The Combination Changing the Neoadjuvant Calculus in Breast Cancer: Coral Omene, MD, PhD

Coral Omene, MD, PhD, program director of breast cancer disparities research and associate professor of medicine at Robert Wood Johnson Medical School, Division of Medical Oncology, Rutgers Cancer Institute, discusses the I-SPY 2.2 trial (NCT01042379) findings on rilvegostomig (R) plus trastuzumab deruxtecan (T-DXd) in breast cancer in an interview with The American Journal of Managed Care^® (AJMC^®) after their presentation at the 2026 American Society of Clinical Oncology (ASCO) conference.

In this study, R, a bispecific checkpoint inhibitor is paired with T-DXd, an antibody-drug conjugate (ADC), and is producing pathologic complete response (pCR) rates that far exceed the benchmarks set by the adaptive platform trial. Overall, the results raise the prospect that selected patients could reach surgery without ever receiving standard cytotoxic chemotherapy. Omene discusses the trial design, the mechanistic rationale behind the combination, and what a 57% pCR rate in hormone receptor–positive (HR+), immune-positive (immune+) patients signals for the future of neoadjuvant therapy.

This transcript has been lightly edited for clarity.

AJMC: Could you briefly outline the trial design and how it tested new combinations of treatments in a streamlined way? What are some of the major notable findings from I-SPY 2.2?

Omene: I-SPY 2.2 is a phase 2 neoadjuvant sequential multiple assignment randomization trial, which evaluates novel agents in up to 3 blocks: A, B, and C. Block A therapy is investigational (here, R + T-DXd). Surgery after Block A is allowed if pCR is likely; otherwise, patients continue to Block B. Here, treatment assignments are based on response predictive subtypes (RPS), which incorporate gene expression–based immune, HR, and HER2 status to classify tumors. If needed, Block B treatment is a taxane plus or minus platinum [chemotherapy] plus pembrolizumab. If pCR is not predicted after Block B, patients proceed to Block C treatment with Adriamycin [doxorubicin] and cyclophosphamide (AC) chemotherapy. The goal is to optimize pCR while minimizing exposure to ineffective chemotherapy.

Several regimens that "graduated" from the I-SPY 2 platform that ultimately influenced clinical practice include the following: (1) pembrolizumab in high-risk breast cancer, particularly triple-negative disease, based on up to 3 times markedly improved pCR rates. The findings anticipated the phase 3 KEYNOTE-522 trial (NCT03036488), and today pembrolizumab plus chemotherapy followed by adjuvant pembrolizumab is a standard neoadjuvant approach for high-risk early-stage triple-negative breast cancer (TNBC). (2) Pertuzumab-based neoadjuvant therapy. Pertuzumab graduated in I-SPY 2 before widespread adoption, and the phase 3 NeoSphere (NCT03595592) and related studies confirmed benefit. Today, dual HER2 blockade with trastuzumab plus pertuzumab is a standard component of neoadjuvant therapy for many patients with stage II-III HER2-positive breast cancer. (3) Neratinib in HER2-positive/HR-negative disease. Neratinib was one of the earliest I-SPY 2 graduates. Although it did not become a standard neoadjuvant regimen, the trial identified a biologically sensitive subgroup, and later studies established a role for neratinib as extended adjuvant therapy in selected HER2-positive patients. The bigger legacy of I-SPY 2 are 3 concepts that are now becoming mainstream: biomarker-selected therapy rather than treating all patients identically, pCR as an early efficacy signal for drug development, and response-adapted treatment escalation and de-escalation.

AJMC: The efficacy results showed a 57% pCR in HR+/immune+ patients against a 15% goal. Putting that into context, what does it mean to blow past a prespecified benchmark by that margin, and how confident are you in these data?

Omene: The combination of R + T-DXd for 4 cycles was highly effective in immune+ patients encompassing both HR+ and HR-negative (HR–) patients with an estimated pCR rate of 57% in HR+ and 52% among HR– patients. This is significant, as the high pCR rate suggests R + T-DXd may allow selected immune+ patients to undergo surgery early without standard chemotherapy, and the wide margin suggests a real possibility for this to show a similar signal in larger phase 3 studies in the neoadjuvant setting to establish the use of this combination.

AJMC: Rilvegostomig blocks both PD-1 and T-cell immunoreceptors, with Ig and Immunoreceptor Tyrosine-based Inhibition Motif (ITIM) domains (TIGIT) at the same time. Why is this distinctive and important?

Omene: R is a bispecific immune checkpoint antibody, a next-generation-type of immunotherapy. Unlike traditional checkpoint inhibitors that target a single immune pathway, R simultaneously binds PD-1 and TIGIT, 2 inhibitory receptors expressed on T cells. Mechanistically, by blocking (1) PD-1, restoring antitumor T-cell activity, similar to other immunotherapy drugs like pembrolizumab that we use in breast cancer, and (2) blocking TIGIT, another immune checkpoint that suppresses T-cell and natural killer (NK)–cell function, it provides synergistic dual checkpoint inhibition through a single molecule.

AJMC: What does the preclinical logic say about why pairing R with T-DXd might work better than each alone?

Omene: Preclinical and clinical data have shown that ADCs, including T-DXd, may synergize with immunotherapy, and indeed several clinical trials have established the efficacy using this combination, and it is now standard of care. For example, the ASCENT 04 study (NCT05382286) in metastatic TNBC. The logic is that ADCs, like T-DXd, may convert an immunologically "cold" tumor into a more "inflamed" and immune-responsive tumor microenvironment, making tumors more susceptible to PD-1 or PD-L1 blockade and TIGIT inhibition. Paradoxically, preclinical studies suggest that ADC-induced immune activation can increase TIGIT expression on activated T cells, and TIGIT signaling may become an important mechanism of immune escape for the tumor after ADC-induced immune activation. Thus, adding R to T-DXd ADC may interrupt this escape pathway.