Dr Yu: At this ASCO annual meeting， you reported the interim results of JAVELIN Bladder 100， which was one of the few late-breaking abstracts. There have not been major advances in first-line bladder cancer treatment for 20-30 years. This study is also the first phase III trial to show OS benefits with immune-oncology (IO) maintenance in platinum sensitive patients. First of all， please share with us the backgrounds and main results of this study.

 

Dr Thomas: Thank you so much for inviting me to talk today. I am really honored to be here. I would like to talk very briefly about the work we presented at ASCO this year. The work is a randomized phase III study in metastatic urothelial cancer. The patient population is patients who have just completed first-line chemotherapy for metastatic disease. We sequenced avelumab， a PD-L1 inhibitor， directly after the chemotherapy. The reason we did it was the following. We feel that second-line immune therapy (pembrolizumab and atezolizumab) is too late， and many patients don’t get second-line therapy. We also felt that single-agent immune checkpoint inhibitors don’t get disease under control in the first-line setting as well as chemotherapy. We feel that chemotherapy is really good for getting initial control of disease， but the problem with chemotherapy is that the cancer always comes back. So what we tried to do was give frontline chemotherapy to get control， and once in control， give the avelumab PD-L1 inhibitor to maintain that control. That is why we called it maintenance therapy – to maintain the control associated with frontline chemotherapy. We randomized 700 patients， and the survival hazard ratio was 0.69， so a 31% reduction in the risk of death associated with avelumab compared to best supportive care. It worked in the whole population. It also worked in the biomarker-positive (PD-L1-positive) population. It also worked across many different subgroups of patients- type of chemotherapy， presence or absence of distal metastases， different PD-L1 subgroups. It was quite a clean study in that respect. It worked across different groups of patients. The last aspect is that the drug was well tolerated， as you would expect with an immune checkpoint inhibitor. The vast majority of patients did well， with only 12% discontinuing for adverse events. In summary， we did a randomized phase III study. It is the first randomized phase III with a survival signal in the first-line setting. It has the lowest hazard ratio for survival we have ever had in urothelial cancer of 0.69 in the ITT population. I think it is a good idea， because second-line is too late， but I do think we need to have the chemotherapy upfront to get control. And for that reason， I feel it is going to be practice changing. We are confident that will be able to give in China， in America， and in Europe as well.

 

Dr Yu: I think the JAVELIN Bladder 100 researchers were very clever by screening patients. Patients receiving avelumab were platinum sensitive patients， similar to studies in ovarian cancer. If patients receive platinum-based chemotherapy and it is effective， it might be that these are the patients who will respond to IO therapy. So it is a clever approach.

 

Dr Thomas: Well it wasn’t all my idea. A lot of different people， in fact 200 people from 200 sites around the world took part. I have had lots of bad ideas as well， but this is one we pursued together with other people. One of the questions that people ask me is， firstly， how does this compare to upfront chemotherapy with immune therapy? Many people feel we should be giving immune therapy and chemotherapy right from the start. I think as things currently stand， there is no evidence that upfront chemotherapy together with immune therapy is better this maintenance approach. As it currently stands， I think the maintenance approach appears to be attractive， and because we have hit a survival signal， I don’t think there are more attractive alternatives for patients.

 

Dr Yu: I agree with you. In clinical practice， when a patient receives chemotherapy， there are two results. One， the tumor is responsive to chemotherapy， and the patient would receive IO maintenance. On the other hand， some patients will progress very rapidly on chemotherapy and never receive IO. How can we assess responsiveness of patients to chemotherapy?

