CADTH Reimbursement Review

Triheptanoin (Dojolvi)

Sponsor: Ultragenyx Canada Inc.

Therapeutic area: Long-chain fatty acid oxidation disorders

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Clinical Review

Executive Summary

An overview of the submission details for the drug under review is provided in Table 1.

Introduction

Long-chain fatty acid oxidation disorders (LC-FAODs) are a heterogenous group of rare autosomal recessive genetic disorders. The disorders are due to mutations in the genes encoding mitochondrial enzymes involved in the conversion of dietary long-chain fatty acids into energy and are associated with chronic energy deficiency and acute crises of energy production.¹ Patients with LC-FAODs are at risk of metabolic decompensation, particularly during times of physiologic stress or when energy intake is reduced (e.g., fasting, vigorous exercise, illness, vomiting, or surgery).² Several types of LC-FAODs have been identified, including carnitine palmitoyltransferase (CPT I or CPT II) deficiency, very long-chain acyl coenzyme A (acyl-CoA) dehydrogenase (VLCAD) deficiency, long-chain 3-hydroxy-acyl-CoA dehydrogenase (LCHAD) deficiency, trifunctional protein (TFP) deficiency, and carnitine-acylcarnitine translocase (CACT) deficiency. The most common is VLCAD deficiency.³ There is considerable variability in the incidence and prevalence estimates for LC-FAODs. Overall, the estimated collective incidence of LC-FAODs is 1 in 5,000 to 10,000 live births. The population prevalence of LC-FAODs range widely, from 1 in 100,000 to 1 in 2,000,000, depending on the specific type.⁴ The sponsor estimates the incidence of LC-FAODs in Canada to be approximately 10 to 15 births per year. Based on a global prevalence of approximately 1 per 100,000, the sponsor estimates that 380 Canadians have LC-FAODs.⁵

The clinical presentation of LC-FAODs can vary, depending on the specific disorder and age of onset, although there are common elements. Acute manifestations of LC-FAODs can include episodes of hypoketotic hypoglycemia, hyperammonemia, or rhabdomyolysis, induced by fasting, exercise, or illness. Patients also develop cardiomyopathy or hepatic dysfunction, which leads to hospitalizations and premature death.¹ According to the clinical experts consulted by CADTH, LC-FAODs can largely be divided into infantile, pediatric, and adult presentations of the disease. Early in life, manifestations include marked metabolic disturbance, with high mortality and morbidity. Newborn screening for LC-FAODs has contributed to early identification and treatment, which has reduced the burden of this disease. Infants have moderate forms, which often involve hepatic and cardiac disease. Juveniles and adults who are diagnosed with LC-FAODs often exhibit neuromuscular symptoms such as rhabdomyolysis, peripheral neuropathy, and retinopathy, which can range from mild to severe and significantly impact quality of life and physical functioning. The spectrum of disease severity can also be correlated with the type of LC-FAOD. For example, manifestations of LCHAD are mainly severe, whereas half of patients with VLCAD can have asymptomatic or mild presentations, as evidenced by milder variants identified by the newborn screening tests.

According to the clinical experts consulted for this review, standard therapy in Canada is largely supportive and individualized based on the needs of the patient. Therapy is often guided by age at diagnosis, severity of clinical presentation, and type of LC-FAOD. Chronic treatment usually includes dietary modification (i.e., low fat, high carbohydrate), avoidance of prolonged fasting, avoidance of activity requiring high exertion, and supplementation with medium-chain triglyceride (MCT)-based products. Some patients may also receive essential fatty acids or carnitine.

Triheptanoin is an MCT consisting of 3 odd-chain 7-carbon fatty acids (heptanoates). As a source of medium odd-chain fatty acids, triheptanoin bypasses the process requiring the specific enzymes that are deficient in patients with LC-FAODs for the conversion of dietary long-chain fatty acids into energy.⁶ Health Canada reviewed triheptanoin under its Priority Review process and approved it as a source of calories and fatty acids for the treatment of adult and pediatric patients with LC-FAODs. The recommended target daily dosage of triheptanoin is up to 35% of the patient’s total prescribed DCI, divided into at least 4 doses, administered at mealtimes or with snacks, at 3-hour to 4-hour intervals or as directed by the health care provider.⁶

The objective of this CADTH reimbursement review is to perform a systematic review of the beneficial and harmful effects of triheptanoin 100% weight per weight (w/w) oral liquid as a source of calories and fatty acids for the treatment of LC-FAODs in adult and pediatric patients.

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient groups that responded to CADTH’s call for patient input and from clinical expert(s) consulted by CADTH for the purpose of this review.

Patient Input

One patient group, MitoAction (Massachusetts, US), responded to the call for patient input for this CADTH reimbursement review. Input was not received from any Canadian patient group. MitoAction has engaged with the patient community through weekly support calls, Facebook groups, and a Mito411 Support line. It has received direct feedback from the patient community in the US about their experience with triheptanoin.

The patient group emphasized that the energy depletion for patients with LC-FAODs can be debilitating, and patients often cannot participate in normal day-to-day activities. Patients must manage their energy exertion throughout the day, because a simple task can physically overwhelm an individual with an LC-FAOD. Limitations to activity can lead to depression, isolation, and other mental health issues, which are very common in patients with a rare disease. Manifestations of LC-FAOD can also lead to hospitalization and organ damage.

Ideal outcomes for the patient community include increased energy levels, which lead to more physical activity, improved cognitive functioning, decreased stress on organ systems, and reduced hospitalizations. This would provide an enhanced quality of life and independence for patients. MitoAction notes that, with proper treatment and disease management, patients with LC-FAODs can lead full and meaningful lives despite their diagnosis.

Clinician Input

Input From Clinical Experts Consulted by CADTH

The clinical experts consulted by CADTH stated that current treatments may help some patients, but there are patients who still experience recurrence of symptoms despite optimized therapy. There is need for more effective treatment for patients with ongoing symptomatic LC-FAODs, particularly those with severe forms of the disease. Supplementation with even-chain MCT has led to a positive response and reduction of complications in some patients. However, tolerability is an issue (due to gastrointestinal [GI] adverse events [AEs]), which, in turn, affects adherence to the treatment regimen.

The experts indicated that, in general, triheptanoin would be reserved for more severe cases of LC-FAODs, or as second-line therapy after even-chain MCT products had been tried. For most patients, the clinical experts anticipate that triheptanoin would be used when there is inadequate response to optimized dietary measures and conventional even-chain MCT supplementation. Triheptanoin may be used as first-line therapy in selected patients (usually neonates or infants) presenting with acute, life-threatening cardiovascular or metabolic decompensation. If such patients respond to triheptanoin, treatment would be expected to continue upon hospital discharge.

According to the clinical experts, in general, it is appropriate for a patient who starts triheptanoin to receive an adequate trial and be evaluated annually for improvement or maintenance of effect, although initial evaluations may be more frequent (e.g., every 3 or 6 months). The clinical experts emphasized that assessing response to treatment should be individualized. Depending on the age of the patient, type of LC-FAOD, presenting symptoms, and clinical severity, the goals of treatment vary (e.g., address rate of progression of left ventricular [LV] dysfunction, frequency of events such as rhabdomyolysis or hospitalization, length of hospital admissions, recurrent episodes of metabolic decompensation, exercise intolerance, muscle pain with exertion, quality of life). For example, in infants presenting with catastrophic events, survival would be a relevant outcome, and follow-up would be performed frequently. In stable older children and adults, follow-up may be performed every 6 to 12 months. The clinical experts stated that the decision to discontinue treatment is made according to individualized parameters based on the patient’s medical history. If parameters used to measure response in the patient return to pre-treatment levels or gains are not maintained, then triheptanoin treatment should be discontinued at the annual assessment. Treatment should also be discontinued if unacceptable side effects develop.

Clinician Group Input

This section was prepared by CADTH staff based on the input provided by clinician groups.

Input was received from the Garrod Association Guideline Committee on the reimbursement review on triheptanoin. The clinician group noted that, currently, the management of patients with LC-FAODs mainly includes medical nutrition therapy. The group commented that this typically includes the restriction of long-chain fatty acids and supplementation with MCT.

The clinician group noted that patients with severe LC-FAODs have a greater unmet need than those with milder LC-FOADs. The group added that this is because patients with severe LC-FOADs can present with symptoms regardless of good compliance with standard treatment. The committee noted that the drug under review will replace and not complement MCT supplements. They recommended that the 2 supplementations (triheptanoin and MCT) should not be given together, owing to a theoretical concern that MCT oil and triheptanoin compete for enzyme activity. The committee noted that patients with moderate-to-severe LC-FAODs are likely to respond to treatment under review and thus would be best suited for treatment.

The group commented that triheptanoin should be used as first- or second-line treatment, based on the clinical judgment of the treating physician. The clinician group added that patients with mild, asymptomatic LC-FAODs who are diagnosed via newborn screening programs would be least suited for treatment with the drug under review. In addition, the clinician group noted that patients diagnosed with LCHAD and mitochondrial TFP deficiencies are at risk of developing retinopathy and peripheral neuropathy. They added that neither MCT supplementations nor triheptanoin treat these symptoms.

Drug Program Input

Input was obtained from the jurisdictions participating in CADTH reimbursement reviews. The following were identified as key factors that could affect the implementation of recommendations:

• availability of tests to diagnose LC-FAOD

• place of triheptanoin in therapy

• eligibility criteria for treatment with triheptanoin

• assessment criteria for measuring therapeutic response.

The clinical experts consulted by CADTH provided responses, which can be found in the Drug Program Input section.

Clinical Evidence

Pivotal Studies and Protocol-Selected Studies

Description of Studies

A total of 3 sponsor-submitted studies were included in this report. Aside from these sponsor-submitted pivotal studies, none of the other identified citations met the inclusion criteria for the CADTH systematic review. Two of the studies (CL201 and CL202) were funded by the sponsor, whereas the third study (Gillingham et al. [2017]) was conducted by an independent investigator.

Study CL201 (N = 29) was a multi-centre, open-label, single-arm phase II study investigating the efficacy and safety of triheptanoin in adults and children (6 months of age and older) exhibiting serious clinical manifestations of LC-FAODs despite current management. Patients must have had severe LC-FAODs with confirmed diagnosis of CPT II, VLCAD, LCHAD, or TFP deficiency, and must have been on stable treatment (including dietary measures). At the baseline visit, any prior MCT was discontinued and treatment with triheptanoin was initiated (i.e., added to standard therapy). The target dosage of triheptanoin was 25% to 35% of DCI or maximum tolerated dosage, and treatment was continued up to 78 weeks (18 months).

Study CL202 (N = 75) is an ongoing, open-label, extension study investigating the long-term safety and efficacy of triheptanoin in patients older than 6 months of age with LC-FAODs. Eligible patients must have had a confirmed diagnosis of CPT I, CPT II, VLCAD, LCHAD, TFP, or CACT deficiency. The study consists of 3 cohorts: patients who had previously participated in CL201 (CL201 rollover cohort, N = 24), patients who failed conventional therapy and continued to exhibit clinical manifestations of LC-FAOD (triheptanoin-naive cohort, N = 20), and patients who participated in other programs to access triheptanoin, such as investigator-sponsored trials (ISTs) or compassionate use (IST or other cohort, N = 31). All 3 were single-arm cohorts; none included a parallel comparator group. The target dosage of triheptanoin was 25% to 35% of DCI, and treatment was continued up to 5 years (60 months) while enrolled in CL202. Data presented in this report reflect an interim analysis with a cut-off date of June 1, 2018; the mean duration of treatment was 25.92 months overall. The mean duration for each treatment cohort was as follows: 23.01 months for the CL201 rollover cohort (excludes CL201 study duration), 15.68 months for the triheptanoin-naive cohort, and 34.77 months for the IST or other cohort.

The study by Gillingham et al. (2017; N = 32) was a double-blind randomized controlled trial (RCT) that investigated whether triheptanoin has a therapeutic advantage over conventional treatment for LC-FAODs. Before study enrolment, patients must have had at least 1 episode of rhabdomyolysis and must have been on a stable diet that included MCT. Adults and children 7 years of age and older with confirmed diagnosis of CPT II, VLCAD, TFP, or LCHAD were randomized 1:1 to a diet containing triheptanoin or trioctanoin (an even-chain fatty acid triglyceride), with both MCTs dosed at 20% of estimated DCI. Randomization occurred separately at 2 investigative sites and was stratified according to diagnosis (CPT II, VLCAD, or TFP and LCHAD). Baseline assessments were completed at enrolment, and patients were admitted to the research centre for 4 days for outcome measurements. Upon discharge, patients continued treatment with assigned diet and MCT supplementation for 4 months. At the end of 4 months, baseline assessments were repeated.

At baseline, the average age of patients in CL201 and CL202 was younger than that of patients enrolled in the Gillingham et al. (2017) trial. The 2 sponsor-funded trials enrolled mainly pediatric patients (< 18 years); the mean age was 12.06 years (standard deviation [SD] = 13.21) in CL201, and 13.87 years (SD = 13.19) in CL202. The mean age in the Gillingham et al. (2017) study was 24.75 years (SD = 14.3). The most common LC-FAODs diagnosed in the patients enrolled in CL201 and CL202 were VLCAD and LCHAD deficiencies. In the Gillingham et al. (2017) study, a similar number of patients was diagnosed with VLCAD, LCHAD and TFP, or CPT II deficiencies. According to available data (i.e., excluding the IST or other cohort of CL202), the majority of patients enrolled in all 3 studies had received prior treatment with an MCT formulation, and all were being treated with dietary measures. In CL201 and CL202, approximately 65% of patients were receiving carnitine supplementation. Prior to enrolment, patients in the CL201 study had received approximately 17% of DCI as medium-chain fat from MCTs.

In Study CL201, patients were prescribed a mean triheptanoin dosage of 31.20% DCI (SD = 8.88). The mean dosage of triheptanoin consumed was 27.5% (SD = 4.58) of DCI. During the study, there was a 10% DCI increase (from average 17.4% to 27.5%) in the amount of medium-chain fat consumed, compared to the pre-triheptanoin period. In Study CL202, the mean dosage of triheptanoin prescribed was 26.95% of DCI (SD = 7.48). The mean triheptanoin dosage (% DCI) actually consumed was not reported, although, on average, most patients consumed more than 90% of their prescribed dosage. In the Gillingham et al. (2017) study, patients consumed 16.62% (SD = 2.66) and 14.83% (SD = 3.40) of DCI from triheptanoin and trioctanoin, respectively.

