Unraveling predictive indicators for therapeutic response in HER2-positive breast cancer

Abstract

Breast cancer is a highly heterogeneous disease, exhibiting significant diversity both in its biological subtypes and clinical manifestations. Biologically directed therapies have revolutionized the treatment landscape for breast cancer in the past three decades, offering more targeted and personalized approaches and have significantly changed the treatment paradigm. This shift in treatment conception has indeed resulted in improved outcomes for patients, driven by extensive research efforts aimed at personalized biomarker exploration.

The aim of my thesis was to identify biomarkers for optimizing therapy selection in HER2- positive breast cancer.

In paper I, we concentrated on current evidence regarding the dynamics of tumor-infiltrating lymphocytes (TILs) during neoadjuvant treatment and examined the fluctuating patterns of TILs, correlating them with treatment prediction and survival outcomes. We observed a consistent decrease in TILs levels after neoadjuvant therapy (NAT) across all breast cancer subtypes, with a numerically larger decrease noted in HER2-positive breast cancer. In TNBC patients, increased TILs during treatment were associated with better disease-free survival (DFS) or recurrence-free survival (RFS), as indicated by pooled hazard ratios from univariate analyses of four eligible studies. However, due to insufficient studies, this analysis was limited to TNBC. Additionally, we identified eight studies reporting on-treatment TILs counts, which was uniformly increased compared to baseline levels. Increased on-treatment TILs compared to bassline were positively associated with pathological complete response (pCR) status in seven out of total eight studies, but no pooled analysis was done due to data heterogeneity. These findings suggest that dynamic monitoring of TILs may serve as a flexible and economical biomarker for treatment de-escalation and future trial design, particularly in HER2-positive and TNBC patients.

In paper II, a comprehensive analysis was conducted to assess the predictive and prognostic significance of baseline and serial levels of serum thymidine kinase (sTK1) in patients with HER2-positive early breast cancer enrolled in the PREDIX HER2 trial. At baseline, no association was found between serum thymidine kinase 1 (sTK1) levels and clinicopathological characteristics such as age, tumor grade, and Ki-67 status. We observed a dramatic increase in TK1 activity in all patients after two cycles of treatment, although neither baseline sTK1 levels nor sTK1 levels at subsequent on-treatment time points were associated with pathological complete response (pCR) status. Furthermore, there was no significant effect of baseline or cycle 2 sTK1 activity on event-free survival (EFS). For patients with residual disease (non-pCR), a higher sTK1 activity at the end of treatment visit appeared to be linked with longer survival time, though the association did not reach statistical significance. Our study provides evidence of sTK1 activity dynamics in a prospective phase II trial, although no prognostic association was identified.

In paper III, we profiled intrinsic molecular subtypes in longitudinally collected tissue material obtained from patients enrolled in the PREDIX HER2 trial and explored their association with treatment response and long-term outcomes. The PAM50 intrinsic subtypes were determined using an SSP-based method. The results revealed that the majority of patients at baseline were classified as HER2-enriched (HER2-E) subtype (55%), as expected. Approximately 40% of patients were categorized as Luminal A or Luminal B types, while the remaining were classified as basal-like (BL) subtype. The baseline HER2-E subtype showed a significant association with better pCR and EFS. Under treatment, intrinsic subtypes exhibited temporal plasticity, with the majority of patients experiencing a subtype switch. Specifically, 71 out of 93 HER2-E patients transitioned to a non-HER2-E subtype from baseline to on-treatment. The present study highlights the potential utility of PAM50 intrinsic molecular subtypes for prognostication in HER2-positive breast cancer, and a prospective validation clinical trial is ongoing. Further exploration of the clinical implications associated with subtype switching during treatment is needed.

In summary, our research provides a comprehensive biomarker exploration aimed at predicting treatment response and adding prognostic value in patients with early HER2- positive breast cancer. Through these investigations, we aim to enhance our understanding of breast cancer heterogeneity and improve treatment escalation and de-escalation strategies for achieving better patient outcomes.