NEWTON, Mass., Dec. 03, 2025 (GLOBE NEWSWIRE) — 4D Path, a company dedicated to personalizing cancer care through a novel, physics-inspired approach to predicting tumor response to therapy, today announced that the company’s Q-Plasia OncoReader (QPOR™) platform has successfully predicted treatment response in patients with early-stage triple-negative breast cancer (TNBC) using digitized baseline biopsy images. Analysis conducted within the Translational Breast Cancer Research Consortium (TBCRC) 030 clinical trial represents the first demonstration of a statistical physics and tumor biology-based computational biomarker predicting post-treatment outcomes from pre-treatment tissue alone, without any clinical or molecular inputs.
Results, published in the Journal of the National Cancer Institute (JNCI), establish a new framework for baseline digital image–based prediction of chemotherapy response and lay the groundwork for future chemo-immunotherapy applications for early-stage TNBC. The multi-institutional collaboration, led by investigators from Dana-Farber Cancer Institute, included researchers from Harvard Medical School, Brigham and Women’s Hospital, 4D Path Inc., and other leading academic centers.
The TBCRC 030 trial previously evaluated Myriad Genetics’ HRD biomarker as a predictor of response to either cisplatin or paclitaxel in TNBC. In this follow-up analysis, investigators applied 4D Path’s QPOR™ algorithm to pre-treatment core biopsies to evaluate tumor-infiltrating lymphocytes (TILs) and computational immune–cell-cycle biomarkers (CmbI, also known as the complex immune response index CIRi), in addition to visual examination (VE) by a pathologist. All assessments were blinded and independently performed on digitized hematoxylin and eosin (H&E) slides.
Key findings include:
- 9X Increased Response Odds of Image-Only Prediction in the Paclitaxel Arm: CmbI predicted RCB 0/1 in paclitaxel-treated patients (OR 9.17, p=0.009; AUC 0.74) using H&E only; CA-TILs and IHI did not.
- Immune Biology Matters: Both visual and computational assessments correlated with response, underscoring the prognostic and predictive role of immune infiltration in TNBC.
- BRCA-Proficient Cohort, Overall Association: In TBCRC 030 (BRCA1/2-proficient), visually assessed-TILs and CmbI associated with response overall; by arm, the signal localized to paclitaxel, not cisplatin.
QPOR’s CmbI is designed to capture the baseline complex immune response as a predictor of post-operative residual cancer burden (RCB) after neoadjuvant chemotherapy. Computed directly from H&E, it integrates—the tumor’s proliferative state, the spatial heterogeneity of immune infiltrates in conjunction with a cell cycle G1/S deregulation signature. This mechanism aligns with the predictive capacity observed with paclitaxel in TBCRC 030 and, importantly, the predictive capacity was also observed in TBCRC 031 (germline BRCA–mutated, HER2-negative breast cancer), where CmbI predicted response to neoadjuvant doxorubicin–cyclophosphamide (AC) and cisplatin monotherapy (ASCO 2024). Together, these observations support potential generalizability of the assay.
“These findings show the complex immune response—a key ingredient of tumor-response biology—can be quantitatively captured directly from routine biopsy slides using first-principles computation as a predictive biomarker of neoadjuvant chemotherapy response, and that G1/S cell-cycle deregulation signatures can be leveraged for a comprehensive quantification of the complex immune response,” said Dr. Satabhisa Mukhopadhyay, Co-founder and Chief Scientific Officer of 4D Path. “It opens the possibility of baseline treatment predictive modeling for treatment selection, trial design, and testing future therapeutic strategies.”
“This is a prospective evaluation of a computational biomarker derived from standard histology that, much like expert visual assessment of TILs, predicts post-treatment response,” said Dr. Erica Mayer, principal investigator of TBCRC 030, and director of breast cancer clinical research at Dana-Farber Cancer Institute. “Developing reproducible predictive signatures from clinical trial pre-treatment biopsies could ultimately improve selection of therapies to optimize patient care.”
The computational analysis produced a quantitative, fully deterministic, and reproducible biomarker predictive of treatment response—by design, intrinsically free of observer variability and machine-learning bias. The study underscores the potential of computational pathology and oncology to complement traditional visual methods, expand biomarker access globally, and facilitate standardized clinical trial workflows.
By relying solely on H&E images, the QPOR™ platform offers a scalable, cost-effective, and interpretable alternative to conventional AI-driven models that require large, annotated datasets. Additionally, this approach provides a foundation for treatment predictive modeling in chemotherapy, potentially applicable to chemoimmunotherapy, and supports the use of physics-based algorithms as complementary tools for patient selection, treatment de-escalation, or early therapy switching decisions.
For additional information, access the full article, “Prognostic value of visually and computationally-assessed tumor-infiltrating lymphocytes in early-stage triple-negative breast cancer (TBCRC-030),” in the Journal of the National Cancer Institute.
About 4D Path
4D Path has created a groundbreaking platform, the patented Q-Plasia OncoReader (QPOR™), designed to directly measure and quantify cell cycle deregulation and tumor immune microenvironment dynamics to predict a patient’s response to therapy. This unprecedented view into tumor dynamics creates new pathways from clinical use to research and development applications— bringing more effective, personalized therapies to patients. 4D Path is focused on a bold vision where every person impacted by cancer is diagnosed quickly and accurately and receives the best, most personalized treatment plan to beat cancer and live a longer, healthier life.
For additional information, please visit www.4dpath.com or follow 4D Path on LinkedIn.