Study CL201 did not explicitly identify primary and secondary efficacy end points; rather, the study grouped end points as key or supportive. Numerous key end points were measured for several disease areas. The following clinical outcomes, relevant to this review, were assessed: major clinical events (MCEs; hospitalizations, emergency department [ED] or acute care visits, or emergency interventions for rhabdomyolysis, hypoglycemia, or cardiomyopathy), exercise intolerance (12-minute walk test [12MWT], cycle ergometry), functional disability and cognitive development (Short Form (10) Health Survey [SF-10], Short Form (12) Health Survey [SF-12]), and cardiac function (echocardiogram).

The primary end point in Study CL202 was the annualized rate of MCEs, including rhabdomyolysis, hypoglycemia, and cardiomyopathy. Annualized duration of total MCEs was considered a secondary efficacy end point, as were the annualized event rate and annualized event days (also referred to as annualized duration rate) of each of the MCEs separately (i.e., rhabdomyolysis, hypoglycemia, and cardiomyopathy).

The primary outcomes in the Gillingham et al. (2017) study included changes in total energy expenditure (TEE), cardiac function (as measured by echocardiography), exercise tolerance (measured by treadmill ergometry), and phosphocreatine recovery following acute exercise.

Efficacy Results

Efficacy results are summarized in Table 2 and Table 3. Results for efficacy outcomes identified in the CADTH review protocol are reported; only the efficacy end points and parameters that were deemed to show favourable changes for triheptanoin according to the trial reports and publications have been included in the tables. In addition, none of the results discussed were adjusted for multiplicity. Consequently, designating differences as statistically significant has been avoided. Of note, none of the 3 studies evaluated the following efficacy outcomes that were identified in the CADTH review protocol: survival, symptom relief, reduction in concomitant medications, or productivity.

Major Clinical Events

MCEs were not measured as part of the efficacy analyses in the Gillingham et al. (2017) study. MCEs were defined in both CL201 and CL202 as rhabdomyolysis, hypoglycemia, or cardiac disease events caused by LC-FAODs, or an intercurrent illness complicated by an LC-FAOD, resulting in any hospitalization, ED or acute care visit, or emergency intervention (any unscheduled administration of therapeutics at home or in the clinic). These measures were presented as annualized event rates and event days (also called duration rate) as an aggregate, as well as major rhabdomyolysis events, hypoglycemia events, and events due to decompensation of cardiomyopathy being presented separately. Of note, the majority of MCEs reported in both CL201 and CL202 were rhabdomyolysis events.

Due to the heterogeneity of clinical manifestations of LC-FAODs, both studies used a retrospective control, comparing MCEs before and during triheptanoin treatment. Retrospective data collection was intended to provide a within-subject comparison for MCEs; thus, each patient acted as his or her own control. In Study CL201, medical history from 18 months (78 weeks) before study entry was collected to establish a pre-triheptanoin baseline and was compared to 78 weeks of triheptanoin treatment. In Study CL202, historical medical data were collected for patients in the CL201 rollover and triheptanoin-naive cohorts. Statistical comparisons were made between data collected from 18 months before triheptanoin treatment and the first 36 months (CL201 rollover cohort, including treatment received during CL201) or 18 months (triheptanoin-naive cohort) of study treatment. No statistical comparisons were made for the IST or other cohort in CL202 due to lack of pre-triheptanoin data.

In Study CL201, a reduction in annualized event rates and event days occurred across all 3 clinical manifestations as a result of triheptanoin treatment but was most favourable for the aggregate measure that included all event types. For total MCEs, including all event subtypes, the difference in the mean annualized event rate was 0.81 events per year, and the difference in mean annualized event days was 2.997 in favour of triheptanoin.

In the CL201 rollover cohort of Study CL202, the most notable improvement due to triheptanoin was in the annualized event rate of total MCEs. For this primary efficacy end point, the difference in the mean annualized event rate of total MCEs, including all event subtypes, was 0.80 events per year in favour of triheptanoin treatment. For the remaining annualized event rates and event days (secondary efficacy end points), triheptanoin treatment resulted in a reduction across all comparisons, but none were significant. In the triheptanoin-naive cohort of CL202, a heavily skewed distribution was observed, which limited the interpretation of results; none of the changes observed in MCEs were significant.

To evaluate the effect of triheptanoin on MCEs in different subgroups, several ad hoc analyses were performed. The following 2 relevant subgroups, identified in the CADTH systematic review protocol, were analyzed in both CL201 and CL202: subgroup by age at triheptanoin initiation (< 6 years, ≥ 6 to < 18 years, and ≥ 18 years) and subgroup by subtype of LC-FAOD (LCHAD, VLCAD, CPT II, and TFP deficiency). For subgroup analyses based on age at treatment initiation, results across different age groups in CL201 were generally consistent with those of the overall population. Inconsistent and variable results were observed in CL202. For subgroup analyses based on subtype of LC-FAOD, results across all diagnosis groups in CL201 were consistent with those of the overall population, except for patients with TFP deficiency. For this subtype only, annualized event rate, but not annualized event duration, was reduced. Similarly, for both CL201 rollover and triheptanoin-naive cohorts of Study CL202, all subtypes except for TFP deficiency had results consistent with those of the overall population. The analyses and interpretability of subgroup data are limited by the small sample sizes of individual subgroups and by skewed data in CL202.

Hospitalizations

Hospitalizations were captured as part of the MCEs in CL201 and CL202. Across both studies, most MCEs before and during triheptanoin treatment were hospitalizations due to rhabdomyolysis. Although few events due to cardiomyopathy occurred during the 2 trials, such events are serious and almost all led to hospitalization.

In Study CL201, annualized hospitalization rates and hospitalization days were reduced across all 3 clinical manifestations as a result of triheptanoin treatment but the reduction was most favourable for the aggregate measure that included all event types. For hospitalizations due to total MCEs, including all event subtypes, the difference in the mean annualized event rate was 0.74 hospitalizations per year and the difference in mean annualized event days was 2.92 in favour of triheptanoin.

In the CL201 rollover cohort of Study CL202, the most notable improvement with triheptanoin treatment was in the annualized hospitalization rate of total MCEs. The difference in the mean annualized hospitalization rate of total MCEs, including all event subtypes, was 0.67 events per year in favour of triheptanoin treatment. For the remaining annualized hospitalization rates and hospitalization days due to specific event subtypes, triheptanoin treatment resulted in a reduction across all comparisons, but none were significant. The exception was hospitalization for major rhabdomyolysis events, for which the mean annualized event days appeared to increase with treatment, although median days decreased. This may be due to the highly skewed distribution of annualized event days observed in this cohort. In the triheptanoin-naive cohort of Study CL202, a heavily skewed distribution was observed, which limited the interpretability of results. None of the changes observed in hospitalizations were significant.

In the study by Gillingham et al. (2017), 7 hospitalizations for acute rhabdomyolysis were reported in each treatment group. There was no difference in length of hospital stay.

Emergency Department Usage

ED usage was not measured as part of the efficacy analyses in the Gillingham et al. (2017) study. ED visits were captured as part of the MCEs in CL201 and CL202. Overall, there were very few ED visits before and during triheptanoin treatment, and all ED visits were due to rhabdomyolysis. In Study CL201, there was no meaningful difference in annualized ED visit rates between the pre-triheptanoin and triheptanoin treatment periods. In Study CL202, no statistical analyses were performed to compare ED visits before and during treatment with triheptanoin. Numerically, the CL201 rollover cohort had an increase in ED visits during triheptanoin treatment, whereas the triheptanoin-naive cohort had a decrease in visits. However, the small number of events and lack of statistical testing preclude drawing any definitive conclusions.

Health-Related Quality of Life

Health-related quality of life (HRQoL) was not measured in the Gillingham et al. (2017) study. In studies CL201 and CL202, changes in HRQoL were measured using the Short Form (10) Health Survey (SF-10) in children 5 to 17 years of age, and the Short Form (12) Health Survey version 2 (SF-12v2) in adults 18 years and older. For both assessments, a score of 50 constituted the normalized base score, and each factor of 10 represented 1 SD above or below the mean. Overall, the population included in the assessments of HRQoL was much smaller than the number of patients enrolled in each study or cohort.

In CL201, the main statistical comparison in HRQoL was the change from baseline at week 24. For pediatric patients (SF-10), mean baseline physical summary score (PHS) indicated impairment, whereas the psychosocial summary score (PSS) was similar to that in the general population. At week 24, no notable changes from baseline were observed in PHS or PSS scores. Beyond week 24, the PHS improved over time as a result of treatment, at week 48 and week 78; however, scores remained below the population norm. For adults (SF-12v2), the mean baseline physical component summary (PCS) score was lower than the population mean; the mental component summary (MCS) score was slightly below the norm. At week 24, there was notable improvement in both PCS and MCS scores as a result of treatment. This benefit was maintained through week 78 for the PCS score, but not MCS. Despite improvement, mean PCS scores remained below the population norm.

In CL202, no statistical tests were performed to compare the change in scores over time. Thus, observations can only be made regarding the general trend in scores with treatment in each of the 3 cohorts.

In the CL201 rollover cohort of CL202, SF-10 PHS scores appeared to decline over the 18 months of treatment during CL202; however, scores remained above the baseline level taken before starting triheptanoin in CL201. The SF-10 PSS scores remained generally stable from baseline through the CL202 study; these scores were similar to the population norm. For SF-12v2, PCS scores during the CL202 study were relatively stable and similar to pre-treatment levels. The MCS scores of SF-12 were also relatively stable during the CL202 study, and mean values remained within the population norm.

In the triheptanoin-naive cohort, the baseline mean PHS scores for SF-10 were lower than the population norm, indicating impairment. Scores appeared to improve over time and were similar to the population average while patients were receiving treatment. The mean SF-10 PSS scores were similar to the population norm at baseline and remained within this range throughout the CL202 study. For SF-12v2, changes in HRQoL were difficult to assess due to the small number of patients in each post-baseline assessment.

In the IST or other cohort, scores for both SF-10 and SF-12v2 remained relatively stable throughout the 18 months of treatment in the CL202 study.

Physical Function and Exercise Tolerance

Physical function and exercise tolerance were measured using the 12MWT and cycle ergometry in Study CL201, and treadmill ergometry and phosphocreatine recovery in the study by Gillingham et al. (2017). Study CL202 did not assess physical function or exercise tolerance.

In Study CL201, the primary analysis for the 12MWT was conducted at week 18, and 8 patients performed the 12MWT at all key assessment points. Although results showed overall improvement with triheptanoin treatment in the various parameters, most were not significant. The mean change from baseline was often associated with wide confidence intervals (CIs), reducing the certainty of the results. The only notable improvement in the 12MWT parameters was in the energy expenditure index (EEI) from baseline to week 18, although baseline EEI was already within the normal range, as identified in the study (0.14 beats/m to 0.89 beats/m).

In Study CL201, the primary analysis for the cycle ergometry test was conducted at week 24, and 7 patients performed the cycle ergometry test at both baseline and the later assessment. At week 24, cycle ergometry workload and duration improved overall, although neither were significant.

In the Gillingham et al. (2017) study, all patients completed the treadmill ergometry test to measure exercise tolerance. After 4 months of treatment, the only notable difference between the 2 treatment groups was in maximum heart rate (HR), with a mean difference in change from baseline of 6.98 beats/min (95% CI, 0.34 to 13.63) in favour of triheptanoin. There was no difference between the 2 treatment groups for VO₂ or peak double product (a marker of cardiac workload measured by multiplying systolic blood pressure by HR); systolic blood pressure remained constant throughout the test.

The study by Gillingham et al. (2017) also measured phosphocreatine recovery after a repetitive lower leg exercise to evaluate muscle adenosine triphosphate (ATP) synthesis. This exercise protocol was completed by 8 adults in the triheptanoin group and 7 adults in the trioctanoin group. After 4 months of treatment, there was no difference between the 2 treatment groups in test results.

Cardiac Function Parameters

Cardiac function was measured using echocardiography in all 3 included studies. In Study CL201, an echocardiogram was performed on all patients at baseline and on 35 patients at week 24. At baseline, mean LV ejection fraction (LVEF) was within the normal rage specified in the study (55% to 70%), and no significant change was observed at week 24. In Study CL202, there were no notable changes overall in the echocardiography parameters. In all 3 cohorts, the mean LVEF at baseline was also within the normal range.

In the Gillingham et al. (2017) study, an echocardiogram was assessed in 21 patients (n = 10 triheptanoin, n = 11 trioctanoin). After 4 months of treatment, there was a difference between the 2 treatment groups in change from baseline in mean LV wall mass as well as mean LVEF. For LV wall mass, the difference in relative change from baseline between the 2 treatment groups was 20% in favour of triheptanoin. For LVEF, the difference between triheptanoin and trioctanoin in change from baseline was 7.4% (95% CI, −0.1% to 15%) in favour of triheptanoin. Of note, all patients except for 1 had normal cardiac function at baseline; the majority of the observed changes were within the normal range.

Harms Results

Safety data are summarized in Table 4. All or almost all (98.7%) of patients enrolled in studies CL201 and CL202, respectively, reported at least 1 treatment-emergent adverse event (TEAE). Although the total number of patients who experienced at least 1 TEAE was not reported in the Gillingham et al. (2017) study, it appears that the majority of patients did experience 1 or more TEAEs; there was a similar frequency of various AEs between the triheptanoin and trioctanoin treatment groups. Of note, complications of the underlying LC-FAOD (e.g., rhabdomyolysis) were also captured as an AE in all 3 studies, which likely contributed to the high rates of reported TEAEs. Overall, the reported TEAEs were similar across studies and generally consistent with the known AE profile of triheptanoin or the underlying LC-FAOD. The most commonly reported TEAEs were rhabdomyolysis, GI-related TEAEs (e.g., diarrhea, vomiting, GI upset), or infections (e.g., upper respiratory tract infections, viral illnesses).

