Combination of triheptanoin with the ketogenic diet in Glucose transporter type 1 deficiency (G1D)

Scientific Reports volume 13, Article number: 8951 (2023) Cite this article

170 Accesses

2 Altmetric

Metrics details

Fuel influx and metabolism replenish carbon lost during normal neural activity. Ketogenic diets studied in epilepsy, dementia and other disorders do not sustain such replenishment because their ketone body derivatives contain four carbon atoms and are thus devoid of this anaplerotic or net carbon donor capacity. Yet, in these diseases carbon depletion is often inferred from cerebral fluorodeoxyglucose-positron emission tomography. Further, ketogenic diets may prove incompletely therapeutic. These deficiencies provide the motivation for complementation with anaplerotic fuel. However, there are few anaplerotic precursors consumable in clinically sufficient quantities besides those that supply glucose. Five-carbon ketones, stemming from metabolism of the food supplement triheptanoin, are anaplerotic. Triheptanoin can favorably affect Glucose transporter type 1 deficiency (G1D), a carbon-deficiency encephalopathy. However, the triheptanoin constituent heptanoate can compete with ketogenic diet-derived octanoate for metabolism in animals. It can also fuel neoglucogenesis, thus preempting ketosis. These uncertainties can be further accentuated by individual variability in ketogenesis. Therefore, human investigation is essential. Consequently, we examined the compatibility of triheptanoin at maximum tolerable dose with the ketogenic diet in 10 G1D individuals using clinical and electroencephalographic analyses, glycemia, and four- and five-carbon ketosis. 4 of 8 of subjects with pre-triheptanoin beta-hydroxybutyrate levels greater than 2 mM demonstrated a significant reduction in ketosis after triheptanoin. Changes in this and the other measures allowed us to deem the two treatments compatible in the same number of individuals, or 50% of persons in significant beta-hydroxybutyrate ketosis. These results inform the development of individualized anaplerotic modifications to the ketogenic diet.

ClinicalTrials.gov registration NCT03301532, first registration: 04/10/2017.

Evidence of reduced brain fuel entry or carbon depletion has long accompanied or characterized neurological disorders such as epilepsy1, dementia2 or trauma3. In man, this can be indirectly inferred from fluorodeoxyglucose-positron emission tomography (PET) or measured via microdialysis of brain tissue. In these disorders, ketogenic diets containing a large proportion of lipids relative to other nutrients are either used as therapy or constitute the subject of clinical investigation4. Although it is unlikely that this scientific and medical interest will ultimately equate with universal efficacy, the initial chance discovery of the effect of fasting ketosis on epilepsy5 has been gradually complemented with the characterization of biochemical mechanisms. As a result, today the best understood value of a ketogenic diet is the provision of alternative substrate capable of fueling the tricarboxylic acid (TCA) cycle when glucose utilization is depressed6.

The principal diet-derived ketone bodies, beta-hydroxybutyrate and acetoacetate, contain 4 carbons and, when metabolized, yield preferentially two molecules of two-carbon acetyl coenzyme A. This dicarbon molecule is also the main glucose oxidation byproduct. Thus, some reactions of glucose metabolism can be replaced by ketone body metabolism from the perspective of acetyl coenzyme A generation and its subsequent flux into the TCA cycle. However, a significant fraction of total brain glycolytic flux, perhaps amounting to 20%, is separately steered into anaplerosis, which is the replenishment of carbon lost in the course of the TCA cycle7. This carbon lost to metabolism ultimately finds its way into excretion byproducts or expired CO2 gas. Most available anaplerotic flux estimates refer to normal brain and their precise values vary depending on investigational methods. Yet, the importance of anaplerosis in the brain is made apparent by reductions in the activity of the enzyme that catalyzes the conversion of pyruvate into oxaloacetate, a key anaplerotic reaction. Individuals with deficiency in this enzyme, named pyruvate carboxylase, can manifest encephalopathy with necrosis of neural tissue8.

Considered from this perspective, the ketogenic diet is metabolically deficient: the alternative fuels provided by ketone bodies, whether ingested as conjugates of other substances or produced after ketogenic diet consumption, lack anaplerotic potential7. This is because dietary fats and their derivative ketone bodies contain an even number of carbons and are fully consumed in the TCA cycle via acetyl coenzyme A formation. In contrast, metabolic substrates containing an odd number of carbons greater than 5 can similarly fuel the TCA cycle through the sequential formation of one or more two-carbon acetyl coenzyme A molecules while also fueling anaplerosis via the additional generation of 3-carbon propionate as the end product of the final 3 carbons of the odd carbon substrate. Propionyl coenzyme A metabolism may then lead to the generation of the TCA cycle intermediate succinate. These reactions result in the supply of net carbon to the TCA cycle, thus compensating for a significant fraction of the carbon loss7.

Therefore, the ketogenic diet can be complemented to mitigate its metabolic insufficiency. To this end, we have studied individuals with partial deficiency of the brain glucose transporter type I (G1D) since they are rapidly and informatively susceptible to dietary metabolic fuel administration9,10. G1D is a prototypic brain carbon depletion state11 associated with synaptic failure that proves only partially treatable with a ketogenic diet10. It commonly adopts the form of a childhood-onset epilepsy refractory to antiseizure drugs which has remained almost inextricably linked to the ketogenic diet therapy for 30 years12,13. In this context, as in other epilepsies, dementia or trauma, which are also characterized by decreased glucose metabolism, the carbohydrate-restricted ketogenic diet leads to an intended but intuitively counterproductive decrease in blood glucose available to the brain. This is because increased glycemia interferes with ketogenesis and reduced glucose favors it.

