This article is more than 5 years old. Information may no longer be current.
Novel PD-L1 antibody demonstrated activity in metastatic urothelial bladder cancer
CHICAGO —Patients with metastatic urothelial bladder cancer and PD-L1–positive disease responded to treatment with the monoclonal antibody MPDL3280A, according to study results presented at the ASCO Annual Meeting.
“We’ve really struggled for the last 30 years to develop bladder cancer therapies. There’s been some effort with some randomized trials, but they haven’t been positive, and there have been no FDA-approved therapies for patients with relapsed disease, which is extraordinary when you consider the frequency of this group of patients,” Thomas Powles, MBBS, MRCP, MD, clinical professor of genitourinary oncology at the Barts Cancer Institute of the Queen Mary University of London, said during a presentation. “Bladder cancer also has a high mutational complexity, which may be important in that this complexity may actually be an attractive target for the immune system due to the potential for many neoantigens.”
Powlesand colleagues evaluated MPDL3280A (Genentech/ Roche) — an anti-PD-L1 monoclonal antibody with an engineered Fc-domain — in 67 patients. They originally enrolled only patients with PD-L1 positivity but later opened the trial to all patients in this setting.
The median age of patients was 65 years (range, 36-86) and 72% were male. A majority of patients had an ECOG performance status of 1 (59%) and had visceral metastases (75%). Seventy-nine percent of patients previously received platinum-based chemotherapy.
Patients received 15 mg/kg MPDL3280A every 3 weeks for a median of 43 days (range, 1-153), and the majority of patients were still receiving treatment at the time of data cut-off.
Median follow-up was 4.2 months (range,1.1-8.5) for PD-L1–positive patients and 2.7 months (range, 0.7-3.6) for PD-L1–negative patients.
Researchers evaluated response according to PD-L1 status. The response rate among 30 patients who were PD-L1 positive — defined as with immunohistochemistry (IHC) 2 or 3 — was 43% (95% CI, 26-63). Among 35 patients who were PD-L1 negative — defined as IHC 0 or 1 — the response rate was 11% (95% CI, 4-26).
“Significantly, PD-L1 expression on immune cells in the tumor may correlate with outcome,” Powles told HemOnc Today. “Therefore, personalized therapy in the medium term is an achievable goal.”
The median time to response was 42 days, and 16 of 17 responders had ongoing responses, Powles said.
Patients with visceral metastases at baseline were among those who responded to MPDL3280A, according to researchers.
Among all patients, the most common grade 1 to grade 4 adverse events associated with MPDL3280A were decreased appetite (12%), fatigue (12%), nausea (12%) and pyrexia (9%). Grade 3 to grade 4 adverse events occurred in 4% of patients, and researchers observed no renal toxicities.
Powles and colleagues then conducted biomarkers analyses. They observed pharmacodynamic markers of activity in the form of transient increases in CD8-positive, Ki-67–positive T cells and plasma proteins, such as interleukin-18, upstream of interferon gamma signaling. Baseline gene analyses indicated patients who progressed on treatment had greater myeloid gene signature, such as IL8 and CCL2.
MPDL3280A has received a breakthrough therapy designation from the FDA, Powles said.
“Providing further studies — including randomized ones — show equally impressive efficacy, it is probable that bladder cancer patients will have a new treatment option in the not too distant future,” Powles said. “While the results are extremely promising, larger studies focusing on specific patient populations such as relapsed metastatic disease are required. These should also include randomized trials to achieve a new global standard of care for specific patient groups.”
For more information:
Powles T. Abstract #5011. Presented at: ASCO Annual Meeting; May 30-June 3, 2014; Chicago.
Disclosure: The researchers report consultant/advisory roles with, honoraria or research funding from, and stock ownership in Bristol-Myers Squibb, Celgene, Genentech, GlaxoSmithKline, Novartis, Pfizer, Roche and US Oncology.
Perspective
Back to Top
The standard for a couple of decades has been just the use of chemotherapy, both in first-line and second-line treatment, for patients with metastatic urothelial carcinoma. Even in the standard of care, which is second-line [docetaxel (Taxotere, Sanofi-Aventis)], for people who failed gemcitabine and cisplatin, the response rate is, on average, around 12% without any survival benefit. It is clear that additional standard chemotherapy is not going to provide much additional benefit.
So we’re really trying to look for some other target, and bladder cancer has been kind of low on the list among people developing immuno-oncology therapeutics. However, the trial that Powles and colleagues conducted is evidence that that may be a potential way to target some second-line response rate in the mid 40% range. That is really pretty striking, and I think it opens the flood gates.
There have also been some other studies at this meeting looking at fibroblast growth factor receptor 3 and receptor targeting, so it may be that we are a decade behind where they are in lung cancer in terms of finding specific phenotypes that may respond to specific agents. It is possible that all bladder cancer isn’t the same, in much the same way that all lung cancer isn’t the same.
I’m not sure we are at the point where we should be testing patients for PD-L1 status yet. First, we see PD-1– or PD-L1–negative agents responding at 11% or 13%, and that’s as good as we do now with standard chemotherapy second line. So what is the appropriate cut-off? Is it 1% of tumor cells or 5% of tumor cells? Should we not be testing the tumor but rather the patient’s immune cells for expression? I don’t think we know, and I’d be very reluctant to exclude patients based on absence or presence of PD-1 or PD-L1 staining. I think the thing to do is to conduct the next set of trials without any such restrictions on entry, and then post hoc go back and look to see if we can determine the appropriate cut-off point and whether there is any relationship. We’re not quite there yet in terms of the biomarker predicting response. [It is] a very interesting drug but, as frequently happens, the biomarker lags behind.
Perspective
Back to Top
PD-L1–positive patients who were PD-L1 expressing on the infiltrating immune cells were positive if they had an immunohistochemistry score of 2 or 3, which is defined by greater than 5% or 10% of these infiltrating cells being PD-L1 positive. Yes, they had better responses, but I will point out that the PD-L1–positive patients were accrued first onto the trial, and they had been on therapy longer. Those who are deemed PD-L1 negative also are having responses, and it would be important to see how their responses compare after they have had similar amounts of therapy. I would not want to be a medical oncologist telling these patients that they can’t have a potentially curable treatment because we deem them PD-L1 negative. This is not the kind of predictive biomarker I think we can use in assaying for selected patients.
Immunotherapy, anti–CTLA-4, anti–PD-1 and anti–PD-L1 elicit enhanced immune responses against tumors that are absolutely incredible, because we have been looking at patients who have had multiple therapies and who are now responding with durable responses. In bladder cancer, these are dramatic responses, and it is very important to develop these strategies because we have not had other treatments approved for bladder cancer. The immune responses are dynamic, and biomarker studies to select patients for treatment based on expression of a certain biomarker may not be feasible. Both PD-L1–negative and PD-L1–positive tumors as defined by various criteria can respond to treatment, and the expression of PD-L1 changes over time. Using PD-L1 expression as a biomarker for selection of patients for treatment would prevent some patients from receiving potentially effective and curable therapies.
I think the era of genomic medicine really provided the concept that biomarkers — for example, BRAF mutation — can be identified and used to select patients for treatment. But this is not applicable to immunotherapy. We cannot be placed in the same box as genomic medicine. Biomarker development for immunotherapy agents may require integration of multiple biologic components as opposed to a single molecule.