Apparent diffusion coefficient; is it an effective index for differentiating between types of lung cancer brain metastases?

Abstract

Abstract Background: The apparent diffusion coefficient (ADC) values of tumors are highly correlated with tumor cellularity and used as a neuroimaging marker with the potential to differentiate between major histological subtypes. Here, we will attempt to determine the sensitivity and specificity of the ADC to distinguish between types of metastatic brain metastases from lung cancer. Methods: One hundred and fifty-six patients (136 [%87, 18] male, 20 [%12.82] female) admitted to our hospital with the diagnosis of primary lung cancer were included in the study. In addition to conventional magnetic resonance imaging sequences, Diffusion-weighted imaging (DWI) and ADC images were evaluated qualitatively and quantitatively. Results: We found hyperintensity in most of the metastatic lesions on a qualitatively evaluated DWI sequence. In quantitative assessment according to ADC value comparisons between the different histologic subtype metastatic lung carcinoma groups, small-cell carcinoma (SCLC) had the highest value (1.93 × 10‒3mm2/s ± 0.95) and nonsmall-cell-combined (NSCCLC) type was the least (0.55 × 10‒3 mm2/s ± 0.46). When we tried to distinguish lung cancer-induced brain metastases into two main groups as SCC and NSC by considering the mean ADC ratios we obtained 0.65 ± 0.14 for SCC and 1.51± 0.30 for NSC. On the other hand, there was no significant statistical difference between the specific histological subtype groups with comparison of ADC values (P > 0.05). Conclusion: Quantitatively quantified DWI-ADC can distinguish metastatic lesions from the normal brain parenchyma. Although we realized whether differentiation of SCLC and non-SCLC in brain metastases can be achieved with DWI, we could not define any correlation between DWI/ADC values and primary histology of the metastatic foci. We believe that more accurate results can be achieved with advanced studies with more patients included and common sequence features.