2026 Theses Doctoral

A theranostic harmonic motion imaging system for breast cancer neoadjuvant systemic treatment response prediction and enhancement

Liu, Yangpei

Breast cancer remains a significant global health challenge, standing as the most frequently diagnosed cancer and the second leading cause of cancer-related mortality among women in the United States. The disease is characterized by substantial biological heterogeneity, classified into intrinsic molecular subtypes, i.e., Luminal A, Luminal B, HER2-enriched, Basal-like, and Claudin-low, which dictate prognosis and therapeutic sensitivity. For high-risk phenotypes such as HER2-positive and triple-negative breast cancer (TNBC), neoadjuvant chemotherapy (NACT) has become the standard of care aimed at down-staging tumors prior to surgery. However, pathologic complete response (pCR), a strong predictor of superior disease-free survival, is achieved in only 10-40% of patients. Because conventional imaging modalities often suffer from low specificity or an inability to detect small-volume residual disease, there is a critical clinical need for non-invasive, cost-effective techniques capable of early treatment monitoring and response prediction to facilitate timely therapeutic adjustments.

The rationale for using elasticity imaging in this context is grounded in soft tissue biomechanics, where pathological progression is linked to alterations in the extracellular matrix, specifically the deregulation of collagen metabolism. This structural reorganization results in malignant tumors exhibiting a Young’s modulus that is significantly higher, often 3 to 13 times greater, than that of normal fibroglandular tissue. While static elastography relies on operator-dependent external compression, dynamic ultrasound elasticity imaging utilizes acoustic radiation force to induce internal vibrations, allowing for the estimation of tissue mechanical properties based on shear wave propagation or on-axis displacement. Although shear wave-based methods provide quantitative stiffness data, they can be susceptible to wave distortion and attenuation in heterogeneous, deep, or obese tissues.

To address these limitations, this dissertation investigates Harmonic Motion Imaging (HMI), an ultrasound-based technique that uses a focused ultrasound transducer to generate an amplitude-modulated (AM) acoustic radiation force. By inducing and simultaneously tracking oscillatory tissue displacement at a specific modulation frequency, HMI filters out bulk motion artifacts and provides high-contrast mechanical assessment of deep tumors. The primary objective of this work is to optimize the clinical HMI system and validate its predictive value for NACT response in clinical breast cancer patients. Furthermore, this research explores the theranostic capabilities of the HMI system under Specific Aim III, developing a method for chemotherapy enhancement via contrast-enhanced-power-Doppler (CEPD)-guided sonoporation. This involves quantifying acute tumor responses and longitudinally monitoring safety and efficacy, specifically assessing whether sonoporation-enhanced regimens can mitigate distant metastasis in breast cancer murine models compared to chemotherapy alone.

While conventional HMI has shown promise in characterizing solid tumors, optimizing its performance remains critical for clinical translation, particularly in achieving high-quality displace-ment estimation and consistent stiffness characterization across heterogeneous tissues. To address the limitations of plane-wave imaging, such as poor contrast and lateral resolution due to the lack of transmit focusing, this dissertation investigated the effects of various tracking beam sequences on HMI image quality. Through extensive in silico simulations and phantom experiments, receive (Rx) parallel tracking with a focused transmit beam was identified as the optimal sequence, demonstrating superior contrast and contrast-to-noise ratio (CNR) compared to plane-wave and coherent plane-wave compounding methods, while maintaining a high frame rate suitable for capturing dynamic tissue motion.

Furthermore, the dependence of inclusion characterization on AM frequency and inclusion dimension was systematically evaluated to ensure accurate stiffness estimation. Phantom studies revealed that the HMI-derived displacement ratio (DR), a surrogate for relative stiffness, is influenced by both the size of the inclusion and the applied AM frequency. Specifically, a lower frequency of 200 Hz provided consistent DR measurements for larger inclusions (> 8.1 mm), whereas a higher frequency of 400 Hz, or a multi-AM sequence combining 200 and 400 Hz, was necessary to accurately characterize smaller inclusions (< 6 mm). This optimization led to a strong linear correlation (𝑅2 = 0.9043) between the inverse DR and the Young’s modulus ratio, validating the method’s capability to quantitatively assess relative tissue stiffness regardless of tumor dimensions. Building on these optimizations, the clinical feasibility of HMI was demonstrated through in vivo imaging of human breast tumors.

The optimized tracking sequence and AM frequency were applied to longitudinally monitor tumor response to neoadjuvant chemotherapy (NACT). Preliminary results from a cohort of fifteen patients (n = 15) indicated that tumor normalized HMI dis-placement could serve as an early biomarker for treatment response, with significant changes in tumor stiffness observed as early as three weeks into treatment (AUC: 0.89, sensitivity: 100%, specificity: 88.9%). Notably, a decrease in tumor stiffness, indicating tumor softening, correlated with pathologic complete response (pCR), while an increase, i.e., stiffening, was indicative of non-response or partial response. Additionally, the development of a novel 3-D HMI system us-ing a row-column-addressed (RCA) array enabled rapid volumetric mapping of tumor mechanical properties, offering a more comprehensive assessment of tumor heterogeneity and residual disease burden compared to traditional 2-D methods.

Finally, CEPD was evaluated as a real-time modality to prospectively guide sonoporation in a 4T1 triple-negative breast cancer murine model. By selectively targeting functionally active tumor vasculature, this approach ensured the effective co-localization of chemotherapy, microbubbles, and ultrasound energy to overcome physical transport barriers. Compared to anatomical B-mode guidance, CEPD-guided sonoporation achieved superior primary tumor control, significantly ex-tended median survival, and markedly reduced metastatic burden in the lungs and liver. Concurrently, longitudinal HMI validated therapy-induced tumor softening as a robust, non-invasive biomarker of treatment response. Systemic profiling further demonstrated favorable immunomodulatory effects, characterized by the expansion of effector T-cell populations and the attenuation of the chemotherapy-induced surge in granulocytic myeloid-derived suppressor cells.