Breast Cancer Highlights From ASCO 2026: Expert Perspectives on Emerging Data

Key Takeaways

• Tersolisib, an investigational mutant-selective PI3Kα inhibitor, resulted in grade ≥3 hyperglycemia in only 5% of patients with baseline prediabetes or diabetes and in none of those with normal baseline glycemic control in the phase I/II PIKALO-1 study in patients with PIK3CA-mutant HR-positive/HER2-negative advanced breast cancer and is being evaluated in the phase III PIKALO-2 trial.
• Sacituzumab govitecan in combination with pembrolizumab is now approved for first-line treatment of adults with unresectable locally advanced or metastatic TNBC whose tumors express PD-L1 (CPS ≥10) based on the phase III ASCENT-04 trial.
• Sacituzumab govitecan and datopotamab deruxtecan are now approved for first-line treatment of adults with unresectable locally advanced or metastatic triple-negative breast cancer who are not candidates for PD-1/PD-L1 inhibitor therapy, based on the phase III ASCENT-03 trial of sacituzumab govitecan and the phase III TROPION-Breast02 trial of datopotamab deruxtecan.

In this commentary, Hope S. Rugo, MD, FASCO, and Nerea Lopetegui-Lia, MD, review key breast cancer studies presented at the 2026 American Society of Clinical Oncology (ASCO) Annual Meeting, with a focus on how new data may influence chemotherapy de-escalation in early disease, biomarker-guided treatment adaptation in advanced disease, and first-line treatment selection in metastatic triple-negative breast cancer (TNBC). Featured studies include OPTIMA, a phase III noninferiority trial of gene expression–directed chemotherapy for early-stage estrogen receptor (ER)–positive/HER2-negative breast cancer; an exploratory biomarker analysis of the phase III NATALEE trial of adjuvant ribociclib; updated results from the phase III SERENA-6 trial of circulating tumor DNA (ctDNA)–guided switching to camizestrant for advanced hormone receptor (HR)–positive/HER2-negative breast cancer; the phase Ib elacestrant plus capivasertib cohort of the ELEVATE umbrella study; updated results from the phase I/II PIKALO-1 study of the investigational agent tersolisib; and progression-free survival after next line of treatment (PFS2) analyses from ASCENT-04/KEYNOTE-D19, ASCENT-03, and TROPION-Breast02.

HR-Positive/HER2-Negative Early Breast Cancer: De-escalation and Biomarker Refinement

OPTIMA
Nerea Lopetegui-Lia, MD:
Tumor gene expression assays are widely used to guide adjuvant chemotherapy decisions in HR-positive/HER2-negative early breast cancer, but the supporting evidence has come largely from postmenopausal patients with no more than 3 involved lymph nodes. OPTIMA evaluated whether the same approach could be applied to a higher-risk patient population. This international, partially blinded, randomized phase III noninferiority trial enrolled 4429 men and women aged 40 years or older with ER-positive/HER2-negative early breast cancer and 0-9 involved axillary nodes, all considered candidates for chemotherapy. Patients were randomly assigned to standard chemotherapy followed by endocrine therapy or to a test-directed strategy in which the Prosigna 50-gene assay determined treatment: A risk of recurrence score >60 directed chemotherapy plus endocrine therapy, and a score of ≤60 directed endocrine therapy alone, with ovarian function suppression for premenopausal patients. The primary endpoint was 5-year invasive breast cancer–free survival (IBCFS), with noninferiority defined by a 3% margin.

At a median follow-up of 4 years, the test-directed strategy met the noninferiority criterion: 5-year IBCFS was 90.3% with endocrine therapy alone compared with 91.8% with standard chemotherapy plus endocrine therapy (adjusted hazard ratio: 1.03; 90% CI: 0.85-1.25; P = .006 for noninferiority), an absolute difference of 1.5% favoring chemotherapy. Among the two thirds of patients with a risk of recurrence score of ≤60, 5-year IBCFS was 93.6% with endocrine therapy alone compared with 94.8% with standard chemotherapy plus endocrine therapy (adjusted hazard ratio: 1.06; 90% CI: 0.80-1.40; P = .003). Outcomes were similar in premenopausal and postmenopausal patients and in those with 1-3 or 4-9 involved nodes, although several subgroups had few events.