Thus, our goal was to investigate whether triheptanoin (C7), an edible triglyceride of 7-carbon heptanoic acid, was compatible with a ketogenic diet using G1D as model disorder. Two main potential limitations to such a combination treatment exist. First, some ketogenic diets contain medium chain triglycerides, which yield the medium chain fatty acids octanoate and decanoate, both of which can compete with C7 metabolism14. Second, since heptanoate metabolism generates acetyl coenzyme A, C7 can potentially stimulate hepatic neoglucogenesis and this could decrease ketosis via insulin release15. Neoglucogenesis stemming from heptanoate has been observed after infusion of heptanoate in G1D mice16. However, the relative amount of heptanoate infused was greater than that derived from the C7 used for G1D subjects, since a large quantity of labeled substrate is necessary to achieve labeling of brain intermediary metabolites in human and mouse 13C metabolic tracer studies17,18,19,20. Nevertheless, uncertainty remains about the magnitude of neoglucogenesis and its potential induction of increased glycemia after C7 ingestion. Both octanoate interference and neoglucogenesis are rapid events occurring within minutes14,16. This informed the duration of our study of compatibility.

We used C7 at the maximum tolerable dose21 which, to our knowledge, has not yet been used in G1D. This dose is considerably higher than previously used9,22. This elevated dose is important not only to maximize any potential benefit in future studies but also to facilitate eliciting any metabolic interference with the ketogenic diet. The ultimate objective was thus to enable future combined or comparative studies because, for one third of G1D patients, the ketogenic diet is insufficient and may thus be partly replaced by C7; conversely, C7 as monotherapy may also prove insufficient in some patients and may thus benefit from the addition of a ketogenic diet13.

Notably, this was not a population or analytical measure distribution study, for addressing those aspects would require a different approach and a sample size unavailable for a relatively infrequent disease. Rather, the goal was to ascertain which one of three possible compatibility scenarios was more likely: (a) noninterference between C7 and the ketogenic diet, where neither the biochemical competition nor the individual metabolic variability reported in other organisms or studies precluded full compatibility, (b) generalized or absolute incompatibility due to biochemical interference, whereby any subject variability, if present, would prove trivial or insufficient to surmount the cited potentially prohibitive biochemical interactions, or (c) compatibility in only a fraction of subjects due to individual variability in one or both kinds of factors. The implication of c is that compatibility is an individual phenomenon and thus future studies or treatments must account for this crucial source of variability. This is what we found.

We followed the Declaration of Helsinki of 1975 criteria as revised in 1983 and received Institutional Review Board approval from the University of Texas Southwestern Medical Center, with ClinicalTrials.gov identifier NCT03301532, first registration 04/10/2017. The inclusion and exclusion criteria are listed in Table 1. Written informed consent was obtained from one participant who was over the age of 18. Written informed consent was obtained from all the subjects or legally authorized representatives. Assent was also documented for cognitively-capable children between 10 and 17 years of age.

The approach included substituting a fraction of ketogenic diet fat with C7, weight by weight, at the maximum tolerable dose (45% of total daily calories21) in individuals receiving a ketogenic diet prior to enrolment as medically prescribed independently of this study. C7 addition and equivalent dietary fat subtraction was calculated to preserve the pre-enrolment fat to protein and carbohydrate ratio. The substitution with C7 was immediate rather than gradual since patients consuming a ketogenic diet are high-fat tolerant and receive well over 45% calories from fat (often as much as 90%), such that C7 replacement was expected to be fully tolerable. To minimize gastrointestinal intolerance from triglyceride consumption, any medium chain triglyceride consumption part of the ketogenic diet was replaced with other dietary fat 24 h before triheptanoin consumption.

Because compatibility can be defined from biochemical, clinical, electroencephalographic or other perspectives, several measures of compatibility were assessed. Additionally, the data are provided or available in full to enable other possible types of compatibility analyses. First, we utilized the consideration that C7 metabolism can potentially interfere with ketosis as the key factor underlying compatibility. Thus, we used blood beta hydroxybutyric level as the primary criteria of compatibility. Second, because G1D individuals exhibit seizures, which often constitute the motivation for the use of the ketogenic diet, an expected outcome, if the ketogenic diet and C7 were compatible, was a lack of change in clinical seizures after C7. Of note, G1D seizures are not precisely quantifiable in EEG recordings lasting even a few days10. Since most G1D seizures are of the absence type, and because absence seizure frequency is often variable on a day-to-day basis and may remain unnoticed by patients10, their frequency and severity was estimated in two complementary ways. Due to the cited electrical-clinical dissociation, these two methods were not expected to necessarily yield the same result. The first method employed the same habitual caretaker for each subject, who spent the study time next to each subject and rated seizures as essentially unchanged, minimally changed (less than 30% change in frequency or duration) or significantly changed (greater than 30% change) throughout an interval that included from at least 2 days prior to the study to the end of the study. Regarding the second method, since the first method may not correlate with electrographic seizures10, we also assessed compatibility based upon the lack of an increase in electroencephalographic (EEG) abnormal activity. This was achieved by a clinical epileptologist (i.e., not automated) comparing of all the clinically significant abnormalities noted in pre-treatment EEGs which, for the 10 subjects studied included slowing, spikes, polyspikes or spike-waves in 48-h recordings. Based on previous G1D data illustrating about a 30% variability for test–retest in prolonged EEG recordings, and taking into consideration time of day variations and meal times10, changes greater than 30% in frequency following treatment measured at sequential 2-h intervals and compared at the same time of the day were considered significant.

The disease features of all subjects included a variable combination of intellectual disability, epilepsy, ataxia, or episodic movement disorder representative of the disease population12. The G1D diagnosis was ascertained via DNA sequence analysis of the Slc2a1 gene, which encodes Glut1, or of the integrity of the appropriate area of chromosome 1 where the gene is located utilizing standard clinical genetic diagnostic criteria23. The reference Slc2a1 transcript was NM_006516.2. Enrolment followed the order of contact made by eligible subjects. Eligible contacts far exceeded enrolment targets, thus reducing potential bias associated with the sampling of small populations. There was no consideration of geographic location (U.S. or abroad) or disease severity. Some subjects were recruited from our Rare Brain Disorders Program at UT Southwestern Medical Center. They included English and Spanish speakers. Medications, including antiseizure drugs, were not allowed to change 90 days prior to or during the study. Subject age, Slc2a1 G1D causative mutation and ketogenic diet ratio are given in Table 2. As previously noted24, exon 4 mutations were predominant.