 

Dr Thomas: I think it is a really good question. In my opinion， the important thing to remember about this is that those patients who progress on chemotherapy do not appear to be the immune therapy responders. If a patient progresses on chemotherapy， and they are given immune therapy， they don’t have spectacular responses to immune therapy. We sort of always knew that. The reason we know that in the pembrolizumab/atezolizumab study， those patients with aggressive disease don’t seem to be salvaged in the second-line setting. We know that in the second-line setting， the patients who do best， tend to be those patients with lymph node disease and those who don’t have rapidly progressing aggressive disease. I think the point you have made is really important. The JAVELIN trial does select patients in that we include patients with stable disease and those that are responding， but we exclude those with primary progression. What do we do with those excluded patients? This is difficult. The reason this is difficult is that those patients don’t seem to respond to chemotherapy or immune therapy， and we don’t currently have any treatment that works well for them. There is drug called enfortumab vedotin， which has very high frontline response rates with pembrolizumab. It may be that we have to use a different class of drug all together to get control of those patients.

 

Dr Yu: In this study， you also used PD-L1 as the main biomarker and obtained significant OS benefits in both the whole cohort and the PD-L1-positive subgroup. It seems that we could use IO in these platinum sensitive patients， regardless of PD-L1 expression status. For immunotherapy of urothelial cancer， in addition to PD-L1， what do you think of other new biomarkers， such as the TCGA subtype， DDR， or TMB?

 

Dr Thomas: It’s a great， really difficult question. We currently don’t have good biomarkers in urothelial cancer. They are OK， but they are very inconsistent. The PD-L1 biomarker is very inconsistent. We know， for example， that it doesn’t appear to work very well in the second-line setting. It does appear to work in the frontline metastatic setting for pembrolizumab/atezolizumab， but it doesn’t appear to work in the adjuvant setting. So the PD-L1 biomarker is inconsistent， and I don’t think it is consistent enough in this trial to be used. What about other markers – tumor mutation burden (TMB)， T- effector signatures， DDR cell cycle genes， TGF-β? These are all biomarkers that hold promise for improving patient selection. But none of them currently look fantastic by themselves. They are all helping to point us in the right direction， but none of them look fantastic as it currently stands.

 

Dr Yu: I have a question regarding the role of PD-L1. We know that PD-L1 status is not associated with the success of avelumab maintenance therapy. But if a patient receives combination therapy， combined with chemotherapy or combined with a TKI， we have seen that the PD-L1-positive patients seem to do better with IO maintenance therapy.

 

Dr Thomas: PD-L1 is a surrogate marker， and there are different ways of measuring PD-L1. It can be measured in tumor cells or immune cells. Immune cell expression of PD-L1 seems to correlate with immune cell signatures – the higher the PD-L1 in the immune component， the higher the T-effector signature. I think that those patients do better in that they have a better prognosis， in urothelial cancers (not necessarily in all cancers). It looks like chemotherapy works slightly better， as we learned in IMvigor211， in PD-L1-positives as well. So this story of predictive biomarkers is extremely complicating， and PD-L1 is not specific for immune therapy. It also seems to be predictive for chemotherapy. Unfortunately， and I wish this was different， as it currently stands in urothelial cancer， it doesn’t appear that PD-L1 is an ideal biomarker. I suspect we may not be using it too much in the future.

 

Dr Yu: According to the study IMvigor130， concurrent treatment with chemotherapy and IO also showed survival benefit. What do you think of the concurrent or maintenance treatment patterns in first-line patients? In addition， what do you think is the optimal duration for IO?

 

Dr Thomas: My feeling is that in the IMvigor130 atezolizumab plus chemotherapy randomized trial， the chemotherapy/immune therapy combination arm showed a significant progression-free survival advantage， but not yet an overall survival advantage. When it does hit that survival advantage (currently 0.83)， if it does at a later analysis become significant， then I think we will have to have a discussion about which approach is best. I don’t believe we are there yet. The point you have made is good. The data looks promising， but it is currently premature. It is not yet significant， and until it does become significant， I don’t think that will supersede the maintenance approach.

 

Dr Yu: It seems to be true in clinical trials， but in clinical practice， when we give patients chemotherapy or IO， we feel that if the patient doesn’t tolerate chemotherapy， we should then use IO. If the patient progresses very rapidly， use of IO is not effective. How do we approach these patient groups? If they cannot tolerate the strategy of chemotherapy， should we stop the chemo and continue immune therapy? From JAVELIN Bladder 100， IO maintenance therapy seems very viable. But are we diminishing the effectiveness of chemotherapy in clinical practice?

