinib and other ideas on a TGF-beta inhibitor, AXL inhibitor or JAK2 inhibitor together with erlotinib, all based on preclinical data; all very good ideas. Of course this is the way we have to move, but again the subgroups are small and we need to have large networks to obtain improvements. 

I don´t know whether BIM is important in other treatments, but BIM has been beamed up in lung cancer: if BIM is not there, the cell cannot undergo apoptosis. So in the case of downregulation of BIM, you can give the best erlotinib treatment or whatever: if the cells cannot die, there is not a good response. This has been looked at in several series, including the EURTAC retrospective analysis, and that to have high BIM is important for erlotinib and for chemotherapy to have benefit.

You know a cell is an EGFR target; some pathway is inhibited but this may result in inhibition of the EGFR pathway, to an expression of all the receptors and all the ligands, it may result in a downregulation of BIM and this is what the group in Spain is intending to study at the forefront. Thinking about re-biopsy, when you give a treatment you cannot re-biopsy in the next day or next 2 days, so they will look at RNA, messenger RNA in plasma in an exploratory study.

In conclusion, there is a lot of progress, and to overcome resistance will be complex. We need to establish some ideas regarding what we should test for in the mutation area in Europe; I know we have broad testing in place to identify the rare patients and develop smart clinical trials together with pharma.