#05 - Dom D'Agostino, Ph.D.: ketosis, n=1, exogenous ketones, HBOT, seizures, and cancer

Dom followed the Eric Kossoff / John Freeman Hopkins textbook protocol precisely, weighing every gram. He tracked blood ketones (Abbott Precision Xtra) obsessively through this period. The 4:1 diet is still the clinical gold standard for pediatric epilepsy at Hopkins and is used as the reference protocol in virtually all clinical trials of the ketogenic diet for cancer. The ratio was not arbitrary: it was empirically derived over decades of clinical observation at Mayo Clinic, then systematized at Hopkins as the minimum fat-to-protein+carb ratio needed to reliably achieve therapeutic ketosis in children.

The modified ketogenic diet + MCT combination is what took Dom's BHB reliably above his classical 4:1 ceiling. The practical daily structure: high-fat breakfast with MCT oil or MCT-enriched coffee, high-fat low-carb meals within a 6–8 hour feeding window for the time-restriction component, exogenous ketone supplements as needed to hit GKI targets. Dom noted that once he incorporated MCT teas (MCTs), his ketone levels exceeded what he had achieved on the pure dairy-fat classical diet.

The press-pulse framework was published by Dom, Thomas Seyfried, and colleagues at Moffitt Cancer Center in Seminars in Cancer Biology (2017) and in Nutrition and Metabolism. Adrian Check at the Barrow Neurological Institute demonstrated in the GL261 mouse GBM model that nutritional ketosis combined with radiation actually cured GBM in that model — a result that spearheaded clinical research. HBOT reverses tumor hypoxia (which drives HIF-1alpha and VEGF-mediated tumor neovascularization and resistance); once the tumor is hyper-oxygenated, its already-damaged mitochondria produce catastrophic superoxide anion via the Fenton reaction, producing hydroxyl radicals and lipid peroxidation-driven necrosis specifically within tumor cells. The average American's GKI of 25–50 versus the target GKI of 1–2 represents a profound metabolic transformation, not a small diet tweak.

Dom's academic entry point into ketosis was through the oxygen-toxicity seizure problem funded by the DOD. The mechanistic link between high O2 and seizures (S-nitrosylation of GAD, shifting glutamate/GABA ratio) is the same pathway the ketogenic diet corrects in epileptic patients. This gives Dom a unique bidirectional lens: the ketogenic diet prevents oxygen-toxicity seizures via the same mechanism it prevents epileptic seizures. Acetone's anticonvulsant role is underappreciated: Dom notes it opens K+ channels (hyperpolarizing membranes) at physiologic concentrations, suggesting that all three ketone bodies contribute to the anti-seizure effect, not just BHB.

Dom has made hundreds if not thousands of paired urine/blood measurements over a decade. His overall conclusion: blood BHB is the reference standard; urine acetoacetate can be useful for screening in and out of ketosis but not for precise titration; breath acetone is a real-time trend indicator. The ClinicTech/Siemens 10-SG combination provides quantitative urine acetoacetate rather than the qualitative color-change of standard keto strips — a meaningful upgrade for research use.

Dom distinguishes the ester from the salt primarily on potency: the ester is the R-1,3-butanediol mono-ester that hydrolyzes in the gut to produce both R-BHB and 1,3-butanediol (which the liver converts to BHB), making it the highest-yielding exogenous ketone per gram. The salts are more practical for daily use. Dom's personal data: sustained exogenous ketone supplementation drops his fasting insulin below the normal reference range floor, consistent with Veech's rodent data showing ~50% insulin reduction.

Dom and Attia discuss GKI as the 'fastest way' to implement the press strategy. Within a 6-hour feeding window, consuming exogenous ketones during the 18-hour fasting window elevates BHB while fasting-induced glucose suppression lowers the numerator. The GKI converts two separate variables (glucose management + ketone achievement) into a single integrated metric reflecting the tumor's fuel environment. Thomas Seyfried popularized this as the metabolic index for comparing ketosis depth across individuals.

The mechanism is dual: ketones provide an alternative brain fuel that is more energetically efficient per unit oxygen consumed, reducing the metabolic need that drives the oxidative cascade; and ketones shift the glutamate/GABA balance back toward inhibition. At 50 feet of sea water on a pure-oxygen rebreather, a Navy diver faces a brain pO2 approximately 2,000× normal — enough to trigger a grand-mal seizure within 10 minutes. Dom's data suggests that being in strong nutritional ketosis, or consuming a ketone ester pre-dive, dramatically extends the time-to-seizure threshold. The application extends beyond diving: anyone doing clinical HBOT who is at seizure risk — including GBM patients — should ideally be in ketosis before each session.