These findings apply to both postmenopausal and premenopausal patients. The results in postmenopausal patients are consistent with data from trials with other gene expression assays. The OPTIMA trial also enrolled premenopausal patients aged 40 years or older, a group for whom prior evidence was more limited and who received ovarian function suppression as part of endocrine therapy. Findings in patients with 4-9 positive nodes are based on smaller numbers of events and shorter follow-up. Three points bear on implementation. First, the NCCN Breast Cancer Panel identifies the 21-gene assay (Oncotype DX) as preferred for prognosis and prediction of chemotherapy benefit and states that other gene expression assays provide prognostic information but that their ability to predict chemotherapy benefit is unknown. Second, the test-directed strategy was studied alongside endocrine therapy that included ovarian function suppression in premenopausal patients. By ensuring that patients assigned to endocrine therapy alone received ovarian function suppression, OPTIMA tests whether robust endocrine therapy can substitute for chemotherapy. This speaks to a persistent interpretive ambiguity in RxPONDER, in which premenopausal women with 1 to 3 involved nodes and a recurrence score below 26 derived an absolute chemotherapy benefit of up to approximately 5%, but it remains unclear whether that reflected true cytotoxic benefit or chemotherapy-induced ovarian failure. Third, the median follow-up of 4 years is short for an ER-positive population in which late recurrences are common, and longer follow-up will be needed to confirm durability.

NATALEE
Hope S. Rugo, MD, FASCO:
The phase III NATALEE trial established an invasive disease–free survival (iDFS) benefit for 3 years of adjuvant ribociclib plus a nonsteroidal aromatase inhibitor (AI) compared with a nonsteroidal AI alone in stage II or III HR-positive/HER2-negative early breast cancer. A prespecified exploratory analysis evaluated whether tumor biology at baseline could identify patients who derive greater benefit with the combination regimen. Baseline surgical samples from 3022 of the 5101 randomized patients were profiled with the NanoString Breast Cancer 360 Panel for PAM50 intrinsic subtype, genomic risk and proliferation signature scores, and selected gene expression markers.

PAM50 subtype was prognostic regardless of treatment assignment. Relative to luminal A tumors, the risk of an iDFS event was 1.39-fold higher with luminal B, 2.62-fold higher with HER2 enriched, and 3.92-fold higher with basal-like disease. Ribociclib benefit was observed across all subtypes, with hazard ratios from 0.77 in luminal A to 0.42 in basal-like disease, and no significant subtype-by-treatment interaction (P = .34). The numerically larger effects in the HER2-enriched (n = 89) and basal-like (n = 91) groups were based on small numbers of patients. Trends toward greater benefit with higher genomic risk (P = .42) and proliferation signature scores (P = .84) were not significant.

These findings do not establish PAM50 testing as a tool for selecting patients for adjuvant ribociclib. Because benefit was observed across subtypes, they are consistent with treating according to NATALEE clinical eligibility. The analysis also reported associations between individual genes and the magnitude of benefit, with higher expression of CEACAM6, NOD2, and GPX3, or lower expression of GATA3, SLC39A6, and MAPT, associated with greater benefit. These are exploratory observations from a single trial and require independent validation.

HR-Positive/HER2-Negative Advanced Breast Cancer: Emerging Resistance Biology and Evolving Evidence

SERENA-6
Nerea Lopetegui-Lia, MD:
ESR1 mutations are a common mechanism of acquired resistance to AI-based therapy and can be detected in ctDNA months before radiographic progression. The phase III SERENA-6 trial evaluated whether changing therapy at the time of that molecular finding, rather than at radiographic progression, improves outcomes. Adults with ER-positive/HER2-negative advanced breast cancer receiving first-line AI plus a CDK4/6 inhibitor for at least 6 months underwent serial ctDNA testing at routine scans. The 315 patients with an emergent ESR1 mutation and no evidence of progression were randomly assigned, in double-blind fashion, to switch the AI to camizestrant 75 mg once daily while continuing the same CDK4/6 inhibitor, or to continue the AI plus the same CDK4/6 inhibitor. The primary endpoint was investigator-assessed progression-free survival (PFS), and key secondary endpoints included PFS2, the time from randomization to progression on the next line of therapy, or death from any cause, as well as OS, safety, and patient-reported outcomes (PROs). Crossover was not permitted.

