The second line may be the optimal setting for patients with HR-positive, HER2-negative advanced breast cancer to receive CDK4/6 inhibitors, according to findings from the phase 3 SONIA trial (NCT03425838) presented at the 2023 ASCO Annual Meeting.

These findings demonstrated that the frontline use of CDK4/6 inhibitors prolongs the time on CDK4/6 inhibitors and could increase costs and toxicities. Additionally, first-line treatment with CDK4/6 inhibition plus aromatase inhibition (AI), followed by fulvestrant (Faslodex), did not improve progression-free survival (PFS), overall survival (OS), or quality-of-life (QOL) vs second-line CDK4/6 inhibition plus fulvestrant after first-line AI therapy.

For patients who received frontline-treatment with the CDK4/6 inhibitor plus AI (n = 524), there were a total of 310 events at data cutoff vs 407 among those who received CDK4/6 inhibition in the second-line. The median PFS was 24.7 months vs 16.1 months (HR, 0.59; 95% CI, 0.51-0.69; P < .0001). When evaluating PFS after 2 lines of therapy (PFS2), the number of events per treatment sequence was 281 with first-line CDK4/6 inhibition and 310 with second-line CDK4/6 inhibition. The median PFS2 was 31.0 months vs 26.8 months, respectively (HR, 0.87; 95% CI, 0.74-1.03; P = .10).

The OS analysis was at 35% data maturity and the number of events in the first- and second-line, respectively, was 184 vs 188. The median OS between the 2 arms, respectively, was 45.9 months vs 53.7 months (HR, 0.98; 95% CI, 0.80-1.20; P = .83).

Moreover, the safety profile was reported to be consistent for CDK4/6 inhibitors. The most common adverse events (AEs) were neutropenia, liver function abnormalities, anemia, and thrombocytopenia. The total number of grade 3 or higher AEs was 2782 when CDK4/6 inhibition was used in the frontline. Comparatively, there were 1620 grade 3 or higher AEs when this treatment was used in the second-line. In total, there was a 42% decreased in the rate of grade 3 or higher AEs.

The PFS2 analysis findings were consistent across subgroups and no test for interaction reached statistical significance.

“SONIA challenges the need for first-line use of a CDK4/6 inhibitor,” Gabe S. Sonke, MD, PhD, a medical oncologist at the Netherlands Cancer Institute, said in a presentation of the findings. “Endocrine monotherapy remains an excellent option, especially in [the] first line.”

As Sonke explained in the presentation, combination treatment has also been shown to increase a patient’s risk of developing a resistant pattern, such as developing an acquired ESR1 mutation. He stated that following initial reports with palbociclib (Ibrance), abemaciclib (Verzenio), and ribociclib (Kisqali), which showed the CDK4/6 inhibitors improving patient outcomes in the frontline, ASCO and ESMO guidelines adapted to recommend these therapeutics are part of a frontline strategy. However, he emphasized that endocrine therapy alone remains an effective strategy for many patients and that, in these trials, many of the patients on the control arm were not able to cross over.

To that end, the phase 3 SONIA trial compared first-line CDK4/6 treatment against second-line treatment for patients with advanced ER+, HER2- breast cancer. The trial enrolled 1050 pre- and postmenopausal women with HR+, HER2-, advanced disease with no visceral crisis. Patients were not allowed to have received prior therapy for advanced disease; however, neoadjuvant and adjuvant therapy was allowed.

Of note, the Dutch government and Dutch insurance companies provided the funding for the trial. All but 1 Dutch hospital partook. Moreover, the Dutch Breast Cancer Patient Organization fully endorsed the trial, and patient advocates were members of the steering committee, actively contributing to the planning and execution of the trial design. According to Sonke, academically funded trials such as SONIA are beneficial because they allow investigators to evaluate the value of therapies. By reducing the number of patients receiving these high-toxicity treatments, they are made more accessible to patients who truly need them and otherwise may not have been able to afford them, he stated.

Participants were randomly assigned to receive either nonsteroidal AI plus a CDK6/6 inhibitor in the first-line, followed by fulvestrant, or to receive front-line nonsteroidal AI, followed by second-line fulvestrant plus a CDK4/6 inhibitor. The protocol did not specify third-line treatments, but patients who had PIK3CA mutations were able to receive alpelisib (Vjoice). The primary end point was PFS2, and the secondary end points included QOL, OS, and cost-effectiveness. Patients were stratified by which CDK4/6 inhibitor they received, visceral disease status, and prior (neo)adjuvant endocrine therapies. Patients underwent tumor assessments every 12 weeks and responses were evaluated using RECIST 1.1 criteria.

The study enrolled patients from November 23, 2017, to September 1, 2021. December 1, 2022, was the data cutoff date, equating to a median follow-up of 37.3 months.

The median patient age between the frontline and second-line treatment arms was 64 vs 63 years. In both arms, 49% of patients had a WHO performance status of 0. Most patients in both arms were postmenopausal (87% vs 86%)

In both arms, 35% of participants had newly diagnosed disease and 47% had experienced a disease-free interval of 24 months or greater. Eighteen percent and 19%, respectively, had experienced a disease-free interval of 24 months or less. Forty percent of patients in both arms had received prior treatment with chemotherapy and 49% of patients in the frontline and 48% of patients in the second-line arm had received prior endocrine therapy.

In both arms, 56% of patients had visceral disease, and 17% had bone-only disease.

Palbociclib was the most common CDK4/6 inhibitor and was administered to 91% of patients on either arm. Ribociclib (8%) and abemaciclib (1%) were less common.

Investigators leveraged the Functional Assessment of Cancer Therapy - Breast (FACT-B) and the European Quality of Life Five Dimension (EQ-5D-5L), 2 validated questionnaires, to assess QOL across 11 timepoints. FACT-B subscores and cost-effectiveness analyses are ongoing, but the primary analysis revealed that there was no difference in FACT-B total scores between the study arms (P = .4). The completions rate for FACT-B was 87% in both arms.

After presenting the data, Sonke asked: “If you were a patient, would you consider a treatment that offers no improvement in QOL and does not improve OS? As a doctor or nurse, would you recommend a treatment to your patients that nearly doubles the incidence of [AEs]? And if you were responsible for covering the costs of these treatments, whether as an individual or a health care insurance, would you consider it worth $200,000?”