Radiomics on baseline CT scans may be able to predict early response to first-line platinum-based chemotherapy in patients with advanced non-small cell lung cancer (NSCLC), researchers reported.
An examination of data from CT imaging of radiomic features from within and around the lung tumor demonstrated that the information can be helpful for predicting overall survival (OS), time to progression (TTP), and response to chemotherapy, reported Muhammadhadi Khorrami, PhD, of Case Western Reserve University in Cleveland, and colleagues in Radiology: Artificial Intelligence.
Currently, only around one in four patients respond to platinum-doublet chemotherapy, and there is no good way to know who those patients will be. The study is the first to show that “computer-extracted patterns of heterogeneity, or diversity, from outside the tumor were predictive of response to chemotherapy,” said co-author Monica Khunger, MD, of the Cleveland Clinic, in a press statement.
She added that identifying patients who are more likely to respond to alternative therapies, such as radiation or immunotherapy, could impact patient outcomes.
The researchers noted that there has been interest in the use of radiomic interrogation to inspect malignant tumor masses, “the rationale being that the tumor microenvironment and habitat might harbor valuable disease-specific prognostic clues.”
The study involved retrospective analysis of data on 125 patients with NSCLC who had been treated with pemetrexed-based platinum doublet chemotherapy at the Cleveland Clinic. Patients were divided randomly so that there were an equal number of responders and nonresponders in the training set (n=53) and the validation set (n=72).
“A machine learning classifier trained with radiomic texture features extracted from intra- and peritumoral regions of non-contrast-enhanced CT images were used to predict response to chemotherapy,” the researchers wrote. “The radiomic risk-score signature was generated by using least absolute shrinkage and selection operator with the Cox regression model: association of the radiomic signature with TTP and OS.”
They reported that several radiomic features combined with a quadratic discriminant analysis classifier yielded a mean maximum area under the receiver operating characteristic curve (AUC) of 0.82 ± 0.09 (standard deviation) in the training set and a corresponding AUC of 0.77 in the independent testing set.
Also, the radiomics signature was also significantly associated with TTP (hazard ratio 2.8, 95% CI 1.95-4.00, P<0.0001) and OS (HR 2.35, 95% CI 1.41-3.94, P=0.0011).
Decision curve analysis showed that the radiomics risk-score signature had a higher overall net benefit in predicting high-risk patients for receiving treatment than the clinical-pathologic measurements across several threshold probability values.
Khorrami said the finding that prognostic accuracy increased when patterns of heterogeneity outside of the tumor, as well as inside the tumor, were unexpected.
“Despite the large number of studies in the CT-radiomics space, the immediate surrounding tumor area, or the peritumoral region, has remained relatively unexplored. Our results showed clear evidence of the role of peritumoral texture patterns in predicting response and time to progression after chemotherapy,” Khorrami said in a press statement.
The researchers noted that previous studies have identified tumor heterogeneity as a predictor of survival in NSCLC patients.
“Although the reason is yet unclear, tumor heterogeneity, which in turn breeds tumor evolution, might be reflective of clonal dominance or genomic heterogeneity within the tumors. Intratumoral genetic heterogeneity also has been shown to be a leading contributor to therapeutic failure,” they stated.
Study limitations included the relatively small sample size and the lack of actionable mutation and PD-1/PD-L1 interaction scores for patients.
“This raises a critical question in that beyond chemotherapy, there is no other viable clinical option to treat these patients,” they wrote, pointing out that the KEYNOTE-189 trial demonstrated the benefit of chemotherapy and immunotherapy over chemotherapy alone in NSCLC.
“In light of these findings, there is an opportunity to refine subgroups of patients with NSCLC who may benefit from such an approach,” the author stated.
The study was funded by the U.S. Department of Defense, the National Cancer Institute, the National Institute of Diabetes and Digestive and Kidney Diseases, the National Center for Research Resources, the Ohio Third Frontier Technology Validation Fund, the Wallace H. Coulter Foundation Program in the Department of Biomedical Engineering, and the Clinical and Translational Science Award Program at Case Western Reserve University.
Khorrami and Khunger declared no relevant relationships with industry. Co-authors disclosed multiple relevant relationships with industry.