At a median follow-up of 23.5 months, switching to camizestrant improved median PFS vs the control arm (16.8 vs 9.2 months, respectively; hazard ratio: 0.45; 95% CI: 0.34-0.59). Final PFS2 also favored the switch to camizestrant vs control (25.7 vs 19.1 months, respectively; hazard ratio: 0.63; 95% CI: 0.46-0.86; P = .00373). Median chemotherapy/ADC-free survival was 22.6 months compared with 18.7 months (hazard ratio: 0.64; 95% CI: 0.47-0.87). In an exploratory analysis, the median reduction in ctDNA mean variant allele fraction from baseline reached 99% by Week 8 in the camizestrant arm, and total ctDNA clearance occurred in 51.0% of patients in the camizestrant arm compared with 1.9% in the control arm, and clearance across both arms was associated with improved overall survival (OS) (hazard ratio: 0.39; 95% CI: 0.19-0.73). OS remains immature (hazard ratio: 0.87; 95% CI: 0.57-1.30), and no new safety signals were reported. Additional benefits of switching to camizestrant plus a CDK4/6 inhibitor included improvements in patient-reported outcomes, including quality of life.

Questions remain about which patients should undergo serial ctDNA testing, at what interval and with which assay, and how a change in therapy at molecular progression affects the choice and sequence of later therapies. Access to oral selective estrogen receptor degrader (SERDs) in subsequent lines of therapy is currently limited in some settings. Camizestrant was approved in the European Union in July 2026, in combination with a CDK4/6 inhibitor, for adults with ER-positive/HER2-negative locally advanced or metastatic breast cancer upon detection of an ESR1 mutation and without disease progression during first-line endocrine therapy plus a CDK4/6 inhibitor. It is also approved in Japan and several other countries. In the United States, the application remains under FDA review following an extension of the decision date.

ELEVATE
Hope S. Rugo, MD, FASCO:
Resistance to first-line endocrine therapy plus CDK4/6 inhibition arises through several mechanisms, including acquired ESR1 mutations and PI3K/AKT/PTEN pathway alterations, which frequently coexist in the same tumor. The phase Ib/II ELEVATE umbrella study is evaluating the oral SERD elacestrant with a range of targeted agents. The analysis presented at ASCO 2026 reported the phase Ib elacestrant plus capivasertib cohort, an all-oral regimen combining ER degradation with AKT inhibition. The objective was to establish safety and a recommended phase II dose (RP2D) in previously treated patients, not to estimate comparative efficacy.

Thirty-one patients were treated across 3 dose cohorts. All had a PIK3CA, AKT1, or PTEN alteration, 55% also had an ESR1 mutation, 94% had received a prior CDK4/6 inhibitor, and 45% had received prior fulvestrant. None had received chemotherapy for advanced disease. Among all 31 patients treated in phase Ib, the most common adverse events of any grade were diarrhea (81%), nausea (74%), and rash (65%). Grade 3/4 rash (27%) and hyperglycemia (20%) occurred only in the highest-dose cohort, which received capivasertib 400 mg twice daily (n = 15). The RP2D was elacestrant 345 mg once daily plus capivasertib 320 mg twice daily for 4 days on and 3 days off. Efficacy was reported for the 9 patients treated at the RP2D. Median PFS was 11.3 months (95% CI: 2.86-not reached), objective response rate was 33.3%, disease control rate was 88.9%, and the 24-week clinical benefit rate was 66.7%. All 3 partial responses occurred in patients with coexisting ESR1 and PIK3CA mutations. Among these 9 patients, no grade 3/4 hyperglycemia was reported, and treatment-emergent adverse events leading to drug withdrawal and to dose reduction were each reported in 1 patient (11%).

