Introduction 1 2 3 4 5 9 10 12 11 l  O 18 l 13 21 21 24 15 16 20 25 28 29 131 188 Materials and methods Patients Twenty-four patients (9 females, 15 males) with a mean age of 49±14 years and histopathologically proven tenascin expression of malignant gliomas (5 anaplastic astrocytomas, 19 glioblastomas) were included in the study. All patients had undergone primary surgery on the tumour or re-operation of tumour recurrence with implantation of an Ommaya reservoir into the resection cavity. Primary surgery had been followed in 23 of the 24 patients by external radiation therapy and in six patients by additional systemic chemotherapy. 131 n 188 n 131 188 PET studies 68 18 14 max Statistical analysis 30 31 n In addition, SN and SP values are given for a visual analysis using nodular versus non-nodular FET uptake as the criterion to distinguish between patients with tumour recurrence and tumour-free patients. max Results n 5 1 max max max 1 max max max Table 1 Patient and tumour characteristics and therapeutic details of tumour-free patients Age (yrs) Sex WHO Surgery RIT cycles Cum. activity (MBq)/nuclide RIT No. of PET scans max max max Survival (mo.) post surgery Serial FET PET monitoring started after the end of RIT 36 f III 05/99 2 131 10/99–11/99 9 – 2.4 (02/01) 1.5 (01/05) 77 (l.c.) 39 m III 10/98 5 131 05/99–01/00 8 – 2.3 (09/01) 1.5 (10/05) 85 (l.c.) 37 f IV 03/00 1 131 11/00 12 – 2.1 (04/02) 1.5 (02/05) 66 (l.c.) 54 f IV 08/98 3 131 04/99–08/99 5 – peak missed 1.6 (05/05) 87 (l.c.) Serial FET PET monitoring during and after RIT 27 f IV 08/01 1 131 03/02 3 2.0 (02/02) 2.2 (04/02) 2.0 (06/02) 35 (l.c.) 22 f IV 01/04 3 131 10/04–02/05 6 1.4 (09/04) 2.1 (04/05) 1.8 (09/05) 21 (l.c.) 40 f IV 11/03 1 188 05/04 7 1.4 (04/04) 1.7 (03/05) 1.4 (09/05) 23 (l.c.) l.c. Fig. 1 131 max  2 max max 2 max max 3 Table 2 Patient and tumour characteristics and therapeutic details of patients with tumour recurrence or regrowth of residual tumour during follow-up Age (yrs) Sex WHO Surgery RIT cycles Cum. activity (MBq)/nuclide RIT No. of PET scans max max Survival (mo.) post surgery Patients without obvious tumour at baseline followed by tumour recurrence 40 m III 10/00 3 131 05/02–10/02 6 1.5 (12/01) 2.5 (03/03) 36 (†) 53 m IV 08/01 1 131 03/02 2 1.7 (01/02) 3.1 (04/02) 19 (†) 46 f IV 11/01 3 131 07/02–01/03 6 1.5 (07/02) 2.0 (09/03) 26 (†) 70 m IV 04/03 2 131 02/04–06/04 4 1.8 (02/04) 2.7 (09/04) 22 (†) 69 m IV 05/04 2 188 07/04–10/04 4 2.1 (07/04) 3.3 (03/05) 13 (l.c.) 55 m IV 07/02 3 131 11/03–03/04 5 2.0 (09/03) 2.1 (09/04) 34 (†) 61 m IV 12/02 1 131 10/03 2 1.6 (08/03) 2.8 (12/03) 17 (l.c.) 56 m IV 04/03 1 131 10/03 7 1.6 (10/03) 2.5 (02/05) 24 (l.c.) 37 f III 04/01 3 131 01/05–05/05 5 1.2 (12/04) 2.4 (07/05) 54 (l.c.) 39 f IV 03/04 2 131 07/04–10/04 6 1.6 (07/04) 2.8 (07/05) 19 (l.c.) Patients with small residual tumour at baseline followed by tumour progression 30 m III 02/01 1 131 07/02 4 2.6 (07/02) 4.0 (12/02) 25 (†) 61 m IV 04/00 3 131 09/01–12/01 3 3.3 (09/01) 3.6 (12/01) 25 (†) 60 m IV 02/01 2 131 12/01–01/02 3 2.5 (12/01) 2.6 (03/02) 28 (†) 59 m IV 05/03 1 188 03/04 3 2.1 (03/04) 2.9 (07/04) 24 (l.c.) 68 m IV 10/02 1 188 05/04 2 2.5 (05/04) 3.1 (06/04) 21 (l.c.) 50 m IV 10/03 1 131 01/04 2 2.1 (01/04) 2.2 (03/04) 12 (l.c.) 68 m IV 08/02 2 131 02/05–07/05 5 2.3 (11/04) 2.9 (09/05) 38 (l.c.)  l.c. Fig. 2 188 max  Fig. 3 131 a b max  4 n 4 max Fig. 4 n a n b  5 p Fig. 5 top bottom max  Visual analysis using a nodular versus a non-nodular uptake pattern as the criterion to distinguish between tumour recurrence and tumour-free status led to correct detection of recurrence in 32 of 34 scans (sensitivity 94%). In 80 of 85 scans performed in a tumour-free status, no additional nodular uptake pattern was present (specificity 94%). In five scans, uptake appeared somewhat focally accentuated owing to circumscribed folding of the cavity wall, and thus mimicked nodular uptake. Restricting visual analysis only to the one scan per patient presenting with the highest FET uptake resulted in correct detection of tumour recurrence in 16 of 17 patients (sensitivity 94%) and of a tumour-free status in five of seven (specificity 71%). In two of seven tumour-free patients, FET uptake that appeared nodular was misleading for reasons mentioned above. Discussion 5 9 10 12 32 131 29 32 33 34 33 34 20 max max max 35 37 max max One possible limitation of the present study was the lack of histological confirmation in 15/24 patients, and especially in five of the seven patients who were considered tumour free. Histopathological confirmation would have been ethically hard to justify in these patients, but its lack gives rise to the question of possible false negative findings. Nevertheless, the unsuspicious long-term clinical follow-up in the patients in question and the fact that the study population consisted only of patients with high-grade gliomas, and in particular glioblastomas (19/24 patients), strongly support the PET results. Another possible limitation is that in four of the 24 patients, follow-up with serial FET PET scans did not start until at least 10 months after the end of RIT, with the consequence that some earlier changes in FET uptake following RIT might have been missed. However, since in three of those four patients FET uptake was still increasing during the initial PET scans, followed by a peak and subsequent decrease, it is unlikely that crucial information was missed. Only in one patient in whom RIT had been completed as long as 33 months prior to the beginning of FET PET monitoring were the post-therapeutic phase of increasing FET uptake and its peak probably missed. 38 Conclusion This study indicates that FET PET is a sensitive tool for monitoring the effects of high local radiation doses given by intracavitary RIT. Homogeneous, slightly increasing FET uptake around the tumour cavity with a peak up to 18 months after RIT, followed by stable or decreasing FET uptake, is indicative of benign, therapy-related changes. These findings are independent of tumour recurrence and must not be misinterpreted as progressive disease. In contrast, focally increased FET uptake is an early and reliable indicator of tumour progression. Since both radiation necrosis and active tumour growth can present with clinical deterioration and are often impossible to distinguish by means of structural brain imaging, FET PET may be a powerful tool for planning further patient management. While radiation necrosis may be treated by steroids or, in extensive cases, by debulking surgery, recurrent tumour requires change of ineffective treatment or palliative care only.