Food grade triheptanoin (Stepan Lipid Nutrition) was consumed 4 times per day (approximately every 6 h) for one day. This time was deemed sufficient for any incompatibility to manifest given the rapidity of the biochemical processes of interest14,16. The C7 dose was determined at 45% of the total daily caloric intake21. This dose was used for replacement of ketogenic fat (weight by weight). To minimize insulin-mediated suppression of ketogenesis from other foods14, each C7 dose was consumed 45–60 min before meals. C7 was optionally mixed with a fat free, sugar free yogurt, pudding, or equivalent low-calorie food.

Figure 1 illustrates the study procedural sequence. Each subject underwent a review of a physical exam from medical records and provided a pretreatment history including seizure frequency and received a daily physical, including neurologic, examination. The subjects were hospitalized in an Epilepsy Monitoring Unit for 48 h for C7 administration and monitoring. Continuous EEG was recorded 12 h preceding, and for the duration of C7 consumption, plus 18 h after the last dose, comprising approximately a total of 48 h. The recordings used an international 10–20 system and were reviewed and analyzed by an epileptologist. The subjects also received a nutritional assessment prior to starting the C7 supplement. Side effects were assessed using the Hague Side Effect Scale25 and the VA Toxicity Scale26.

Study procedures. BHB beta-hydroxybutyric acid, Glu glucose.

An analytical laboratory evaluation was obtained and reviewed on day 1, which was the day prior to starting C7. This included a comprehensive metabolic panel (glucose, blood urea nitrogen (BUN), creatinine, sodium, potassium, chloride, CO2, anion gap, calcium, total protein, albumin, alkaline phosphatase, aspartate aminotransferase (AST), alanine aminotransferase (ALT), total bilirubin), lipid panel, lactate, complete blood count, and beta-hydroxybutyrate. Plasma glucose and beta-hydroxybutyrate were then measured twice on days 2 and 3 approximately 2 h after the first C7 dose and 10 min after the last C7 dose on day 2. Hospital admission concluded on day 3. A complete blood count, comprehensive metabolic panel, lactate, and beta-hydroxybutyrate labs were obtained again on day 4. The subjects were also surveyed for clinical manifestations on day 5 and via telephone on day 30. Additional determinations of the C5 ketones β-hydroxypentanoate and β-ketopentanoate were performed in 5 subjects (participants 6, 7, 8, 9 and 10) at baseline and approximately 2 h approximately after the first C7 dose and 10 min after the last C7 dose on day 2, as previously described27.

One method to judge C7-ketogenic diet compatibility from the perspective of beta-hydroxybutyric acid level changes relies, as prerequisite, on an estimation of beta-hydroxybutyric acid level variability in independence of C7. Clinical practice indicates that individuals who receive a ketogenic diet may exhibit significant fluctuation in blood ketone levels irrespective of the indication for the diet28. The same is likely to apply to G1D individuals (J.M.P. observations in n = 122 G1D subjects not treated with C7). This may impact the assessment of C7 effects on ketonemia in our C7 compatibility study subjects by introducing a source of normal variability unrelated to C7. For example, fluctuations or coefficients of variation as high as 44%29 and 46%30 (defined as the standard deviation divided by the mean of serial measurements in individual subjects) in blood beta-hydroxybutyric acid levels have been observed in individuals with epilepsy treated with a ketogenic diet.

Thus, since there were no previously reported data in G1D, a second group of subjects from our Rare Brain Disorders Program was used to estimate the variability of blood beta-hydroxybutyric acid levels in G1D subjects who consume a ketogenic diet. Thus, Institutional Review Board approval was obtained from UT Southwestern Medical Center for a retrospective chart review. This second group variability was measured in the absence of C7 consumption and was thus used as a normative or reference range for the estimation of C7 compatibility in the first group of G1D subjects. To this end, we analyzed the variation in blood beta-hydroxybutyric acid levels in 20 G1D individuals unrelated to the first group but comprising a similar age range, who were studied in our Rare Brain Disorders Program between May 2013 and November 2022. These individuals provided a total of 224 blood beta-hydroxybutyric acid values, ranging from 6 to 28 measurements per individual and spanning a minimum and maximum period of measurement of 1 month and 4 years for the individuals of this entire group. To allow for as an ample degree of variability as possible, we included subjects who consumed a ketogenic diet at the same ratio range of the subjects studied here for C7 compatibility and did not account for changes in ratio within this range for any particular individual during his or her serial measurements, nor for the time of the day when the measurements were made or time between measurements, nor for the degree of clinical therapeutic efficacy attributable to the diet. This approach was intended to provide a broad, non-experimentally controlled measure of blood beta-hydroxybutyric acid level variation consistent with fluctuations commonly observed in standard clinical practice, thus minimizing the likelihood of incorrectly attributing any observed variation after C7 to incompatibility with the ketogenic diet rather than to potentially normal variation. Because this fluctuation was measured in all cases over a significantly longer time period than the duration of our compatibility study, we reasoned that it is appropriate to use this variability as the maximum normal (i.e., C7-unrelated) variability that could be expected during our compatibility study.

To further reduce sources of C7-unrelated variability in beta-hydroxybutyric acid level comparisons pre and post C7, we also separated the analysis of the values obtained from C7-treated individuals when their level of ketosis pre-C7 treatment was below 2 mM, since any fluctuation of these values after C7 would be overshadowed by the greater degree of fluctuation estimated from the above averaged percent variability in non-C7 treated individuals. A detailed justification for this level based on the values obtained in our subjects and on other studies is discussed below.

Since future studies may determine a different degree of normal ketosis variability in G1D subjects depending on ketogenic diet ratio, time and number of blood ketone measurements or other factors, and since other numeric estimates of compatibility are also possible, we provide all of the individual values from our study to enable such future methods of analysis.

Ten individuals with genetically confirmed G1D were enrolled to study C7-ketogenic diet compatibility (Table 2). As previously noted, Slc2a1 exon 4 mutations were common24. Median age at enrollment was 10 years. 6 of 10 subjects were female. Most subjects identified themselves as white (9) and non-Hispanic (9), 1 subject identified as white Hispanic and another as Asian. All of the recruited participants completed the study.

