Multiple b value diffusion-weighted MRI based on intravoxel incoherent motion model applied to lymphoma-pilot study

Abstract

Objective To evaluate treatment effect of nodal lymphoma by using multiple b value diffusion-weighted MRI based on intravoxel incoherent motion (IVIM) model. Methods From November 2012 to November 2013, 22 patients with pathology confirmed lymphoma in Guangdong General Hospital were chosen. Patients were examined on a 1.5 T MR scanner with plain MRI scan and multiple b value diffusion-weighted MRI scan before and after cycle two of chemotherapy. According to chemotherapy response which evaluated after cycle 2 of chemotherapy, lymphoma nodes were divided into four groups: complete response (CR), partial response (PR), stable disease (SD), progressive disease (PD). Kruskal-Wallis H test was used to compare prechemotherapy D value, f value among CR, PR, SD groups. Nemenyi test was used to compare prechemotherapy D value between the two of CR, PR, SD groups. One-way ANOVA was used to compare D* value among CR, PR, SD groups. Paired-sample t test was used to compare D, f, D* value between before and after chemotherapy in PR group. CR and PR group were ascribed to curative group, and SD group ascribed to poor response group. ROC curve was used to evaluate the predictive efficiency of parameters derived from IVIM. Results Twenty-two lymphoma patients were scanned before chemotherapy and 21 patients were scanned after cycle 2 of chemotherapy. There were 49 lesions in CR group, 17 lesions in PR group, 8 lesions in SD group and no lesions in PD group. Prechemotherapy D value of CR, PR, SD group were (0.63±0.26)×10-3, (0.57±0.10)×10-3, (0.42±0.04)×10-3 mm2/s, respectively. There was significant difference among the three groups (H=12.944, P=0.002). There was no statistically difference of prechemotherapy D value between CR and PR group (χ2=0.072, P=0.965). Prechemotherapy D value was lower in SD group than that in CR group (χ2=12.090, P=0.002) and PR group (χ2=10.684, P=0.005). There was no statistically difference of prechemotherapy f value among CR, PR, SD groups (χ2=2.312, P=0.315) or D* value (F=0.535, P=0.588). D value significantly increased after chemotherapy in PR group [(1.03±0.37)× 10-3 vs.(0.63±0.26)×10-3 mm2/s, t=-4.781, P=0.001]. f value significantly increased after chemotherapy in PR group [(9.39±4.52)% vs. (6.44±3.25)%, t=2.294, P=0.036]. D* value slightly increased after chemotherapy but with no statistical difference in PR group [(99.72±42.12)×10-3 vs. (90.37±45.33)× 10-3 mm2/s, t=-0.579, P=0.570]. When a D value of 0.48 × 10-3 mm2/s was used as the threshold value for predicting chemotherapy response, the best results were obtained with sensitivity of 100.00% and specificity of 75.76%. Conclusions Prechemotherapy D value can predict chemotherapy response and D value can monitor chemotherapy response in lymphoma. Key words: Magnetic resonance imaging; Lymphoma; Comparative effectiveness research