OS6.5 ERGO2: A prospective randomized trial of a 9-day schedule of calorically restricted ketogenic diet and fasting or standard diet in addition to re-irradiation for malignant glioma

BACKGROUND

Ketogenic diet (KD) and fasting have anticancer effects in tumor models, possibly due to a differential stress response with sensitization of tumor cells and protection of normal tissue. We therefore set up ERGO2 (NCT01754350), the first randomized clinical trial of calorically-restricted KD and intermittent fasting (KD-IF) in addition to re-irradiation for recurrent malignant gliomas.

MATERIAL AND METHODS

Patients were randomized 1:1 to re-irradiation combined with either calorically unrestricted diet (standard diet, SD) or KD-IF. The KD-IF schedule included 3 days of KD (21–23 kcal/kg/d), followed by 3 days of fasting and again 3 days of KD. The primary endpoint was progression-free survival (PFS) rate at 6 months (PFS6). Secondary endpoints were PFS, local control, overall survival (OS), frequency of epileptic seizures, rate of ketosis and quality of life.

RESULTS

50 patients were included. Four patients quit the trial before treatment and three patients stopped KD-IF prematurely. Of the 20 patients who completed KD-IF, 17 patients developed ketosis at day 6, and glucose levels declined significantly. KD-IF was well-tolerated with a modest weight loss of -2.1±1.8 kg. No severe adverse events attributable to the diet occurred. There was no difference in PFS6 between the two groups (KD-IF: 20%, SD: 16%). Similarly, no difference in PFS, local PFS6 and OS were observable. Explorative analysis revealed that among patients of the KD-IF group, those who achieved ketosis of at least 1.5 mmol/l had significantly longer PFS compared to those with lesser or no ketosis.

CONCLUSION

KD-IF is feasible and effective in inducing ketosis in heavily pretreated patients with recurrent glioblastoma. However, the short schedule reported here failed to increase the efficacy of re-irradiation.

Very late relapses in glioblastoma long-term survivors

e13017

Background: Long-term survival of patients with histologically confirmed glioblastoma is a rare event with figures in the range of 2–3% for 5-year survival. Prognosis and further clinical course of these patients beyond 5 years after diagnosis are in essence unknown with only anecdotal reports of patients surviving for 10 years or more. Methods: We here report on the extended follow-up (mean, 139.4 months) of a cohort of 10 glioblastoma long-term survivors. Retrospective central histology, clinical monitoring, and regular magnetic resonance imaging were done. Results: A total of 5 patients of the 9 patients alive at a previous analysis have died during the extended follow-up. Four of these patients died from recurrent tumor. One patient died from leukoencephalopathy-associated complications without evidence of tumor progression. Details concerning the occurrence of relapses and their treatment are displayed in table 1. Notably, very late relapses occurred in three patients after 118, 124, and 126 months of progression-free survival. Thus, from the original cohort of 10 patients only 3 remain free from recurrence after the extended follow up of more than 10 years. Conclusions: Very late relapses pose a serious threat for glioblastoma long-term survivors and call for continuous vigilance. The maintenance of a tight control schedule even in patients surviving for more than 10 years should be considered.

[Table: see text]

No significant financial relationships to disclose.