Choi J W, Lee D, Hyun S H, Han M, Kim J-H, Lee S J
Department of Radiology, Ajou University School of Medicine, Suwon, Republic of Korea.
Department of Pathology, Ajou University School of Medicine, Suwon, Republic of Korea.
Clin Radiol. 2017 Jun;72(6):482-489. doi: 10.1016/j.crad.2017.01.019. Epub 2017 Mar 9.
To evaluate the association between the tumour-stroma ratio and intratumoural heterogeneity measured using 2-[F]-fluoro-2-deoxy-d-glucose (FDG) positron-emission tomography (PET) and magnetic resonance imaging (MRI), and further investigate the prognostic significance of imaging biomarkers in head and neck squamous cell carcinoma (HNSCC).
Textural-based imaging parameters of the primary tumour were extracted in 44 patients. In addition, the difference between the minimum and maximum apparent diffusion coefficient (ADC) values (ADCdiff) was calculated on MRI. The relationships between the tumour-stroma ratio and imaging parameters were evaluated. The associations between imaging parameters and recurrence-free survival (RFS) were assessed using Cox proportional hazard regression models.
Coarseness (r=-0.382) on PET and ADCdiff (r=0.534) on MRI were significantly correlated with the proportion of stroma. The best imaging biomarkers for the 2-year RFS prediction were coarseness (AUC=0.741) and ADCdiff (AUC=0.779). Multivariate analysis showed that coarseness (hazard ratio=10.549, 95% confidence interval=2.544-43.748, p=0.001) was an independent prognostic factor for RFS.
Heterogeneity imaging parameters are significantly associated with the tumour-stroma ratio. These imaging biomarkers may help to facilitate the risk stratification for tumour recurrence in HNSCC.
评估使用2-[F]-氟-2-脱氧-D-葡萄糖(FDG)正电子发射断层扫描(PET)和磁共振成像(MRI)测量的肿瘤-基质比与肿瘤内异质性之间的关联,并进一步研究成像生物标志物在头颈部鳞状细胞癌(HNSCC)中的预后意义。
提取了44例患者原发肿瘤基于纹理的成像参数。此外,在MRI上计算最小和最大表观扩散系数(ADC)值之间的差异(ADCdiff)。评估肿瘤-基质比与成像参数之间的关系。使用Cox比例风险回归模型评估成像参数与无复发生存期(RFS)之间的关联。
PET上的粗糙度(r = -0.382)和MRI上的ADCdiff(r = 0.534)与基质比例显著相关。预测2年RFS的最佳成像生物标志物是粗糙度(AUC = 0.741)和ADCdiff(AUC = 0.779)。多变量分析显示,粗糙度(风险比 = 10.549,95%置信区间 = 2.544 - 43.748,p = 0.001)是RFS的独立预后因素。
异质性成像参数与肿瘤-基质比显著相关。这些成像生物标志物可能有助于促进HNSCC肿瘤复发的风险分层。