Abstract
Purpose: Since diffusion-weighted MR (DW-MR) has shown promise in differentiating benign from malignant disease in several oncologic applications, we aimed to evaluate the potential role of this technique in differentiating benign from malignant lung lesions.
Material and methods: 50 patient staged with PET-CT and operated because of proven lung cancer or having a suspicious lung opacity were included. DW-MR was performed one day before surgery. DW-MR was evaluated first by visual inspection of all MR images by a chest radiologist and second by calculating the ADC values. Both PET/CT and DW-MR findings were correlated with pathology.
Results: Good correlation was found between DW-MR and pathology (κ=0.56, p<0.0001), whereas PET/CT performed worse (κ=0,20, p=0.1457). In total, 33 patients were diagnosed correctly with PET/CT, 7 incorrectly and 10 undetermined. DW-MR staged 45 patients correctly and 5 incorrectly. The 10 undetermined cases on PET/CT were correctly diagnosed on DW-MR. Pure ADC-average based diagnose showed an optimal threshold of 0.00152 mm2/s between benign and malignant lesions, with sensitivity and specificity of 91 and 57% respectively.
Conclusion: DWI-MR could become an appropriate diagnostic instrument for preoperative lung cancer patients in the near future because it has a high accuracy for differentiating benign from malignant lung lesions.
- © 2011 ERS