Trastuzumab Deruxtecan with Nivolumab in HER2-Expressing Metastatic Breast or Urothelial Cancer: Analysis of the Phase Ib DS8201-A-U105 Study

HER2 overexpression occurs in many cancers, including breast and urothelial cancers (1, 2). Approximately 15% to 20% of metastatic breast cancers (mBC) have HER2 overexpression (IHC 3+ or IHC 2+/ISH⁺; ref. 3), and nearly 60% of mBC tumors that were categorized as HER2 negative could be classified as HER2-low (IHC 1+ or IHC 2+/ISH⁻; ref. 4). In advanced urothelial cancer/metastatic urothelial cancer (mUC), approximately 9% to 37% of tumors have HER2 overexpression (IHC 3+ or 2+; refs. 1, 5, 6). HER2 overexpression has been associated with poor prognosis in patients with mBC and mUC, and some HER2-directed therapies have shown poor results in mUC (5, 7, 8).

Trastuzumab deruxtecan (T-DXd) is an antibody–drug conjugate (ADC) composed of an anti-HER2 antibody, a tetrapeptide-based linker, and a topoisomerase I inhibitor payload (9–12). In April 2024, T-DXd was approved for the treatment of adults with unresectable or metastatic HER2-positive (IHC 3+) solid tumors who have received prior systemic treatment and have no satisfactory alternative treatment options, supported by results from three phase II trials, including the DESTINY-PanTumor02 (NCT04482309) trial that analyzed T-DXd across multiple tumor types, including bladder cancer (13, 14). The current study is one of the first analyses of T-DXd that include patients with HER2-expressing (IHC 3+/2+ or 1+) mUC. Additionally, nivolumab is an anti–programmed cell death 1 protein (PD-1) antibody approved for the treatment of mUC and other tumor types (15–17). Nivolumab acts as an immune checkpoint inhibitor by preventing binding of the PD-1 ligands, PD-L1 and PD-L2 (17), and emerging data have shown that patients with high levels of CD8⁺ T cells have a better response to checkpoint inhibitor therapy. Because the cytotoxic effect of ADCs increases the recruitment of CD8⁺ T cells to the tumor, combining ADCs with an immuno-oncology drug may expand the therapeutic benefit (18). Moreover, most patients with estrogen receptor–positive mBC do not respond to immunotherapy; the tumor microenvironment is generally immunosuppressive and contains few tumor-infiltrating lymphocytes and reduced potential for tumor neoantigens versus HER2-positive and triple-negative breast cancer. Thus, combination therapies are needed to enhance the activity of checkpoint inhibitors in breast cancer, especially in the estrogen receptor–positive subtype (19, 20). Last, in preclinical studies of mice with human HER2-expressing tumors, T-DXd in combination with an anti–PD-1 antibody showed greater antitumor activity compared with T-DXd alone (21). We report the primary efficacy, safety, pharmacokinetic (PK), and biomarker data of T-DXd in combination with nivolumab in patients with HER2-expressing mBC and mUC in a phase Ib study.

DS8201-A-U105 was an open-label, two-part (dose escalation/expansion), nonrandomized, phase Ib study conducted to evaluate T-DXd in combination with nivolumab in patients with HER2-expressing mBC or mUC from 21 sites across the United States and Europe (NCT03523572). The study was approved by the institutional review board at each site and conducted in accordance with the International Council for Harmonization Good Clinical Practice guidelines, the Declaration of Helsinki, and local regulations with regard to the conduct of clinical research. All patients provided written informed consent.

The dose-escalation study (part 1) was conducted to determine the MTD or the recommended dose for expansion (RDE) of T-DXd when used in combination with nivolumab; the dose-expansion phase (part 2) evaluated the efficacy, safety, and tolerability of the combination treatment at the RDE. Eligible patients were ≥18 years of age, had an Eastern Cooperative Oncology Group performance status score of 0 or 1, and the presence of at least one measurable lesion based on RECIST v1.1. Patients were enrolled in four cohorts during dose-expansion part 2; patients meeting the cohort-specific eligibility criteria for any of the four cohorts in part 2 were eligible for enrollment in part 1. Cohort 1 included patients with HER2-positive mBC who had received prior trastuzumab emtansine, cohort 2 included patients with HER2-low mBC who had exhausted all treatments that could confer any clinically meaningful benefit, cohort 3 included patients with HER2-high advanced urothelial cancer/mUC who had received prior platinum-based chemotherapy, and cohort 4 included patients with HER2-low advanced urothelial cancer/mUC who had received a prior platinum-based chemotherapy regimen.

