Saturday, 4 July 2020

Reirradiation + ketogenic diet + intermittent fasting, randomized trial

ERGO2: A prospective randomized trial of calorie restricted ketogenic diet and fasting in addition to re-irradiation for malignant glioma    (full PDF from sci-hub)   (link to journal abstract, full PDF requires a payment)

Half the patients were randomized to re-irradiation plus standard diet (SD), the other half were randomized to re-irradiation plus ketogenic diet plus intermittent fasting (KD-IF).

  • PFS6 was not significantly different between the two groups (KD-IF: 20%, SD: 16%). 
  • Similarly, no difference in PFS, local PFS6 and OS was observable.
  • Explorative analysis revealed that patients of the KD-IF group who had a glucose of less than the median (83.5 mg/dl) on day 6 had a significantly longer PFS and OS compared to those above the median


  1. I haven't even read the full study yet, I just wanted to get the links out there as I know many readers of this blog are interested in therapeutic dietary alterations.

  2. I’m confused. The top line results say that there was no difference between the groups but for a subset there was a difference in outcome?? How big is the difference and how big is the subset do we know?

    1. Hi Stu, if you'd like to download the full study, you can click on the first link in my post (sci-hub). I've copied and pasted figure 3 from the study into the post.

      For the group that got ketogenic diet and intermittent fasting, those who had a glucose below the median (83.5 mg/dL), they had a better PFS and OS (see figure 3c and 3d). When looking at all the patients together, the gap narrowed and there was only a trend towards better PFS, but it wasn't statistically significant (figure 3e and 3f).

      So why would lower glucose be beneficial for those on ketogenic diet and intermittent fasting but not as beneficial for those on standard diet? This says to me there were other factors involved besides just blood glucose levels. No patients in the standard diet group had ketone levels above 0.5 mmol/L, so perhaps it's a combination of elevated ketones and low glucose that was beneficial and not simply lower glucose alone?

      Figure 2 of the study also shows that a handful of patients in the ketogenic diet group still had higher glucose levels on day 6 than the median level in the standard diet group. So it's important to look at the subgroups that actually achieved low glucose, as they did.

      Looking at figure 2b, it looks like the subgroup analysis that showed improved PFS and OS was based on only 10 patients in the KD-IF group who had glucose below the median. It would have been nice to have had higher numbers of patients. But nevertheless this is probably the highest degree of evidence so far to recommend ketogenic diet during radiation, and provides a target level for glucose and ketone bodies (apart from glucose levels there was a trend for better survival and PFS in those patients with ketones above 0.5 mmol/L. With so few patients it is more difficult to achieve statistical significance, whereas a bigger trial with more patients might have shown statistically significant improvements in outcome with higher ketone levels).

    2. And as for the size of the difference, in the ketogenic diet group, median PFS was 111 days for those with lower than median glucose levels, and 42 days for those with higher than median glucose.

      Median survival was 348 days for those with lower than median glucose levels, and 142 days for those with higher than median glucose levels.

      These are the number corresponding to figures 3c (PFS) and 3d (OS).