Background: Identifying reliable biomarkers to assess treatment response and improve disease monitoring remains an unmet need in breast cancer (BC). Thymidine kinase, an enzyme for DNA synthesis upregulated during cell proliferation, has emerged as a promising biomarker in advanced HR+/HER2- BC. Higher baseline serum thymidine kinase activity (TKa) correlates with worse prognosis. Serial TKa assessment during the first cycle of a CDK4/6i with endocrine therapy (ET) revealed three response patterns – suppressed, rebound, and unsuppressed. The latter two are associated with inferior outcomes but the relationship between TKa response and tumor genomics is unexplored. This case series evaluates TKa patterns in relation to baseline circulating tumor DNA (ctDNA) profiles in patients with HR+/HER2- metastatic BC (MBC) treated with a CDK4/6i and ET.

Methods: We evaluated a group of patients from two cancer centers with HR+/HER2- MBC initiating treatment with a CDK4/6i (ribociclib, abemaciclib, palbociclib) and ET (an aromatase inhibitor/AI or the selective estrogen receptor degrader/SERD, fulvestrant). Triple therapy with the PI3K inhibitor, inavolisib, was permitted for PIK3CA-mutated tumors. A subset of patients are participants in the TK IMPACT study (NCT04968964, Washington University). Baseline ctDNA was assessed within one month of treatment start using Guardant360 or Tempus xF. TKa was measured via the DiviTum® TKa assay (Biovica, USA) at baseline prior to initiation of CDK4/6i, Cycle 1 Day 15 (C1D15), and C2D1. A TKa level of 50 DuA (DiviTum unit of Activity) is the lower limit of quantification of the assay. Suppressed pattern was defined as TKa levels that declined to <50 DuA at C1D15 of treatment and remained <50 at C2D1; rebound was defined as TKa that declined to <50 DuA at C1D15 but increased to >50 at C2D1; unsuppressed remained >50 DuA at C1D15 and C2D1. Progression-free survival (PFS) was assessed.

Results: Twenty-two patients were included; 32% with de novo MBC, 68% with a metastatic recurrence. Most patients were on first-line therapy (95%). Ribociclib was the most common CDK4/6i (86%) while 14% received palbociclib. Most patients were treated with an AI (77%); a smaller subset received fulvestrant (23%). Baseline ctDNA revealed several potential ET and CDK4/6i resistance mediators including AKT/PTEN mutations (18%), TP53 (14%), RAS/RAF pathway mutations (14%), ERBB2 (14%), ESR1 (9%), RB1 loss (5%), FGFR1 (5%) and CCNE1 amplification (5%). One patient had undetectable baseline TKa and no identifiable mutations in ctDNA. The following TKa patterns were observed: 7 suppressed, 11 rebound, and 3 unsuppressed. Notable cases included a patient with an unsuppressed TKa pattern while on ribociclib/AI who experienced progression within 3 months. Baseline ctDNA revealed an ERBB2 mutation and ATM, BRCA2, and CHEK2 copy number loss. Another patient with RB1 loss and multiple co-occurring mutations (ESR1, PIK3CA, KRASG12V, ATM and BRCA2 loss) had a reduction in TKa from 242 at baseline to 68 at C2D1 after starting palbociclib, inavolisib, and fulvestrant.

Conclusions: TKa is a novel biomarker with important clinical implications. Understanding the genomic drivers underlying treatment response is an area of active investigation. This hypothesis-generating study compares baseline ctDNA profiles with TKa response patterns in patients with HR+/HER2- MBC starting treatment with CDK4/6i-based therapy. Longitudinal TKa monitoring alongside imaging will continue, including evaluation at progression for acquired genomic changes. Updated findings in these and additional patients will be presented at the meeting.

VIEW THE PUBLICATION