Treatment-emergent serious adverse events (SAEs) were reported in 65.5% of patients in CL201 and 76.0% of patients in CL202; these rates included MCEs that were also reported as an SAE. In Study CL202, the proportion of patients who experienced at least 1 SAE during the study was similar across the 3 cohorts. The most common SAEs were related to the underlying LC-FAOD (e.g., rhabdomyolysis) or acute infectious diseases, including GI infections. The study by Gillingham et al. (2017) did not categorize TEAEs by severity or seriousness. In Study CL201, 4 patients discontinued triheptanoin treatment due to TEAEs, most of which were GI-related. Treatment was discontinued due to TEAEs in 1 patient in Study CL202 (non-serious rhabdomyolysis) and in none of the patients in the Gillingham et al. (2017) study. Across the 3 studies, 2 deaths were reported, both in Study CL201; neither was considered to be due to triheptanoin. Although weight gain was identified as a notable harm in the CADTH review protocol, it was not reported as an AE in any of the 3 studies. According to growth measures collected throughout the study, there were no clinically significant changes in z scores for weight (in pediatric patients for CL201 and CL202). In the Gillingham et al. (2017) study, no difference in body composition or weight gain was noted between the 2 treatment groups.

Additional Information

As part of the sponsor’s feedback on this CADTH reimbursement review report, the sponsor sent CADTH a summary of updated analysis for certain outcomes in Study CL202 (Appendix 5). Due to the brief and selective nature of the information provided, CADTH could not use the summary to update all the relevant CL202 interim data and is unable to provide critical appraisal of the updated analysis.⁷

Critical Appraisal

A few major limitations and sources of bias are provided. Further details for each point, as well as a complete list of limitations and sources of bias, are available in the Critical Appraisal section of this report.

• Studies CL201 and CL202 were single-arm, phase II trials that did not include a parallel treatment comparator. Analyses of MCEs were conducted using a before-after design. The MCEs were evaluated before and after initiation of triheptanoin, with each patient serving as his or her own control, using data collected retrospectively from medical records. Due to inherent limitations in the study design (e.g., lack of relevant comparator as a control, no blinding of treatment, potential influence of concurrent therapies, impact of growth and maturation of patients on test performance), results from these 2 trials could be considered supportive but cannot offer solid evidence of treatment benefits. The comparative efficacy of triheptanoin compared to even-chain MCTs was investigated only in the Gillingham et al. (2017) trial.

• The effects of triheptanoin as first-line treatment in patients who have not received any form of prior MCT supplementation require further investigation. The majority of patients (≥ 90%) in studies CL201 and the CL201 rollover and triheptanoin-naive cohorts of CL202 had received prior treatment with an MCT formulation. As per inclusion criteria, all patients enrolled in the Gillingham et al. (2017) study had received prior supplementation with MCT.

• Study results cannot be generalized to patients with CACT or CPT I deficiencies due to low enrolment numbers in Study CL202, and these patients were excluded from the CL201 and Gillingham et al. (2017) trials. Notably, in Canada, the CPT IA P479L variant is prevalent in Indigenous communities (e.g., British Columbia First Nations and Inuit populations) and the CPT IA G710E variant is prevalent in Hutterite communities, but data on the efficacy of triheptanoin in these groups are lacking.^11,12 However, the clinical experts consulted on this review note that patients with these CPT IA variants typically have mild disease or are asymptomatic and generally do not require active treatment with MCTs.

• In all 3 trials, the sample size of each study and treatment group and cohort were small. As a result, differences in 1 or 2 patients can have a substantial impact on results, leading to a high degree of uncertainty due to imprecise estimates. Nevertheless, because of the rarity of this disease, such a small sample size is not unusual.

• None of the 3 trials employed a hierarchical testing procedure or strategy to control for the overall type I error rate; no adjustments were made for multiple testing among any of the outcomes analyzed. Consequently, statistically significant results should be interpreted with consideration of the potential for inflated type I error.

• The evaluation of patient-reported outcomes (e.g., HRQoL), exercise tests that depend on patient effort, or AEs in studies CL201 and CL202 may have been influenced by the unblinded treatment regimens, resulting in reporting bias. Furthermore, an estimated minimally important difference (MID) has not been identified in the LC-FAOD population for SF-10 or SF-12, nor have these tests been validated in patients with LC-FAODs. For these reasons, along with the small sample sizes, the clinical significance of the HRQoL findings is unclear.

• Confounding due to changes in diet and MCT dosage cannot be ruled out. For example, in Study CL201, there was an increase of approximately 10% DCI in the dosage of MCT when patients transitioned from MCT oil to triheptanoin after study enrolment. For studies CL202 (except for the CL201 rollover cohort) and Gillingham et al. (2017), no baseline dietary treatment information, including dosage of prior MCT supplementation, was available.

• The efficacy of triheptanoin on survival, peripheral neuropathy, or retinopathy is unknown, as none of the studies measured these important clinical outcomes. As well, the majority of MCEs documented in studies CL201 and CL202 were due to rhabdomyolysis. The small number of events and patients who had cardiomyopathy or experienced hypoglycemia limits the interpretation of efficacy for MCEs other than rhabdomyolysis.

• The RCT by Gillingham et al. (2017) did not include end points that were deemed important by clinicians and patient groups, including survival, clinical events, symptoms such as fatigue, or HRQoL. Thus, the relative efficacy of triheptanoin compared to even-chain MCTs (i.e., trioctanoin) for these important outcomes is unknown, and available data do not provide evidence to support the use of triheptanoin over trioctanoin to prevent or reduce clinical events.

Indirect Comparisons

No indirect comparisons were submitted by the sponsor or identified in the literature search.

Other Relevant Evidence

No other relevant evidence was included in this review.

Conclusions

LC-FAODs are a complex group of diseases with a wide spectrum of disease manifestations and heterogenous presentations. Based on the 3 sponsor-submitted studies, current evidence does not adequately address the clinical question of whether triheptanoin improves relevant outcomes compared to current standard of care in patients with LC-FAOD who require treatment.

The 2 single-arm, phase II, open-label trials (CL201 and CL202) appear to show a reduction in annualized rate and duration of MCEs, mainly driven by hospitalizations due to rhabdomyolysis, when comparing events that occurred during triheptanoin treatment to a retrospective pre-treatment period. The clinical experts consulted by CADTH noted that these results are clinically meaningful; however, not all patients responded favourably. Due to the significant risk of bias, potential confounding factors, and statistical uncertainty, it cannot be stated with confidence whether any benefits observed in these trials are attributable to triheptanoin treatment.

The double-blind RCT by Gillingham et al. (2017) appears to show some benefit of triheptanoin over trioctanoin, an even-chain MCT, in exercise tolerance (i.e., maximum HR on treadmill ergometry), as well as cardiac parameters (i.e., LVEF and LV mass on echocardiography). However, the relationship between the modest difference in HR and longer-term exercise tolerance is unknown, and the clinical relevance of cardiac parameter findings is difficult to determine, considering that patients had normal cardiac function at baseline. The short duration of treatment (4 months) and the lower dose of triheptanoin than that recommended in the product monograph further add to the limitations in generalizability of the study results to clinical practice. At this time, there is no evidence showing superiority of triheptanoin over other sources of MCTs for the clinically relevant end points of mortality, morbidity such as reduction in clinical events or hospitalization, or HRQoL. As a result, firm conclusions on the clinical benefit of triheptanoin over even-chain MCTs cannot be made.

Because of its odd-carbon-chain structure, triheptanoin is thought to act as an anaplerotic compound that directly addresses the tricarboxylic acid (TCA) cycle deficiencies that arise in LC-FAODs, which are not addressed by even-chain MCTs.¹ However, based on currently available aggregate data, it is unclear whether the advantage at the cellular level translates to definitive clinical benefit. Overall, whether triheptanoin will improve the lives of patients with LC-FAOD, compared to even-chain MCTs, has not been adequately addressed with available data. It is highly uncertain whether triheptanoin is better than the alternative MCT formulations currently available. Evidence gaps also remain for other clinical manifestations of LC-FAODs that have not been investigated; for example, retinopathy or peripheral neuropathy. The clinical experts consulted on this review emphasized the unmet need in previously undiagnosed patients who present with acute, life-threatening cardiovascular or metabolic decompensation. In these crisis situations, the experts anticipate patients will start on triheptanoin as inpatients and continue on treatment upon discharge if they have a good response.

Introduction

Disease Background

Fatty acid oxidation disorders are inborn errors of metabolism caused by the disruption of fatty acid substrate entry into mitochondria or a defect in their beta-oxidation in the mitochondrial matrix. These metabolic disorders are primarily categorized as medium- or long-chain, based on the length of the fatty acid chain, with LC-FAODs affecting metabolism of fatty acids greater than 8 to 12 carbons. LC-FAODs are a heterogenous group of rare, autosomal recessive genetic disorders.¹ Six types of LC-FAODs have been identified: carnitine palmitoyltransferase (CPT I or CPT II) deficiency, very long-chain acyl coenzyme A (acyl-CoA) dehydrogenase (VLCAD) deficiency, long-chain 3-hydroxy-acyl-coA dehydrogenase (LCHAD) deficiency, and trifunctional protein (TFP) deficiency. The most common is VLCAD deficiency.³

The mutations associated with this disorder occur in genes that encode enzymes involved in the carnitine shuttle, which transports long-chain fatty acids into the mitochondria, or in beta-oxidation for conversion of the long-chain fatty acids into energy (Figure 1).³ LC-FAODs are characterized by chronic energy deficiency, with episodes of acute crises of energy production. During times of reduced intake, prolonged fasting, or increased demand for energy due to illness, fatty acids are released from storage in the adipose tissues. Normally, fatty acids are then metabolized in the mitochondria via beta-oxidation and used as a major source of energy for the myocardium, skeletal muscles, and liver.^1,2 With exercise, glycogen from muscle is used as the main source of energy for the initial 20 to 30 minutes; however, prolonged and high-intensity exercise depends on fatty acids for sources of energy. Because patients with LC-FAOD are unable to convert long-chain fatty acids into energy, they are at risk of metabolic decompensation, particularly during times of physiologic stress or when energy intake is reduced (e.g., fasting, vigorous exercise, illness, vomiting, or surgery).² Potentially toxic fatty acid intermediates can also accumulate in the blood and organs.¹

The clinical presentation of LC-FAODs can vary depending on the specific disorder and age of onset, although there are common elements. Acute manifestations of LC-FAODs can include episodes of hypoketotic hypoglycemia, hyperammonemia, or rhabdomyolysis induced by fasting, exercise, or illness. Patients also develop cardiomyopathy or hepatic dysfunction, which leads to hospitalizations and premature death. Hypoglycemia is experienced more often in infants and younger children, whereas myopathy is more frequent in adults or children older than 6 years of age. Peripheral neuropathy and pigmentary retinopathy are unique to LCHAD and TFP deficiencies.¹

According to the clinical experts consulted by CADTH, LC-FAODs can largely be divided into infantile, pediatric, and adult presentations of the disease. Early in life, manifestations include marked metabolic disturbance, with high mortality and morbidity. Newborn screening for LC-FAODs has contributed to early identification and treatment, which has reduced mortality rates. Infants have moderate forms, which are often implicated with hepatic and cardiac disease. Juveniles and adults who are diagnosed with LC-FAODs often exhibit neuromuscular symptoms such as rhabdomyolysis, peripheral neuropathy, and retinopathy, which can range in severity from mild to severe and significantly impact quality of life and physical functioning. The spectrum of disease severity can also be correlated with the type of LC-FAOD. For example, manifestations of LCHAD are mainly severe, whereas half of patients with VLCAD can have asymptomatic or mild presentations, as evidenced by milder variants identified by the newborn screening tests.

Before the introduction of newborn screening, mortality was as high as 60% to 90%. Newborn screening for LC-FAODs and early intervention have reduced mortality rates; however, patients continue to experience morbidity due to episodes of metabolic decompensation. These recurrent symptoms and hospitalizations have a negative impact on the quality of life of patients and caregivers.¹

There is considerable variability in the incidence and prevalence estimates for LC-FAODs. Overall, the estimated collective incidence of FAODs is 1 in 5,000 to 10,000 live births. The prevalence of FAODs ranges widely, from 1 in 100,000 to 1 in 2,000,000, depending on the specific type.⁴ Although detailed epidemiology of this rare disorder across Canada is limited, some estimates can be derived from provincial data. For example, based on the 2019 report from Newborn Screening Ontario, the following disease prevalence was reported for individual deficiencies: 1 in 323,173 for CPT I, 1 in 387,808 for CPT II, 1 in 161,587 for LCHAD, and 1 in 71,816 for VLCAD.¹³ The sponsor estimates the incidence of LC-FAODs in Canada to be approximately 10 to 15 births per year. Based on a global prevalence of approximately 1 per 100,000, the sponsor estimates that 380 Canadians have LC-FAODs.⁵ FAODs occur in both males and females and all ethnic populations.⁴

Standards of Therapy

The current standard therapy for LC-FAODs mainly involves nutritional and symptomatic management. With nutritional management, the primary goal is to limit the use of long-chain fat as a substrate for energy production. Part of this is achieved by restricting the amount of dietary long-chain fatty acids. However, an adequate amount for normal growth and development is still required. Additionally, patients are counselled to avoid fasting to prevent beta-oxidation and catabolism of long-chain fatty acids into required energy. Further dietary adjustments, such as increasing carbohydrate intake, may also necessary during times of strenuous exercise or illness. Supplementation with carnitine or essential fatty acids (e.g., docosahexaenoic acid) may also be required.¹

Patients are also given supplementation with MCTs as a source of fatty acids and calories.^1,14 Other than triheptanoin, there are no therapies for treatment of LC-FAODs approved in Canada. Regulation of even-chain MCT supplements in Canada falls under the Natural and Non-prescription Health Products Directorate (NNHPD).¹⁵ As such products, even-chain MCTs can be obtained without a prescription but are reimbursed in some jurisdictions under specialized publicly funded drug programs.¹⁶ The commercially available non-prescription MCT formulations consist of a mix of even-chain fatty acids, usually 8-carbon (octanoate), 10-carbon (decanoate), and 12-carbon (dodecanoate) fatty acids, and the proportions of fatty acids may vary from lot to lot. MCTs are used as supplementation in patients with LC-FAODs, as they pass through a different metabolic pathway than long-chain fatty acids. For example, MCTs enter the mitochondria directly and do not depend on the carnitine-based transport system. Furthermore, MCTs are metabolized by medium-chain-specific enzymes and bypass the need for long-chain-specific transport and enzymes.

