Identification of epidermal growth factor receptor mutations in pulmonary adenocarcinoma using dual-energy spectral computed tomography.

OBJECTIVES

To explore the role of dual-energy spectral computed tomography (DESCT) quantitative characteristics for the identification of epidermal growth factor receptor (EGFR) mutation status in a cohort of East Asian patients with pulmonary adenocarcinoma.

MATERIALS AND METHODS

Patients with lung adenocarcinoma who underwent both DESCT chest examination and EGFR test were retrospectively selected from our institution's database. The DESCT visual morphological features and quantitative parameters, including the CT number at 70 keV, normalized iodine concentration (NIC), normalized water concentration, and slopes of the spectral attenuation curves (slope λ HU [Hounsfield unit]), were evaluated or calculated. The patients were divided into two groups: the EGFR mutation group and EGFR wild-type group. Statistical analyses were performed to identify the DESCT quantitative parameters for diagnosis of EGFR mutation status.

RESULTS

EGFR mutations were detected in 66 (55.0%) of the 120 enrolled patients. The univariate analysis revealed that sex, smoking history, CT texture, NIC, and slope λ HU were significantly associated with EGFR mutation status (p = 0.037, 0.001, 0.047, 0.010, and 0.018, respectively). The multivariate logistic analysis revealed that smoking history (odds ratio [OR] = 3.23, p = 0.005) and NIC (OR = 58.026, p = 0.049) were the two significant predictive factors associated with EGFR mutations. Based on this analysis, the smoking history and NIC were combined to determine the predictive value for EGFR mutations with the area under the curve of 0.702.

CONCLUSIONS

NIC may be a potential quantitative DESCT parameter for predicting EGFR mutations in patients with pulmonary adenocarcinoma.

KEY POINTS

• DESCT can provide multiple quantitative image parameters compared to conventional CT. • Identification of the radio-genomic relation between DESCT and EGFR status can help to define molecular subcategories of lung adenocarcinoma, which is valuable for personalized clinical targeted therapy. • NIC may be a potential DESCT quantitative parameter for predicting EGFR mutations in pulmonary adenocarcinoma.