Adjuvant Pembrolizumab Improves DFS, But Not OS, in Locally Advanced/Muscle-Invasive Urothelial Carcinoma

Andrea B. Apolo, MD

Adjuvant pembrolizumab (Keytruda) generated a statistically significant and clinically meaningful disease-free survival (DFS) benefit compared with observation in patients with muscle-invasive and locally advanced urothelial carcinoma, regardless of PD-L1 status, according to interim findings from the phase 3 AMBASSADOR Alliance A031501 trial (NCT03244384), which were presented at the 2024 Genitourinary Cancers Symposium.¹

At a median follow-up of 22.3 months (range, 0.03-48.9), the median DFS with pembrolizumab in the intention-to-treat population was 29.0 months (95% CI, 21.8-not reached [NR]) vs 14.0 months (95% CI, 9.7-20.2) with observation (HR, 0.69; 95% CI, 0.54-0.87; P = .001).

However, at a median follow-up of 36.9 months (range, 0-63.9), the median overall survival (OS) was 50.9 months (95% CI, 43.8-NR) with pembrolizumab vs 55.8 months (95% CI, 53.3-NR) with observation (HR, 0.98; 95% CI, 0.76-1.26; P = .884).

“The trial met its efficacy end point and is a positive trial,” Andrea B. Apolo, MD, of the National Institutes of Health in Bethesda, Maryland, stated in a presentation of the data. “The OS end point was not met at interim analysis. It may have been impacted by the high number of patients in the observation arm receiving a checkpoint inhibitor.”

Although the standard of care for patients with muscle-invasive urothelial carcinoma (MIUC) is radical surgery with or without neoadjuvant cisplatin-based chemotherapy, many patients are not eligible for cisplatin or have persistent muscle-invasive disease after neoadjuvant chemotherapy and surgery. Furthermore, cisplatin-based therapy in the adjuvant setting is not recommended for patients who have received neoadjuvant chemotherapy and is associated with toxicities and administration difficulties post-radical surgery in patients not treated with neoadjuvant chemotherapy.^2,3

The randomized, open-label, multicenter AMBASSADOR trial assessed the PD-1 checkpoint inhibitor pembrolizumab vs observation in patients with high-risk MIUC, including cancers of the bladder, renal pelvis, urethra, and ureter.¹ Patients must have received radical surgery, including cystectomy, nephrectomy, nephroureterectomy, or ureterectomy more than 4 but no more than 16 weeks before trial initiation. Eligible patients included those who had received post-neoadjuvant chemotherapy and had at least pT2 and/or N+ positive margins or those who refused or were ineligible for cisplatin and had at least pT3 and/or pN+ positive margins. Patients were stratified by PD-L1 status, neoadjuvant chemotherapy status (yes vs no), and pathological stage (pT2/3/4aN0 vs pT4aN0 vs pT4bNx/N1-3 vs +surgical margins).

AMBASSADOR planned to enroll 734 patients. At an accrual closure date of August 24, 2021, the trial was 96% accrued, with 702 patients. AMBASSADOR closed early because of the August 2021 FDA approval of adjuvant nivolumab (Opdivo) for patients with high-risk urothelial carcinoma.⁴

Patients in AMBASSADOR were randomly assigned 1:1 to either pembrolizumab (n = 354) at 200 mg every 3 weeks for 1 year (18 cycles) or observation (n = 348).¹

DFS and OS served as the dual primary end points. Key secondary end points included DFS and OS by PD-L1 status and safety. Correlative end points included DFS and OS by circulating tumor DNA (ctDNA) status; DFS and OS by immune gene signatures; DFS and OS by tumor molecular subtype; DFS and OS by T-cell receptor clonality; and quality of life. The trial would be considered positive if it met either primary end point.

In the pembrolizumab arm, 24 patients never started treatment, and 207 patients were censored for DFS, 166, of whom were alive and disease-free at follow-up, 5 of whom received alternative therapy, and 36 of whom withdrew from the trial or clinical follow-up without a DFS event. In total, 147 DFS events occurred, 121 of which were progression/recurrence and 26 of which were death. In the observation arm, 4 patients never started observation, and 176 patients were censored for DFS, 117, of whom were alive and disease-free at follow-up, 11 of whom received alternative therapy, and 48 of whom withdrew from the trial or clinical follow-up without a DFS event. In total, 172 DFS events occurred, 143 of which were progression/recurrence and 29 of which were death.

