The proposal for ketogenic metabolic therapy as a baseline for glioblastoma treatment is compelling. However, a few points warrant further discussion:
The article emphasizes the utility of the Glucose-Ketone Index (GKI) for monitoring therapeutic efficacy, yet the variability of GKI responses across patient populations with differing metabolic baselines (e.g., insulin resistance or cancer cachexia) remains underexplored. Could the authors elaborate on how GKI targets are adapted for such heterogeneous groups, particularly in relation to individual caloric and macronutrient needs?
While the study highlights KMT’s synergy with chemoradiotherapy, the potential risks of delaying standard-of-care treatments for interim KMT evaluation might raise ethical concerns. How do the authors propose addressing patient safety and long-term outcomes during such delays, especially for aggressive tumor subtypes?
The reliance on self-reported dietary compliance or intermittent biological markers in earlier studies is noted as a limitation. Could the authors suggest specific technologies (e.g., wearable glucose-ketone monitors) or protocols that future studies should adopt to ensure real-time, high-fidelity data collection?
Thank you for your interest in our work.
The glucose-ketone index (GKI) is less variable than its comprising parameters (glycemia/ketonemia), but is still expected to change dynamically under different patient populations. As we have addressed in the manuscript and in the peer-review reports (which are also available along the article), the GKI is a proxy marker of the underlying biological processes of interest, which are not easily accessible for repeated sampling outside laboratory settings.
The goal of dietary KMT is to mimic the metabolism of fasting, leading to an overall reduction and stability in glycemia and corresponding increase in ketonemia, increased ecological competition for nutrients between non-tumoral and cancer cells, as well as reduced insulin/growth signaling and inflammation, leading to a metabolic state that is inhospitable for tumor growth. Therefore, the GKI would be a suitable marker when following a long-term eucaloric/calorically-restricted ketogenic diet or during fasting, but not as a single measurement, for example, after exogenous ketone supplementation under a high-carbohydrate diet. Naturally, several physiological processes can influence the GKI beyond food selection (e.g., stress responses, sleep quality, physical activity, drug treatments). Accordingly, patients should aim for the lowest, physiologically sustainable GKI over extended periods of time; many factors may increase variability, but the overall trend should be towards a chronic reduction from baseline.
Connecting with your last question: indeed, continuous measurement using both continuous glucose monitors (CGM) and continuous ketone monitors (CKM) would be the ideal monitoring tools to capture the area under the curve (AUC) of glycemia/ketonemia and associations with survival outcomes. This has been emphasized multiple times in the manuscript: “Depending on the trial budget, real-time CGM ± CKM is preferred, delivering more robust data collection and biofeedback”. CGM and CKM sensors are already available for both medical applications and over-the-counter (e.g., Freestyle Libre, SiBio KS1). Integrated, dual CGM/CKM sensors are being developed.
Lastly, we agree that delaying selected standard therapies to evaluate diet-drug KMT (particularly radiotherapy due to potential antagonism) has ethical implications that need to be addressed; changes in SOC timing would be considered for testing combined dietary and pharmacological KMT, not dietary KMT alone. The 5-year survival of patients diagnosed with GBM is less than 5%, which has not improved significantly in the last 50 years. While the contribution of chemoradiotherapy to this long-term survival is very small, it may still provide desirable transient disease control, which may be lower, equal or higher than diet-drug KMT. Beyond GBM, other tumor types and stages have unique risk/benefit considerations. However, we note that it is very difficult to design clinical trials when an established standard therapy is biologically antagonistic with the novel proposed therapy, which may be only applicable to specific timings and contexts (e.g., immunotherapy and corticosteroid use, or, as in this case, KMT and brain radiotherapy). Consequently, we propose that judicious SOC delay and/or dose-reduction with frequent monitoring may be the best compromise. Ultimately, we believe this should be the decision of the informed patient.