GU Cancers 2021 | POUT: extended follow-up results
Alison Birtle, MRCP, FRCR, MD, Lancashire Teaching Hospitals, Preston, UK, discusses the updated outcomes of the randomized Phase III POUT trial (NCT01993979) investigating peri-operative chemotherapy versus surveillance in upper tract urothelial cancer (UTUC). 261 patients with UTUC who had undergone nephroureterectomy were enrolled and randomized 1:1 to gemcitabine-carboplatin or surveillance. Previous follow-up at 30.3 months found that adjuvant chemotherapy improved disease-free survival (DFS) for patients with UTUC. At a median follow-up of 48.1 months this DFS benefit was maintained, with no observed detrimental long-term toxicity. Overall survival at three years was 67% for surveillance and 79% for chemotherapy, however this was not statistically significant. Dr Birtle also discusses results from the CheckMate 274 trial (NCT02632409) investigating nivolumab compared with placebo in patients with muscle invasive bladder cancer or upper urinary tract urothelial cancer. This interview took place during the 2021 Genitourinary Cancers Symposium.
Transcript (edited for clarity)
Hello, I’m Alison Birtle and I’m a consultant oncologist at the Rosemere Cancer Center in the Northwest of England. And I’m the chief investigator on the POUT study which of course we presented the original data, the first data from this ASCO GU in 2018. And I’m delighted this year at the virtual ASCO GU ’21 to present the updated results with the extended follow-up period and for the first time overall survival data...
Hello, I’m Alison Birtle and I’m a consultant oncologist at the Rosemere Cancer Center in the Northwest of England. And I’m the chief investigator on the POUT study which of course we presented the original data, the first data from this ASCO GU in 2018. And I’m delighted this year at the virtual ASCO GU ’21 to present the updated results with the extended follow-up period and for the first time overall survival data.
So of course you remember that this was an unmet need for patients. Previously, there was no randomized data at all looking at systemic therapy for patients with completely resected upper tract urothelial carcinoma. We had a sense that chemotherapy would probably work because we knew from the EORTC 30994 study of patients who’d had cystectomy that there was a disease-free survival benefit in giving adjuvant chemotherapy. And there was no point in looking at new sexy drugs unless we tried chemotherapy first. And back in 2012 when the study first opened, there of course were no new novel sexy drugs in bladder cancer at that time. This is the pre-immunotherapy era.
So POUT originally looked at randomizing patients who had had a nephroureterectomy and were within 90 days of surgery. And for that to be eligible, they had to have T2 to T4 A disease. They could be node positive but if they’d had nodes that were visualized on the preoperative scan, those had to be removed macroscopically at the time of surgery and then they had to have a negative post-operative CT scan. There’ve been some queries about why we didn’t mandate a lymph node dissection. And that’s because that really is still not standard of care with level one evidence in UTUC although some people do it. And we actually did a survey across urologists the length of the UK before opening POUT to see what they would do in this situation. And there was no support for a lymph node dissection, extended lymph node dissection in these patients.
So the patients are randomized very simply on a one-to-one basis between either four cycles of adjuvant chemotherapy or surveillance. And the chemotherapy was gemcitabine with either carboplatin or cisplatin. And the only reason to give carboplatin was based on GFR as we’ve previously discussed. And the follow-up was standard, looking at disease-free survival as the primary endpoint of the study. And we chose that because patients with locally advanced upper tract urothelial carcinoma have a high rate of relapse really within the first three years. And this had also been chosen as the primary endpoint of the muscle invasive bladder cancer adjuvant study, the 30994.
And, so, we originally presented data looking at the primary endpoint of disease-free survival three years ago. We actually presented that early because the Independent Data Monitoring Committee met in November 2017 and they deemed that we’d met the primary endpoint early. We’d actually met it at two years rather than three years which is what we’d be originally been powered to do. So, at that point, we had 261 patients recruited into the study and randomized out of an initial sample population of 346 patients. But we’d met the primary endpoint of disease-free survival. And this was from patients randomized from 75 recruiting centers across the length and breadth of the UK.
What it showed at that point when we presented it in 2018 and then subsequently published all of this data in the Lancet in March 2020, was that there was a significant improvement in disease-free survival at three years and also metastasis-free survival in favor of chemotherapy in the order at that time of about 17% difference between the two groups with a strong hazard ratio. We looked at quality of life at that point as well and other than a temporary dip for patients who are still on chemotherapy, quality of life recovered quite quickly to baseline in the chemotherapy group.