The mean beta-hydroxybutyric acid levels in the 20 C7-untreated G1D subjects receiving a ketogenic diet was 3.34 ± 1.86 mM (mean and SD). In these non-C7 treated subjects, the fluctuation in blood beta-hydroxybutyric acid levels was about 50%, consistent with reports in non-G1D subjects receiving a ketogenic diet29,30. This value also implied that, in the 10 subjects to be treated with C7, pretreatment blood beta-hydroxybutyric acid levels below 2 mM were considered as potentially too low to allow for reliable evaluation of changes relative to the larger C7-unrelated variability expected and thus this subset of subjects merited additional disaggregated analysis. A blood beta-hydroxybutyric acid level greater than 2 mM is also associated with significant ketonuria, a common clinical indicator of ketosis29.

In addition to standard Epilepsy Monitoring Unit constant and offline video supervision, all 10 subjects were almost continuously observed by a primary caretaker including at night. Parents were asked to document any apparent seizure for subsequent EEG analysis. As previously established, subjects or caretakers in general did not notice a significant fraction of electrographic seizures10. No subject displayed significantly increased observable seizures during the study. Two subjects exhibited complete cessation of observable seizures and one displayed a 75% reduction in seizure frequency while receiving C7. In these three subjects, seizures returned upon C7 discontinuation, as also attested by the EEG in two of them.

Of the 10 individuals the 48 h continuous video EEG was normal in 3 individuals and abnormal in 7. These results are presented in Table 3 including, for reference, previous EEG findings for each subject. As noted, the EEG remained unchanged in 7 of the 10 subjects, improved in 1 and worsened in 2, but only after discontinuation of C7.

Considering all the subjects as a group, the mean fasting blood glucose about 15 h before C7 was 81.2 ± 13.6 mg/dl (mean and SD). This value did not significantly change 30–60 min after ingestion of the first dose of C7 (79.6 ± 7.3, t test, p > 0.05) nor at any of the subsequent determinations (Table 4).

Figure 2 and Table 5 display the impact of C7 addition to the ketogenic diet on beta-hydroxybutyric acid. The impact, when present, was rapid, as expected from biochemical principles15 and was fully discernible by the first beta-hydroxybutyric acid level determination on day 3. Although a specific minimum level of ketosis has not been defined for the treatment of G1D, two subjects (1 and 9) displayed reduced levels pre-C7 relative to the other subjects. Based on these data, particularly on the beta-hydroxybutyric acid levels following 4 doses of C7, we estimated compatibility in subjects 3, 4, 7 and 10 and poor or no compatibility in subjects 2, 5, 6 and 8 (Fig. 2). There was no statistically discernible pattern in the two subjects (1 and 9) whose beta-hydroxybutyric acid levels were below 2 mM on day 1. There was also no obvious correlation between these levels and the subjects’ clinical manifestations and degree of treatment efficacy afforded by the ketogenic diet prior to enrolment.

Beta-hydroxybutyrate levels independent of, before, during and after C7 ingestion. Left panel: Non C7: Percent change in 224 beta-hydroxybutyrate levels from 20 G1D individuals not receiving C7 and not treated in this study; C7 Compatible: change in G1D individuals that exhibited beta-hydroxybutyrate values indicative of compatibility with C7. C7 Non compatible: change in G1D individuals where the ketogenic diet was estimated non compatible with C7; The variability in sample-to-sample change for non-C7 treated subjects was not significantly different from the variability in G1D patients who received C7 in whom C7 was compatible with the ketogenic diet. Welch's ANOVA (for unequal variances) with Dunnett's correction for multiple comparisons. *p < 0.05 and **p < 0.01. The Browne-Forsythe test indicated similar significant differences. Right panel: Evolution of beta-hydroxybutyrate levels over time of the day for subjects where the two treatments were compatible (purple circles) or non-compatible (black circles). Compatibility from this perspective was defined as less than 50% change in beta-hydroxybutyrate levels post C7 by day 3 in relation to pre-C7 levels.

Overall, beta-hydroxybutyrate levels changed from the fasting state 3.82 ± 1.96 mM to 2.67 ± 1.61 mM in the final post C7 determination on day 4. 9 of the 10 individuals presented decreased beta-hydroxybutyrate values after the first dose of C7, and all of them did so after the last dose on day 2. Over the following 48 h (days 3 and 4), beta-hydroxybutyrate increased in all individuals, but did not reach initial pre-C7 values in any of the subjects. One subject (number 2) did not receive the fourth C7 dose due to safety considerations stemming from low beta-hydroxybutyrate levels.

In summary, we found that ketosis (defined as blood beta-hydroxybutyrate level), clinical seizures, glycemia, and EEG, utilized as measures responsive to reduction of the ketogenic state, were acceptably altered by our compatibility criteria for 4 individuals supplemented with C7.

There were no serious or unexpected adverse events. Seven individuals experienced mild digestive discomfort that resolved without intervention. There was no difference in frequency or severity of G1D-related symptoms or in physical examination in either C7-ketogenic diet compatible or non-compatible subjects. One subject received one dose of ondansetron after vomiting once, which had been previously used sporadically as needed for nausea and vomiting since the initiation of the ketogenic diet by this subject.

One subject improved significantly (stopped having seizures for the first time in several years) after C7 administration. Two of them exhibited worsening of seizures on day 3, after C7 discontinuation (subjects 4 and 6, Table 3). Subject 4 displayed electroclinical worsening (increased frequency of interictal paroxysms manifested as polyspike and spike and slow wave discharges as well as increased frequency of electroclinical absence seizures) 11 h after the last C7 dose, in association with a beta-hydroxybutyrate levels of 2.2 mM (from a pre-C7 level of 3.2 mM). Subject 6 also manifested electroclinical worsening approximately 11 h after the last C7 dose with increased frequency of interictal paroxysms (polyspike and spike and slow wave discharges) as well as increased frequency of electroclinical myoclonic absence seizures in relation to a beta-hydroxybutyrate level of 0.9 mM (pre-C7 level 3.3 mM).

On days 5 and 30 all appreciable clinical changes had reverted to the pre-treatment state.

