HCRN GU20-444 was a phase II trial looking at the ability of using pembrolizumab after TURBT followed by clinical restaging for a potential bladder-sparing approach. So radical cystectomy with or without perioperative systemic therapy is a traditional approach for the treatment of muscle-invasive bladder cancer. Radical cystectomy is a life-changing operation, it involves urinary diversion. It’s associated with significant morbidities, quality of life issues, 90-day mortality rates that approach 10% particularly in patients of more advanced age. We know now, especially with more advanced systemic therapies, that we can achieve a pathologic complete response in a subset of patients, and many of those patients will go onto have disease-free survival without any other therapy, so that’s suggesting that a subset of patients are potentially cured prior to having that surgery.
We even know that immunotherapy such as pembrolizumab alone after TURBT has achieved pathologic complete response of about 40%, which is similar to what we’ve seen with cisplatin-based neoadjuvant chemotherapy.
We previously reported a study HCRN GU16-257 where we showed that the combination of cisplatin-based chemotherapy with PD-1 blockade can achieve clinical complete response in a subset of patients and many of those patients were able to safely, in an individualised manner, retain their bladder without metastatic recurrence. We identified some limitations to that study, notably many patients either do not want or cannot for medical reasons receive cisplatin-based chemotherapy. So we designed HCRN GU20-444, a phase II trial looking at pembrolizumab alone after TURBT followed by clinical restaging, and then potential bladder-sparing.
What were the key eligibility criteria and endpoints for this phase 2 study?
The co-primary endpoints were clinical complete response rate and the ability of clinical complete response to predict benefit from treatment which was defined as two-year metastasis free survival in patients that achieved a clinical complete response and forgo cystectomy.
I’ll mention as well the design of the trial. This was a phase II investigator-initiated study. Patients were eligible if they had clinical T2 to T3 N0 muscle-invasive bladder cancer. They were either cisplatin-ineligible or declined cisplatin and chemotherapy, and were eligible with a creatinine clearance of at least 30.
Patients that enrolled underwent two cycles of pembrolizumab; it was dosed at 400mg intravenously every six weeks. Then after those two cycles they underwent a uniform and stringent clinical restaging assessment. That involved cystoscopy with either resection of any visible tumour, biopsy of the location at which the original tumour was resected, or biopsies of a pre-defined template in the bladder, as well as urine cytology and MRI of the bladder. Patients that were deemed to have achieved a clinical complete response, meaning that whole assessment showed no evidence of disease, received an additional seven cycles of maintenance pembrolizumab. Patients that did not achieve a clinical complete response for any one of those criteria underwent immediate cystectomy or chemoradiation and then were offered an additional seven cycles of adjuvant pembrolizumab.
What were the primary outcomes, and how effective was pembrolizumab in achieving bladder-sparing responses?
Of those 46 patients that enrolled, all 46 patients underwent clinical restaging. Of those 46 patients, 20 patients achieved a clinical complete response, so that’s 43 percent of the ITT cohort achieved a clinical complete response. All 20 of those patients omitted upfront cystectomy. Of the remaining patients that did not achieve a clinical complete response, the majority underwent immediate cystectomy or chemoradiation which is definitive local therapy to the bladder, other than the patients that either declined or had metastatic recurrence. That is the primary endpoint that we’re presenting at this meeting.
The co-primary endpoint of the ability of clinical complete response to predict benefit is not yet mature.
Were there any notable safety concerns or adverse events observed in the trial?
No, all of the safety profile was in line with the known safety that we have come to know from pembrolizumab. There were two grade 5 adverse events that we did unfortunately see. One was a cardiac arrest that was felt to be not treatment-related. One was a death due to fungal pneumonia from a patient that received systemic steroids for immune-related hepatitis. Both were felt to be non-cancer-related causes of death.
How might these results influence standard-of-care treatment for muscle-invasive bladder cancer?
I think that we've come a long way in this field. Our systemic therapy, particularly for muscle-invasive bladder cancer has improved rapidly and we importantly need to listen to what patients want. Very often what patients will say is they want to retain their bladder, they want less toxic therapies, they don’t want to undergo a morbid surgery if they can potentially safely forgo it. We’ve come to a time where our therapeutics and our diagnostics now are at the point where we can try to make a more individualised and safe treatment algorithm.
I’ll say as well from this study, we also incorporated circulating tumour DNA. We saw that patients who had undetectable circulating tumour DNA at baseline had a higher rate of achieving a clinical complete response, and of all the patients that did achieve a clinical complete response, all of those patients had undetectable ctDNA by cycle two of treatment.
I think the field has grown rapidly, and incorporating all of these elements, including a stringent and uniformly assessed clinical restaging, will ultimately be what’s best for our patients.
