Abstract Introduction: Failure to achieve a pathologic complete response (pCR) to neoadjuvant treatment informs risk-adapted adjuvant therapies, which reduce recurrence risk and improve outcomes. However, pCR-based assessments are limited given most patients with residual disease do not relapse and tissue visualization cannot define molecular residual disease (MRD). While liquid biopsies can define MRD, circulating tumor DNA is undetectable in most patients with pathologic residual disease, even with the most sensitive technologies. Methods: Motivated by the critical need to define which patients with residual disease are at risk of relapse and characterize their post-neoadjuvant genomic landscapes, we developed a novel, personalized, ultra-sensitive, tissue-based strategy to define MRD. We characterized the resection tissues of 29 diverse breast cancers using a joint tumor-informed and tumor-naive approach, capturing both persistent and emergent lesions. To interrogate genomic alterations under neoadjuvant selection at the level of genes and cytobands, we derived a statistical framework ranking each lesion’s position in a clonal hierarchy based on relative pre- and post-neoadjuvant abundance and associated alterations with selection and survival. We focused on recurrent ciliary alterations and evaluated their enrichment and survival associations, which we validated in external breast cancer datasets. Results: We identified recurrent alterations in genes and pathways involved in cell migration, adhesion, and tumor invasion under selection. Across the cohort, we also identified alterations in ciliary genes with diverse roles, including intraflagellar transport (IFT88), ciliary vesicle recruitment (CEP89), microtubule stabilization (CSPP1), and structural components of the ciliary base (EHD1), distal appendage/centriole (CEP350), and dynein arm (DNAAF4, DNAH14, DYNC2H1). We found enrichment of a broader panel of ciliary alterations in the post-neoadjuvant tissues of relapsing patients and that these lesions were under selection. In public breast cancer datasets, we validated ciliary alterations as under neoadjuvant selection and associated with relapse ( Table 1 Table 1. Association of ciliary alterations with post-neoadjuvant abundance and relapse. Study cohort: Association of ciliary lesions with relapse Ciliary alteration (N # alterations) Non-ciliary alteration (N) P-value (Chi-square) Relapse 70 647 0.00001 No relapse 42 1045 Study cohort: Association of ciliary lesions with selection Ciliary alteration (N # alterations) Non-ciliary alteration (N) P-value (Chi-square) Selected 58 68 0.00001 Non-selected 53 924 External cohort: Association of ciliary lesions with relapse Ciliary alteration (N # patients with alteration) Non-ciliary alteration (N) P-value (Chi-square) Inferior relapse-free survival 125 48 0.00001 Non-inferior relapse-free survival 631 755 External cohort: Association of ciliary lesions with selection and relapse Selected (N # patients with alteration) Non-selected (N) P-value (Chi-square) Inferior relapse-free survival 36 84 0.00001 Non-inferior relapse-free survival 46 540 ). Conclusions: Loss of the cilium may not only be an early breast cancer event, but also potentially a novel mechanism of taxane chemoresistance. We hypothesize that cilia, cellular antennae that are the primordial remnants of motile flagella, promote physical dispersion and metastatic potential of invading tumor cells through aberrant cellular signaling, polarity, and mechanosensation. Because hedgehog and PI3K pathway signaling rely on cilia, adjuvant targeting of these pathways in tumors bearing ciliary lesions may represent a novel approach to eradicate micrometastatic disease in the post-neoadjuvant window. We suggest the need for functional study of ciliary alterations in breast tumors. Citation Format: J. D. Ransohoff, M. Carleton, S. Miron Barroso, N. Phillips, J. Demeter, J. Goldstein, A. Maltos, M. Khodadoust, J. Ford, A. Kurian, G. Bean, M. Diehn, P. Jackson, M. Telli, D. Kurtz, A. Alizadeh. Recurrent ciliary alterations under neoadjuvant selection are associated with inferior breast cancer outcomes abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS2-11-20.
Ransohoff et al. (Tue,) studied this question.