C5 ketogenesis is variable across individuals21. To investigate if the degree C5 ketosis influenced compatibility with the ketogenic diet, we studied several subjects serially in relation to C7 administration. Figure 3 illustrates C5 ketosis for beta-hydroxy pentanoate and beta-keto pentanoate in select subjects as a function of C7 administration. These values were commensurate with previous measurements27 and with determinations made at the maximum tolerable dose21. There was no correlation between these values and previous beta-hydroxybutyrate levels, or between these values and compatibility with the ketogenic diet.

C5 ketonemia in select subjects before and after C7 administration. Left panel: Beta-hydroxy pentanoate values in relation to C7 administration times (4 doses, administered on day 2). Right panel: values for beta-keto pentanoate.

Triheptanoin was previously used at a lower dose as a triglyceride food supplement to a regular human diet in individuals with Glucose transporter 1 deficiency (G1D)9. The goal was to replenish depleted brain carbon31 in a glucose-independent or complementary manner. The biochemical basis for this intervention were substantiated by multiplet 13C-nuclear magnetic resonance (NMR) spectroscopy32, gas chromatography-mass spectrometry (GC–MS) and liquid chromatography-mass spectrometry (LC–MS) studies of the metabolism of infused [5,6,7-13C3]heptanoate in a G1D mouse model faithful to the most common human disorder phenotype16. This metabolic substrate is primarily metabolized in the liver, producing blood [3,4,5-13C3] C5 ketones27. This work revealed enrichment in heptanoate-derived plasma glucose via neoglucogenesis and increased cerebral acetyl coenzyme A and glutamine concentrations. The appearance of 13C label in specific carbons of glutamate, glutamine and GABA was consistent with metabolism of heptanoate and or derivative C5 ketones in glia in a flux-sensitive manner16,17,18,33. Thus, the β-oxidation of carbons 1 to 4 of heptanoate generates two molecules of acetyl coenzyme A and one molecule of propionyl coenzyme A derived from carbons 5 to 7. The latter can enter the TCA cycle through propionyl coenzyme A carboxylase.

However, these studies, in which a regular diet was maintained, could not be directly extended to individuals who receive a ketogenic diet due to several potential limitations. First, stimulated blood ketone body levels are normally variable in persons28. Second, upon triheptanoin consumption, the concentrations of its two derivative C5 ketone bodies can be variable across individuals34 and exhibit no correlation in particular individuals21. Third, C5 ketone blood levels are not necessarily proportional to biological effect in the brain and therefore cannot be used alone as indicators of compatibility between triheptanoin and a ketogenic diet. This is due to the avid uptake of C5 ketones by several tissues rich in 3-oxoacid-coenzyme A transferase, including the brain35,36,37. This makes efficacy dependent upon not only blood level but also brain uptake affinity. Further, as noted in the mouse, there are several possible metabolic effects derived from heptanoate, or its byproducts, in the brain16. This is due to the brain fuel potential of C5 ketones, heptanoate itself, and glucose from neoglucogenesis, all of which would be difficult to mechanistically separate without co-infusion labeling or other complex studies given the uncertainties about the magnitude of some of the relevant metabolic reactions33.

These considerations justify our direct human investigation of compatibility from several perspectives. Previously, 14 G1D subjects on a regular diet studied in the fasting state9 exhibited a mean blood glucose that did not significantly change 30–60 min after a smaller C7 dose than used in this study. Beta-hydroxybutyrate levels were also unchanged in the fasting state relative to the post-C7 state. While this argued against significant neoglucogenesis from this reduced dose, the natural level of ketosis was too modest to allow inferences. The present data, obtained under the maximum tolerable C7 dose21, corroborate that C7 metabolism does not appreciably interfere with blood glucose level when taken simultaneously with a ketogenic diet in the time frame of this study.

50% or 4 of 8 subjects with initial beta-hydroxybutyrate levels greater than 2 mM demonstrated a significant reduction in ketosis after C7. This allowed us to deem C7 not compatible with the ketogenic diet in these subjects. The restoration of beta-hydroxybutyrate levels started about 30 h after the first C7 dose. However, ketone levels may not be the most valuable measure for add-on triheptanoin tolerability estimation since other factors stemming from the addition of C7 may compensate for any decrease in beta-hydroxybutyrate level.

4 of the subject families expressed overall satisfaction with symptom amelioration after C7. In all of them, C7 was deemed compatible with the ketogenic diet. Besides seizure cessation, these symptoms eluded precise quantification, since they involved facilitation of thought, expressive communication and limb coordination. Our study, however, did not evaluate these aspects. One individual (subject 6) experienced cessation of seizures after C7 addition in the context of blood glucose increase and beta-hydroxybutyrate level decrease, with a return of seizures after C7 discontinuation. This individual, who likely benefited from a slight increase in glycemia due to neoglucogenesis, may exemplify the therapeutic value of modest glucose elevations in G1D10. Rather than a case of incompatibility, this may be considered an instance of therapeutic substitution. Two other individuals experienced either a 75% reduction (subject 3) and cessation (subject 4) of EEG abnormalities following consumption of C7, which suggests that addition of C7 may almost immediately augment the therapeutic effect of a ketogenic diet, thus warranting further investigation.

This study justifies expanding the study of triheptanoin in two contexts. First, there is a need for an alternative or a supplement to the commonly-used ketogenic diet38, not only to facilitate tolerability39,40,41,42 but also to fulfill biosynthetic demands or anaplerosis, which is deficient in numerous disease states43,44. Second, G1D is most symptomatic in early childhood, when brain growth parallels a robust stimulation of cerebral glucose and protein metabolism, which rely on net carbon deposition or, in short, anabolism and anaplerosis45. This underscores the importance of a carbon donor-rich diet, whereas the ketogenic diet is restricted in this regard. Relatively unbiased tools such a whole-exome DNA sequencing, comprehensive genomic hybridization and Sanger gene panels23 are increasingly uncovering G1D in young infants, for whom the ketogenic diet remains insufficiently tested46 and poor in anaplerotic potential during this period of rapid brain growth47.

We did not study long-term compatibility between C7 and the ketogenic diet, which may be potentially influenced by adaptation or other poorly understood processes. These hypothetical mechanisms stand in contrast with the relatively short time scale (minutes to hours) of the known relevant biochemical processes14,16. Therefore, the shorter time frame was selected for our investigation. Second, it is possible to define compatibility based on criteria different from ours. For example, clinical changes alone, more prolonged EEG recordings or alternative forms of data analysis such as weighted combinations of biochemical and clinical effects. Lastly, there are distinct metabolic processes for different routes and forms of administration of C7 and ketone bodies14. Thus, compatibility may differ under these other conditions.

