A randomized controlled phase 2 study of the ketogenic diet for patients with newly diagnosed glioblastoma in combination with standard-of-care treatment - ABSTRACT
Median glioblastoma (GBM) survival is ~15 months and is little changed the past 50 years. Better therapies are
needed. Preclinical work from us and others has shown that a high-fat / low-carbohydrate ketogenic diet (KD)
can slow cancer growth in multiple models including GBM. Like many cancers, GBM depends on glycolysis for
the biosynthetic, bioenergetic, and signaling needs of oncogenesis. By limiting glycolytic flux and altering
metabolism, KD may have anti-cancer effects, including abrogating dysregulated insulin-PI3K signaling,
suppressing the pro-inflammatory tumor environment, and facilitating an anti-tumor immune response. Given the
dismal prognosis, many GBM patients adopt a quasi-KD despite the lack of high-quality data supporting its use.
To fill this gap, we conducted a pilot KD study for brain tumor patients, wherein we saw improvements in fatigue,
mood, cognitive function, and seizure control, as well as radiographic responses. We recently completed a phase
1 KD trial for patients with newly diagnosed GBM. The primary objectives of the phase 1 study - safety and
feasibility – were met. Importantly, we found an impressive median Overall Survival (OS) of 29.1 months, a 99%
improvement vs. historic control from the phase 3 trial that established the current standard of care (SOC). We
are now poised to conduct the first ever well-powered randomized controlled phase 2 study of KD + SOC vs.
SOC for patients with newly diagnosed GBM.
We propose a 170-person multicenter randomized controlled phase 2 study of KD plus SOC vs. SOC for patients
newly diagnosed with GBM. Our primary objective is to test efficacy between KD + SOC vs. SOC as assessed
by improved OS. We will assess quality of life (QoL), cognition, and physical function, as benefits in any of these
would be impactful to patients’ lives. We will take full advantage of tumor, blood, and stool samples that we will
be collecting to understand the effects of KD on tumor and host metabolism and immune response.
Desperate patients are not waiting. We thus have an urgent responsibility to prospectively and comprehensively
evaluate KD and understand its effects, good or bad. If negative, patients can be counseled accordingly, and
spend their precious time and energy on other endeavors, though QoL benefits (if seen) may still persuade some
to go on a KD. If positive, we have the potential to establish a new SOC that can be broadly applied without the
cost or toxicity of most new cancer therapies. We will seize the opportunity to gain a greater mechanistic
understanding of the effects of KD on GBM, which may shape future efforts to identify early markers/predictors
of response and optimize therapeutic combinations. These data can also help manage GBM patients who
choose to go on a KD either for OS or QoL benefits (if seen). Thus, regardless of outcome, this study will shape
clinical practice.