In the pembrolizumab arm, 223 patients were censored for OS, 36 of whom withdrew from the trial and 187 of whom were alive and in follow-up at data cutoff. In the observation arm, 222 patients were censored for OS, 42 of whom withdrew from the trial and 179 of whom were alive and in follow-up at data cutoff. A total of 131 and 126 OS events occurred in the pembrolizumab and observation arms, respectively.

Patients in the pembrolizumab arm received a mean of 11 treatment cycles (range, 1-18). All patients enrolled in the trial are currently not receiving study treatment.

Patients in the pembrolizumab arm had a median age of 69.0 years (range, 22.0-92.0). In total, 91.2%, 4.0%, 1.4%, 0.6%, and 2.8% of patients were White, Black/African American, Asian, American Indian/Alaskan Native, and not reported/unknown, respectively. Most patients were male (76.6%), had received neoadjuvant therapy (65.3%), and had PD-L1–positive disease (57.1%). Regarding pathologic stage, 2.5%, 50.9%, 41.2%, and 5.4% of patients had positive surgical margins, pT-any N+ (any) disease, pT2/3N0 or NX disease, and pT4N0 or NX disease, respectively. Primary tumor sites included the bladder (75.4%), urethra (1.7%), and upper tract (renal pelvis and ureter; 22.9%). In total, 16.9% of patients had variant histology, defined as mixed urothelial histology excluding neuroendocrine carcinoma.

Patients in the observation arm had a median age of 68.0 years (range, 34.0-90.0). In total, 89.1%, 3.2%, 2.9%, 0.6%, and 4.3% of patients were White, Black/African American, Asian, American Indian/Alaskan Native, and not reported/unknown, respectively. Most patients were male (72.7%), had received neoadjuvant therapy (62.6%), and had PD-L1–positive disease (57.8%). Regarding pathologic stage, 2.3%, 48.8%, 43.1%, and 5.8% of patients had positive surgical margins, pT-any N+ (any) disease, pT2/3N0 or NX disease, and pT4N0 or NX disease, respectively. Primary tumor sites included the bladder (75.9%), urethra (3.4%), and upper tract (renal pelvis and ureter; 20.7%). In total, 14.7% of patients had variant histology.

All prespecified patient subgroups benefitted from pembrolizumab vs observation, except for those with a primary tumor site of the upper tract (HR, 1.05; 95% CI, 0.61-1.82).

“The upper tract subgroup I think needs further investigation, and the magnitude of benefit is unclear,” Apolo noted.

In patients with PD-L1–positive disease, defined as those with a PD-L1 combined positive score (CPS) of at least 10%, the median DFS was 32.8 months (95% CI, 28.1-NR) with pembrolizumab vs 20.7 months (95% CI, 13.5-NR) with placebo (HR, 0.77; 95% CI, 0.57-1.04; P = .091). In patients with PD-L­1–negative disease, the median DFS was 22.1 months (95% CI, 13.8-NR) with pembrolizumab vs 9.1 months (95% CI, 7.0-15.3) with placebo (HR, 0.61; 95% CI, 0.44-0.84; P = .002). Regarding OS differences between prespecified patient subgroups, the HR in the subgroup whose primary site was in the upper tract was 1.45 (95% CI, 0.75-0.99).

The median OS in patients with PD-L1–positive disease was NR (95% CI, 46.5 months-NR) with pembrolizumab vs 56.1 months (95% CI, 55.2-NR) with observation (HR, 1.09; 95% CI, 0.77-1.54; P = .624). The median OS in patients with PD-L1–negative disease was 43.8 months (95% CI, 35.9-NR) with pembrolizumab vs 36.7 months (95% CI, 29.3-NR) with observation (HR, 0.83; 95% CI, 0.59-1.18; P = .297).

“PD-L1 positivity using the CPS was associated with a better prognosis but was not predictive for treatment efficacy,” Apolo emphasized. “PD-L1 status should not be used to select patients for treatment.”