Despite treatment, patients with all LC-FAODs still experience episodes of hypoglycemia, cardiomyopathy, and rhabdomyolysis, and those with LCHAD and TFP deficiencies experience peripheral neuropathy and pigmentary retinopathy as well. Patients also continue to experience recurring hospitalizations as well as high rates of morbidity and mortality.¹

Input From Clinical Experts Consulted by CADTH

Standard Therapy in Canada

The overall goals of treatment include prevention of acute metabolic or cardiac decompensations and prevention of long-term complications and disabilities associated with the course of disease. Specific goals can vary with each individual and may include different areas of focus, such as preventing episodes of rhabdomyolysis or hypoglycemia, reducing hospital admissions or duration of hospitalization, slowing progression of cardiomyopathy, improving HRQoL, or increasing measures of physical function, either in terms of activities of daily living or the ability to participate in activities and sports.

Standard treatment is largely supportive and is individualized based on the needs of the patient, often guided by age of diagnosis, severity of clinical presentation, and type of LC-FAOD. Asymptomatic adults generally do not automatically receive treatment; however, asymptomatic early-onset forms (i.e., upon diagnosis on newborn screening) are likely to be treated with standard measures. Chronic treatment usually includes dietary modification (i.e., low fat, high carbohydrate), avoidance of prolonged fasting, avoidance of activity requiring high exertion, and supplementation with medium chain–based products. Some patients may also receive essential fatty acids or carnitine.

Drug

Triheptanoin is an MCT consisting of 3 odd-chain 7-carbon fatty acids (heptanoates). As a source of medium odd-chain fatty acids, triheptanoin bypasses the process requiring the specific enzymes that are deficient in patients with LC-FAODs for the conversion of dietary long-chain fatty acids into energy.⁶ Triheptanoin is available as an oral liquid containing 100% w/w of triheptanoin as an active ingredient. Each mL of triheptanoin oral liquid provides 8.3 kcal.⁶

As an odd-chain MCT, triheptanoin is thought to differ from even-chain MCTs in the way it undergoes oxidation (Figure 1). The benefits of odd-carbon triheptanoin are thought to be due to its anaplerotic properties (production of metabolic intermediates), which resupply TCA cycle intermediates. Upon oral administration, triheptanoin is extensively hydrolyzed in the intestines by pancreatic lipases to heptanoate and glycerol. Once in the mitochondria, heptanoate is metabolized by small- and medium-chain beta-oxidation enzymes. By undergoing 1 or 2 cycles of beta-oxidation, the heptanoate produces 2 units of 2-carbon acetyl coenzyme A (CoA) and 1 unit of 3-carbon propionyl-CoA product, or 1 unit of acetyl-CoA and 1 unit of 5-carbon pentanoyl-CoA products.³ Acetyl-CoA is used as a substrate in the TCA cycle as well as for ketone synthesis. Ketones serve as an alternate form of energy for the brain, myocardium, muscle, kidney, and other tissues, and are important when glucose is unavailable during fasting or physiologic stress.¹ Propionyl-CoA is further metabolized, ultimately to succinyl-CoA and succinate, which resupplies the TCA cycle intermediates to increase ATP production and supports continued gluconeogenesis, as well as mitochondrial energy production. In the liver, pentanoyl-CoA also serves as an anaplerotic substrate to support generation of ketones, which can be used by peripheral tissues.³ Thus, in addition to bypassing the deficient enzymes, triheptanoin is thought to act as an anaplerotic compound to directly address the TCA cycle deficiencies in LC-FAODs, which are not addressed by administration of even-chain MCTs.¹ Unlike triheptanoin, even-carbon fatty acids are thought to have limited effectiveness due to the depletion of odd-chain TCA cycle intermediates.³

Health Canada reviewed triheptanoin under its Priority Review process. A Notice of Compliance was issued on February 15, 2020, which approved triheptanoin as a source of calories and fatty acids for the treatment of adult and pediatric patients with LC-FAODs. The requested reimbursement criteria from Ultragenyx Canada Inc., the sponsor, aligns with the Health Canada indication. On June 30, 2020, the US FDA approved triheptanoin for the treatment of pediatric and adult patients with molecularly confirmed LC-FAODs.

Figure 1: Mechanism of Action of Odd-Carbon (C7) or Even-Carbon (C8) MCTs

CACT = carnitine-acylcarnitine translocase; CoA = coenzyme A; CPT-1 = carnitine palmitoyltransferase 1; CPT-2 = carnitine palmitoyltransferase 2; LCFAO = long-chain fatty acid oxidation; MCFAO = medium-chain fatty acid oxidation; VLCAD = very long-chain acyl-CoA dehydrogenase; TFP = trifunctional protein.

Note: From original figure: “Model for the proposed benefit of triheptanoin (C7) compared to trioctanoate (C8) among patients with LC-FAODs. Trioctanoate provides three 8-carbon fatty acids (C8) that, once imported into the mitochondria, are oxidized to 4 acetyl-CoA molecules. Triheptanoate provides three 7-carbon fatty acids (C7) that, once imported into the mitochondria, are oxidized to produce 2 acetyl-CoA and 1 propionyl-CoA molecule. Propionyl-CoA is converted to D-methylmalonyl-CoA by mitochondrial propionyl-CoA carboxylase, followed by conversion to succinyl-CoA by D-methylmalonyl-CoA isomerase and L-methylmalonyl-CoA mutase. Succinyl-CoA is an intermediate of the citric acid cycle (CAC) and can increase intermediate pool size of carbon substrates.” Diagram is from the Gillingham et al. (2017) study, in which trioctanoin was used as the source of even-carbon (chain) MCT. The red and blue enzymes represent the enzymes affected by LC-FAOD; the red enzymes are the LC-FAOD types (CPT II, VLCAD, TFP and LCHAD deficiency) included in the trial.

Source: Triheptanoin versus trioctanoin for long-chain fatty acid oxidation disorders: a double blinded, randomized controlled trial. Gillingham MB, Heitner SB, Martin J, et al. J Inherit Metab Dis. 2017;40(6):831-843. © John Wiley and Sons 2017. Reprinted with permission.

Stakeholder Perspectives

Patient Group Input

This section was prepared by CADTH staff based on the input provided by patient groups.

About the Patient Groups and Information Gathered

One patient group, MitoAction (Massachusetts, US), responded to the call for patient input for this CADTH reimbursement review. Input was not received from any Canadian patient group(s).

MitoAction is a non-profit organization founded by patients, parents, and Boston hospital health care leaders who had a vision of improving quality of life for children and adults with mitochondrial disease. MitoAction’s mission is to improve the quality of life for children, adults, and families living with mitochondrial disease through support, education, outreach, advocacy, and clinical research initiatives and by granting wishes for children affected by mitochondrial disease.

MitoAction has engaged with the patient community through weekly support calls, Facebook groups, and a Mito411 Support line. It has received direct feedback from the patient community in the US about their positive experience with triheptanoin (Dojolvi). Since the US approval of triheptanoin, MitoAction has interacted with dozens of adult patients and parents who have shared their feedback.

Disease Experience

MitoAction provided information that patients with LC-FAODs have trouble breaking down fat to produce usable energy. Symptoms of LC-FAODs include lethargy, irritability, noticeably enlarged liver, abnormal heart rhythms, cardiac failure, cardiopulmonary failure, poor muscle tone, and periodic severe muscle pain caused by rhabdomyolysis. Patients must be on a strict diet to manage fat intake and energy reserves. They often need to take breaks when performing simple activities and take naps during the day. Therefore, patients are often unable to participate in normal day-to-day activities, as this becomes too draining and causes extreme exhaustion, which can lead to hospitalization and damage to their organs. As energy levels become depleted, organ function can become significantly impacted and severe muscle weakness can occur; this is known as a “mito crash.” Patients must manage their energy exertion throughout the day, so a simple task can physically overwhelm an individual with LC-FAOD. Limitations to activity can lead to depression, isolation, and other mental health issues, which are very common in patients with a rare disease. LC-FAODs are a progressive disease. Hopefully, with proper treatment and disease management, patients can lead full and meaningful lives, despite their diagnosis.

Experience With Treatment

The patient group indicated that, before triheptanoin was available, patients’ only option was the use of over-the-counter MCT oils. In the US, MCT oils are not regulated by the FDA, and the dosing and quality vary among manufacturers. Compliance is difficult to manage. These products can also be very costly for families.

A high school student shared that she was able to enjoy her senior-year activities because of using triheptanoin. Prior to treatment, she would not have been able to do such activities, but now she even looks forward to not using her wheelchair and walking at her graduation. Parents of patients have shared that, after treatment with triheptanoin, their children are able to participate in extracurricular activities and may not require as much rest time during the day. Patients with LC-FAODs have to be very cognizant of their energy usage, often taking naps throughout the day.

The patient group noted that triheptanoin provides a treatment that allows patients to properly break down fats and therefore produce energy to function. In addition to increased energy levels, patients reported decreased hospitalization as compared to when they were not taking triheptanoin. Patients shared that, since rhabdomyolysis is associated with elevated creatine kinase (CK) levels, while on triheptanoin, their CK levels have not been elevated to critical levels. Another patient shared that, even after the first treatment with triheptanoin, there was a “calm of the storm.” Her health had been severely declining and she was using a wheelchair, fully dependent upon a caregiver, with significant muscle weakness and stress on the heart. Within 3 weeks of taking triheptanoin, she began to have muscle twitches, indicating that her muscles were “coming back to life.” She was able to build upon this returning function and, within 2.5 months, was able to stand on her own. Within 6 months she was able to walk outside her house with a walker. The severe muscle decline had caused significant pain and nausea, and she indicated that, for the first time in a long time, she was able to tolerate food and actually felt hungry. The impact of treatment was life-altering.

MitoAction stated that, with the availability of triheptanoin, patients can now be managed by their clinician using a regulated therapy with controlled dosing and product quality. This treatment is readily accessible in the US with a prescription. Under financial assistance programs, triheptanoin is accessible to every patient in the US who may benefit from the therapy.

Triheptanoin was the first therapy for patients affected by LC-FAODs approved by the FDA. According to MitoAction, this was monumental for this rare disease community, and triheptanoin truly has shown a tremendous impact in the quality of life for patients with LC-FAODs.

Improved Outcomes

The patient group emphasized that the energy depletion for patients with LC-FAODs can be debilitating. The level of exhaustion is almost incomprehensible for someone who has never experienced this level of fatigue. A good analogy is to consider running a house on 1 AA battery that never charges beyond 20%. The disease also has a devastating effect on organs and body functions.

Ideal outcomes for the patient community include increased energy levels, leading to boosted physical activity, improved cognitive functioning, decreased stress on organ systems, and reduced hospitalizations. These outcomes would provide an enhanced quality of life and independence for patients.

Clinician Input

Input From Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise regarding the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, and providing guidance on the potential place in therapy). In addition, as part of the triheptanoin review, a panel of 5 clinical experts from across Canada was convened to characterize unmet therapeutic needs, assist in identifying and communicating situations where there are gaps in the evidence that could be addressed through the collection of additional data, promote the early identification of potential implementation challenges, gain further insight into the clinical management of patients living with a condition, and explore the potential place in therapy of the drug (e.g., potential reimbursement conditions). A summary of this panel discussion is presented.

Unmet Needs

Current treatments may help some patients, but there are patients who still experience recurrence of symptoms despite optimized therapy. For example, musculoskeletal manifestations (e.g., rhabdomyolysis) are often poorly controlled in adults with severe disease. There is a need for more effective treatment for patients with ongoing symptomatic LC-FAODs, particularly those with severe forms of the disease. While metabolic markers of disease and genotype and phenotype correlation may predict some patients with severe forms of LC-FAODs, the correlation is not perfect. Furthermore, currently available therapies do not effectively treat retinopathy or peripheral neuropathy associated with LC-FAODs.

Supplementation with even-chain MCT has led to positive response and reduction of complications in some patients. However, tolerability (i.e., GI AEs) affects adherence to the treatment regimen.

Place in Therapy

In general, triheptanoin would be reserved for more severe cases of LC-FAODs or used as second-line therapy after even-chain MCT products had been tried. For most patients, the panellists anticipate that triheptanoin would be used when there is inadequate response to conventional even-chain MCT supplementation. Triheptanoin is much more costly than even-chain MCT products, and certain patients will still respond to the latter. All patients should be on, and adherent to, appropriate dietary management. Triheptanoin should be considered as first-line treatment in selected patients presenting with life-threatening symptoms of LC-FAODs. These occur most often in infants but are not limited to infants.

Patient Population

It is important for patients to have a confirmed diagnosis that is based on clinical, biochemical, or molecular parameters. The patient population eligible for triheptanoin treatment would be those with an inadequate response despite receiving optimized therapy with dietary measures and even-chain MCT products. It is expected that patients with chronic symptoms would try even-chain MCT products before switching to triheptanoin.

An exception would be previously undiagnosed patients (usually neonates or infants) presenting with acute, life-threatening cardiovascular or metabolic decompensation. In these crisis situations, patients may be started on triheptanoin as an inpatient. If patients respond to triheptanoin, treatment would be expected to continue upon discharge, without requiring a trial of even-chain MCT products.

Assessing Response to Treatment

Depending on the presenting symptoms, goals of treatment vary (e.g., rate of progression of LV dysfunction, frequency of events such as rhabdomyolysis or hospitalization, length of hospital admissions, recurrent episodes of metabolic decompensation, exercise intolerance, muscle pain with exertion, quality of life). Often, the age of the patient, type of LC-FAOD, and clinical severity also influence the goals of treatment. Thus, the definition and assessment of response must be individualized based on the patient’s history. For example, in infants presenting with catastrophic events, survival would be a relevant outcome, and follow-up would be frequent. In older children and adults with stable symptoms, follow-up may be performed every 6 months to 12 months.

