REPORTING FROM GUCS 2018
SAN FRANCISCO (FRONTLINE MEDICAL NEWS) – Emerging data from phase 3 clinical trials are better clarifying the efficacy and safety of immunotherapy in advanced urothelial cancer and helping identify patients most likely to benefit.
An updated analysis of the KEYNOTE-045 trial showed that, compared with chemotherapy, pembrolizumab (Keytruda), an antibody to programmed death-1 (PD-1), almost doubled the 2-year survival rate in patients with recurrent or advanced urothelial cancer. No cumulative toxicity was seen.
Biomarker analyses from the IMvigor 211 trial showed that, compared with chemotherapy, atezolizumab (Tecentriq), an antibody to programmed cell death ligand 1 (PD-L1), prolonged survival by more than 7 months in patients with platinum-treated locally advanced or metastatic disease whose tumors were positive for this ligand and had a high mutational burden. The difference translated to a halving of the risk of death.
Results of both trials were reported at the 2018 Genitourinary Cancers Symposium sponsored by the American Society of Clinical Oncology, ASTRO, and the Society of Urologic Oncology.
“It is clear that PD-1 and PD-L1 targeted immunotherapy now has a role in most patients with advanced urothelial cancer,” Robert J. Jones, MBChB, PhD , professor of clinical cancer research, University of Glasgow, Beatson West of Scotland Cancer Centre, commented in an invited discussion. “It is also true that cytotoxic chemotherapy maintains a role for many, and that both modalities are often ineffective and may be toxic.”
KEYNOTE-045 trial update
Initial results of KEYNOTE-045, at a median follow-up of 14.1 months, provided level I evidence for the safety and efficacy of pembrolizumab over chemotherapy as second-line therapy for recurrent, advanced urothelial cancer ( N Engl J Med. 2017;376:1015-26 ), leading to approval for that indication. (Pembrolizumab is also approved as first-line therapy for cisplatin-ineligible patients).
Lead investigator Joaquim Bellmunt, MD, PhD, an associate professor of medicine at Harvard Medical School and director of the Bladder Cancer Center, Dana-Farber Cancer Institute, both in Boston, reported a trial update, now at a median follow-up of 27.7 months.
Results among 542 patients showed that the initial overall survival benefit of pembrolizumab over chemotherapy (vinflunine, paclitaxel, or docetaxel) (hazard ratio, 0.73; P = .0022) was sustained and in fact now somewhat greater with longer follow-up (HR, 0.70; P = .00017).
Median overall survival was 10.3 months and 7.3 months, respectively. Corresponding 1-year survival rates were 44.4% and 29.8%, and corresponding 2-year survival rates, 27.0% and 14.3%.
At 24 months, 60.6% of patients in the chemotherapy arm had received an immunotherapy agent, including some who had received pembrolizumab as part of crossover, Dr. Bellmunt noted.
Overall survival benefit was generally similar across subgroups, including patients with PD-L1–positive tumors (defined as a combined positive score of 10% or higher) (HR, 0.56; P = .00153) and patients with PD-L1–negative tumors (HR, 0.75; P = .00859).
The lack of a progression-free survival benefit of pembrolizumab over chemotherapy in the initial analysis (HR, 0.98; P = .420) persisted in the updated analysis (HR, 0.96; P = .317).
The overall response rate now was 21.1% with pembrolizumab and 11.0% with chemotherapy. In the former group, the rate of complete response had increased from 7.0% to 9.3% with the longer follow-up. Median time to response was identical, at 2.1 months, but duration of response was longer with pembrolizumab (not reached vs. 4.4 months).
“We haven’t seen any signals of cumulative toxicity with subsequent follow-up,” Dr. Bellmunt reported. Similar to findings of the initial analysis, the most common grade 3-5 treatment-related adverse events were pruritus, fatigue, and diarrhea with pembrolizumab, and neutropenia, anemia, and fatigue with chemotherapy. As expected, the pembrolizumab group had higher rates of hypothyroidism, pneumonitis, hyperthyroidism, and colitis.
“The overall survival benefit and superior safety of pembrolizumab versus chemotherapy in this second-line patient population is maintained after 2 years of follow-up. At 24 months, 27% of patients are alive, and this is similar to what we are seeing with other immune-sensitive diseases like melanoma,” he concluded. “Results in patients with PD-L1–positive or –negative tumors were consistent with the intent-to-treat population. We have seen some hints with this biomarker, but they are not very striking.”
IMvigor 211 trial biomarker analyses
The IMvigor 211 trial enrolled 931 patients with locally advanced or metastatic urothelial carcinoma who had experienced progression during or after platinum-based chemotherapy and had received at most two prior lines of therapy.