For mBC, HER2 expression was centrally confirmed prospectively by analysis of archival tissue according to the guidelines from the American Society of Clinical Oncology/College of American Pathologists (22). Accordingly, HER2 positivity was defined as HER2 IHC 3+ or IHC 2+/ISH⁺. HER2-low was defined as HER2 IHC 1+ or IHC 2+/ISH⁻. For mUC, HER2 IHC status was determined by central assessment, and scoring followed the American Society of Clinical Oncology/College of American Pathologists gastric cancer guidelines (23). HER2-high mUC was defined as HER2 expression of IHC 2+ or 3+, whereas HER2-low mUC was defined as HER2 expression of IHC 1+. The VENTANA/PATHWAY anti-HER2 (4B5) antibody (RRID: AB_3101913) was the assay used.

Patients were excluded if they had prior nivolumab or T-DXd, anti–PD-1, anti–PD-L1, or anti–CTL-associated protein 4 antibody, or any other antibody or drug specifically targeting T-cell co-stimulation or checkpoint pathways; had received prior anti-HER2 therapy (except for patients with HER2-positive breast cancer); or had active central nervous system (CNS) metastases. Patients with treated/stable CNS metastases (i.e., patients who received prior CNS-directed therapy for their CNS metastases and whose CNS disease is stable) and did not require treatment with corticosteroids or anticonvulsants were allowed to enroll (24).

Part 1 (dose escalation) of this study enrolled patients using a 3 + 3 + 3 design (Supplementary Fig. S1). Patients were treated with a starting dose of T-DXd 3.2 mg/kg in combination with a fixed dose of nivolumab 360 mg intravenously every 3 weeks; escalation/de-escalation of T-DXd to the next dose was based on an acceptable safety signal from the earlier dose cohort. The dose-limiting toxicity observation period was every two cycles (6 weeks). In part 2 (dose expansion), patients received the RDE determined in part 1 (T-DXd 5.4 mg/kg plus nivolumab 360 mg intravenously every 3 weeks). All patients received treatment until disease progression, occurrence of unacceptable adverse events, withdrawal of consent, or death.

The primary objective of part 1 was to determine the MTD or RDE of T-DXd in combination with nivolumab. The primary efficacy endpoint for each cohort of part 2 was confirmed objective response rate (cORR) by independent central review (ICR) per RECIST v1.1 and included patients treated at T-DXd 5.4 mg/kg during part 1 per study protocol. Secondary endpoints for part 2 included median duration of response (DOR), disease control rate (DCR), median progression-free survival (PFS) by ICR, time to response (TTR) by ICR, best percentage change in the sum of diameters of measurable tumors, and overall survival (OS). Tumor imaging assessments were performed per RECIST v1.1 at screening and every 6 weeks during the first year and every 12 weeks thereafter until discontinuation criteria were met by protocol. PK endpoints including serum concentrations of T-DXd, total anti-HER2 antibody, and DXd were measured using validated assays. Serum samples were taken at 8 hours before infusion, within 15 minutes after the end of infusion of T-DXd, and 5 hours from the start of T-DXd infusion (±2 hours) at cycles 1 and 2, day 1; and within 30 minutes of infusion on cycle 3, day 1. Patient demographics and baseline characteristics were summarized by cohort.

The planned sample size was between approximately 99 and 108 patients. Assuming an observed objective response rate (ORR) of 63.3% for 30 patients in cohort 1, the 80% confidence interval (CI) would be 50.0% to 75.2%, and assuming an observed ORR of 30.0% in cohort 3, the 80% CI would be 19.0% to 43.2%. Cohorts 2 and 4 were exploratory, and enrollment of 15 patients in each cohort was planned. The primary database lock occurred 6 months after the last patient’s first visit. For the primary analyses and in each respective cohort, patients enrolled in part 1, who were treated at the RDE, were pooled with patients enrolled in part 2.

The cORR by ICR and DCR estimates and two-sided 95% exact CI were calculated. cORR was defined as the proportion of patients who achieved a best overall response of a confirmed complete response (CR) or partial response (PR) as defined by RECIST v1.1. The DOR, PFS, and OS Kaplan–Meier estimates of medians and their respective 95% CIs were determined using the Brookmeyer and Crowley method (25); summary statistics were provided for TTR. Best response was calculated based on RECIST v1.1 criteria. Additional methods, including the definitions of each endpoint, exclusion criteria, assessments, population analysis sets, biomarker analyses, and statistical analyses are included in the Supplementary Methods.

Anonymized individual participant data (IPD) on completed studies and applicable supporting clinical study documents may be available upon request at https://vivli.org/. In cases where clinical study data and supporting documents are provided pursuant to our company policies and procedures, Daiichi Sankyo Companies will continue to protect the privacy of company and our clinical study subjects. Details on data sharing criteria and the procedure for requesting access can be found at this web address: https://vivli.org/ourmember/daiichi-sankyo/.