Time to response to triheptanoin treatment and assessment intervals will vary according to the indication. Several indications, such as recurrent rhabdomyolysis and episodes of metabolic decompensation (i.e., hypoglycemia, hyperammonemia) can be assessed annually, with the goal of reducing the frequency or duration of hospitalization from clinical events. For chronic progressive cardiomyopathy, assessments (i.e., echocardiograms) can be performed every 6 months to determine whether progression is slowing. A deterioration on at least 2 consecutive echocardiograms may indicate a nonresponse. Other symptoms may show improvement sooner, and therefore can be assessed several months after triheptanoin initiation. As an example, in patients with myopathy and/or exercise intolerance, dynamic testing such as the 12MWT should be performed at baseline and 6 months, with the goal of improvement being similar to those reported in clinical trials. For patients with muscle pain upon exertion, improvement can occur within 1 month to 2 months, and gains can be assessed through HRQoL measures, work or school productivity, participation in exercise, or need for analgesics.

Not all patients will respond to triheptanoin treatment. In general, it is appropriate for a patient who starts triheptanoin to receive an adequate trial and be evaluated annually for improvement or maintenance of effect.

Discontinuing Treatment

Patients who respond to triheptanoin are expected to continue treatment indefinitely, as long as it is tolerated. Treatment may be discontinued according to individualized parameters based on the patient’s medical history. As previously discussed,, the goals of treatment for each patient vary, depending on the presenting symptoms and indication. If parameters used to measure response in the patient return to pre-treatment levels, or gains are not maintained, then triheptanoin treatment should be discontinued at the annual assessment.

Lack of adherence to therapy or assessments may also be an indication to discontinue triheptanoin treatment. After attempts to mitigate the adverse effects and improve tolerance to triheptanoin (e.g., reduce dosage or alter administration schedule), if nonadherence continues, physicians must assess whether continuing triheptanoin is warranted.

Prescribing Conditions

The diagnosis and treatment of LC-FAODs, as well as prescription of triheptanoin for the treatment of LC-FAODs, should be restricted to qualified specialists in pediatric or adult inherited metabolic diseases. Such physicians may also be affiliated with different specialties; for example, medical genetics, biochemical genetics, pediatrics, and endocrinology.

In most cases, triheptanoin would be started in an outpatient specialty clinic. In some cases, for example, in a previously undiagnosed patient presenting to a hospital with acute metabolic or cardiac decompensation, triheptanoin would be started on an inpatient basis by a metabolic disease specialist; upon the patient’s discharge, the treatment would be continued under the guidance of a specialty clinic.

Additional Considerations

The panellists noted that patients with LC-FAODs present with varying symptoms that range in severity. Thus, it is important for metabolic specialists to identify patients with phenotypes that are severe and can benefit from triheptanoin treatment. The notion of an “n of 1” trial was also emphasized; when there is an absolute indication for triheptanoin, the treatment plan must be individualized. Furthermore, panellists highlighted the importance of assessing patients at regular intervals to ensure benefit is maintained and for continuity of treatment.

Clinician Group Input

This section was prepared by CADTH staff based on the input provided by clinician groups.

One clinician group input was received from the Canadian Association of Centres for the Management of Hereditary Metabolic Disorders or the Association canadienne des centres de traitement pour les maladies métaboliques héréditaires, also known as the Garrod Association, on the review of triheptanoin (Dojolvi) for LC-FAODs. The Garrod Association is a national body formed to help coordinate the management of inherited metabolic disorders in Canada.

The Garrod Association Guideline Committee formulates clinical guidelines, endorses existing guidelines, and writes statements on specific inherited metabolic diseases for approval by the Garrod Executive. The committee may also assemble task forces to review specific diseases and formulate recommendations for review by the committee. The Garrod Association Guideline Committee is composed of 6 to 10 members of the Garrod Association with interest, experience, and expertise in clinical practice, laboratory practice, and/or research involving metabolic disorders. Different perspectives are provided through members having diverse job profiles, geographic locations, and career stages; information has been compiled into a resource for continuous updating or for publication. Currently, the committee includes 6 clinicians, a biochemical geneticist, and 2 methodologists.

For this reimbursement review, the Garrod Association provided a statement based on clinical judgment and expertise; it did not conduct a formal review of the evidence. Therefore, the clinical group notes that only the clinician members of the Guideline Committee, with input from additional metabolic physicians from across Canada with interest and experience in LC-FAODs, participated in preparing the input document. Information was gathered based on both clinician experience and literature review. Keywords used for the literature search include “fatty acid oxidation disorders, treatment, dietary treatment, management guidelines, clinical trials, triheptanoin.” A draft prepared by 1 clinician was circulated to the other members for their comments and feedback, followed by approval from the president of the Garrod Association.

Unmet Needs

The clinician group noted that, currently, treatment available for the management of patients with LC-FAODs mainly includes medical nutrition therapy. The group commented that this therapy typically involves the restriction of long-chain fatty acids (LCFs) and supplementation with MCTs. The group noted that MCT supplementation is administered as a powdered infant formula high in MCTs and low in long-chain triglycerides (e.g., Lipistart, Portagen, Monogen) or MCT Procal, MCT oil, and Liquigen in older children. In addition, the Guideline Committee noted that treatment is tailored to a specific condition, age, and disease severity. Limiting fasting and emergency treatment regimens are the cornerstones of therapy.

The Guideline Committee also noted that treatment is directed to reduce cardiac complications and rhabdomyolysis episodes as well as to prevent episodes of acute metabolic decompensation. The group added that the restriction of LCFs and supplementation with MCT has been shown to reverse cardiomyopathy and reduce disease complications.

As noted by the clinician group, some currently available therapies for LC-FAODs include the use of L-carnitine; however, the group added that this use is controversial. The clinician group commented that triheptanoin was available only in clinical trials until recently and that, in Canada, it was given to patients with severe disease as compassionate treatment. In this case, the group noted that the treatment was targeting improvement in symptoms to reduce the frequency and the severity of acute disease manifestations, such as hyperammonemia. Docosahexaenoic acid (DHA) supplements and antioxidants were considered helpful for retinopathy in cases of LCHAD and TFP deficiencies. The group noted that no other medications are commonly used.^1,2,17-20

The Guideline Committee noted that the goals of treatment include early detection through newborn screening and early initiation of treatment to reduce morbidity and mortality of patients with LC-FAODs. The clinician group added that treatment aims to improve or prevent disease complications, including cardiac disease, liver disease, skeletal myopathy, hypoglycemia, retinopathy, and rhabdomyolysis. The clinician group also commented that treatment is directed to prevent or reduce the frequency of acute metabolic decompensation episodes (including hypoketotic hypoglycemia, liver dysfunction, hyperammonemia, rhabdomyolysis, arrhythmias, and cardiomyopathy). It noted that treatment with triheptanoin is expected to decrease disease-related complications, improve quality of life, and prolong the life expectancy of patients with LC-FAODs.

The clinician group noted that the following needs are not being met by currently available treatment: not all patients respond to available treatments; better-tolerated treatments are needed; patients become refractory to current treatment options; and treatments are needed to improve compliance.^2,21,22

The clinician group noted that patients with severe LC-FAODs have greater unmet needs than those with milder LC-FOADs. The group added that this is because patients with severe LC-FOADs can present with symptoms regardless of good compliance with standard treatments. These symptoms include repeated episodes of acute metabolic decompensation, episodes of rhabdomyolysis, and worsening of cardiopathy, leading to recurrent hospital admissions, reduced quality of life, and shortened life expectancy.² The Guideline Committee added that the drug under review would address some of these unmet needs.

Place in Therapy

The Garrod Association Guideline Committee noted that the drug under review will replace and not complement MCT supplements. They recommended that the 2 supplementations (triheptanoin and MCT) should not be given together, owing to a theoretical concern that MCT oil and triheptanoin compete for enzyme activity. The clinician group also noted that the drug under review is not the first treatment approved to address the underlying disease process; however, it might add a treatment benefit, as triheptanoin has 7-carbon fatty acids that are metabolized to both acetyl-CoA and propionyl-CoA. The group noted that this provides an anaplerotic effect to replenish deficient TCA cycle metabolites, an effect that is not provided by standard MCT-based supplementation.²³

In addition, the group noted that the mildest cases of LC-FAODs do not require specific dietary therapy and that moderate-to-severe cases of LC-FAODs require dietary modifications and supplementation with MCT or triheptanoin. The group commented that, in a case of neonate or an infant, a formula enriched in MCT and restricted in long-chain triglycerides may be the first-line treatment. In older children and adults, the group noted that stand-alone MCT-containing products (oil or powder) or triheptanoin can be used as a first-line treatment along with dietary modifications. The group commented that triheptanoin may also be considered a first-line treatment in patients presenting with cardiomyopathy. Therefore, taking into consideration these points, the clinician group noted that they anticipate that the drug under review may cause a shift in the current treatment paradigm.

With respect to whether it would be appropriate to recommend that patients try other treatments before initiating treatment with the drug under review, the clinician group noted that, in neonates and infants, it is appropriate to recommend starting treatment with MCT-enriched, LCF-restricted infant formulas. A large portion of the diet for neonates and infants is based on infant formula that supplies a complete diet, including protein, carbohydrates, vitamins and minerals, trace elements, and essential fatty acids. The group commented that, after weaning infant formulas, the patients might be switched to triheptanoin instead of MCT supplements. The clinician group added that, in cases of predicted severe disease and based on clinical judgment of the treating physician, triheptanoin could be tried as first-line treatment in neonates and infants if the risk of trying standard MCT first is too great. Regardless of first-line or second-line use, the clinician and metabolic dietician will need to work together to come up with an individualized prescribed infant formula to provide all needed nutrients during treatment with triheptanoin.

With respect to sequencing of therapies, the clinical group noted that, when treatment with MCT supplementation fails, treatment with triheptanoin should be tried and vice versa. The group added that infants on complete MCT formulas might be switched to triheptanoin after weaning from infant formula feeding. The group commented that this is a significant departure from the sequence employed in current practice, as there are currently no other treatment options.

Patient Population

The Garrod Association Guideline Committee noted that patients with moderate-to-severe LC-FAODs are likely to respond to triheptanoin and are thus best suited for treatment. The group added that patients with moderate-to-severe disease with partial or poor response to standard treatment with MCT supplementation are most in need of the drug under review and that these patients tent to be on the severe end of the disease spectrum. The group added that patients with mild forms of the disease are usually not treated with diet modifications, i.e., MCT supplements. These patients would not need to be treated with triheptanoin.

The clinician group noted that patients best suited for treatment with triheptanoin will be identified by the following criteria:

• laboratory tests, including glucose, CK, aspartate transaminase, alanine transaminase, and acylcarnitine profile

• genetic testing

• enzymatic testing (when available and indicated)

• cardiac echocardiogram and electrocardiogram

• number of metabolic decompensations and hospital admissions for acute metabolic decompensation

• clinical judgment of a physician experienced in metabolic diseases.

With respect to diagnosis, the clinician group noted that it is not challenging to diagnose LC-FAODs. This group of conditions is diagnosed in routine clinical practice via provincial newborn screen programs and, in rare cases where the newborn screen was not available (for example, in the case of new immigrants), it is diagnosed based on clinical symptoms and suggestive biochemical markers. In addition, the group noted that late-onset presentation of LC-FAODs is diagnosed based on symptoms in patients who were born before initiation of newborn screening.

The clinician group added that there are no major issues related to diagnosis and that biochemical testing, including acylcarnitine profile and urine organic acids, is widely available. The group also noted that molecular testing is available via provincial newborn screening programs or diagnostic laboratories (in or outside of the province). Additionally, the group commented that specific enzyme activity testing based on the specific LC-FAODs diagnosed is available as an out-of-country test, entailing additional costs, but could be available in Canada soon. All of these tests are currently performed in clinical practice.

The clinician group noted that underdiagnosis of LC-FAODs is unlikely, as most patients are diagnosed via newborn screening programs in Canada. However, there might be underdiagnosis in the adult population born before 2007 to 2010 (when LC-FAODs newborn screening was started in most provinces in Canada) presenting with late-onset disease. The group noted that pre-symptomatic patients with LC-FAODs are currently treated with MCT supplementation if biochemical markers (acylcarnitine profile, enzyme activity), molecular results, or cardiac function evaluation (echocardiography, electrocardiography) suggest more significant disease.

The group commented that triheptanoin should be used as first- or second-line treatment based on the clinical judgment of the treating physician. The clinician group added that patients with mild, asymptomatic disease who are diagnosed via newborn screening programs would be least suited for treatment with triheptanoin.

With respect to identifying the patients who are most likely to exhibit a response to treatment, the clinician group noted that most patients will be identified based on clinical presentation. The group added that patients with severe disease might present in the first few weeks of life with cardiomyopathy, arrhythmias, hypoglycemia, and liver disease. The group noted that these patients might present before the availability of newborn screening results. They added that differentiation between mild disease variants and classical disease is often possible through biochemical and/or genetic testing. The group noted that patients with moderate disease who do not respond optimally to standard treatment with MCT supplementation would also likely exhibit a response to treatment with triheptanoin. Finally, as noted by the clinician group, patients are identified and referred to metabolic treating centres via provincial newborn screen programs or, among adults, following symptomatic presentation.

