Parameters of MR perfusion-weighted imaging predict the response and prognosis to high-dose methotrexate-based chemotherapy in immunocompetent patients with primary central nervous system lymphoma

To investigate the effectiveness of dynamic susceptibility contrast-perfusion weighted imaging (DSCPWI) in predicting the progression-free survival (PFS) and chemotherapeutic responsiveness of primary central nervous system lymphoma (PCNSL) before high-dose methotrexate-based chemotherapy. DSCPWI was used to analyze 35 patients who had pathology-confirmed PCNSL. Relative cerebral blood volume (rCBV), relative cerebral blood flow (rCBF), relative mean transit time (rMTT) and relative time to peak (rTTP) were measured on parameter maps. The 5th, 50th and 95th percentile values of every parameter were recorded for enhanced tumors and compared with the parameters of the normally contralateral hemisphere. The ratio of each PWI parameter (rrCBV, rrCBF, rrMTT, rrTTP) was obtained. The influence of parameters on responsiveness and PFS was investigated by univariate Kaplan-Meier analysis and logistic regression and Cox regression for multivariate analysis with a stepwise method. Differences in PWI parameters between the higher and lower vascular endothelial growth factor (VEGF) groups were assessed by the Mann-Whitney U test. Eighteen patients achieved a complete response (CR_i) after four initial cycles of chemotherapy_. Patients with lower age (p = 0.011), higher VEGF (p < 0.001), higher Karnofsky Performance Status (KPS) (p < 0.001), higher rrCBV_95% (p < 0.001), higher rrCBV_50% (p = 0.016), higher rrCBF_95% (p < 0.001), higher rrCBF_50% (p = 0.002) showed better PFS; there was difference on age(p = 0.044), KPS (p < 0.001), VEGF (p < 0.001), rrCBV_95% (p = 0.018), rrCBF_95% (p = 0.018), rrCBF_50% (p = 0.007) between CR_i and nonCR_i. Multivariate analysis demonstrated that rrCBF_95% (p = 0.037, 95% confidence interval: 1.065-7.206) was significantly associated with PFS. rCBV and rCBF may be used to assess the responsiveness and prognosis of PCNSL, and rCBF_95% may be a better predictor.