But obviously, we were also concerned about a safety signal and there was no real additional side effects other than what you might expect with chemotherapy because this is a chemotherapy study. So, there were toxicities related to that but no treatment-associated deaths. So, POUT was a positive study but at the time, we couldn’t present secondary endpoints such as overall survival because we needed enough events.
And the update this year was looking at an extended follow-up. So now, we have a follow-up of just under 50 months. So over four years with a follow-up and it has shown that the disease-free survival has actually strengthened. This is now a 21% difference in favor of chemotherapy at three years and a 17% at five years. And we see a 19% difference in metastasis-free survival both at three and five years in favor of chemotherapy.
Now we did a pre-planned sub-group analyses and these were the stratification factors in POUT. And those were whether patients were node-positive or not, whether they had microscopic positive margins, whether they had carboplatin or cisplatin and overall stage. And what it showed was that if we look at the subgroup analysis, we can see that this is positive in all of these subgroups. There has been some debate about carboplatin. What I will say with that is that when you have a pre-planned sub-group analysis, it’s not all about the p-value. It’s about you do a test for interaction and if that’s negative then you have no reason to suspect that the primary endpoint of the study is any different in any of the subgroups.
So overall survival, well, this was not statistically significant but it was a 12% difference between the two groups. We can speculate on why it wasn’t statistically significant but I think the main reason is because we didn’t get the 346 patients that the study was powered for. So, because we’d met the primary disease-free survival endpoints early. And so, the study is obviously underpowered for overall survival.
The second thing is if you think that we started recruiting patients back in May 2012 and we stopped recruiting in November 2017, the majority of those patients are going to have been in the pre-immunotherapy era. And that’s really important because if you think about patients that progressed who will have been in the chemotherapy arm, then you might have had even more striking benefit for chemotherapy if those patients had been allowed to go on and have a second-line immunotherapy but of course they wouldn’t have been able to do that because this was for the majority of patients in the pre-IO era in the UK.
The other data that was new at ASCO GU this time was quickly extended quality of life data which we have now out to two years. This shows that you do get a temporary dip in quality of life and this was down really to some nausea and vomiting and fatigue but it recovered quite quickly to baseline. And interestingly, if you look at the quality of life curves, what you can see is that the patients in the surveillance arm, it starts dropping down and that’s possibly due to these patients recurring.
I really think disease-free survival is very important because if you wait until these patients progress, they’re probably going to be costlier to both themselves and to health economics because you’re more likely to give them more chemotherapy. So adjuvant treatment is four cycles treatment for metastatic disease is usually a standard of six cycles. And we had a very good debate at the poster session at ASCO GU about how many cycles you should give before moving onto maintenance immunotherapy. And it still seems to be the thought that you would still try and get as many cycles of chemotherapy in as you could if someone’s responding in metastatic disease. So, if you wait until these people relapse and they’ve got metastatic disease, you’re going to be giving them more chemotherapy and then potentially maintenance immunotherapy when you’ve got something which has got a strong disease-free survival advantage from giving it earlier. So, this is why the data based on the original publication from March of last year has been now incorporated into the EAU guidelines and is really what we would advocate as standard of care.
One of the studies that was also presented at ASCO GU was Matt Galsky presented the updated data from adjuvant nivolumab. The patients who’d had a cystectomy for muscle invasive bladder cancer. And you remember that this study was originally just for patients with bladder cancer but because it was recruiting quite slowly, similarly to the adjuvant atezolizumab study, there was a protocol amendment to add in up to 10% of patients who had upper tract urothelial carcinoma resected. Although the global study is, as a whole, was probably underpowered to be able to fully analyze that group. But from my reading of this, the subgroup analysis shows that the nivolumab benefit was not really seen at all in terms of disease-free survival for the patients with UTUC, unlike the PD-L1 positive muscle invasive bladder cancers where there was a much stronger signal in favor of adjuvant nivolumab.
So, what we’re dealing with are two completely different disease entities and it’s completely wrong to actually lump bladder cancer together with upper tract urothelial cancer just because they share some similarities in terms of the baseline pathology. There are multiple differences between the two groups.