Multidelay Pseudocontinuous Arterial Spin Labelling and Conventional Single-delay Pseudocontinuous Arterial Spin Labelling In Nasopharyngeal Carcinoma: Correlation with Dynamic Contrast-enhanced Magnetic Resonance Imaging.

The purpose of this prospective study was to investigate the potential correlations among three perfusion methods-multidelay pseudocontinuous arterial spin labelling (MD-PCASL), conventional single-delay pseudocontinuous arterial spin labelling (SD-PCASL), and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI)-in patients with locoregionally advanced nasopharyngeal carcinoma (ANPC). Forty-seven patients with locoregional ANPC were prospectively recruited. Two observers independently calculated the MD-PCASL parameters, including the ATT-corrected tumour blood flow (c-TBF), arterial transit time (ATT), and arterial tumour blood volume (aTBV); the SD-PCASL parameters, including the noncorrected tumour blood flow (nc-TBF); and the DCE-MRI quantitative parameters, including Ktrans, Kep, Ve, and Vp. The intra- and interobserver agreement was evaluated via the intraclass correlation coefficient (ICC) and BlandAltman analysis. Relationships between the MD-PCASL, SD-PCASL and DCE-MRI parameters were assessed via Spearman's rank correlation. The intra- and interobserver reproducibility were excellent, with ICCs ranging from 0.81 to 0.997 and a narrow width of 95% limits of agreement. A good positive correlation was observed between MD-PCASL-related perfusion parameters (c-TBF, aTBV) and DCE-MRI parameters (Ktrans, Kep) (Spearman's rank correlation coefficients: 0.61 to 0.71, p < 0.001). The SD-PCASL (nc-TBF) values were fairly correlated with Ktrans (p<0.001, r = 0.53) and Kep (p<0.001, r=0.51). In addition, there was a good correlation between the MD-PCASL-related perfusion parameters (c-TBF) and the SD-PCASL parameters (nc-TBF) (Spearman's rank correlation coefficients: 0.70, p < 0.001). No correlation was detected between MD-PCASL-related perfusion parameters (c-TBF and aTBV), SD-PCASL (nc-TBF) and DCE-MRI parameters (Ve and Vp) (Spearman's rank correlation coefficients: 0.07 to 0.145, p > 0.05). MD-PCASL multiparametric perfusion, SD-PCASL single-parameter perfusion and quantitative DCE-MRI parameters were correlated and reproducible in the assessment of ANPC. Although SD-PCASL demonstrated a fair correlation with DCE-MRI, the correlation was inferior to that of MD-PCASL with multiple parameters. MD-PCASL, as a noninvasive perfusion imaging tool, may become an alternative for assessing ANPC perfusion levels in the future.