The ZEST phase III clinical trial evaluating the PARP inhibitor niraparib (Zejula) for preventing breast cancer recurrence in patients with measurable residual disease (MRD), defined by circulating tumor DNA (ctDNA) presence, was terminated early due to low randomization rates, revealed at the San Antonio Breast Cancer Symposium (SABCS) 2024. The study's design, which allowed enrollment of low-risk patients, resulted in insufficient ctDNA detection, impeding the trial's ability to assess niraparib's efficacy.
Trial Design and Patient Population
The ZEST trial (NCT04915755) aimed to assess niraparib's potential in delaying or preventing breast cancer recurrence in patients with MRD, indicated by ctDNA presence post-treatment. Patients with stage I-III triple-negative breast cancer (TNBC) or BRCA-mutated HER2-negative breast cancer were eligible. ctDNA testing was initiated post-definitive treatment, with HR-positive patients allowed concurrent endocrine therapy. Upon ctDNA detection and exclusion of metastatic disease, patients were randomized to niraparib or placebo.
Key Findings
Of 2,746 prescreened patients, 1,901 underwent ctDNA testing, with 147 (8%) testing positive (135 with TNBC, 12 with BRCA-mutated HR-positive BC). Only 40 patients were randomized (18 to niraparib, 22 to placebo) due to detected ctDNA, radiologic recurrence, or unmet inclusion/exclusion criteria. While underpowered, the niraparib arm showed a numerically longer recurrence-free interval.
ctDNA detection rates were highest within three months post-treatment (6.7%), decreasing thereafter. In TNBC, 60% of ctDNA-positive cases were detected within six months, aligning with the early recurrence pattern typical of TNBC.
Insights and Future Directions
Nicholas Turner, MD, PhD, from the Royal Marsden Hospital, suggested initiating ctDNA testing earlier in the disease trajectory of TNBC due to the high early recurrence rate at ctDNA positivity. Ian Krop, MD, PhD, of Yale School of Medicine, cautioned against clinical use of ctDNA testing before formal validation, highlighting potential harms like increased toxicity and patient anxiety.
An exploratory analysis showed a potentially longer recurrence-free interval in patients with low ctDNA levels at baseline. Patients with TNBC and high ctDNA levels treated with placebo (n = 13) had a median recurrence of 3.2 months (95% CI, 0.3-9.3) compared to 5.7 months (95% CI, 0.6-14.1) with niraparib (n = 7; HR, 0.91; 95% CI, 0.34-2.44). Those with low ctDNA levels at baseline had a median recurrence of 7.4 months (95% CI, 2.6-NE) with placebo (n = 9) versus 15.9 months (95% CI, 8.2-NE) with niraparib (n = 11; HR, 0.60; 95% CI, 0.20-1.81).
Implications for Clinical Trials
The ZEST trial's challenges highlight the need for refined patient selection in ctDNA-guided trials, focusing on higher-risk individuals to enhance ctDNA detection rates. Turner emphasized the importance of designing studies to enroll a relatively high-risk patient population to ensure sufficient conversion of patients into the study. Starting ctDNA testing earlier, potentially during neoadjuvant therapy, may also improve trial feasibility and outcomes, especially in aggressive subtypes like TNBC.