Seven patients with mBC were enrolled in part 1; of these patients, three were treated with T-DXd 5.4 mg/kg plus nivolumab 360 mg intravenously every 3 weeks. Results from these three patients were then pooled with part 2 data, depending on centrally determined HER2 expression and indication. The remaining four patients were treated with T-DXd 3.2 mg/kg plus nivolumab 360 mg intravenously every 3 weeks and were only included in the part 1 analysis; of these four patients with mBC, there were two with HER2 IHC 3+ and one each with HER2 IHC 2+/ISH⁺ and IHC 1+ (Supplementary Table S1). The median treatment duration of these four patients was 3.4 months (range, 1.0–11.0 months) with T-DXd and 2.8 months (range, 1.0–4.0 months) with nivolumab. Of the four patients, two patients and one patient experienced a treatment-emergent adverse event (TEAE) related to T-DXd or nivolumab, respectively. There were no dose-limiting toxicities observed in part 1 of the study (Supplementary Table S2). A TEAE associated with study drug discontinuation was experienced by one patient and was related to T-DXd. Based on the safety data of part 1, the RDE for T-DXd was 5.4 mg/kg in combination with nivolumab 360 mg intravenously every 3 weeks. Overall, among these four patients, one patient with HER2 IHC 3+ mBC achieved a confirmed PR, one patient with HER2 IHC 2+/ISH⁺ mBC achieved stable disease (SD), and two patients with breast cancer (HER2 IHC 1+ and IHC 3+, respectively) were not evaluable for response (Supplementary Table S3).

Eighty-two patients [48 with mBC (inclusive of the three patients with HER2-positive mBC who received T-DXd 5.4 mg/kg in part 1); 34 with mUC] were enrolled as of the data cutoff (July 22, 2021). Thirty-two patients with HER2-positive (IHC 3+ or IHC 2+/ISH⁺) mBC were enrolled in cohort 1 [hormone receptor–positive (HR⁺), 20 patients; HR⁻, 12 patients] and 16 patients with HER2-low (IHC 1+ or IHC 2+/ISH⁻) mBC were enrolled in cohort 2 (HR⁺, 13 patients; HR⁻, three patients). Thirty patients with HER2-high (IHC 3+/2+) mUC were enrolled in cohort 3. Only four patients with HER2-low (IHC 1+) mUC were enrolled in cohort 4 because slow enrollment led to premature closing, with the low prevalence of HER2-low (IHC 1+) mUC being a key factor.

At data cutoff (July 22, 2021), seven patients (21.9%), one patient (6.3%), four patients (13.3%), and zero patients remained on T-DXd in cohorts 1 to 4, respectively (Supplementary Table S4). Most patients discontinued T-DXd treatment because of progressive disease (PD) in the mBC cohorts 1 and 2 (28.1% and 68.8%, respectively) and in the mUC cohorts 3 and 4 (50% each). Six patients (18.8%), zero patients, four patients (13.3%), and one patient (25.0%) in cohorts 1 to 4, respectively, remained on nivolumab treatment (Supplementary Table S4). Most patients with mBC in cohorts 1 and 2 (25.0% and 68.8%, respectively) and with mUC in cohorts 3 and 4 (43.3% and 50.0%) discontinued nivolumab treatment because of PD. Baseline characteristics and prior therapy are summarized in Table 1. Representativeness of the study participants is shown in Supplementary Table S5. The median number of unique regimens in the metastatic setting in mBC cohorts 1 and 2 was 5.0 lines (range, 0–14.0 lines) and 4.0 lines (range, 0–10.0 lines), respectively; in mUC cohorts 3 and 4, it was 1.0 line (range, 0–3.0 lines) and 0.5 lines (range, 0–1.0 line). The median duration of follow-up in mBC cohorts 1 and 2 was 18.7 months (range, 1.7–26.9 months) and 12.7 months (range, 1.7–21.4 months), respectively; in mUC cohorts 3 and 4, it was 9.2 months (0.3–21.3 months) and 6.2 months (2.6–15.0 months), respectively (Table 2).

cORR by ICR, the primary endpoint for the dose-expansion study, DCR, clinical benefit rate, DOR, TTR, median PFS, median OS, and landmark OS were assessed for each cohort, with the exception of cohort 4 due to the small sample size (Fig. 1A–C; Table 2). For cohorts 1 to 3, antitumor activity of T-DXd in combination with nivolumab was observed, as shown by changes in tumor size from baseline over time (Supplementary Fig. S2A–S2C).

The cORR by ICR was 65.6% (21/32 patients; 95% CI, 46.8%–81.4%), including 3 CRs (9.4%) and 18 PRs (56.3%). The median DOR was not evaluable (NE; 95% CI, 7.9 months–NE), and the TTR was 1.6 months (range, 1.2–5.5 months). The median PFS was 11.6 months (95% CI, 6.9 months–NE). The median OS was NE (95% CI, 20.8 months–NE).

