The phase 2 adaptive I-SPY 2.2 trial (NCT01042379) investigated the efficacy of datopotamab deruxtecan (Dato-DXd) in combination with durvalumab (Imfinzi) in patients with stage II/III high-risk HER2-negative breast cancer. Presented at the 2024 ESMO Congress, the study highlighted the variability in pathologic complete response (pCR) rates across different response predictive subtypes (RPS).
The overall observed pCR rate with the Dato-DXd plus durvalumab treatment strategy was 50%. Meghna S. Trivedi, MD, MS, lead study author and assistant professor of medicine at Columbia University Irving Medical Center, noted that further investigation is warranted in randomized controlled trials, specifically within immune-positive and hormone receptor (HR)-negative/DNA damage repair deficiency (DRD)-negative subtypes.
Rationale and Background
The rationale for exploring this combination in the early-stage setting stems from promising results observed in the phase 1b/2 BEGONIA trial (NCT03742102). Preclinical data suggested an enhanced antitumor immune response when combining topoisomerase-1 inhibitors with PD-L1 therapy. The BEGONIA trial demonstrated a confirmed objective response rate of 79% (95% CI, 66.8%-88.3%), a median progression-free survival of 13.8 months, and a median duration of response of 15.5 months as frontline therapy in patients with metastatic triple-negative breast cancer.
The I-SPY 2.2 trial design incorporates sequential treatment blocks assigned based on a unique set of biomarkers called RPS. Patients received experimental treatment in block A, RPS-guided therapy in block B, and salvage chemotherapy in block C. Early treatment escalation was permitted in blocks A and B if needed, and de-escalation was also an option, allowing patients to proceed directly to surgery if deemed appropriate.
Study Design and Patient Population
The trial aimed to identify novel agents and individualize neoadjuvant breast cancer treatment to achieve a pCR. The RPS considers immunotherapy benefit, DRD, HR status, and HER2 status throughout treatment. Patient response was evaluated using breast MRI and core biopsy as part of the predicted RCB or pre-RCB assessment, guiding decisions for therapy de-escalation or early escalation.
Eligible patients for the Dato-DXd/durvalumab arm had anatomic stage II or III HER2-negative MammaPrint high-risk breast cancer. They received 6 mg/kg of intravenous (IV) Dato-DXd plus 1120 mg of IV durvalumab on day 1 of each 3-week cycle for up to 4 cycles in block A.
Baseline characteristics of the overall population (n = 106) showed a median age of 50.5 years (range, 25.0-77.0), with most patients being non-Hispanic/Latino (76.4%) and White (59.4%). The HR-positive, HER2-negative population (n = 42) was balanced between HR-positive/HER2-negative/immune-negative/DRD-negative (n = 25) and HR-negative/HER2-negative/immune-negative/DRD-negative (n = 23) subtypes. Among the HR-negative, HER2-negative population (n = 64), most had HER2-negative/immune-positive disease (n = 47) compared to HER2-negative/immune-negative/DRD-positive disease (n = 11).
Treatment Outcomes by Subtype
Of the 106 patients treated in block A, 64 proceeded to block B, and 25 to block C. The HR-positive/immune-negative/DRD-negative and immune-negative/DRD-positive subtypes did not outperform the dynamic control. Specifically, the pCR rate in the HR-positive/immune-negative/DRD-negative population was 14% (95% CI, 0%-28%) vs 15% in the dynamic control (95% CI, 6%-24%; P > .41). The immune-negative/DRD-positive population (n = 10) had a pCR rate of 50% (95% CI, 25%-70%) vs 72% in the dynamic control (95% CI, 52%-92%; P > .10). The HER2-negative/immune-positive population (n = 47) also did not exceed the model rate of pCR, with a rate of 77% (95% CI, 65%-89%) vs 78% in the dynamic control (P >.46).
However, the HR-negative/HER2-negative/immune-negative/DRD-negative population (n = 23), historically known for low response rates, achieved a pCR rate of 44% (95% CI, 24%-64%) compared to 16% (95% CI, 6%-26%) with the dynamic control (P > .99), meeting graduation criteria.
Timing of pCR and Toxicity
Notably, in the immune-positive and triple-negative subtypes, over 50% of the observed pCRs were achieved by the end of block A, potentially avoiding standard chemotherapy. Nearly all pCRs were achieved by the end of block B, avoiding doxorubicin and cyclophosphamide (AC).
Across all patients (n = 106), cumulative observed pCR rates by the end of blocks A, B, and C were 47% (n = 25/53), 89% (n = 47/53), and 100% (n = 53/53), respectively. In the HER2-negative/immune-positive population (n = 47), these rates were 54% (n = 20/37), 92% (n = 34/37), and 100% (n = 37/37), respectively, for a total pCR rate of 79%. In the HR-negative/HER2-negative population (n = 63), excluding one patient who did not receive pembrolizumab in block B, the respective rates were 54% (n = 21/39), 92% (n = 36/39), and 100% (n = 39/39), for a total pCR rate of 62%.
The toxicity profile of the combination was consistent with prior studies. While various adverse effects (AEs) associated with cytotoxic chemotherapy could be avoided if a pCR was achieved in block A, one participant experienced a grade 5 cardiac arrest in block B. Immune-related AEs across all blocks aligned with previous findings, and no reports of adrenal insufficiency were noted in patients who exited after block A.