Among the patients in the pembrolizumab arm who did not withdraw consent, 99 received alternative systemic therapy. Patients were permitted to report multiple alternative treatments. In the post-DFS setting, 82 patients received alternative systemic therapy, including a checkpoint inhibitor (n = 19), chemotherapy (n = 38), antibody-drug conjugates (ADCs; n = 45), and other therapy (n = 4). In the pre-DFS setting, 17 patients received alternative systemic therapy, including a checkpoint inhibitor (n = 6), chemotherapy (n = 4), and ADCs (n = 3).

Among the patients in the observation arm who did not withdraw consent, 123 received alternative systemic therapy. In the post-DFS setting, 104 patients received alternative systemic therapy, including a checkpoint inhibitor (n = 76), chemotherapy (n = 34), ADCs (n = 21), and other therapy (n = 3). In the pre-DFS setting, 19 patients received alternative systemic therapy, including a checkpoint inhibitor (n = 18), chemotherapy (n = 6), and ADCs (n = 1).

In the pembrolizumab arm, 48.4% of patients experienced adverse effects (AEs) of grade 3 or higher. In total, 38.6%, 4.9%, and 4.9% of patients had grade 3, 4, and 5 AEs, respectively. Grade 3 and 4 hematologic AEs were observed in 5.5% and 0.9% of patients, respectively, and no grade 5 hematologic AEs were reported. Among the non-hematologic AEs observed, 37.7%, 4.6%, and 4.9% were grade 3, 4, and 5, respectively.

In the observation arm, 31.8% of patients experienced AEs of grade 3 or higher. In total, 23.3%, 3.8%, and 4.7% of patients had grade 3, 4, and 5 AEs, respectively. Grade 3 hematologic AEs were observed in 2.6% of patients, and no grade 4/5 hematologic AEs were reported. Among the non-hematologic AEs observed, 22.7%, 3.8%, and 4.7% were grade 3, 4, and 5, respectively.

Among the patients who received pembrolizumab, the most common treatment-related AEs were fatigue (any-grade, 47%; grade 3, 2%), pruritus (22%; 1%), diarrhea (21%; 3%), hypothyroidism (20%; 0%), maculo-papular rash (17%; 2%), increased creatine (15%; 1%), anorexia (15%; 1%), anemia (14%; 2%), increased lipase (13%; 3%), dyspnea (11%; 0%), increased serum amylase (9%, 2%), hyperthyroidism (8%; 0%), and increased alanine aminotransferase (6%; 1%). Additionally, 1% of patients experienced grade 4 lipase increase.

“These results support adjuvant pembrolizumab as a new therapeutic option for patients MIUC at high risk of recurrence,” Apolo concluded.

Follow-up is ongoing for the final DFS/OS analysis of AMBASSADOR, as well as the ctDNA analyses and additional correlative analyses.

Editor’s Note: Dr Apolo reports no relationships to disclose.

References

1. Apolo AB, Ballman KV, Sonpavde GP, et al. AMBASSADOR Alliance A031501: Phase III randomized adjuvant study of pembrolizumab in muscle-invasive and locally advanced urothelial carcinoma (MIUC) vs observation. J Clin Oncol. 2024;42(suppl 4):LBA531. doi:10.1200/JCO.2024.42.4_suppl.LBA531
2. Advanced Bladder Cancer (ABC) Meta-analysis Collaborators Group. Adjuvant chemotherapy for muscle-invasive bladder cancer: a systematic review and meta-analysis of individual participant data from randomised controlled trials. Eur Urol. 2022;81(1):50-61. doi:10.1016/j.eururo.2021.09.028
3. Sternberg CN, Skoneczna I, Kerst JM, et al. Immediate versus deferred chemotherapy after radical cystectomy in patients with pT3-pT4 or N+ M0 urothelial carcinoma of the bladder (EORTC 30994): an intergroup, open-label, randomised phase 3 trial. Lancet Oncol. 2015;16(1):76-86. doi:10.1016/S1470-2045(14)71160-X
4. US Food and Drug Administration approves Opdivo (nivolumab) for the adjuvant treatment of patients with high-risk urothelial carcinoma. News release. Bristol Myers Squibb. August 20, 2021. Accessed January 27, 2024. https://news.bms.com/news/details/2021/U.S.-Food-and-Drug-Administration-Approves-Opdivo-nivolumab-for-the-Adjuvant-Treatment-of-Patients-with-High-Risk-Urothelial-Carcinoma/default.aspx