The cORR by ICR was 50.0% (8/16 patients; 95% CI, 24.7%–75.3%); all were PRs (eight patients). Of the 13 patients with HR⁺/HER2-low mBC, seven achieved confirmed PR, three achieved SD, one achieved PD, and two were NE. Of the three patients with HR⁻/HER2-low mBC, one achieved confirmed PR, one achieved SD, and one achieved PD. The median DOR was 5.5 months (95% CI, 2.8–8.0 months), and the TTR was 3.7 months (range, 2.6–9.8 months). The median PFS and OS were 7.0 months (95% CI, 2.3–10.8 months) and 19.5 months (95% CI, 2.7 months–NE), respectively.

The cORR by ICR was 36.7% (11/30 patients; 95% CI, 19.9%–56.1%), including four CRs (13.3%) and seven PRs (23.3%). The median DOR was 13.1 months (95% CI, 4.1 months–NE), and the TTR was 1.9 months (range, 1.2–6.9 months). The median PFS and OS were 6.9 months (95% CI, 2.7–14.4 months) and 11.0 months (95% CI, 7.2 months–NE), respectively. Among a subgroup of patients with HER2 IHC 3+, five of eight (62.5%) patients had a confirmed objective response.

In cohort 4 (four patients), two patients achieved PRs, one patient achieved SD, and the other achieved PD as the best response. The median PFS and OS were both NE. cORR and TTR were not summarized because of the small sample size.

The median treatment duration in patients with mBC in cohort 1 (n = 29) was 8.9 months (range, 1–23 months) with T-DXd and 5.5 months (range, 1–23 months) with nivolumab (Table 3). In cohort 1, 15 patients (51.7%) experienced grade ≥3 TEAEs. The most common TEAEs by preferred term were nausea (16 patients; 55.2%), fatigue (14 patients; 48.3%), and constipation (14 patients; 48.3%; Supplementary Table S6). TEAEs associated with T-DXd discontinuation occurred in 12 patients (41.4%), with the most common being respiratory disorders [pneumonitis, three patients (10.3%); interstitial lung disease (ILD), two patients (6.9%); acute ILD, one patient (3.4%)]. Fourteen patients (48.3%) experienced TEAEs associated with nivolumab discontinuation, with the most common being pneumonitis and ILD [two patients (6.9%) each]. TEAEs associated with death occurred in two patients (6.9%), of which one (3.4%) was adjudicated by the committee as drug-related grade 5 ILD/pneumonitis and the other was unrelated disease progression.

In cohort 2 (n = 16), the median treatment duration with T-DXd and nivolumab was 6.9 months (range, 1–21 months) and 5.9 months (range, 1–14 months), respectively (Table 3). Overall, seven patients (43.8%) experienced grade ≥3 TEAEs. The most common TEAEs by preferred term were nausea (10 patients; 62.5%), vomiting (8 patients; 50.0%), and anemia (8 patients; 50.0%; Supplementary Table S6]. TEAEs associated with T-DXd discontinuation occurred in one patient (6.3%) who experienced pneumonitis. TEAEs associated with nivolumab discontinuation occurred in two patients (12.5%) and included neutropenia and aspartate aminotransferase increased [one patient (6.3%) each]. TEAEs associated with death occurred in two patients (12.5%; disease progression), of which neither was drug related.

In patients with mBC who received T-DXd 5.4 mg/kg in part 2, adjudicated drug-related ILD/pneumonitis was observed in six patients (20.7%) in cohort 1, with five grade 2 (17.2%) and one grade 5 (3.4%) ILD events (Table 4). No drug-related cases were observed in cohort 2. The median time to first adjudicated drug-related ILD/pneumonitis event was 193.0 days (range, 40–250 days), with a median duration of 71.0 days (range, 29–102 days) in cohort 1. Per protocol, patients with ILD/pneumonitis received systemic steroid treatment. Left ventricular dysfunction occurred in one patient (3.4%; grade 2) in cohort 1 and no patients in cohort 2 (Supplementary Table S7).

In patients with mUC in cohort 3 (n = 30), the median treatment duration with T-DXd and nivolumab was 3.9 months (range, 1–21 months) and 4.1 months (range, 1–20 months), respectively (Table 3). Overall, 21 patients (70.0%) experienced grade ≥3 TEAEs. The most common TEAEs by preferred term were nausea (22 patients; 73.3%), fatigue (16 patients; 53.3%), and vomiting (14 patients; 46.7%; Supplementary Table S6). TEAEs associated with T-DXd discontinuation occurred in five patients (16.7%) and included pneumonitis (four patients; 13.3%) and one case (3.3%) each of alanine and aspartate aminotransferase increased. TEAEs associated with nivolumab discontinuation were experienced by eight patients (26.7%), with the most common being pneumonitis (three patients; 10.0%) and aspartate aminotransferase increased (two patients; 6.7%). TEAEs associated with death occurred in seven patients (23.3%), of which only one (3.3%) was a drug-related TEAE (grade 5 ILD/pneumonitis).