All the data are publicly available from the corresponding author.

Triheptanoin

Glut1 deficiency

Glucose transporter protein type 1

Fluorodeoxyglucose-positron emission tomography

Sacktor, B., Wilson, J. E. & Tiekert, C. G. Regulation of glycolysis in brain, in situ, during convulsions. J. Biol. Chem. 241, 5071–5075 (1966).

Article CAS PubMed Google Scholar

Alavi, A. et al. Regional cerebral glucose metabolism in aging and senile dementia as determined by 18F-deoxyglucose and positron emission tomography. Exp. Brain Res. Suppl. 5, 187–195. https://doi.org/10.1007/978-3-642-68507-1_26 (1982).

Article ADS MathSciNet CAS Google Scholar

Langfitt, T. W. et al. Computerized tomography, magnetic resonance imaging, and positron emission tomography in the study of brain trauma. Preliminary observations. J. Neurosurg. 64, 760–767. https://doi.org/10.3171/jns.1986.64.5.0760 (1986).

Article CAS PubMed Google Scholar

Stafstrom, C. E. & Rho, J. M. The ketogenic diet as a treatment paradigm for diverse neurological disorders. Front. Pharmacol. 3, 59. https://doi.org/10.3389/fphar.2012.00059 (2012).

Article CAS PubMed PubMed Central Google Scholar

Guelpa, G. & Marie, P. A lutte contre l’epilepsie par la desintoxication et par la reeducation alimentaire. Revue de Therapie Medico Chirurgicale 78, 8–13 (1911).

Google Scholar

Cervenka, M. et al. Metabolism-based therapies for epilepsy: New directions for future cures. Ann. Clin. Transl. Neurol. 8, 1730–1737. https://doi.org/10.1002/acn3.51423 (2021).

Article PubMed PubMed Central Google Scholar

Marin-Valencia, I., Roe, C. R. & Pascual, J. M. Pyruvate carboxylase deficiency: Mechanisms, mimics and anaplerosis. Mol. Genet. Metab. 101, 9–17. https://doi.org/10.1016/j.ymgme.2010.05.004 (2010).

Article CAS PubMed Google Scholar

Pascual, J. M. in Rudolph's pediatrics (ed. Mark W. Kline) (McGraw-Hill In press).

Pascual, J. M. et al. Triheptanoin for glucose transporter type I deficiency (G1D): modulation of human ictogenesis, cerebral metabolic rate, and cognitive indices by a food supplement. JAMA Neurol. 71, 1255–1265. https://doi.org/10.1001/jamaneurol.2014.1584 (2014).

Article PubMed PubMed Central Google Scholar

Rajasekaran, K. et al. Metabolic modulation of synaptic failure and thalamocortical hypersynchronization with preserved consciousness in Glut1 deficiency. Sci. Transl. Med. 14, eabn2956. https://doi.org/10.1126/scitranslmed.abn2956 (2022).

Article CAS PubMed Google Scholar

Pascual, J. M., Van Heertum, R. L., Wang, D., Engelstad, K. & De Vivo, D. C. Imaging the metabolic footprint of Glut1 deficiency on the brain. Ann. Neurol. 52, 458–464. https://doi.org/10.1002/ana.10311 (2002).

Article CAS PubMed Google Scholar

Pascual, J. M. & Ronen, G. M. Glucose transporter type I deficiency (G1D) at 25 (1990–2015): presumptions, facts, and the lives of persons with this rare disease. Pediatr. Neurol. 53, 379–393. https://doi.org/10.1016/j.pediatrneurol.2015.08.001 (2015).

Article PubMed PubMed Central Google Scholar

Hao, J., Kelly, D. I., Su, J. & Pascual, J. M. Clinical aspects of glucose transporter type 1 deficiency: Information from a global registry. JAMA Neurol. 74, 727–732. https://doi.org/10.1001/jamaneurol.2017.0298 (2017).

Article PubMed PubMed Central Google Scholar

Deng, S., Zhang, G. F., Kasumov, T., Roe, C. R. & Brunengraber, H. Interrelations between C4 ketogenesis, C5 ketogenesis, and anaplerosis in the perfused rat liver. J. Biol. Chem. 284, 27799–27807. https://doi.org/10.1074/jbc.M109.048744 (2009).

Article CAS PubMed PubMed Central Google Scholar

Newsholme, E. A. Carbohydrate metabolism in vivo: Regulation of the blood glucose level. Clin. Endocrinol. Metab. 5, 543–578 (1976).

Article CAS PubMed Google Scholar

Marin-Valencia, I., Good, L. B., Ma, Q., Malloy, C. R. & Pascual, J. M. Heptanoate as a neural fuel: Energetic and neurotransmitter precursors in normal and glucose transporter I-deficient (G1D) brain. J. Cereb. Blood Flow Metab. 33, 175–182. https://doi.org/10.1038/jcbfm.2012.151 (2013).

Article CAS PubMed Google Scholar

Marin-Valencia, I. et al. High-resolution detection of (1)(3)C multiplets from the conscious mouse brain by ex vivo NMR spectroscopy. J. Neurosci. Methods 203, 50–55. https://doi.org/10.1016/j.jneumeth.2011.09.006 (2012).

Article CAS PubMed Google Scholar

Marin-Valencia, I. et al. Cortical metabolism in pyruvate dehydrogenase deficiency revealed by ex vivo multiplet (13)C NMR of the adult mouse brain. Neurochem. Int. 61, 1036–1043. https://doi.org/10.1016/j.neuint.2012.07.020 (2012).

Article CAS PubMed Google Scholar

Marin-Valencia, I. et al. Glucose metabolism via the pentose phosphate pathway, glycolysis and Krebs cycle in an orthotopic mouse model of human brain tumors. NMR Biomed. 25, 1177–1186. https://doi.org/10.1002/nbm.2787 (2012).

