Olaparib falls short as add-on to durvalumab in mUC, but combo may benefit HRR+ subgroup

Results from the phase 2 BAYOU trial showed that adding olaparib (Lynparza) to frontline durvalumab (Imfinzi) did not lead to a significant progression-free survival (PFS) benefit in patients with platinum-ineligible metastatic urothelial carcinoma; however, the combination may boost PFS in patients harboring homologous recombination repair (HRR) mutations.1

The BAYOU trial (NCT03459846) data, which were presented during the 2022 Genitourinary Cancers Symposium, showed that the median investigator-assessed PFS with the doublet (n = 78) in the intent-to-treat (ITT) population was 4.2 months (95% CI, 3.6-5.6) vs 3.5 months (95% CI, 1.9-5.1) with durvalumab alone (n = 76; HR, 0.94; 95% CI, 0.64-1.39; log-rank P = .789).

However, data from a prespecified secondary analysis suggested that olaparib plus durvalumab had a potential PFS benefit in a subgroup of patients with homologous recombination repair (HRR) mutations. In this population, the median PFS achieved with the doublet (n = 17) was 5.6 months (95% CI, 1.9-8.1) vs 1.8 months (95% CI, 1.7-2.2) with single-agent durvalumab (n = 14; HR, 0.18; 95% CI, 0.06-0.47; log-rank P < .001).

“The PFS results suggest a potential role for PARP inhibitors in the treatment of metastatic urothelial carcinoma with HRR mutations, and further investigation is warranted,” Jonathan E. Rosenberg, MD, chief of the Genitourinary Oncology Service in the Division of Solid Tumor Oncology, and Enno W. Ercklentz Chair at Memorial Sloan Kettering Cancer Center, said in an oral presentation on the data.

Patients with metastatic urothelial carcinoma are known to have a poor prognosis, particularly those who cannot tolerate platinum-based chemotherapy, Rosenberg said. For these patients, atezolizumab (Tecentriq) and pembrolizumab (Keytruda) are given, irrespective of PD-L1 status; however, a high unmet need for additional therapeutic options remains.

Defects in DNA damage repair occur in approximately 24% of patients with urothelial carcinoma, and the dysregulation of DNA repair genes has been associated with poorer outcomes. Notably, these defects are known to make cancer cells sensitive to PARP inhibitors. As such, investigators hypothesized that HRR mutations and PARP inhibition could strengthen the antitumor response achieved with immune checkpoint inhibitors.

To this end, BAYOU sought to enroll approximately 150 patients with unresectable, stage IV urothelial carcinoma, a transitional cell carcinoma of the bladder, renal pelvis, ureter, or urethra. Patients needed to be at least 18 years of age and have an ECOG performance status ranging from 0 to 2.

Study participants were randomized 1:1 to receive either intravenous (IV) durvalumab at a dose of 1500 mg every 4 weeks plus oral olaparib at a twice-daily dose of 300 mg or durvalumab alone. Key stratification factors included HRR status (mutant vs wild-type) and Bajorin risk index (0 vs 1 vs 2).

The primary end point of the trial was PFS by RECIST v1.1 criteria and per investigator assessment. Key secondary end points comprised PFS by RECIST v1.1 criteria in the subgroup of patients with HRR mutations, overall survival (OS), duration of response (DOR), objective response rate (ORR), and PFS at 6 months by RECIST v1.1 criteria in both the ITT population and the HRR-mutated subset. Other end points of interest included safety and tolerability.

The data cutoff date was October 15, 2020. The median follow-up in the investigative arm was 9.8 months (range, 0.0-29.0) vs 10.7 months (range, 1.0-29.0) in the control arm.

Those who received the doublet were slightly older than those who received durvalumab monotherapy, at a median of 79 years (range, 47-89) vs 72 years (range, 45-88). Slightly more than half of patients in both arms were male. In the investigative arm, 51% of patients were White, 49% were Asian, and none were Black or African American; in the control arm, these rates were 46%, 51%, and 3%, respectively.

Regarding ECOG performance status, 15% of patients who received olaparib plus durvalumab had a status of 0, 38% had a status of 1, and 44% had a status of 2. Of those who were given durvalumab monotherapy, 18% had an ECOG performance status of 0, 45% had a status of 1, and 37% had a status of 2. Most patients in both arms had pure transitional cell carcinoma.

Moreover, 67% of patients who received the doublet had visceral metastases, and 33% had lymph-node only disease; these rates were 63% and 37%, respectively, in those who received the monotherapy. In the investigative and control arms, 11.5% and 10.5% of patients, respectively, received previous therapy.

In the investigative arm, 21% of patients did not have any Bajorin risk factors, 49% had 1 risk factor, and 31% had 2. In the control arm, 24% did not have Bajorin risk factors, 47% had 1, and 29%, and 2. Seventy-eight percent of patients in the doublet arm had HRR wild-type disease, and 22% had HRR-mutated disease; in the control arm, these rates were 82% and 18%, respectively.