In cohort 4, all patients experienced grade ≥3 TEAEs (Table 3). The most common TEAEs by preferred term were nausea (three patients; 75.0%) and fatigue (two patients; 50.0%; Supplementary Table S6). TEAEs associated with T-DXd or nivolumab discontinuation were experienced by 50.0% and 25.0% of patients, respectively. There were no TEAEs associated with death in cohort 4.

In patients with mUC, adjudicated drug-related ILD/pneumonitis was observed in six patients (20.0%) in cohort 3 and two patients (50.0%) in cohort 4 (Table 4). There were four grade 2 (13.3%), one grade 3 (3.3%), and one grade 5 (3.3%) ILD events in cohort 3 and two grade 1 (50.0%) ILD events in cohort 4. The median time to first adjudicated ILD/pneumonitis event was 135.5 days (range, 42–342 days) and 56.0 days (range, 35–77 days) in cohorts 3 and 4, respectively. A median duration of 174.5 days (range, 162–187 days) was observed in cohort 3, and it was not estimable in cohort 4. Per protocol, patients with ILD/pneumonitis received systemic steroid treatment. Left ventricular dysfunction occurred in one patient (3.3%; grade 2) in cohort 3 and no patients in cohort 4 (Supplementary Table S7).

PK concentrations across sampling time points on day 1 of cycles 1 to 3 (i.e., at end of infusion and 5 hours after the end of infusion on day 1 of cycles 1 and 2 and before infusion on day 1 of cycles 2 and 3) were comparable between patients with mBC (cohorts 1 and 2) and those with mUC (cohorts 3 and 4) in part 2 for T-DXd, total mAb, and DXd (Supplementary Tables S8–S10). No patients tested positive for treatment-emergent antidrug antibodies (ADAs) or neutralizing ADAs. The prevalence of ADA positivity was generally low across the cohorts, with only 10 patients (12.2%) testing positive at baseline [cohort 1: two patients (6.9%); cohort 2: four patients (25.0%); cohort 3: four patients (13.3%); and cohort 4: zero patients]. Overall, 9 of the 10 patients had baseline positive ADA samples but no positive ADA samples after the first dose of study drug regimen.

In patients with mBC, antitumor activity with T-DXd plus nivolumab treatment was observed regardless of PD-L1 IHC status or blood tumor mutational burden (bTMB) levels (Supplementary Tables S11 and S12). Of note, there were five patients with mBC who had high bTMB ≥20 mut/Mb, three from cohort 1 (HER2-positive) and two from cohort 2 (HER2-low). Of these, one of three patients with HER2-positive mBC and both patients with HER2-low mBC achieved a PR to T-DXd plus nivolumab (Supplementary Table S12).

In patients with mUC in cohort 3, tumor cell scoring did not show any differences in ORR with respect to PD-L1 levels (Supplementary Table S13). However, there were a higher number of responders in patients with a visual combined positive score (vCPS) cutoff ≥1; in patients with PD-L1 IHC vCPS <1, no patient (0/6) responded, whereas in patients with PD-L1 IHC vCPS ≥1, 57.1% (8/14; 95% CI, 28.9%–82.3%) of patients responded. Moreover, there were a numerically higher number of responders in the bTMB high-expression subgroups [ORR, 38.9% (95% CI, 17.3%–64.3%) for bTMB ≥10 mut/Mb; ORR, 60.0% (95% CI, 26.2%–87.8%) for bTMB ≥20 mut/Mb] versus low-expression groups [ORR, 25.0% (95% CI, 3.2%–65.1%) and 18.8% (95% CI, 4.0%–45.6%), respectively; Supplementary Table S12].

In this study, a T-DXd dose of 5.4 mg/kg, the RDE determined in part 1, in combination with nivolumab (360 mg) intravenously every 3 weeks was tolerable and showed antitumor activity in patients with heavily pretreated HER2-positive mBC (ORR, 65.6%) or HER2-high (IHC 3+/2+)–expressing mUC (ORR, 36.7%), especially those with HER2 IHC 3+ mUC [five of eight patients (62.5%) had a confirmed objective response]. This antitumor activity is similar to that observed in previous studies of T-DXd monotherapy in these patient populations (13, 26–28). Although data from the HER2-low (IHC 1+ or 2+/ISH⁻) mBC and HER2-low (IHC 1+) mUC cohorts 2 and 4, respectively, also showed antitumor activity, the sample size was insufficient to conclusively determine the effects of nivolumab combined with T-DXd in this setting. The benefit of adding nivolumab to T-DXd in HER2-positive or HER2-low mBC also could not be clearly determined, in terms of antitumor activity, in this nonrandomized phase Ib trial. However, because immuno-oncology therapy does not contribute as much in an immunosuppressive late-line setting and may be more effective in earlier stages of cancer and lines of treatment, the efficacy observed in heavily pretreated patients in the metastatic setting in the current study may be primarily due to T-DXd activity (18, 29, 30). With regard to mUC, the efficacy results show promising and durable antitumor activity, particularly in patients with high levels of HER2 expression (IHC 3+; ref. 13). Overall, the current study shows that T-DXd may be combined with nivolumab in patients with HER2-expressing mBC and mUC, and further exploration of the T-DXd plus nivolumab combination in other indications is warranted. However, the added value of this combination would need to be explored in a randomized setting to appropriately characterize the contribution of immunotherapy to the known antitumor activity of T-DXd.