Article CAS PubMed PubMed Central Google Scholar

Maher, E. A. et al. Metabolism of [U-13 C]glucose in human brain tumors in vivo. NMR Biomed. 25, 1234–1244. https://doi.org/10.1002/nbm.2794 (2012).

Article CAS PubMed PubMed Central Google Scholar

Malaga, I. et al. Maximum dose, safety, tolerability and ketonemia after triheptanoin in glucose transporter type 1 deficiency (G1D). Sci. Rep. 13, 3465. https://doi.org/10.1038/s41598-023-30578-z (2023).

Article ADS CAS PubMed PubMed Central Google Scholar

Striano, P. et al. A randomized, double-blind trial of triheptanoin for drug-resistant epilepsy in glucose transporter 1 deficiency syndrome. Epilepsia 63, 1748–1760. https://doi.org/10.1111/epi.17263 (2022).

Article CAS PubMed PubMed Central Google Scholar

SoRelle, J. A., Pascual, J. M., Gotway, G. & Park, J. Y. Assessment of interlaboratory variation in the interpretation of genomic test results in patients with epilepsy. JAMA Netw. Open. 3, e203812. https://doi.org/10.1001/jamanetworkopen.2020.3812 (2020).

Article PubMed PubMed Central Google Scholar

Pascual, J. M. et al. Structural signatures and membrane helix 4 in GLUT1: Inferences from human blood-brain glucose transport mutants. J. Biol. Chem. 283, 16732–16742. https://doi.org/10.1074/jbc.M801403200 (2008).

Article CAS PubMed PubMed Central Google Scholar

Carpay, J. A. et al. Parent-reported subjective complaints in children using antiepileptic drugs: What do they mean?. Epilepsy Behav. 3, 322–329. https://doi.org/10.1016/s1525-5050(02)00047-1 (2002).

Article PubMed Google Scholar

Guy, W. ECDEU Assessment Manual for Psychopharmacology. (US Department of Health, Education, and Welfare, Public Health Service, 1976).

Kallem, R. R., Primeaux, S., Avila, A., Pascual, J. M. & Putnam, W. C. Development and validation of a LC-MS/MS method for quantitation of 3-hydroxypentanoic acid and 3-oxopentanoic acid in human plasma and its application to a clinical study of glucose transporter type I deficiency (G1D) syndrome. J. Pharm. Biomed. Anal. 205, 114335. https://doi.org/10.1016/j.jpba.2021.114335 (2021).

Article CAS PubMed PubMed Central Google Scholar

Saudubray, J. M. et al. Variation in plasma ketone bodies during a 24-hour fast in normal and in hypoglycemic children: Relationship to age. J. Pediatr. 98, 904–908. https://doi.org/10.1016/s0022-3476(81)80583-5 (1981).

Article CAS PubMed Google Scholar

Gilbert, D. L., Pyzik, P. L. & Freeman, J. M. The ketogenic diet: Seizure control correlates better with serum beta-hydroxybutyrate than with urine ketones. J. Child Neurol. 15, 787–790. https://doi.org/10.1177/088307380001501203 (2000).

Article CAS PubMed Google Scholar

SuntrupIii, D. J., Ratto, T. V., Ratto, M. & McCarter, J. P. Characterization of a high-resolution breath acetone meter for ketosis monitoring. PeerJ 8, e9969. https://doi.org/10.7717/peerj.9969 (2020).

Article Google Scholar

Marin-Valencia, I. et al. Glut1 deficiency (G1D): Epilepsy and metabolic dysfunction in a mouse model of the most common human phenotype. Neurobiol. Dis. 48, 92–101. https://doi.org/10.1016/j.nbd.2012.04.011 (2012).

Article CAS PubMed PubMed Central Google Scholar

Marin-Valencia, I. et al. High-resolution detection of (13)C multiplets from the conscious mouse brain by ex vivo NMR spectroscopy. J. Neurosci. Methods. https://doi.org/10.1016/j.jneumeth.2011.09.006 (2011).

Article PubMed PubMed Central Google Scholar

Jeffrey, F. M. et al. Modeling of brain metabolism and pyruvate compartmentation using (13)C NMR in vivo: Caution required. J. Cereb. Blood Flow Metab. 33, 1160–1167. https://doi.org/10.1038/jcbfm.2013.67 (2013).

Article CAS PubMed PubMed Central Google Scholar

Roe, C. R. & Brunengraber, H. Anaplerotic treatment of long-chain fat oxidation disorders with triheptanoin: Review of 15 years experience. Mol. Genet. Metab. 116, 260–268. https://doi.org/10.1016/j.ymgme.2015.10.005 (2015).

Article CAS PubMed PubMed Central Google Scholar

Williamson, D. H., Bates, M. W., Page, M. A. & Krebs, H. A. Activities of enzymes involved in acetoacetate utilization in adult mammalian tissues. Biochem. J. 121, 41–47. https://doi.org/10.1042/bj1210041 (1971).

Article CAS PubMed PubMed Central Google Scholar

Fukao, T. et al. Enzymes of ketone body utilization in human tissues: Protein and messenger RNA levels of succinyl-coenzyme A (CoA):3-ketoacid CoA transferase and mitochondrial and cytosolic acetoacetyl-CoA thiolases. Pediatr. Res. 42, 498–502. https://doi.org/10.1203/00006450-199710000-00013 (1997).

Article CAS PubMed Google Scholar

Leclerc, J. et al. Metabolism of R-beta-hydroxypentanoate and of beta-ketopentanoate in conscious dogs. Am. J. Physiol. 268, E446-452. https://doi.org/10.1152/ajpendo.1995.268.3.E446 (1995).

Article CAS PubMed Google Scholar

Klepper, J. et al. Seizure control and acceptance of the ketogenic diet in GLUT1 deficiency syndrome: A 2- to 5-year follow-up of 15 children enrolled prospectively. Neuropediatrics 36, 302–308. https://doi.org/10.1055/s-2005-872843 (2005).

Article CAS PubMed Google Scholar

Stewart, W. A., Gordon, K. & Camfield, P. Acute pancreatitis causing death in a child on the ketogenic diet. J. Child Neurol. 16, 682 (2001).

