Researchers have used artificial intelligence (AI) to predict the sensitivity of tumours to three systemic cancer therapies, an advance that may lead to new ways of finding better cures for the malignant disease.
The study, published in the journal Clinical Cancer Research, assessed computed tomography (CT) X-ray scans from patients with advanced non-small cell lung cancer (NSCLC).
According to the researchers, including Laurent Dercle from the Columbia University Irving Medical Center in the US, the interpretation of CT scans from cancer patients treated with systemic therapies is inherently subjective.
"The purpose of this study was to train cutting-edge AI technologies to predict patients' responses to treatment, allowing radiologists to deliver more accurate and reproducible predictions of treatment efficacy at an early stage of the disease," said Dercle.
To determine if patients with NSCLC are responding to systemic therapy, radiologists currently quantify changes in tumour size, and the appearance of new tumor lesions, Dercle noted.
However, the researchers said, this type of evaluation can be limited, especially in patients treated with immunotherapy, who can display atypical patterns of response and progression.
More From This Section
"Newer systemic therapies prompt the need for alternative metrics for response assessment, which can shape therapeutic decision-making," Dercle said.
In the study, Dercle and his colleagues utilised data from multiple clinical trials that evaluated systemic treatment in patients with NSCLC.
They said these patients were treated with one of three agents -- the immunotherapeutic agent nivolumab (Opdivo), the chemotherapeutic agent docetaxel (Taxotere), or the targeted therapeutic gefitinib (Iressa).
The researchers also analysed standard-of-care CT images from 92 patients receiving nivolumab in two trials.
They also did the same for 50 patients receiving docetaxel in one trial, and 46 patients receiving gefitinib in one trial.
The scientists then used the CT images taken at baseline and on first-treatment assessment to develop a computer model.
They classified the tumours as either treatment-sensitive or treatment-insensitive based on the reference standard of each trial.
Among all three cohorts, patients were randomized into training or validation groups.
The researchers then used a machine learning model to predict the treatment sensitivity in the training cohort.
Across all cohorts, the scientists said, a total of eight radiologic features were used to build the three prediction models.
These included changes in tumour volume, heterogeneity, shape, and margin.
They said both the nivolumab and gefitinib models used four radiologic features, and the docetaxel model used one.
"We observed that similar radiomics features predicted three different drug responses in patients with NSCLC," Dercle said.
"Further, we found that the same four features that identified EGFR treatment sensitivity for patients with metastatic colorectal cancer could be utilized to predict treatment sensitivity for patients with metastatic NSCLC," she said.
Dercle said that radiomic signatures offer the potential to enhance clinical decision-making.
"With AI, cancer imaging can move from an inherently subjective tool to a quantitative and objective asset for precision medicine approaches," he said.
"Because AI can continuously learn from real-world data, using AI on larger patient datasets will help us to identify new patterns to build more accurate prediction models," Dercle added.