Hong Kong PolyU Develops MRI Biomarker to Predict Breast Cancer Drug Response
The Tumour Heterogeneity Index assesses patient response to AKT inhibitors noninvasively before surgery
By Letizia Giugliano
2 min readHuman-Edited, AI-Assisted
Researchers at The Hong Kong Polytechnic University have introduced a medical innovation that predicts how breast cancer patients will respond to targeted therapy before treatment starts. Using advanced radiomics on routine magnetic resonance scans, the team developed an imaging biomarker called the Tumour Heterogeneity Index. The method identifies patients likely to respond to AKT inhibitors, sparing others from ineffective treatments and toxic side effects.
Reading tumor complexity through magnetic resonance data
Two breast cancer tumors of equal size often behave differently when treated with targeted drugs. To identify these differences before surgery, a team led by Cai Jing, head of the Department of Health Technology and Informatics at The Hong Kong Polytechnic University (PolyU), turned to dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI). The imaging technique is already used in clinical practice for diagnosis and surgical planning, as reported by MedicalXpress.
Working with Dr. Teng Xinzhi, research assistant professor at PolyU, the team processed the imaging data with advanced radiomics algorithms. The resulting metric, the Tumour Heterogeneity Index (THI), detects structural variations inside the tumor that remain invisible to normal human inspection. This diagnostic innovation allows doctors to examine the immune cells and the vascular environment surrounding the lesion without performing an invasive procedure.
“THI is more than an imaging analysis tool. It quantifies the complexity of a tumor's internal structure, transforming assessment that has traditionally relied on experience into an objective, data-driven approach.”
— Cai Jing, chair professor of medical physics and intelligent oncology at PolyU
Predicting response to AKT inhibition in difficult subtypes
The research focused on patients with high-risk, early-stage breast cancer treated with the investigational AKT inhibitor MK2206, which was added to neoadjuvant chemotherapy before surgery. The AKT/mTOR/PI3K signaling pathway promotes tumor survival and growth. While MK2206 previously increased response rates in the I-SPY2 clinical trial, its overall performance varied across aggressive subtypes such as hormone receptor-negative, HER2-negative and triple-negative breast cancer, according to MedicalXpress.
In the study published in the journal Breast Cancer Research, patients with a high THI score achieved an estimated pathological complete response rate of 56.1%. That figure was significantly higher than the response rate recorded among patients with low index values. The biomarker also showed strong predictive power in triple-negative cases, an aggressive subtype that typically responds poorly to standard treatments.
A noninvasive alternative to tissue biopsies and genetic tests
Conventional assessment relies on immunohistochemistry tests and tissue biopsies. These procedures only sample a small portion of the tumor, creating the risk of sampling errors caused by differences in biopsy location. In contrast, the THI index analyzes the entire tumor mass from standard imaging scans, as documented by MedicalXpress.
The PolyU researchers verified the repeatability and reproducibility of their imaging method across different scanning machines and healthcare facilities. Cai noted that the noninvasive invention could reduce reliance on expensive genetic sequencing in routine clinical practice, helping medical teams determine patient risk and drug resistance earlier in therapy.
Read also on The Patent: Microfluidic chip screens leukemia drug therapies in children