The overall safety profile was generally consistent with previous studies for T-DXd monotherapy in patients with mBC, and the addition of nivolumab did not seem to result in any overall increased toxicity, although the rate of drug discontinuation due to a TEAE was numerically higher in this study versus studies of T-DXd monotherapy for mBC in a similar patient population (26–28, 31). In the current study, 28.9% of patients with mBC (cohorts 1 and 2) experienced a TEAE leading to T-DXd discontinuation, which included adjudicated ILD/pneumonitis in cohort 1 (two patients; 6.9%). In previous studies of T-DXd in mBC, the incidence of TEAEs that led to T-DXd discontinuation ranged from 13.6% to 20.0% (26–28, 31). Furthermore, the risk of ILD/pneumonitis in the mBC cohorts was as expected based on previous studies of T-DXd in mBC; combination therapy with immunotherapy did not seem to increase the risk of ILD/pneumonitis. The incidence of all-grade adjudicated drug-related ILD/pneumonitis was 20.7% in cohort 1 and 0% in cohort 2; all but one ILD/pneumonitis events were grade 2. One patient in cohort 1 died from adjudicated drug-related ILD/pneumonitis (grade 5). Because ILD/pneumonitis is an important identified risk for T-DXd (26, 27, 31), which requires careful monitoring and prompt intervention, the assessment of ILD/pneumonitis events continues to be an important analysis in studies of T-DXd.

In patients with mUC, the safety profile and incidence of adjudicated drug-related ILD/pneumonitis were generally consistent with those reported in previous studies of T-DXd in other cancer types (32) and with the overall safety of nivolumab in patients with mUC (15, 16). The incidence rate of all-grade adjudicated drug-related ILD/pneumonitis in patients with HER2-high mUC (cohort 3) was 20.0%; the majority of events were grade 2, with 1 (3.3%) grade 5 event; two of four patients with HER2-low mUC (cohort 4) had ILD/pneumonitis, which were both grade 1. This rate of ILD/pneumonitis has been consistent with a previous analysis of T-DXd, which reported a rate of 20.6% in tumor types other than breast, gastric, lung, or colorectal cancer (32). Further exploration (i.e., larger sample size) would be needed to further assess the potential risk for ILD/pneumonitis in patients with mUC.

PK and ADA results in this study were comparable with those of previous reports (26, 33, 34). In this study, the prevalence of ADA positivity was generally low, and there were no treatment-emergent neutralizing anti–T-DXd antibodies detected. An exploratory analysis showed that patients with HER2-expressing mBC (cohorts 1 and 2) and mUC (cohort 3) experienced response to treatment with T-DXd plus nivolumab regardless of PD-L1 IHC status and in patients with high bTMB. Nevertheless, these exploratory biomarker results among patients with mBC and mUC should be interpreted with caution because of the small sample sizes of patients with PD-L1–positive status and the fact that markers such as PD-L1 expression may be prognostic rather than predictive of response (35–39).

A limitation of this study was the small sample size of patients with HER2-low mBC (16 patients) and HER2-low mUC (four patients), as well as the nonrandomized study design. Moreover, outcomes cannot be interpreted with respect to triple-negative breast cancer because more than 80% of patients with HER2-low disease in this study had HR⁺ disease. However, other clinical trials are exploring the use of T-DXd in combination with immunotherapy in this patient population (40). T-DXd as monotherapy is also being investigated in other HER2-expressing solid tumors in DESTINY-PanTumor02, including biliary tract cancer, cervical cancer, endometrial cancer, ovarian cancer, pancreatic cancer, and other tumors (13). In summary, T-DXd in combination with nivolumab demonstrated antitumor activity in patients with HER2-expressing mBC or mUC and had a manageable safety profile consistent with the known profile of T-DXd; no new safety signals were identified.