These are early data from a small, nonrandomized cohort and the combination remains investigational. Enrollment in the phase II trial is complete (N = 64), and I look forward to those data, which are anticipated by the end of 2026.

PIKALO-1 and PIKALO-2
Hope S. Rugo, MD, FASCO:
PIK3CA mutations occur in approximately 40% of HR-positive/HER2-negative advanced breast cancers, but the use of PI3Kα inhibitors has been limited by toxicities attributed to inhibition of wild-type PI3Kα, particularly hyperglycemia, and patients with diabetes or prediabetes have often been excluded from these trials. Tersolisib is an oral, allosteric, central nervous system–penetrant, pan-mutant–selective PI3Kα inhibitor designed to suppress mutant PIK3CA signaling while sparing wild-type PI3Kα. The phase I/II PIKALO-1 trial is evaluating whether that selectivity of tersolisib is associated with a lower incidence of hyperglycemia while retaining antitumor activity. This ongoing study is evaluating tersolisib alone, tersolisib with fulvestrant, and tersolisib with endocrine therapy plus an approved CDK4/6 inhibitor. Patients with controlled type 2 diabetes were eligible, including those with an HbA1c below 8% and a fasting blood glucose below 140 mg/dL. The ASCO 2026 report presented updated safety and efficacy across the doublet and triplet cohorts.

Across the doublet and triplet cohorts (n = 199), treatment-emergent hyperglycemia occurred in 51 patients (26%) and was predominantly grade 1/2. Grade ≥3 hyperglycemia occurred only in patients with baseline prediabetes or diabetes (n = 21 in each cohort), in 5% of the fulvestrant doublet cohort and 5% of the ribociclib triplet cohort, with no such events among patients with normal baseline glycemic control. Other class toxicities, including rash and stomatitis, were reported at low rates. Among patients with measurable disease treated at doses of ≥60 mg, objective responses were observed in 31% of the fulvestrant doublet cohort (n = 36) and in 33% of the palbociclib (n = 46) and 44% of the ribociclib (n = 9) triplet cohorts, including among patients previously treated with a CDK4/6 inhibitor, although several cohorts were small and follow-up was short.

One practical question is whether patients with baseline dysglycemia should receive prophylactic antihyperglycemic therapy or be treated only if hyperglycemia develops, which is being evaluated prospectively. Tersolisib remains investigational, and PIKALO-1 is ongoing. The randomized, double-blind, placebo-controlled phase III PIKALO-2 trial is recruiting patients and is comparing first-line tersolisib plus endocrine therapy and a CDK4/6 inhibitor with placebo plus the same backbone in approximately 920 adults with PIK3CA-mutated HR-positive/HER2-negative advanced breast cancer and no prior systemic therapy for advanced disease. It includes endocrine-sensitive and endocrine-resistant populations, with investigator-assessed PFS as the primary endpoint.

First-line Metastatic TNBC: TROP-2 ADCs Expand Treatment Options

ASCENT-04/KEYNOTE-D19
Hope S. Rugo, MD, FASCO:
For patients with previously untreated PD-L1–positive advanced TNBC, the phase III ASCENT-04/KEYNOTE-D19 trial compared the TROP-2–directed ADC sacituzumab govitecan with chemotherapy, each given with pembrolizumab. It randomly assigned 443 adults with a PD-L1 combined positive score (CPS) ≥10 to sacituzumab govitecan plus pembrolizumab or to chemotherapy plus pembrolizumab, and the primary analysis showed a significant PFS improvement (11.2 vs 7.8 months; hazard ratio: 0.65; 95% CI: 0.51-0.84). Patients in the control arm could cross over to sacituzumab govitecan at progression, and the ASCO 2026 analysis reported PFS2 and time to subsequent therapies.

Median PFS2 was not reached compared with 21.0 months with chemotherapy plus pembrolizumab (hazard ratio: 0.67; 95% CI: 0.48-0.95). Among the 119 control-arm patients who received any subsequent therapy, 81% received sacituzumab govitecan. Median time to first subsequent therapy was 17.3 months compared with 9.8 months (hazard ratio: 0.59; 95% CI: 0.46-0.76). OS remains immature.