Assessing Response to Treatment

The clinician group noted that the following outcomes are used to determine whether a patient is responding to treatment:

• Clinical outcomes: number of acute metabolic decompensation events (requiring hospitalization and ED visits) and reduction in chronic symptoms such as myalgia, myopathy, reduced exercise endurance

• Patient-reported improvement in quality of life

• Improvement or stability in biochemical marker (glucose, CK, aspartate transaminase, alanine transaminase, and acylcarnitine) profile and imaging studies (echocardiogram)

The group added that the outcomes used in clinical practice are aligned with the outcomes typically used in clinical trials.^10,24

The group also noted that a clinically meaningful response to treatment will be considered as follows:

• Stability or improvement in periodic echocardiograms

• Stability or improvement in laboratory markers, including glucose, CK, aspartate transaminase, alanine transaminase, and acylcarnitine profile

• Stability or improvement in reported clinical symptoms, including myalgia, myopathy, exercise endurance

• Reduction in acute metabolic decompensation events leading to hospital admissions

According to the clinician group, treatment response should be assessed based on a few parameters, similar to current disease monitoring for patients receiving MCT supplementation. These include clinical monitoring and biochemical markers (glucose, CK, aspartate transaminase, alanine transaminase, and acylcarnitine profile) conducted every few weeks to every few months, depending on the type of LC-FAOD, disease severity, and patient age, as well as annual cardiac echocardiography and electrocardiography, although the frequency might change based on disease severity.

Discontinuing Treatment

The clinician group noted the following factors that should be considered when deciding to stop treatment:

• Clinical and/or biochemical deterioration while on the treatment; increased number of hospital admissions with acute metabolic decompensation episodes; worsening episodes of hypoketotic hypoglycemia and liver disease (as noted by documented hypoglycemia and/or worsening in aspartate transaminase and alanine transaminase profiles); rhabdomyolysis (worsening in CK levels); or worsening cardiac disease, as noted in repeated echocardiograms

• Significant adverse reactions, hypersensitivity, or intolerance to the medication

• Nonadherence with the medication

Prescribing Conditions

With respect to the settings most appropriate for treatment with triheptanoin, the clinician group noted that a hospital outpatient clinic is the most appropriate setting for monitoring. They added that patients will be evaluated every few months (depending on disease type, severity, and patient’s age) by a metabolic physician and metabolic dietician. In addition, the group noted that a clinical biochemical geneticist and a metabolic physician or a clinical geneticist are required to diagnose, treat, and monitor patients who might receive triheptanoin.

Additional Considerations

The Garrod Association Guideline Committee noted that patients diagnosed with LCHAD and TFP deficiencies are at risk of developing retinopathy and peripheral neuropathy. They added that neither MCT supplementation nor triheptanoin treat these symptoms. They commented that docosahexaenoic acid (DHA) supplementation and antioxidants can be used to treat retinopathy. However, there is no specific treatment available for peripheral neuropathy.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review by identifying issues that may affect their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 5.

Clinical Evidence

The clinical evidence included in the review of triheptanoin is presented in 3 sections. The first section, the Systematic Review, includes pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those studies that were selected according to an a priori protocol. The second section includes indirect evidence selected from the literature that met the selection criteria specified in the review. The third section includes additional relevant studies that were considered to address important gaps in the evidence included in the systematic review. Of note, no indirect evidence or other additional relevant studies were identified in the literature search.

Systematic Review of Pivotal and Protocol-Selected Studies

Objectives

To perform a systematic review of the beneficial and harmful effects of triheptanoin 100% w/w oral liquid as a source of calories and fatty acids for the treatment of LC-FAODs in adult and pediatric patients.

Methods

Studies selected for inclusion in the systematic review included pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those meeting the selection criteria presented in Table 6. Outcomes included in the CADTH review protocol reflected outcomes considered to be important to patients, clinicians, and drug plans.

The literature search for clinical studies was performed by an information specialist using a peer-reviewed search strategy according to the PRESS Peer Review of Electronic Search Strategies checklist.²⁵

Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946–) via Ovid and Embase (1974–) via Ovid. All Ovid searches were run simultaneously as a multi-file search. Duplicates were removed using Ovid deduplication for multi-file searches, followed by manual deduplication in Endnote. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concept was Dojolvi (triheptanoin). Clinical trials registries were searched: the US National Institutes of Health’s clinicaltrials.gov, WHO International Clinical Trials Registry Platform (ICTRP) search portal, Health Canada’s Clinical Trials Database, and the European Union Clinical Trials Register.

No filters were applied to limit the retrieval by study type. Retrieval was not limited by publication date or by language. Conference abstracts were excluded from the search results. Refer to Appendix 1 for the detailed search strategies.

The initial search was completed on April 28, 2021. Regular alerts updated the search until the meeting of the CADTH Canadian Drug Expert Committee on August 18, 2021.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool For Searching Health-Related Grey Literature tool.²⁶ Included in this search were the websites of regulatory agencies (US FDA and European Medicines Agency). Google was used to search for additional internet-based materials. Refer to Appendix 1 for more information on the grey literature search strategy.

These searches were supplemented by reviewing bibliographies of key papers and through contacts with appropriate experts. In addition, the manufacturer of the drug was contacted for information regarding unpublished studies.

Two CADTH clinical reviewers independently selected studies for inclusion in the review based on titles and abstracts, according to the predetermined protocol. Full-text articles of all citations considered potentially relevant by at least 1 reviewer were acquired. Reviewers independently made the final selection of studies included in the review, and differences were resolved through discussion.

A focused literature search for network meta-analyses regarding “fatty acid oxidation” was run in MEDLINE All (1946–) on April 27, 2021. No limits were applied.

Findings From the Literature

A total of 5 publication from 3 studies were identified from the literature for inclusion in the systematic review (Figure 2). None of the studies met the CADTH systematic review criteria; included studies are pivotal trials submitted by the sponsor. The included studies are summarized in Table 5. A list of excluded studies is presented in Appendix 1.

Figure 2: Flow Diagram for Inclusion and Exclusion of Studies

Description of Studies

Three sponsor-submitted trials (CL201, CL202, and Gillingham et al. [2017]) were included in this review. Details of all 3 trials are provided in Table 7.

Study CL201 (N = 29) was a multi-centre, prospective, open-label, single-arm phase II study investigating the efficacy and safety of triheptanoin in adults and children (6 months of age and older) exhibiting serious clinical manifestations of LC-FAODs despite current management. The primary objective of the study was to evaluate the effects of triheptanoin on acute clinical pathophysiology (including exercise tolerance and HRQoL) associated with chronic energy deficiency due to LC-FAODs following 24 weeks of treatment. Additionally, the objective following 78 weeks of treatment was to evaluate the impact of triheptanoin on MCEs (i.e., metabolic crises). CL201 was conducted between February 4, 2014, and August 25, 2016, at 10 investigative sites across the US and UK. The study consisted of a 4-week run-in, 24-week treatment, and 54-week treatment extension period (Figure 3). Inclusion criteria for each patient was assessed based on available medical history of MCEs during the prior 24 months. As well, a medical history from the prior 18 months (78 weeks) was collected to establish a pre-triheptanoin comparison. In this study, assessments collected during the run-in period and from historical data were also used as comparators for the treatment phase. Due to the heterogeneity of clinical manifestations of LC-FAODs, the retrospective data collection was intended to provide a within-subject comparison for MCEs; thus, each patient acted as his or her own control for the comparison of 78 weeks of conventional management (pre-triheptanoin) to the 78 weeks of triheptanoin treatment. During the 4-week run-in phase, enrolled patients continued current management to establish a stable baseline. Thereafter, at the baseline visit, any use of MCT was discontinued and treatment with triheptanoin was initiated (i.e., added to standard therapy). Patients received up to a total of 78 weeks (18 months) of treatment with triheptanoin, which included the 24-week treatment and an additional 54-week treatment extension period if indicated by the investigator. After initiation of treatment, patients were followed every 4 to 6 weeks during the 24-week treatment period and every 12 weeks to 18 weeks during the extension period to assess long-term safety and efficacy. For patients who did not continue on a separate extension protocol, a follow-up visit occurred 4 weeks after the week 78 visit.

Additional dietary assessments were performed for patients enrolled in CL201; data were obtained from each patient’s medical record. Although dietary data were not recorded and collected during the entire retrospective period (78 weeks before triheptanoin start), they were captured at run-in (4 weeks before baseline) as well as at baseline and were considered representative of the entire retrospective period.

Figure 3: Study Design for CL201

MCT = medium-chain triglyceride; HRQoL = health-related quality of life.

Source: CL201 Clinical Study Report.⁸

Study CL202 (N = 75) is an ongoing open-label extension study investigating the long-term safety and efficacy of triheptanoin in patients older than 6 months of age with LC-FAODs. The study comprises 3 cohorts: patients who had previously participated in CL201 (CL201 rollover cohort, N = 24), patients who failed conventional therapy and continued to exhibit clinical manifestations of LC-FAODs (triheptanoin-naive cohort, N = 20), and patients who participated in other programs to access triheptanoin, such as ISTs or compassionate use (IST or other cohort, N = 31). All 3 cohorts were single-arm; none included a parallel comparator group. The first patient initiated the study on December 9, 2014, and the study is being conducted at 10 investigative centres across the US and UK. The study consists of a baseline visit, then a treatment period of up to 5 years (60 months); for patients in the CL201 rollover cohort, the baseline visit was the final week 78 visit. In patients for whom historical medical data had already been collected as part of a prior triheptanoin treatment protocol, the effect of triheptanoin on MCEs during CL202 was compared with the available historical information. Data from 18 months before initiation of triheptanoin were considered the retrospective period, and patients acted as their own internal controls for efficacy assessment during the CL202 treatment period. Although the CL201 rollover cohort and triheptanoin-naive cohorts had retrospective periods, patients in the IST or other cohort had limited historical information on medical management before triheptanoin. As a consequence, data while on treatment with triheptanoin were not compared to a retrospective period in the IST or other cohort. Efficacy and safety of triheptanoin were reported across 3 study periods in CL202 (Figure 4). Study period 1 (main study period) started on day 1 and continued until the last visit day in CL202; study period 2 had the same ending as study period 1 but started from day 1 of triheptanoin treatment and was used in selected analyses for CL201 rollover patients. Study period 3 started from 18 months preceding start of triheptanoin and ended 1 day before day 1 of triheptanoin treatment; this was also referred to as the retrospective period for MCE data collection and was used for selected analyses of CL201 rollover and triheptanoin-naive patients. Data presented in this report reflect an interim analysis with a cut-off date of June 1, 2018. After initiation of treatment, an initial assessment via phone call was conducted at week 2. Patients were then assessed at the clinic at 6-month intervals from month 6 to month 60. Between clinic visits, phone assessments were also made at 6-month intervals between month 3 and 57 to assess for MCEs, AEs, SAEs, and changes to concomitant medications. After study treatment was discontinued, a safety follow-up call was conducted 30 to 35 days after the last dose of the study drug.

Figure 4: Study Design and Cohorts for CL202

IST = investigator-sponsored trial.

Source: CL202 Clinical Study Report.⁹

The study by Gillingham et al. (2017, n = 32) was a double-blind RCT that investigated whether triheptanoin has a therapeutic advantage over conventional treatment for LC-FAODs. Adults and children 7 years of age and older were randomized 1:1, using a permuted block randomization scheme, to a diet containing triheptanoin or trioctanoin (an even-chain fatty acid triglyceride). Patient enrolment occurred during November 2011 and February 2015; randomization occurred separately at 2 investigative sites in the US (Oregon Health and Science University and the University of Pittsburgh) and was stratified according to diagnosis (CPT II, VLCAD, or TFP and LCHAD). The randomization table was generated by a study statistician, and all study personnel were blinded to treatment assignment, except for the research pharmacy dispensing the study oil, the kitchen preparing meals, and the primary study coordinator, until data collection was complete. Baseline assessments of TEE, response to a moderate-intensity treadmill exercise test, cardiac function, and LCF acid oxidation capacity were completed at enrolment. Patients were admitted to the research centre for 4 days for outcome measurements. Upon discharge, patients continued treatment with assigned diet and MCT supplementation for 4 months. At the end of 4 months, baseline assessments were repeated (Figure 5).

Figure 5: Study Design for Gillingham et al. (2017)

Source: Triheptanoin versus trioctanoin for long-chain fatty acid oxidation disorders: a double blinded, randomized controlled trial. Gillingham MB, Heitner SB, Martin J, et al. J Inherit Metab Dis. 2017;40(6):831-843. © John Wiley and Sons 2017. Reprinted with permission.

Both CL201 and CL202 were sponsored by Ultragenyx Pharmaceutical Inc., whereas the study by Gillingham et al. (2017) was conducted by an independent investigator and funded through a grant from the Orphan Drug Development Program by the US FDA.

Populations

Inclusion and Exclusion Criteria

To be eligible for Study CL201, patients had to be at least 6 months of age and have a confirmed diagnosis of CPT II, VLCAD, LCHAD, or TFP deficiency; patients diagnosed with CPT I or CACT deficiency were excluded. Patients must also have had severe LC-FAOD and must have been receiving stable treatment (i.e., for the past 60 days), which may have included dietary measures and/or MCT. Severe LC-FAOD was defined as having 1 of the following significant clinical manifestations despite management: chronic elevated CK with MCEs, episodic elevated CK with reported muscle dysfunction, highly elevated CK but asymptomatic, frequent severe major medical episodes, severe susceptibility to hypoglycemia, or evidence of functional cardiomyopathy (detailed definitions found in footnote of Table 7).

To be eligible for Study CL202, patients were also required to be at least 6 months of age; however, the trial enrolled patients with CACT or CPT I deficiency in addition to CPT II, VLCAD, LCHAD, and TFP deficiencies. Because this was an extension study, patients must have participated in a clinical study assessing triheptanoin treatment for LC-FAODs as part of clinical programs with the 2 study sites (Oregon Health and Science University or the University of Pittsburgh) and/or study sponsor (e.g., CL201). The study also enrolled, at the discretion of the sponsor, patients who had received triheptanoin through other clinical studies, ISTs, or expanded access or compassionate use programs, as well as treatment-naive patients who had failed conventional treatment and had documented severe unmet need.

In the Gillingham et al. (2017) study, age of eligibility was 7 years or older to ensure the patient was able to comply with the protocol, such as completing a 45-minute exercise treadmill test. Each individual was evaluated to determine whether they could complete the study protocol. Patients must have had a confirmed diagnosis of CPT II, VLCAD, LCHAD, or TFP deficiency. Similar to CL201, patients with CPT I or CACT deficiencies were not included in the study. To be eligible for participation, patients must also have had at least an episode of rhabdomyolysis and be on a stable diet that included MCT. Patients with anemia, history of myocardial infarction, or peripheral neuropathy that limited ability to complete treadmill studies were excluded from the trial.