The study established the foundational framework Dom and Attia build all their clinical recommendations on: the brain is not glucose-dependent, it is energy-dependent. BHB crosses the blood-brain barrier via monocarboxylate transporters that are upregulated during starvation or nutritional ketosis. Cahill measured the arteriovenous difference across the brain for glucose and ketones, giving a direct readout of fuel utilization. The ethical Wild West of Harvard in the 1960s would never pass modern IRB review — no animal ethics committee today allows >48h fasting, let alone 40-day fasts followed by insulin challenges.

The physics: AFM uses a monoatomic-tip probe scanning nanometrically across a live cell surface; membrane roughness correlates with lipid peroxidation. Dom measured membrane perturbations alongside MDA production (TBARS assay) to confirm oxidative membrane damage. Normal cells at 3.25 ATA O2 showed no dramatic damage; GBM cells showed mitochondrial fragmentation and necrotic cell death. The insight was that tumor hypoxia has paradoxically made tumor mitochondria more vulnerable: they are chronically underpowered, structurally compromised, and stripped of the normal antioxidant defenses that healthy cells maintain.

MCT (8–14 carbon medium-chain triglycerides, particularly C8 caprylic acid) bypass lymphatic packaging — unlike long-chain fats that enter chylomicrons, MCTs go directly to the liver via hepatic portal circulation, where the mitochondria-dense liver rapidly oxidizes them and drives ketogenesis. This produces a bolus effect that longer-chain fats cannot replicate. The modified ketogenic diet + MCT combination is what Dom recommends as the practical daily foundation for the press-pulse cancer protocol and for cognitive performance optimization.

Water removes heat 200× faster than air. Even in the upper-80s Florida Keys water, 6-hour EVAs produced progressive hypothermia that Dom described as 'shaking uncontrollably' by hour 4–5. His blood glucose in those post-dive windows was consistently in the 30s–40s mg/dL while on ketosis — a level that would cause incapacitation in a normoglycemic person. The testosterone drop (~25%) was attributed to caloric deficit plus sleep disruption (averaging 6 hours, with disproportionate deep sleep via Oura ring) in a circadian-light-disrupted environment 60 feet underwater. These data feed into the next planned NEEMO 23 mission with his wife as crew member.

The cancer-specific rationale for avoiding antioxidants: (1) Many chemotherapeutic drugs work via oxidative stress — antioxidants blunt them; (2) Radiation's tumor killing is 80% ROS-dependent; (3) HBOT's pulse effect is entirely ROS-dependent (cancer cell mitochondria exploding under hyperoxic conditions); (4) IV vitamin C at millimolar concentrations acts as a pro-oxidant via the Fenton reaction, the opposite of its behavior at nanomolar dietary concentrations. The corollary is that antioxidants may be appropriate for healthy-tissue protection during non-cancer aging contexts — the problem is context specificity.

Achieving a GKI of 1–2 practically: a time-restricted eating window of 6 hours (e.g., 2–8 PM), a ketogenic diet within that window, and exogenous ketones (salts or esters) consumed in the fasting window to simultaneously drive up BHB and drive down glucose via the ketone-induced insulin rheostat. Exogenous ketones provide a non-fermentable calorie source during the fasting window — cancer cells cannot use them as fuel the way they use glucose or glutamine. Low-dose metformin (500–2,000 mg/day) as an adjunct presses the GKI further by activating AMPK and modestly suppressing hepatic glucose output.

Dom measures his own fasting insulin every 1–2 months. His baseline on a standard ketogenic diet is at the low end of the reference range. After several weeks of significant exogenous ketone supplementation, his insulin drops below the reference range floor. The mechanism may involve: (1) ketone-induced pancreatic beta cell suppression (ketones act as a satiety signal via GPR109A); (2) improved insulin sensitivity (Richard Veech's finding); (3) direct hepatic ketone signaling reducing gluconeogenic insulin demand. The practical implication: exogenous ketones may be particularly valuable for cancer patients who cannot achieve very low insulin through diet alone because of appetite or GI constraints from their disease or treatment.