The 20 most frequent gene mutations demonstrated a pattern that proved to be consistent with what has previously been observed in bladder cancer, according to Rosenberg. Across all patients, the most common HRR gene mutations were loss of p53 function (59.5%) and TERT promoter mutations (55.6%).

Utilizing the validated FoundationOne assay, investigators tested tumor samples for loss of function alterations in the following 15 prespecified HRR genes: ATM, BARD1, CHEK1, PALB2, RAD51C, BRCA1, BRIP1, CHEK2, PPP2R2A, RAD51D, BRCA2, CDK12, FANCL, RAD51B, and RAD54L. HRR mutations were found in 7 of the 15 genes analyzed. The most frequent included ATM, which was identified in 8.5% of all tumors and in 41.9% of patients in the HRR-mutated subset, as well as BRCA2, which was found in 4.6% of all tumors and in 22.6% of those in the HRR-mutated subgroup.

Exploratory subgroup analyses also showed that PFS favored olaparib/durvalumab over durvalumab alone in those with a Bajorin risk index of 0 (HR, 0.42; 95% CI, 0.18-0.92). However, PFS appeared to favor single-agent durvalumab over the doublet in those with HRR wild-type disease (HR, 1.29; 95% CI, 0.85-1.96), in those with a Bajorin risk index of 1 (HR, 1.90; 95% CI, 1.07-3.49), and in those with a Bajorin risk index of 2 (HR, 1.19; 95% CI, 0.63-2.28).

Data from the prespecified secondary analysis of OS showed that olaparib plus durvalumab did not result in improved OS vs durvalumab monotherapy in the ITT population. Specifically, the median OS with the doublet 10.2 months (95% CI, 7.0-13.9) vs 10.7 months (95% CI, 7.2-17.3) with durvalumab alone (HR, 1.07; 95% CI, 0.72-1.61; log-rank P = .728).

The same was true in the subgroup of patients with HRR wild-type disease. The median OS with olaparib plus durvalumab (n = 61) was 10.9 months vs 13.7 months with single-agent durvalumab (n = 62; HR, 1.35; 95% CI, 0.85-2.16). However, the doublet (n = 17) was found improve median OS in the subset of patients with HRR-mutated disease vs the monotherapy (n = 14), at 8.6 months vs 5.8 months, respectively (HR, 0.56; 95% CI, 0.25-1.23).

The addition of olaparib to durvalumab resulted in an ORR of 28.2% in the ITT population vs 18.4% with durvalumab alone. In the investigative arm, 3.8% of patients achieved a complete response, 24.4% had a partial response (PR), 34.6% achieved stable disease (SD) for at least 7 weeks, 32.1% experienced disease progression, and 5.1% were not evaluable. The median DOR in the investigative arm was 8.9 months (range, 4.7-12.1) vs 14.8 months (range, 7.5-17.2) in the control arm.

In the HRR-mutated subset, the combination produced an ORR of 35.3% vs 0% with durvalumab alone. Among those who responded to the doublet, 35.3% achieved a PR and 29.4% had SD for 7 weeks or longer. Moreover, 29.4% had progressive disease and 5.9% were not evaluable. The median DOR with olaparib/durvalumab was 6.7 months (range, 3.7-12.0) in this subset.

No new safety signals were observed with the combination of olaparib and durvalumab, according to Rosenberg. Any-grade treatment-related adverse effects were experienced by 72.4% of patients who received the doublet (n = 76) vs 60.5% of those given the monotherapy (n = 76).

The most common toxicities observed in the investigative arm included anemia (any grade, 23.7%; grade 3 or 4, 6.6%), nausea (any grade, 19.7%; grade 3 or 4, 1.3%), fatigue (any grade, 15.8%; grade 3 or 4, 1.3%), decreased appetite (any grade, 10.5%), diarrhea (any grade, 10.5%), and hypothyroidism (any grade, 9.2%). In the control arm, the most common toxicities included anemia (any grade, 13.2%; grade 3 or 4, 1.3%), pruritus (any grade, 10.5%), fatigue (any grade, 6.6%), rash (any grade, 5.3%), nausea (any grade, 3.9%), hypothyroidism (any grade, 3.9%).

One death was reported in the durvalumab monotherapy arm, and this was because of anemia.


1. Rosenberg JE, Park SH, Dao TV, et al. BAYOU: a phase II, randomized, multicenter, double-blind, study of durvalumab (D) in combination with olaparib (O) for the first-line treatment of platinum-ineligible patients with unresectable, stage IV urothelial carcinoma (UC). J Clin Oncol. 2022;40(suppl 6):437. doi:10.1200/JCO.2022.40.6_suppl.437.