Baseline Characteristics

There were notable differences in baseline characteristics across the trials, particularly between the 2 sponsor-funded trials and the Gillingham et al. (2017) study, as shown in Table 8. In general, patients in CL201 and CL202 were younger than those in the Gillingham et al. (2017) trial, and the former trials involved mainly pediatric patients (< 18 years); the mean age was 12.06 years (SD = 13.21) in CL201, 13.87 (SD = 13.19) in CL202, and 24.75 (SD = 14.3) in the Gillingham et al. (2017) study. Half of the patients in CL201 were children younger than 6 years of age. In CL202, most patients were younger than 18 years of age, although the median age in the IST or other cohort (14.41 years) was older than that in the other 2 cohorts. More male patients (approximately 60%) were enrolled in CL201 and CL202, with the exception of the IST or other cohort of CL202, in which approximately 60% were female. Approximately 60% of patients enrolled in the Gillingham et al. (2017) study were female. Most patients enrolled in CL201 and CL202 were White; no data on race or ethnicity were reported in the Gillingham et al. (2017) study.

In CL201 and CL202, most patients were diagnosed with an LC-FAOD during infancy (often < 3 months of age); a few patients were diagnosed in adulthood, with the exception of the triheptanoin-naive cohort in CL202, in which none were diagnosed as adults. Common methods of diagnoses included acylcarnitine profile, mutation analysis, and fatty acid oxidation probe studies. The most common LC-FAOD subtypes that patients had in CL201 and CL202 were VLCHAD deficiency and LCHAD deficiency. In the Gillingham et al. (2017) study, LCHAD deficiency and TFP deficiency were combined; thus, the exact breakdown of patients with either specific subtype is unknown. Otherwise, there was a similar number of patients diagnosed with VLCAD, LCHAD and TFP, or CPT II deficiencies. Most patients enrolled in CL201 and CL202 had experienced common manifestations of LC-FAODs, including rhabdomyolysis, muscle pain, exercise intolerance, hypoglycemia, muscle weakness, and cardiomyopathy.

Compared to studies CL201 and CL202, there was little information available on the medical history of patients enrolled in the Gillingham et al. (2017) study. According to the limited data, the mean age of patients in the triheptanoin treatment group (27.2 years, SD 15.9) was slightly older than patients in the trioctanoin group (22.3 years, SD 12.7). Gender and type of LC-FAOD was generally well-balanced between the 2 treatment groups.

According to available data (i.e., excluding the IST or other cohort of CL202), the majority of patients enrolled in all 3 studies had received prior treatment with an MCT formulation, and all were being treated with dietary measures. In CL201 and CL202, approximately 65% of patients were receiving carnitine supplementation. The type of MCT formulation varied; examples provided in the CL201 Clinical Study Report include Enfaport, Lipistart, Liquigen, MCT Procal, Monogen, Portagen, and Pregestimil Lipil. Before enrolment, patients in the CL201 study had received approximately 17% of DCI as medium-chain fat from MCTs. A detailed breakdown of dietary management before triheptanoin treatment was available only for patients enrolled in CL201 (Table 9).

Interventions

In all 3 studies, any prior MCT product taken by patients was discontinued before starting study treatment. Because CL201 and CL202 were single-arm trials, all patients received triheptanoin treatment, as outlined in Table 10. In the Gillingham et al. (2017) study, patients were randomized in a 1:1 ratio to 1 of 2 treatment arms. Study treatment included either triheptanoin (C7, odd-chain MCT) or trioctanoin (C8, even-chain MCT). In all 3 studies, triheptanoin was provided as an oil. In the Gillingham et al. (2017) study, trioctanoin was synthesized in a comparable manner to eliminate fatty acid variation in the composition of commercial even-chain MCT products.

Outcomes

A list of efficacy end points identified in the CADTH review protocol that were assessed in the clinical trials included in this review is provided in Table 11. These end points are further summarized. A detailed discussion and critical appraisal of the outcome measures are provided in Appendix 4.

Of note, the CL201 study did not explicitly identify primary and secondary efficacy end points. Rather, the study categorized end points as either key or supportive. Key efficacy end points demonstrated efficacy of the triheptanoin, and supportive efficacy end points provided supplemental information on the efficacy evaluation of triheptanoin.

Evaluation Time Points for Study CL201

In CL201, data analyses were conducted in 2 stages to assess the effects of triheptanoin treatment in patients with LC-FAODs. The first analysis, conducted after the 24-week (6-month) treatment period, evaluated the effects of triheptanoin on energy physiology. Specifically, clinical and biologic disease during the 24 weeks of treatment was assessed in 3 relevant disease areas: skeletal myopathy (e.g., hypotonia, exercise intolerance, motor development, functional disability, non-acute CK levels not associated with rhabdomyolysis), hepatic disease (e.g., interventions and complications related to hypoglycemia, hepatomegaly, markers of hepatic function), and cardiac disease (e.g., LVEF, LV shortening fraction, cardiac biomarkers). Comparisons were made primarily against the baseline 4-week run-in period for each patient. The second analysis, which evaluated the effects of triheptanoin treatment on MCEs associated with LC-FAODs, such as hospitalization, medical interventions, and related clinical manifestations, was conducted after the total 78-week (18-month) treatment extension period. Comparisons were made against the historical data from 18 months to 24 months before study entry (or from birth, for patients less than 18 months of age).

Clinical Events

In both CL201 and CL202, major events were defined as musculoskeletal (rhabdomyolysis), hepatic (hypoglycemia), and cardiac disease (cardiomyopathy) events caused by LC-FAODs, or an intercurrent illness complicated by LC-FAODs, resulting in any hospitalization, ED or acute care visit, or emergency intervention (any unscheduled administration of therapeutics at home or in the clinic). In both studies, major rhabdomyolysis events, hypoglycemia events, and events due to decompensation of cardiomyopathy were measured separately, with the number and duration of events captured for each. Although various biomarkers, clinical evaluations, and laboratory parameters were measured to characterize these clinical events, an exact definition for each event was not provided in the studies, except for clinically important hypoglycemia, which was defined in CL201 as serum glucose levels less than 60 mg/dL (3.3 mmol/L). The total number and duration of each major event (ED or acute care visits, hospitalizations, or emergency intervention) were recorded. Data for the 3 MCEs were presented as annualized event rates and annualized event days (also referred to as annualized duration rate). In Study CL202, the annualized event rate and annualized event days were also calculated for MCEs overall. The number of MCEs was annualized for each patient by dividing the total number of MCEs of interest by the total duration of data collection divided by 365.25. The duration of MCEs was also annualized by dividing the total duration (days) of MCEs of interest by the duration of data collection divided by 365.25.

In Study CL201, the key efficacy end points included the annualized event rate, annualized event days, total number of events, and total duration (days) for each of the clinical event types (rhabdomyolysis, hypoglycemia, or cardiac events) resulting in hospitalization, ED visit, or emergency intervention. Of note, the end point measuring annualized event rate and annualized event days for total MCEs (i.e., for the combination of rhabdomyolysis, hypoglycemia, and cardiomyopathy events) to evaluate the aggregate effect of triheptanoin on metabolic crises was not pre-specified and was added after the final database lock. The MCEs captured over 78 weeks of triheptanoin treatment were compared with MCEs those during 78 weeks before triheptanoin initiation, collected through a retrospective review of medical records.

In Study CL202, the primary efficacy end point was the annualized LC-FAOD major event rate, inclusive of skeletal myopathy (rhabdomyolysis), hepatic (hypoglycemia), and cardiomyopathy events. Annualized duration of all MCEs was considered a secondary efficacy end point, as were the annualized event rate and annualized event days of each of the MCEs (rhabdomyolysis, hypoglycemia, and cardiomyopathy). Two cohorts (CL201 rollover and triheptanoin-naive) out of the 3 in CL202 had retrospective data on MCEs collected from medical records, which were compared to MCE data collected during triheptanoin treatment. For patients in the CL201 rollover cohort, pre-triheptanoin data captured from retrospective medical record reviews during CL201 were used for comparison; MCEs were collected during CL202 using the same methodology as CL201. Retrospective data were compared to MCEs collected over 36 months of triheptanoin treatment in the CL201 rollover cohort and over 18 months in the triheptanoin-naive cohort.

In both CL201 and CL202, all MCEs were identified, recorded, and analyzed in the same manner, both before and during triheptanoin treatment. To reduce potential differences in MCE identification before or during treatment with triheptanoin, each patient was assessed by the same investigator. For collection of retrospective data in CL201 and 2 cohorts in CL202 (i.e., CL201 rollover and triheptanoin-naive), at the screening visit, MCEs that occurred 18 to 24 months before study initiation were identified by investigators through a review of medical records of patients at their clinical sites. The events were source-verified from medical charts using discharge summaries, history and physical examinations, and progress notes. For patients younger than 18 months of age at the time of study drug initiation, retrospective data collection included all MCEs reported from the date of birth until the day before initiation of triheptanoin. During triheptanoin treatment, the same events were captured through reporting by subjects or caregivers at clinic visits and telephone calls; data were confirmed with review of medical records and source document verification in the same manner as the pre-triheptanoin period. Data on the event type, the number of days in hospital or the intensive care unit, and the type and number of days of treatment and intervention were recorded. MCEs that occurred at the same time or during the same hospitalization were counted separately.

The trial by Gillingham et al. (2017) did not investigate MCEs as a part of the study’s outcomes.

Hospitalizations Due to Clinical Events

Details on hospitalizations due to an MCE were captured as part of the composite measure of MCEs, as previously described in studies CL201 and CL202. The Gillingham et al. (2017) study reported the number of hospitalizations due to acute rhabdomyolysis; no other hospitalization event was captured in the study.

Emergency Department Utilization

Details on ED visits due to an MCE were captured as part of the composite measure of MCEs, as previously described in studies CL201 and CL202. The Gillingham et al. (2017) study did not capture ED usage as part of the study’s outcomes.

Health-Related Quality of Life

In both CL201 and CL202, HRQoL was measured as part of the assessment for functional disability. Age-appropriate questionnaires (SF-10 or SF-12) were used to measure HRQoL, and the change from baseline in the scores measured by these instruments were key efficacy end points in CL201 and exploratory end points in CL202. In CL201, the surveys were administered at baseline and weeks 12, 24, 48, and 78 (or early termination) visits. The primary statistical comparison was the change from baseline at week 24. In CL202, the surveys were administered at baseline and then at 6-month intervals thereafter (or the early termination visit). Refer to Appendix 4 for further description and appraisal of SF-10 and SF-12. The changes from baseline in t scores were measured at these time points. HRQoL was not measured in the Gillingham et al. (2017) study.

Short Form (10) Health Survey

The SF-10 is a 10-item questionnaire completed by caregivers. It is a generic measure of HRQoL and is designed for healthy and ill adults. In the studies, the 10 items used a 4-week recall period. Responses were used to generate 2 component summary scores: PHS and PSS. Higher global scores are associated with better quality of life. The 2 summary measures were calculated from raw scores to have a mean of 50 and SD of 10. In the 2 pivotal studies,^8,9 the t score-based scoring was used to score the SF-10 summary scales. The scale scores were centred, so that a score of 50 corresponds to the average score in a comprehensive 2006 US sample (a combination of general population and supplemental disability and chronic condition samples). According to both CL201 and CL202 study protocols, the SF-10 was administered to children 6 years through 17 years at the time of informed consent. However, during the study, the survey was conducted and assessed in some younger children (5 years through 17 years). According to the study report, the instrument is valid for this age range; however, information on validity of SF-10 in patients with LC-FAODs was not identified in the CADTH literature search. An estimated MID was not identified from the literature for patients with LC-FAODs.

Short Form (12) Health Survey Version 2

The SF-12 version 2 is a 12-item interview and self-administered questionnaire. It is a generic measure of HRQoL designed for healthy and ill adults. The standard version with a 4-week recall period was used in the studies. The 12 items in the SF-12 are a subset of the items in the SF-36 and measure the following 8 domains: physical functioning, role limitations due to physical health problems, bodily pain, general health, vitality (energy and fatigue), social functioning, role limitations due to emotional problems, and mental health. Responses were used to generate 2 component summary scores for physical health (PCS) and mental health (MCS). The raw scores range from 0 to 100, with higher global scores associated with better health and quality of life. Domain and component summary scores were calculated from raw scores to have a mean of 50 and SD of 10. Scoring the SF-12 version 2 was accomplished using t score-based scoring, which is standardized using the means and SDs from the 2009 US general population. Thus, a score of 50 constitutes the normalized base score, while each factor of 10 (± 5) represents an SD above or below the mean (i.e., of the US general population t score). The SF-12 was administered to patients 18 years of age and older at the time of informed consent. An estimated MID was not identified from the literature for patients with LC-FAODs.

Measures of Physical Function or Exercise Tolerance

12-Minute Walk Test

In Study CL201, the 12MWT was a key efficacy end point. In CL202, the 12MWT efficacy assessment was originally included as part of the study’s exploratory end points. However, this was removed from the protocol in an amendment (protocol amendment 4) to reduce the burden on patients and investigational sites, with the rationale that continuing to use the 12MWT assessment in the open-label, single-arm extension was unlikely to yield informative data. The Gillingham et al. (2017) study did not conduct a 12MWT as part of the outcome measures.