E. Hamilton reports grants and other support from Daiichi Sankyo during the conduct of the study, as well as grants and other support from Accutar Biotechnology, Arvinas, AstraZeneca, Daiichi Sankyo, Ellipses Pharma, Gilead Sciences, Eli Lilly and Company, Mersana, Novartis, Olema, Pfizer, Roche/Genentech, and Stemline Therapeutics, grants from AbbVie, Acerta Pharma, ADC Therapeutics, Akesobio Australia, Amgen, Aravive, ArQule, Artios, Atlas Medx, BeiGene, Black Diamond, Bliss Biopharmaceutical, Boehringer Ingelheim, Bristol Myers Squibb, Cascadian Therapeutics, Clovis, Compugen, Context Therapeutics, Cullinan, Curis, CytomX, Dana-Farber Cancer Institute, Dantari, Deciphera, Duality Biologics, eFFECTOR Therapeutics, Eisai, Elucida Oncology, EMD Serono, Fochon Pharmaceuticals, Fujifilm, G1 Therapeutics, H3 Biomedicine, Harpoon, Hutchison MediPharma, ImmunoGen, Immunomedics, Incyte, Infinity Pharmaceuticals, Inspirna, InventisBio, Jacobio, Karyopharm, K-Group Beta, Kind Pharmaceuticals, Leap Therapeutics, Loxo Oncology, Lycera, MabSpace Biosciences, MacroGenics, MedImmune, Merus, Millennium, Molecular Templates, Myriad Genetic Laboratories, NuCana, OncoMed, Oncothyreon, ORIC Pharmaceuticals, Orinove, Orum Pharmaceuticals, PharmaMar, Pieris Pharmaceuticals, Pionyr Immunotherapeutics, Plexxikon, Prelude Therapeutics, ProfoundBio, Radius Health, Regeneron, Relay Therapeutics, Repertoire Immune Medicines, RGENIX, Seagen, Sermonix Pharmaceuticals, Shattuck Labs, Silverback Therapeutics, Stemcentrx, Sutro, Syndax, Syros, Taiho, TapImmune, Tesaro, Tolmar, Torque Therapeutics, Treadwell Therapeutics, Verastem, Zenith Epigenetics, and Zymeworks, and other support from Circle Pharma, Entos, Fosun Pharma, Janssen, Jazz Pharmaceuticals, Jefferies LLC, Johnson and Johnson, Medical Pharma Services, Tempus Labs, Theratechnologies, Tubulis, Verascity Science, and Zentalis Pharmaceuticals outside the submitted work. M.D. Galsky reports personal fees from Astellas, Bristol Myers Squibb, Merck, Genentech, AstraZeneca, EMD Serono, Seagen, Janssen, Numab, Dragonfly, GSK, Basilea, UroGen, Rappta Therapeutics, Fujifilm, Curis, Gilead Sciences, Asieris, AbbVie, Daiichi Sankyo, and Aktis outside the submitted work. S. Ochsenreither reports other support from Daiichi Sankyo during the conduct of the study; personal fees from Bristol Myers Squibb, Genmab, Immunocore, Ipsen, Janssen, Merck, MSD, and Pfizer; grants from Bayer outside the submitted work; and a patent on T-cell Therapy Target licensed to Fred Hutchinson Cancer Research Center, Seattle, WA. G. Del Conte reports other support from MSD Italia Srl, Janssen-Cilag SpA, and Daiichi Sankyo outside the submitted work. M. Martín reports grants and personal fees from Roche and Novartis and personal fees from Pfizer, Puma, Eli Lilly and Company, Daiichi Sankyo, AstraZeneca, GSK, and MSD outside the submitted work. E.Y. Yu reports grants from Daiichi Sankyo during the conduct of the study, as well as personal fees from Bristol Myers Squibb and grants from Seagen outside the submitted work. A. Williams reports personal fees from Ellipses Pharma and Hospital Corporation of America International outside the submitted work. M. Gion reports grants from AstraZeneca, Gilead, Pfizer, and Novartis outside the submitted work. A.R. Tan reports other support from Merck, Genentech, and GlaxoSmithKline and personal fees from AstraZeneca outside the submitted work. L. Agrawal reports personal fees from Eli Lilly and Company, Pfizer, AstraZeneca, and Breast Cancer Index outside the submitted work. J.-P. Machiels reports other support from Pfizer, Roche, Bayer, Merck Serono, Boehringer Ingelheim, Bristol Myers Squibb, Novartis, Incyte, Cue Biopharma, ALX Oncology, iTeos, etherna, Nektar, F-Star, Seagen, Genmab, Astellas, CureVac, MSD, GSK, Merus, Ipsen, Amgen, Gilead, and Sanofi outside the submitted work. P.R. Debruyne reports grants from Pfizer and personal fees from Bristol Myers Squibb, Ipsen, Merck, Astellas, and Bayer outside the submitted work, as well as stock ownership in Alkermes, Mural Oncology, and Biocartis NV. V. Moreno reports other support from AbbVie, Roche, Bayer, Bristol Myers Squibb, Janssen, Syneos, Affimed, and AstraZeneca outside the submitted work, as well as being a principal investigator receiving institutional funding from AbbVie, Achilles, Adaptimmune, ADC Therapeutics, Ascendis Pharma, AstraZeneca, Bayer, BeiGene, Bicycle Therapeutics, BioInvent, Biomea Fusion, BioNTech, Bristol Myers Squibb, Boehringer Ingelheim, C4 Therapeutics, Calico Life Sciences LLC, Celgene, Constellation, Crescendo Biologics, Cullinan, Daiichi Sankyo, Debiopharm, Dragonfly, Enliven Therapeutics, Epizyme, Exelixis, FameWave, F-star Beta Limited, Genentech, Genmab, Gilead, Grey Wolf Therapeutics, GSK, Hexal AG & Sandoz, HiFiBiO, Hookipa Biotech, HUTCHMED, IGM Biosciences, Imcheck Therapeutics, Immunocore, Immutep, Incyte iOmx Therapeutics, Iovance, Italfarmaco, iTeos, Janssen, Light Chain Bioscience, Eli Lilly and Company, Loxo Oncology, Merck, Merus, Miltenyi Biomedicine, Monta Biosciences, MSD, Mythic Therapeutics, Ningbo Newbay, Novartis, Oxford BioTherapeutics, Pfizer, PharmaMar, PMV Pharma, Prelude Therapeutics Inc., Pyxis Oncology, Regeneron, Relay Therapeutics, Repare Therapeutics, Revolution, Roche, Schrödinger, Scorpion Therapeutics, Seagen, Shattuck Labs, Synthorx, Takeda, Tango Therapeutics, Tesaro, Totus Medicines, Turning Point Therapeutics, and Vividion Therapeutics. A. Minchom reports other support from Daiichi Sankyo during the conduct of the study, as well as personal fees from Janssen, Pfizer, AstraZeneca, Immutep, Chugai, Faron, GSK, Seagen, Takeda, Amgen, Merck, MSD, and Genmab and grants from Astex, Merck, and MSD outside the submitted work. F. Valdes-Albini reports grants from Daiichi Sankyo during the conduct of the study, as well as personal fees and other support from AstraZeneca outside the submitted work. D. Petrylak reports grants and personal fees from Seattle Genetics, Merck, Bicycle Therapeutics, Pfizer, Corbus, Astellas, and Gilead, grants from Arvinas, Exelixis, and Daiichi Sankyo, and personal fees from Bristol Myers Squibb, Bayer, Janssen, ProGenics, and Sanofi outside the submitted work. S. Hurvitz reports grants and nonfinancial support from Daiichi Sankyo during the conduct of the study, as well as grants from Ambrx, Amgen, Bayer, CytomX, Celcuity, Dantari, G1 Therapeutics, GSK, MacroGenics, Orinove, Orum, Phoenix Molecular Designs, Pieris, Radius/Stemline/Menarini, and Sanofi, grants and nonfinancial support from Arvinas, AstraZeneca, Genentech/Roche, Gilead/Immunomedics, Greenwich LifeSciences, Novartis, OBI Pharma, Eli Lilly and Company/Loxo, Pfizer, Puma, Seagen, and Zymeworks/Jazz Pharma, and other support from BriaCell outside the submitted work. No disclosures were reported by the other authors.

