A phase II study of palbociclib plus letrozole plus trastuzumab as neoadjuvant treatment for clinical stages II and III ER+ HER2+ breast cancer (PALTAN)

Abstract

Patients with ER+/HER2+ breast cancer (BC) are less likely to achieve pathological complete response (pCR) after chemotherapy with dual HER2 blockade than ER-/HER2+ BC. Endocrine therapy plus trastuzumab is effective in advanced ER+/HER2+ BC. Inhibition of CDK4/6 and HER2 results in synergistic cell proliferation reduction. We combined palbociclib, letrozole, and trastuzumab (PLT) as a chemotherapy-sparing regimen. We evaluated neoadjuvant PLT in early ER+/HER2+ BC. Primary endpoint was pCR after 16 weeks. Research biopsies were performed for whole exome and RNA sequencing, PAM50 subtyping, and Ki67 assessment for complete cell cycle arrest (CCCA: Ki67 ≤ 2.7%). After 26 patients, accrual stopped due to futility. pCR (residual cancer burden-RCB 0) was 7.7%, RCB 0/I was 38.5%. Grade (G) 3/4 treatment-emergent adverse events occurred in 19. Among these, G3/4 neutropenia was 50%, hypertension 26.9%, and leucopenia 7.7%. Analysis indicated CCCA in 85% at C1 day 15 (C1D15), compared to 27% at surgery after palbociclib was discontinued. Baseline PAM50 subtyping identified 31.2% HER2-E, 43.8% Luminal B, and 25% Luminal A. 161 genes were differentially expressed comparing C1D15 to baseline. MKI67, TK1, CCNB1, AURKB, and PLK1 were among the genes downregulated, consistent with CCCA at C1D15. Molecular Signatures Database gene-sets analyses demonstrated downregulated processes involved in proliferation, ER and mTORC1 signaling, and DNA damage repair at C1D15, consistent with the study drug's mechanisms of action. Neoadjuvant PLT showed a pCR of 7.7% and an RCB 0/I rate of 38.5%. RNA sequencing and Ki67 data indicated potent anti-proliferative effects of study treatments. ClinicalTrials.gov- NCT02907918.

Fig. 1. Patient-reported outcomes collected at baseline, after cycle 1, and after neoadjuvant therapy using NCI PRO-CTCAE.

A Severity, B interference, and C frequency.

Fig. 2. Boxplots comparing Ki67 scores in samples at baseline, C1D15, and surgery.

The center line represents the median (middle quartile), the bounds of the boxes represent the upper and lower quartile, and the whiskers represent the minimum and maximum values.

Fig. 3

Image of tumor sample with Ki67 scoring of ≤2.7% and >2.7% by immunohistochemistry (IHC).

Fig. 4. Tumor somatic mutations by WES occurring in at least three patient samples according to pathological status.

The vertical axis represents the patient ID, tumor molecular subtype, and genes with somatic mutations. Each column represents a unique patient’s sample.

Fig. 5. Differentially expressed genes at C1D15 versus Baseline.

The volcano plot color codes the genes that are statistically significant and have a log2 fold change of +2 or −2.

Fig. 6. Significantly up- and down-regulated biological processes at C1D15 versus Baseline.

Generally applicable Gene-set Enrichment analysis of Molecular Signatures Database Hallmark gene-sets comparing C1D15 and baseline samples.

Fig. 7. Clinical trial schema.

Diagram of a clinical trial showing chemotherapy regimen and biospecimen collection time points.