At a median follow-up of 17.3 months, the trial did not meet its primary endpoint of significantly better overall survival with atezolizumab versus chemotherapy (vinflunine, docetaxel, or paclitaxel) in patients having PD-L1–positive tumors, defined as immunohistochemical staining of 2 or 3 (HR, 0.87; 95% confidence interval, 0.63-1.21; P = .41) ( Lancet. 2018;391:748-57 ).
Findings were similar in the entire intention-to-treat population (HR, 0.85; 95% CI, 0.73-0.99) and in the subset with tumors having PD-L1–negative tumors, defined as immunohistochemical staining of 0 or 1 (HR, 0.84; 0.71-1.00). In an unexpected finding, PD-L1 positivity was associated with better outcome in both treatment arms.
“So PD-L1 is a classic prognostic, not predictive, biomarker,” commented lead investigator Thomas Powles, MBBS, MRCP, MD , a clinical professor of genitourinary oncology with Barts Health NHS Trust, St. Bartholomew’s Hospital, London. The investigators therefore conducted a series of analyses in the intent-to-treat population to identify predictive biomarkers.
Results were essentially the same with the tumor gene expression 3 (tGE3) signature, an RNA signature that captures expression of the genes encoding interferon gamma, a chemokine ligand, and PD-L1, and that is a marker for preexisting T-cell immunity. And they were also similar with the DNA damage response (DDR) biomarker.
However, a different pattern was seen with tumor mutational burden (TMB), a FoundationOne panel. Patients with TMB-high tumors had a substantial gain in overall survival from atezolizumab versus chemotherapy (HR, 0.68; 95% CI, 0.51-0.90), whereas those with TMB-low tumors did not (HR, 1.00; 95% CI, 0.75-1.32).
“TMB appears to be a predictive but not a prognostic biomarker,” Dr. Powles said. “It’s not perfect. Complete and partial responses and long overall survival were seen in both arms. Nevertheless, it seems like a step in the right direction.”
Finally, with insight on the nature and relationships of the various biomarkers, the investigators assessed the combination of TMB with PD-L1, finding that atezolizumab had a marked overall survival benefit in patients with TMB-high and PD-L1–positive tumors (17.8 vs. 10.6 months; HR, 0.50; 95% CI, 0.29-0.86).
“I think by using combinations of biomarkers, first-generation and second-generation, we may actually be able to better select patients for treatment in the future,” Dr. Powles concluded.
New data, new insights
“We already know the KEYNOTE-045 trial is positive,” commented the invited discussant, Dr. Jones. “This longer follow-up data is important because we need to better describe the magnitude of benefit.”
The updated survival findings are exciting because they resemble those seen with ipilimumab in melanoma, with about a fifth of patients still alive several years out, he said. Longer follow-up is needed, but “the data we see today are in keeping with the possibility that we might be seeing a similar long-term tail – the so-called immuno-oncology tail – of survival for patients with urothelial cancer as well.”
Although not statistically significant, the difference in progression-free survival is clinically important, according to Dr. Jones. This benefit is driven by both higher response rate and longer duration of response with pembrolizumab.
Long-term toxicity, especially immune-related toxicity, is also a consideration. “There is a small increment in the number of some of these toxicities, but essentially it hasn’t changed. So it would appear that there is certainly not an unexpected peak of latent immunotoxicity with this treatment,” he noted.
“Our confidence in the data for second-line pembrolizumab, if it needed to be further increased, is increased. This does help our patients make an informed decision about whether or not to accept this treatment,” he summarized.
Turning to the IMvigor 211 biomarker study, Dr. Jones said, “I would argue that the choice of second-line therapy in advanced urothelial cancer is not a clinically important decision. The reason for saying that is the case for second-line chemotherapy is poor. We’ve all used it, but we’ve never had high level evidence of benefit. … But it’s important because when we consider moving into the first-line setting, where there are active alternatives, or maybe even more into the perioperative setting, it actually could be vitally important.”
Studies of immunotherapies and targeted therapies in other cancers suggest that a clinically useful predictive biomarker will identify patients who will derive at least a doubling of favorable outcome or a halving of unfavorable outcome when given the drug as compared with a control. “So it appears that the bar has been set quite high,” Dr. Jones said.
“These IMvigor 211 data are exploratory, and they would need further independent validation, in my view,” he said. Nevertheless, “this may provide us the opportunity to use combinations of biomarkers where now we are seeing a hazard ratio of 0.50 [for risk of death] combining PD-L1 with high tumor mutational burden (TMB). That hazard ratio is bringing it into the area where it may be of a magnitude big enough to use in clinical practice, if this could be validated.”
“I would say that none of these data support a role for second-line cytotoxics after failure of platinum, at least not in preference to a checkpoint inhibitor,” Dr. Jones concluded. “There is still no biomarker to inform a choice in second-line treatment. However, TMB, either alone or in combination with other markers, shows promise, which we need to validate in future randomized trials.”