Tafadzwa Chaunzwa, MD, MHS

Developing an Imaging Biomarker for Lung Cancer
 

Tafadzwa Chaunzwa, MD, MHS, and his team at Dana-Farber Brigham Cancer Center are using artificial intelligence to develop imaging-based biomarkers that will help improve treatment selection for patients with advanced non-small cell lung cancer (NSCLC). Immunotherapy boosts a person’s immune system to find and destroy cancer cells and has emerged as an important treatment for lung cancer, particularly when the disease has spread locally or to other parts of the body. However, only a minority of patients (10-20%) benefit from immunotherapy long term, so identifying these individuals as well as patients that are unlikely to respond to immunotherapy and need intensified treatment is essential to providing personalized care. Existing biomarkers for advanced NSCLC lack specificity and sensitivity or require invasive tissue sampling, and new tools for selecting the most effective treatments are needed.

In prior research, Dr. Chaunzwa and mentors Hugo Aerts, PhD, and Raymond Mak, MD, have shown that deep learning based radiomics, which utilizes artificial intelligence to analyze quantitative features of medical images captured during routine care, can be used to predict tumor histology and outcomes for patients with NSCLC, including survival and likelihood of the tumor to respond to chemoradiation. “Given that immunotherapy shows great promise for treating lung cancer but is only effective for a relatively small number of patients, we need better tools to help us understand who will benefit. We are building on our earlier work and aiming to develop novel cancer imaging biomarkers that will more accurately predict response to immunotherapy alone or in combination with other treatments, including stereotactic ablative radiotherapy (SABR) or chemotherapy,” says Dr. Chaunzwa.

With a James D. Cox Research Award from ROI, Dr. Chaunzwa’s team is:

• Analyzing computed tomography (CT) scans of the chest routinely acquired before and during treatment for NSCLC to identify radiomic signatures that help predict response to immunotherapy.
• Combining the image analysis with clinical and demographic risk factors, such as tumor genetics, age, gender and smoking status, to develop an algorithm that can more accurately predict which patients are likely to respond to immunotherapy and which patients could benefit from additional treatments.  

They aim to develop an advanced and inexpensive AI software that can increase personalization of care for patients with advanced NSCLC. A non-invasive tool to optimize lung cancer treatment selection could also help bring emerging technologies like immunotherapy and SABR to low-resource settings by supporting cost-effective resource utilization. Dr. Chaunzwa’s research promises to improve outcomes for patients with lung cancer by helping to ensure they receive the most effective treatments.

The James D. Cox Research Award is a special recognition for a resident pursuing a career in research that is generously supported by Ritsuko Komaki-Cox, MD, FASTRO.

Presentations
 

• AI-Derived CT Body Composition in Advanced Non-Small Cell Lung Cancer: A Multicohort Study was presented at the 2023 ASTRO Annual Meeting. 
• Immunotherapy Prognostication with Thoracic CT Contrastive Self-Supervised Learning in Patients with Non-Small Cell Lung Cancer was presented at the 2023 World Conference on Lung Cancer. 
• Contrastive Self-Supervised Learning of Lung Tumor Imaging Predicts Immunotherapy Response was presented at ESTRO 2023.