Article CAS PubMed Google Scholar

Berry-Kravis, E., Booth, G., Taylor, A. & Valentino, L. A. Bruising and the ketogenic diet: Evidence for diet-induced changes in platelet function. Ann. Neurol. 49, 98–103 (2001).

3.0.CO;2-2" data-track-action="article reference" href="https://doi.org/10.1002%2F1531-8249%28200101%2949%3A1%3C98%3A%3AAID-ANA13%3E3.0.CO%3B2-2" aria-label="Article reference 40" data-doi="10.1002/1531-8249(200101)49:13.0.CO;2-2">Article CAS PubMed Google Scholar

Kielb, S., Koo, H. P., Bloom, D. A. & Faerber, G. J. Nephrolithiasis associated with the ketogenic diet. J. Urol. 164, 464–466 (2000).

Article CAS PubMed Google Scholar

Best, T. H., Franz, D. N., Gilbert, D. L., Nelson, D. P. & Epstein, M. R. Cardiac complications in pediatric patients on the ketogenic diet. Neurology 54, 2328–2330 (2000).

Article CAS PubMed Google Scholar

Brunengraber, H. & Roe, C. R. Anaplerotic molecules: Current and future. J. Inherit. Metab. Dis. 29, 327–331 (2006).

Article PubMed Google Scholar

Borges, K. in Ketogenic Diet and Metabolic Therapies: Expanded Roles in Health and Disease (ed. Susan Masino) 336–345 (Oxford University Press, 2016).

Wang, D. et al. Glut-1 deficiency syndrome: Clinical, genetic, and therapeutic aspects. Ann. Neurol. 57, 111–118. https://doi.org/10.1002/ana.20331 (2005).

Article CAS PubMed Google Scholar

Klepper, J. et al. Introduction of a ketogenic diet in young infants. J. Inherit. Metab. Dis. 25, 449–460 (2002).

Article CAS PubMed Google Scholar

Settergren, G., Lindblad, B. S. & Persson, B. Cerebral blood flow and exchange of oxygen, glucose, ketone bodies, lactate, pyruvate and amino acids in infants. Acta Paediatr. Scand. 65, 343–353 (1976).

Article CAS PubMed Google Scholar

Download references

We thank the G1D individuals and their families for their participation. We acknowledge the late Dr. Charles R. Roe for discussion of biological mechanisms and for clinical investigative advice. We are grateful to the Glut1 Deficiency Foundation for assistance with individual recruitment. We also thank Drs. Adam Hartman (NIH), Marc C. Patterson, Gabriel Ronen, Salvatore DiMauro, Eric H. Kossoff and Robert Rapaport (Glut1 Deficiency Foundation) for serving in the Data Safety Monitoring Committee of this study. We thank Jim Butterwick at Stepan Lipid Nutrition for a gift of food grade triheptanoin.

Glut1 Deficiency Foundation (to J.M.P.), NIH grants NS094257, NS102588 and NS077015 (to J.M.P.) and Alicia Koplowitz Foundation (Ayuda de Estancia Corta to I.M.).

These authors contributed equally: Adrian Avila, Ignacio Málaga and Deepa Sirsi.

Rare Brain Disorders Program, The University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Mail Code 8813, Dallas, TX, 75390, USA

Adrian Avila, Ignacio Málaga, Sharon Primeaux, Gauri A. Kathote, Vikram Jakkamsetti & Juan M. Pascual

Department of Neurology, The University of Texas Southwestern Medical Center, Dallas, TX, 75390, USA

Adrian Avila, Ignacio Málaga, Deepa Sirsi, Saima Kayani, Sharon Primeaux, Gauri A. Kathote, Vikram Jakkamsetti & Juan M. Pascual

Department of Pediatrics, The University of Texas Southwestern Medical Center, Dallas, TX, 75390, USA

Deepa Sirsi, Saima Kayani & Juan M. Pascual

Department of Pharmacy Practice and Clinical Pharmacology, Experimental Therapeutics Center, Texas Tech University Health Sciences Center, Dallas, TX, 75235, USA

Raja Reddy Kallem & William C. Putnam

Department of Pharmaceutical Science, Texas Tech University Health Sciences Center, Dallas, TX, 75235, USA

William C. Putnam

Department of Pathology, The University of Texas Southwestern Medical Center, Dallas, TX, 75390, USA

Jason Y. Park

Departments of Neurology and Pediatrics, Albert Einstein College of Medicine, Bronx, NY, 10467, USA

Shlomo Shinnar

Department of Physiology, The University of Texas Southwestern Medical Center, Dallas, TX, 75390, USA

Juan M. Pascual

Eugene McDermott Center for Human Growth & Development/Center for Human Genetics, The University of Texas Southwestern Medical Center, Dallas, TX, 75390, USA

Juan M. Pascual

You can also search for this author in PubMed Google Scholar

Conception and design: J.M.P. Acquisition of data: J.M.P. (participant scoring pre, during and post hospitalization), A.A. (participant scoring and procedural coordination pre, during and post hospitalization), S.P. (participant scoring and procedural coordination pre, during and post hospitalization), D.S. (participant exam during hospitalization), S.K. (participant exam during hospitalization), R.R.K. (analytical measurements), T.P. (analytical measurements). Analysis and data interpretation: J.M.P. (all data modalities), A.A. (all data modalities), S.P. (all data modalities), I.M. (all data modalities), D.S. (EEG analysis), R.R.K. (analytical measurements), T.P. (analytical measurements), G.K. (statistical analysis and presentation), V.J. (statistical analysis and presentation). Drafting of the manuscript: A.A., I.M., J.M.P. Revision of the manuscript for intellectual content: All authors. Obtaining funding: J.M.P. Supervision: J.M.P.

Correspondence to Juan M. Pascual.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

Avila, A., Málaga, I., Sirsi, D. et al. Combination of triheptanoin with the ketogenic diet in Glucose transporter type 1 deficiency (G1D). Sci Rep 13, 8951 (2023). https://doi.org/10.1038/s41598-023-36001-x

Download citation

Received: 04 March 2023

Accepted: 27 May 2023

Published: 02 June 2023

DOI: https://doi.org/10.1038/s41598-023-36001-x

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.