 

Dr Thomas: You have articulated a problem really well. The reality is， we don’t know. I think deep down， we need more and better drugs and better combinations. Clearly， as you said， there are a group of patients who are not benefiting from our current approach. The maintenance approach appears to be the most pragmatic and the most efficacious with the data we currently have. It is not perfect. There are those patients who progress that we are not treating. But I am unsure what drugs can help that group. What you have said is right. Do we have to give the full six cycles of chemotherapy? I don’t know the answer to that question. I suspect we don’t need to. Four cycles would be enough. That is my gut feeling. Additionally， there are some patients that we simply can’t help. Immune therapy is not great at getting control of disease. Chemotherapy is not great at stopping the cancer from coming back in the future. If we can sequence them， patients do better. But we need more drugs and we need better drugs.

 

Dr Yu: If a patient receives chemotherapy and IO in first-line， which strategy should be adopted after progression? Is IO after IO reasonable? 

 

Dr Thomas: If you progress on maintenance avelumab， with all the options available in the world， I would switch to enfortumab vedotin (which is licensed in America， but not yet in Europe). If they have an FGF alteration， I would give erdafitinib， an FGF inhibitor. If there is a long period since the last chemotherapy (> 1year)， I would rechallenge with chemotherapy. We have seen good re-responses with chemotherapy after immune therapy provided there is a period of relapse. So first， enfortumab vedotin， FGF inhibitors like erdafitinib where there are FGF alterations， or rechallenge with chemotherapy. I would not be retesting immune therapy. We have not seen re-responses with immune therapy. I don’t think it is a great idea.

 

Dr Yu: You mentioned the FGFR3 inhibitors. This mutation is not common in China. But we do have a special situation. Chinese patients are presenting with more and more upper urinary tract cancers. We think this is a special situation whereby patients are seeing increasing rates of FGFR mutations.

 

Dr Thomas: I think we are doing really good work in bladder cancer right now looking at the molecular biology of the disease. We have overtaken kidney cancer. We have FGF and PD-L1 as biomarkers we can use. We are distinguishing between upper tract tumors and urothelial tumors. We have TGA classifications being used. I believe you are right – there is a distinction between upper tract tumors and bladder urothelial cancer. Those upper tract tumors do have higher FGF alterations， and this is a real opportunity for advances. One thing I will say as a caveat to that， we have combined FGF inhibitors with drugs like pembrolizumab， atezolizumab and avelumab in these patients with FGF alterations， without spectacular responses. While the FGF component responds OK， it doesn’t look like combinations result in a big jump in responses. It is really interesting.

 

Dr Yu: I hope in the near future， we can cooperate in the area of upper tract cancers. What do you think of precision medicine in urothelial cancer? What is the optimal time to recommend next-generation sequencing (NGS) testing?

 

Dr Thomas: There are two components to your question. Firstly， with tumor heterogeneity， do you need to have contemporary tissue at the time of a decision? Or can you use archive tissue from the past. In most of the trials we have done， we have allowed archive tissue. As I have mentioned previously， we have not been spectacularly successful with biomarker research. Although， FGF alterations appear to occur quite early in the disease process and seem quite stable. Additionally， mutations appear more stable then gene or protein expression. So， from an FGF perspective， I think it is reasonable to use archive tissue. But the future is around circulating tumor material. Using different technologies， we can be looking for circulating tumor DNA (ctDNA). From a blood draw， we can identify FGF alterations. My feeling is that that is going to be the future. We are going to be leaving tissue behind and using circulating tumor material， perhaps in the not-too-distant future. It is a really exciting area of research. We are also measuring chemokines and cytokines. We have a recent publication with IL6 and IL8 looking at that and the response to atezolizumab. This is an incredibly exciting field， because how can we use the drugs we have better? What we have done so far is to combine drugs together and not always got the results we wanted. We have ended up sequencing drugs， which appears to work better in my opinion. But we need to advance to the next level of using biomarkers properly.