E. Hamilton: Conceptualization, investigation, methodology, project administration. M.D. Galsky: Investigation, project administration. S. Ochsenreither: Investigation. G. Del Conte: Investigation. M. Martín: Investigation. M.J. De Miguel: Investigation. E.Y. Yu: Investigation. A. Williams: Investigation. M. Gion: Investigation. A.R. Tan: Investigation. L. Agrawal: Investigation. A. Rutten: Investigation. J.-P. Machiels: Investigation. S. Cresta: Investigation. P.R. Debruyne: Investigation. A. Hennequin: Investigation. V. Moreno: Investigation. A. Minchom: Investigation. F. Valdes-Albini: Investigation. D. Petrylak: Investigation. L. Li: Formal analysis. Z. Tsuchihashi: Conceptualization, formal analysis. F. Suto: Conceptualization, formal analysis. F.-C. Cheng: Conceptualization, formal analysis, methodology. M. Kandil: Conceptualization, methodology. D. Barrios: Conceptualization, formal analysis, methodology, project administration. S. Hurvitz: Conceptualization, investigation, methodology, project administration.

This study was sponsored by Daiichi Sankyo Co., Inc., and AstraZeneca. For data collection and analysis, the study was designed and led by Daiichi Sankyo. AstraZeneca entered into a collaboration agreement with Daiichi Sankyo Co., Ltd, for T-DXd in March 2019. Daiichi Sankyo and AstraZeneca were both involved in study oversight and data collection. Nivolumab was provided by Bristol Myers Squibb. We thank the patients who participated in this study, as well as their families and caregivers. We also thank the staff and investigators at all the study sites. We also thank Dr. Yusuke Kuwahara for contributions to the informatic analysis and Dr. John Allard for contributions to the biomarker analysis. Under the guidance of the authors, assistance in medical writing and editorial support was provided by Andre Wang, PharmD, and Jill Shults, PhD, of ApotheCom, and was funded by Daiichi Sankyo, Inc.