A prospective diffusion-weighted MRI study for differentiation of squamous cell carcinoma and adenocarcinoma in lung cancer

Abstract

Abstract Background Non-small cell lung cancer (NSCLC) management depends on accurate histologic classification, yet tissue sampling may be limited by procedural risk and intratumoral heterogeneity. Diffusion-weighted MRI (DWI)-derived apparent diffusion coefficient (ADC) offers a non-invasive biomarker potentially reflecting tumor cellularity and enabling subtype differentiation. This study aimed to assess the diagnostic performance of ADC values for differentiating lung adenocarcinoma from squamous cell carcinoma (SCC) in CT-detected lung masses suspicious for primary malignancy. Methods In this prospective study, 32 patients with CT-detected parenchymal lung mass lesions underwent thoracic MRI on a 1.5-T system, including DWI ( b = 0 and 800 s/mm 2 ) and automated ADC map generation. A blinded radiologist placed ROIs within the solid tumor component, avoiding areas of necrosis and adjacent atelectasis. The lesion-level ADC (× 10 −3 mm 2 /s) was calculated as the mean of three repeated measurements. Histopathology from CT-guided, ultrasound-guided, or bronchoscopic biopsy served as the reference standard. Results The cohort comprised 18 adenocarcinomas (56.3%) and 14 SCCs (43.8%); mean age was 61 ± 10 years, and 26 patients were male (81.3%). Adenocarcinoma demonstrated higher mean ADC than SCC (1.311 ± 0.044 versus 1.217 ± 0.065 × 10 −3 mm 2 /s; P < 0.001). ADC showed good discrimination for adenocarcinoma versus SCC with an AUC of 0.869 (95% CI, 0.709–1.000). At a cutoff of > 1.25 × 10 −3 mm 2 /s, sensitivity was 94.44% and specificity 85.71%, with PPV 89.5% and NPV 92.3%. Conclusions DWI-derived ADC provides a good non-invasive discrimination between lung adenocarcinoma and SCC and represents a non-invasive adjunct to histopathology for lung cancer subtype assessment.