Radiomics via PET imaging may identify cell mutations in NSCLC patients

Liz Meszaros, MDLinx | April 11, 2017

Through an advanced image analysis technique known as radiomics via positron emission tomography (PET), researchers may be able to identify genetic cell mutations in patients with non-small cell lung cancer (NSCLC), according to results from a proof-of-concept study published in the April issue of The Journal of Nuclear Medicine.


Tumors harbor different somatic mutations

From left to right are patients with EGFR mutation, KRAS mutation, and EGFR– and KRAS– tumors, respectively. Stage I and III tumors are shown in the top and bottom rows, respectively. Arrows indicate the locations of the lung tumors. Credit: Stephen S.F. Yip, PhD, and Hugo Aerts, PhD, Dana-Farber Cancer Institute, Brigham and Women’s Hospital, and Harvard Medical School, Boston Massachusetts; John Kim, MD, University of Michigan Health System, Ann Arbor, MI.

“To our knowledge, this is the first study to investigate the relationship between somatic mutations and the metabolic phenotypes, which may provide valuable information for developing non-invasive imaging biomarkers for determining mutation status,” explains Stephen Yip, PhD, Harvard Medical School, Boston, Massachusetts. “Identifying mutation status in NSCLC patients is an important component of selecting an optimal treatment plan for the patient. The current standard of care uses molecular testing based on biopsies of tumor tissue or surgical resection to identify mutation status. Molecular testing, however, can be limited by invasive procedures and long processing times. In addition, tissue samples are not always readily available.”

For the study, Dr. Yip and colleagues included 348 patients with NSCLC, who underwent diagnostic F18-fluorodoxyblucose PET (F18-FDG PET) scans and were tested for genetic mutations.

In all, 13% had an epidermal growth factor receptor (EGFR) mutation and 28% had a Kristen rat sarcoma viral (KRAS) mutation. After assessing 21 imaging features, including 19 independent radiomic features to quantify phenotypic traits, and 2 conventional features, researchers found that eight radiomic features and both conventional features were significantly associated with EGFR mutation status (FDRWilcoxon=0.01–0.10). For predicting EGFR mutation status, one radiomic feature (normalized inverse difference moment) outperformed all other features (EGFR+ vs EGFR-negative, AUC = 0.67, FDRNoether = 0.0032), as well as for differentiating between KRAS-positive and EGFR+ (AUC = 0.65, FDRNoether = 0.05). None of the features was associated with, or predictive of, KRAS mutation status (KRAS-positive vs KRAS-negative, AUC = 0.50–0.54).

“This study may thus help develop an imaging biomarker that can non-invasively and accurately identify EGFR mutation status using PET imaging to complement, but not to replace, molecular testing,” Dr. Yip concluded.

This study was supported by the National Institutes of Health (Award Number U01CA190234 and U24CA194354) and research seed funding grant from the American Association of Physicists in Medicine. The authors also thank the PROFILE team for help with somatic mutation testing.