Introduction 1 5 3 6 7 8 9 10 11 8 12 14 The aim of this study was to determine whether DSC-MRI obtained through routine pre-therapy MR diagnostic examinations may be used to distinguish molecular subtypes of oligodendroglial tumours, and to investigate the influence of tumour blood volume on outcome following procarbazine, lomustine and vincristine (PCV) chemotherapy compared to molecular genetics. Methods Patient selection 7 15 18 19 DCS-MRI 1 Fig. 1 8 9 a b c d e f  arrows a  c  e b  d  f  20 2 tumour normal brain 21 Molecular genetics 16 18 p53 PCV chemotherapy, response and clinical outcome 22 2 2 2 18 23 n n Information regarding current management, follow-up and outcome was collected prospectively. Progression-free survival (PFS) and overall survival (OS) were calculated from the start of PCV to progression, last seen, or death as appropriate. Statistical analysis 24  t P Results The study included 30 patients with primary tumours without previous therapy and 7 with recurrent or persistent disease following radiotherapy. Median time between previous radiotherapy and MR imaging was 6.5 years (range 1.1–14.2 years). Median age was 44 years (range 28–71 years). Interobserver and intraobserver variability 1 2 2 2 Fig. 2 a  b c  Histopathology, genotype and rCBV 3 3 3 Fig. 3 a P b P c P  t  1 4 p53 1 Table 1 rCBV, molecular genetics and clinical characteristics in the series and primary tumours   All tumours Primary tumours n Number of tumours with high rCBV  P a n Number of tumours with high rCBV  P a Loss of 1p36 and 19q13 Yes 19 17 (89%) 0.000* 16 14 (88%) 0.000 No 18 4 (22%)  14 2 (14%)  Loss of 17p13 Yes 11 2 (18%) 0.003* 8 0 (0%) 0.001 No 26 19 (73%)  22 16 (73%)  Loss of 10q 22–26 ±10p11–15 Yes 3 3 (100%) 0.238 1 1 (100%) 1.0 no 33 17 (52%) 29  15 (52%)  p53 Yes 11 2 (18%) 0.004* 8 0 (0%) 0.001 No 25 18 (72%) 21  15 (71%)  Patient age (years) <50 25 13 (52%) 0.491 20 9 (45%) 0.260 >50 12 8 (67%)  10 7 (70%)  Histology subtype Oligodendroglioma 13 10 (77%) 0.091 11 8 (73%) 0.142 Oligoastrocytoma 24 11 (46%)  19 8 (42%)  WHO histology grade II 23 11 (48%) 0.191 9 9 (43%) 0.118 III 14 10 (71%)  21 7 (78%)  Contrast enhancement Absent 8 2 (25%) 0.046 8 2 (25%) 0.092 Present 28 19 (68%)  21 14 (68%)  Tumour location Temporal 14 7 (50%) 0.733 13 6 (46%) 0.713 Nontemporal 23 14 (61%)  17 10 (59%)  Tumour status Primary 30 16 (53%) 0.674    Recurrent 7 5 (71%)     Response to PCV  All tumours Responders 21 14 (67%) 0.142 17 11 (65%) 0.097 Nonresponders 11 4 (36%)  8 2 (25%)   Tumours with loss of 1p/19q Responders 16 14 (88%)  13 11 (85%)  Nonresponders 0   0    Tumours with intact 1p/19q Responders 5 0 (0%) 0.245 4 0 (0%) 0.515 Nonresponders 11 4 (36%)  8 2 (25%)  a Fig. 4  arrow  CR  PR  MR  PD  SD  P  R  rCBV and outcome following PCV chemotherapy 4 23 4 1 4 2  P 2 5 5  P  P  P Table 2 n  nc  nr   n Progression-free survival (months) Overall survival (months) Median (95%CI) P a P b Median (95%CI) P a P b All tumours Loss of 1p36 and 19q13 Yes 16 >56 (nr) 0.005  >56 (nr) 0.020  No 17 8 (0–21)   >49 (nr)   Histopathology subtype Oligodendroglioma 11 >56 (nr) 0.055 0.187 >56 (nr) 0.113 0.288 Oligoastrocytoma 21 8 (9–41)   50 (9–91)   Histopathology grade II 19 >56 (nr) 0.004 0.002 >56 (nr) 0.007 0.012 III 14 4 (8–26)   21 (8–34)   rCBV Low 14 >46 (nr) 0.766 0.003 >49 (nr) 0.585 0.002 High 19 46 (nc)   >56 (nr)   Primary tumours Loss of 1p36 and 19q13 Yes 14 >56 (nr) 0.037  >56 (nr) 0.056  No 12 >39 (nr)   49   Histopathology subtype Oligodendroglioma 10 >56 (nr) 0.055 0.137 All censored 0.092 0.209 Oligoastrocytoma 16 46 (6–86)   50 (6–101)   Histopathology grade II 17 >56 (nr) 0.047 0.009 >56 (nr) 0.048 0.041 III 9 6 (13–37)   50 (nc)   rCBV Low 12 >46 (nr) 0.681 0.007 >49 (nr) 0.611 0.003 High 14 >56 (nr)   >56 (nr)   a b Fig. 5 a b c d a c b d  n  n  n  n  n  n  n  n  Discussion 15 18 8 20 25 26 21 25 27 28 29 26 30 32 19 28 33 33 8 34 35 12 14 13 1 23 8 9 5 34 36 37 12 28 11 Conclusion In conclusion, we have shown that high rCBV is more likely in oligodendroglial tumours with the −1p/−19q genotype and may be a useful aid to diagnosis; however, rCBV does not predict chemosensitivity. Lack of correlation with histopathology grade and the data suggesting that the prognostic significance of rCBV may differ in patients with or without the −1p/−19q genotype, indicates that different interpretation of rCBV data may be necessary for oligodendroglial compared with astrocytic gliomas. Further study is essential to fully evaluate the role of DSC-MRI in the non-invasive diagnosis and prediction of prognosis in gliomas with an oligodendroglial phenotype.