ASCENT-03
Nerea Lopetegui-Lia, MD:
ASCENT-03 evaluated sacituzumab govitecan monotherapy in patients ineligible for PD-1/PD-L1 inhibitors, for whom chemotherapy has been the standard first-line option. It randomly assigned 558 adults with previously untreated metastatic TNBC to first-line sacituzumab govitecan or investigator’s choice of chemotherapy, with a median PFS of 9.7 months compared with 6.9 months with standard chemotherapy (hazard ratio: 0.62; 95% CI: 0.50-0.77) in the primary analysis. Median PFS2 was 18.2 months compared with 14.0 months (hazard ratio: 0.70; 95% CI: 0.55-0.90). Among the 179 control-arm patients who received subsequent therapy, 82% received sacituzumab govitecan. Median time to first subsequent therapy was 11.2 months with sacituzumab govitecan compared with 7.9 months with chemotherapy (hazard ratio: 0.61; 95% CI: 0.50-0.75). OS remains immature.

These analyses complement the primary PFS results that supported the June 24, 2026, FDA approvals of sacituzumab govitecan as first-line monotherapy for adults with unresectable locally advanced or metastatic TNBC who are not candidates for PD-1 or PD-L1 inhibitor–based therapy. Given the extensive crossover to sacituzumab govitecan, which may delay a mature OS readout, the PFS2 and the prolongation of time to first and second subsequent therapy provide supportive interim evidence that first-line sacituzumab govitecan extended durable initial disease control and delayed the need for subsequent therapy.

TROPION-Breast02
Hope S. Rugo, MD, FASCO:
A second TROP-2–directed ADC, datopotamab deruxtecan, is also now approved in the immunotherapy-ineligible advanced TNBC setting. The phase III TROPION-Breast02 trial randomly assigned 644 adults with previously untreated, locally recurrent inoperable or metastatic TNBC for whom immunotherapy was not an option to datopotamab deruxtecan or investigator’s choice of chemotherapy, and met both coprimary endpoints, with a median PFS of 10.8 months compared with 5.6 months (hazard ratio: 0.57), respectively, and a median OS of 23.7 months compared with 18.7 months (hazard ratio: 0.79). On May 22, 2026, the FDA approved datopotamab deruxtecan for adults with unresectable or metastatic TNBC who are not candidates for PD-1/PD-L1 inhibitor therapy.

At ASCO 2026, additional efficacy endpoints were reported at a median follow-up of 27.5 months. Median PFS2 was 15.6 months with datopotamab deruxtecan compared with 11.8 months with chemotherapy (hazard ratio: 0.61; 95% CI: 0.50-0.74), median time to first subsequent therapy was 10.9 months compared with 5.6 months (hazard ratio: 0.49; 95% CI: 0.41-0.59), and median time to second subsequent therapy was 16.7 months compared with 12.6 months (hazard ratio: 0.67; 95% CI: 0.55-0.81). Subsequent ADC exposure was lower than in the ASCENT-03 trial, at 21% and 41% with datopotamab deruxtecan and investigator’s choice of chemotherapy, respectively, as datopotamab deruxtecan was not approved in TNBC while the trial was ongoing and crossover was not provided. This difference should be considered when comparing PFS2 results across these trials.

Sacituzumab govitecan and datopotamab deruxtecan have not been compared directly, and the trials enrolled different populations, so cross-trial efficacy comparisons are not informative. Treatment selection is guided by label eligibility and PD-L1 status, which distinguish the 2 agents in PD-L1–positive disease. The regimens also differ in schedule, with sacituzumab govitecan given on Days 1 and 8 of a 21-day cycle compared with every 3 weeks for datopotamab deruxtecan, and in toxicity profile, with neutropenia and diarrhea reported more frequently with sacituzumab govitecan and stomatitis, ocular toxicity, and interstitial lung disease or pneumonitis requiring monitoring with datopotamab deruxtecan. Prior therapy, disease tempo, comorbidities, and patient preference are also considerations when making recommendations for patients with newly diagnosed advanced TNBC.

Your Thoughts
How will findings presented at ASCO 2026 affect your practice? Leave a comment to join the conversation!