In Study CL201, parameters of muscle function were evaluated using the 12MWT, which is a variation of the 6-minute walk test (6MWT) that is used to assess endurance through walking. The longer 12MWT was selected based on a hypothesis that longer walking times would put greater demands on the muscle, by stressing the physiologic deficit caused by LC-FAODs. Thus, the impact and potential treatment response on muscle function in patients with LC-FAODs may be better distinguished with the 12MWT than the 6MWT. The 12MWT was administered to patients 6 years of age or older who were able to safely perform the assessment. In children who turned 6 years old during the study or achieved mastery of all Peabody Developmental Motor Scales, 2nd edition (PDMS-2) skill sets measuring gross motor development, the 12MWT could be administered at the next scheduled visit. Patients who met the safety criteria were instructed to walk the length of a pre-measured 20 m course in a hallway for 12 consecutive minutes. The lap distance was standardized using traffic cones, and patients walked around the cones until the end of the time period. Instructions and encouragement were given according to American Thoracic Society guidelines. The distance walked at the end of 12 minutes was recorded in metres, along with the distance walked after the first and second 6 minutes. The percentage of predicted normal values was calculated for the 6MWT distance using age-appropriate reference data, and the patient’s HR and blood pressure were checked before and after the test. The efficacy assessment following the 12MWT also included additional measures of perceived exertion (before and after the 12MWT) using the OMNI Scale as well as the perceived muscle pain (before and after the 12MWT) using a visual analogue scale (patients ≥ 18 years) or Faces Pain Scale – revised (patients < 18 years). To measure walking efficiency, the EEI (the ratio of HR per metre walked, in beats/m) was also calculated for the 12MWT.

The test was administered to patients during the run-in period (week 4) and at weeks 8, 18, 36, and 60 (or at the early termination visit). Approximately 2 hours before the test, patients were fed a standardized macronutrient meal including either MCT oil (at the run-in visit, if applicable) or triheptanoin (all visits post-baseline). The test was administered by a trained clinician according to American Thoracic Society guidelines established for the 6MWT. The 12MWT was not performed if there were any concerns with the patient completing the test reliably and safely. For example, the test was not performed if significant muscle pain was reported before starting the test, and the test could be discontinued at any time if there were any concerns about induction of a major safety event, such as rhabdomyolysis.

The CL201 statistical plan noted that the impact of 24 weeks of triheptanoin treatment on muscle function would be evaluated by comparing EEI and distance travelled during the 12MWT at weeks 8 and 18 to baseline (last assessment during the run-in period). A similar analysis was conducted to evaluate changes from baseline at weeks 36 and 60. According to the Clinical Study Report, the primary analysis for the 12MWT was assessed at week 18. As muscle function increases, EEI during the 12MWT is expected to decrease, whereas the distance travelled is expected to increase. An estimated MID was not identified from the literature for patients with LC-FAODs.

Cycle Ergometry

In CL201, cycle ergometry was a key efficacy end point. In the CL202 and Gillingham et al. (2017) studies, cycle ergometry was not included as part of the efficacy assessment.

In CL201, exercise tolerance was assessed using cycle ergometry, a submaximal exercise test. The aerobic exercise testing to measure workload was performed in all patients aged 6 years or older at the screening visit. Patients who turned 6 years old within 6 months of the baseline visit could also be considered for the cycle ergometry test at screening. If valid testing was feasible, additional testing at baseline and weeks 4, 12, 24, 48, and 78 (or at the early termination visit) was performed. The main end point of the cycle ergometry test was energy expenditure. Gas exchange variables, respiratory exchange rate (RER), blood pressure, HR, pain and exertion, blood lactate, acylcarnitine, and CK levels were also measured. A 40-minute cycle ergometry protocol at 60% of age-predicted maximal HR was performed in patients who were able to safely perform the test. To stress the physiologic deficit in LC-FAODs, a multi-modal approach (HR, perceived exertion, and perceived pain) was used to target a zone of exercise intensity in which fat would be the preferential energy source, if available. Workload at a fixed HR, duration of test protocol, and RER (an indicator of whether carbohydrates or fats are being consumed as fuel) were evaluated. The workload performed at the baseline visit was repeated for all subsequent cycle ergometer tests. Approximately 2 hours before the test, patients were fed a standardized macronutrient meal, including either MCT oil (at screening and baseline, if applicable) or triheptanoin (all visits after baseline). To standardize the administration of the test, cycle ergometry was administered in an exercise testing laboratory by trained clinicians. The cycle ergometry test was not performed if there were any concerns with the patient completing the test reliably and safely. For example, the test was not performed if significant muscle pain was reported before starting the test, and the test could be discontinued at any time if there were any concerns about induction of a major safety event, such as rhabdomyolysis. Cycle ergometry was performed at separate time points from the 12MWT to ensure patients had time to recover.

In CL201, the impact of 24 weeks of triheptanoin treatment on exercise intolerance was evaluated by comparing the total workload, time-adjusted area under the curve for RER, and actual duration of exercise during cycle ergometry at weeks 4, 12, and 24 to baseline. The primary statistical comparison was between baseline and week 24 for all 3 parameters. A similar analysis was conducted to evaluate changes from baseline at week 78. Total workload and duration of exercise during cycle ergometry are expected to increase as intolerance to exercise decreases.

Treadmill Ergometry

In the Gillingham et al. (2017) study, exercise tolerance was considered 1 of the primary outcomes. It was measured using a 45-minute moderate-intensity treadmill test with continuous ECG monitoring and collection of respiratory gases. In the CL201 and CL202 studies, treadmill ergometry was not included as part of the efficacy assessment.

Treadmill ergometry in the Gillingham et al. (2017) study was performed 2 hours after the patient finished a standardized lunch. At baseline, all patients received trioctanoin in their lunch and again as an oral bolus 20 minutes before the exercise test. At 4 months, patients repeated the treadmill ergometry and received either triheptanoin or trioctanoin before exercise based upon treatment randomization. The bolus dose was 0.3 g oil/kg lean body mass.

During the test at baseline, the rate and incline of the treadmill were increased until the patient’s HR was 50% to 60% of his or her predicted maximum. The same grade, speed, and duration were repeated at the follow-up visit to keep the workload (and therefore oxygen consumption) constant. Perceived physical exertion using the Borg perceived exertion scale (a scale of 1 to 20 of increasing exertion), blood pressure, HR, and ventilation at a given workload were measured during the treadmill exercise. Levels of blood lactate and CK concentrations before and after exercise were also measured as indicators of rhabdomyolysis. The exercise test was terminated early if any unexpected adverse symptoms developed.

Phosphocreatine Recovery

In the Gillingham et al. (2017) study, phosphocreatine recovery following acute exercise was considered 1 of the primary outcomes. In the CL201 and CL202 studies, phosphocreatine recovery was not measured as part of the efficacy assessment.

Phosphocreatine recovery, a measure of muscle ATP synthesis, was measured after a bout of acute phosphocreatine-depleting exercise using the lower leg. The Gillingham et al. (2017) employed a defined exercise protocol that involved repetitive leg or ankle flexion and extension against resistance (e.g., foot peddle or latex band) while evaluating the isolated working muscle using phosphorus spectroscopy. The exercise was performed approximately 3 hours after lunch, and patients practised the movement before starting the test; during the test, patients were instructed to push their foot repeatedly as fast and hard as they were able to for 30 seconds. Magnetic resonance spectroscopy was used to evaluate high-energy phosphate metabolism, to measure how quickly muscles make energy (ATP) after a depleting exercise. In addition to imaging, absolute concentrations of phosphocreatine, ATP, inorganic phosphate, and other phosphorus metabolites in the tibialis were measured. Data were collected from the muscle at rest and then before, during, and after the exercise. Phosphocreatine synthesis was measured during the recovery phase of each patient at baseline and was compared to results after 4 months of treatment. Not all subjects completed this exercise, and phosphocreatine recovery was not measured in children younger than 16 years of age.

Symptom Relief

In CL201, a diary was used to capture daily incidence of muscle weakness and fatigue by the patient and/or caregiver. The number of episodes of muscle pain and leg cramps each day was also recorded, along with the activity level corresponding with each incidence of muscle pain or leg cramps and weakness or fatigue. In CL202, the incidence and frequency of fatigue, exercise tolerance, muscle pain, and activity level were reported weekly by the patients or caregiver in a diary. This information was gathered as part of the study’s exploratory end points. However, the patient diary was removed from the protocol in an amendment (protocol amendment 4) to reduce the burden on patients, with the rationale that continuing to record patient-reported symptoms of muscle weakness and fatigue in the open-label, single-arm extension was unlikely to yield informative data. New or worsened events of fatigue and muscle pain were still collected as AEs. In the Gillingham et al. (2017) study, symptom relief from triheptanoin treatment was not captured as an efficacy end point; however, musculoskeletal pain, cramping, or elevated CK were collected as part of data on AEs.

Cardiac Function Parameters

In CL201, the change from baseline in LVEF was considered a supportive end point. In CL202, cardiomyopathy and cardiac function, including ventricle size, ejection fraction, and shortening fraction at each visit and change from baseline were among the secondary efficacy end points. In both studies, cardiac disease was assessed by echocardiography and electrocardiography at the baseline visit. Thereafter, in CL201, the tests were repeated at week 24 (or at the early termination visit if that visit was before the week 24 assessment), whereas, in CL202, cardiac function was measured annually (or at the early termination visit if it had not been performed within 6 months before termination). In both studies, additional tests could be performed if any abnormalities were detected or if medically indicated, and clinically significant changes from baseline in echocardiograms were also recorded as an AE, if appropriate.

In the Gillingham et al. (2017) study, cardiac function, as measured by echocardiography, was 1 of the primary outcomes. At baseline and after 4 months of treatment, a comprehensive echocardiographic evaluation was performed in the echocardiography laboratories at each institution. Cardiac function was assessed at baseline and after treatment, approximately 1 hour after lunch, and was measured at rest. Right and left ventricle chamber volumes, ejection fraction, and cardiac muscle strain were evaluated. Analysis of echocardiographic variables was blinded. Intra-observer concordance correlation (correlation for repeated echocardiographic analysis by 1 evaluator) was r = 0.75 (95% CI, 0.09 to 0.95), and inter-observer concordance correlation (between 2 evaluators) was r = 0.89 (95% CI, 0.51 to 0.98).

Concomitant Medications

In the CL201 study protocol, clinical outcomes for cardiac disease included cardiac medication requirement for maintenance treatment (e.g., diuretics, digitalis), although it is unclear whether this was considered as part of key or supportive efficacy end points. Interventions that were required for the maintenance treatment of cardiomyopathy were assessed at each treatment visit, in conjunction with the review of concomitant medications taken by the patient, and through the extension period (week 78). Start and stop dates, dosage, and dosing interval were recorded. Changes in the use of concomitant medications were not explored in CL202 or the Gillingham et al. (2017) studies.

Productivity

The effect of triheptanoin treatment on productivity at school or work was not measured in any of the 3 studies included in this review.

Safety

In CL201, safety and tolerability of triheptanoin were routinely recorded at each visit and monitored through weekly phone calls by the study coordinator from baseline to week 4, and then once between each subsequent study visit. Safety was also monitored throughout the study by a data monitoring committee and included the collection of AEs and SAEs. Specifically, TEAEs, serious TEAEs, fatal AEs, TEAEs leading to discontinuation, treatment-related TEAEs, and treatment-related serious TEAEs were collected. GI TEAEs categorized using the standardized MedDRA query (SMQ) were also assessed. Selected clinical laboratory parameters and changes from baseline values, electrocardiograms, echocardiograms, height, weight, and body mass index were also included as part of the safety end points. AEs were assessed at all visits and throughout the course of the study. An AE was considered as treatment-emergent if it occurred on or after the date of the first treatment of triheptanoin.

In CL202, there was no formal data monitoring committee; conduct of the study and safety of the patients were monitored by the sponsor regularly. AEs were recorded at each visit, as well as through weekly phone calls by the study coordinator from baseline to month 6, and then once between each subsequent study visit through month 60. Safety assessments included the collection of AEs and SAEs; the safety end points in CL202 included incidence and severity of TEAEs (primary safety end points), vital signs, incidence of laboratory abnormalities, and concomitant medications. Specifically, the incidence, frequency, severity, and relatedness of AEs and SAEs were collected, and these included clinically significant changes from baseline to scheduled time points in vital signs, weight, physical examination, and clinical laboratory evaluations. An AE was considered treatment-emergent if it occurred on or after the first triheptanoin dose during the CL202 study period.

In the Gillingham et al. (2017) study, safety, and efficacy were monitored by a data safety and monitoring board. Patients were monitored for AEs and SAEs throughout the study. During weekly communication with the study coordinator, patients and/or parents were asked about AEs from treatment. Strategies to minimize GI upset were also discussed.

For all 3 studies, the severity of AEs was graded using the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) version 4.0. During regular communications with study personnel, patients were solicited for information specifically on the following potential AEs: GI upset, steatorrhea or frequent loose stools, and weight gain in CL201 and CL202; and GI upset, steatorrhea or frequent loose stools, weight gain, muscle pain, and lethargy in the Gillingham et al. (2017) study.

Statistical Analysis

Study CL201

In Study CL201, approximately 30 patients with severe LC-FAODs were planned for enrolment. This planned sample size was not based on powering for hypothesis testing of a specific end point. Rather, the sample size was intended to provide the maximum amount of information regarding triheptanoin tolerability and dosing regimen, and to provide long-term safety and efficacy data following 78 weeks’ maintenance treatment with triheptanoin. No further details on recruitment or determination of this sample size were provided in the statistical plan. A pre-specified interim analysis was planned for week 12 per the protocol; however, this analysis was not performed, and the reason was not provided in the Clinical Study Reports. An interim analysis was also scheduled for week 48 but was removed in a protocol amendment, as this time point was deemed premature to accurately perform an analysis on the rate of MCEs. To assess the effects of triheptanoin treatment, analyses of data were conducted in 2 stages: the week 24 assessment was the main analysis for non-acute clinical disease (i.e., energy physiology through biologic and clinical assessments focusing on skeletal myopathy, hepatic disease, and cardiac disease) associated with LC-FAODs, where comparisons were based primarily on a change from the baseline (or run-in) value. Analyses were performed to assess MCEs following 78 weeks of treatment compared to historical data collected for the 18- to 24-month before study entry.

All statistical tests were 2-sided, and results showing a P value of less than or equal to 0.05 were considered statistically significant. No adjustments for multiplicity were made.

For MCEs, the annualized event rate and annualized event days of over the 78 weeks of triheptanoin treatment were compared with the pre-triheptanoin period using the paired t-test. For end points measuring change from baseline over time (e.g., 12MWT, cycle ergometry, and HRQoL), assessments were analyzed using a generalized estimation equation (GEE) model that included time as the categorical variable and adjusted for baseline measurement. Compound symmetry, which specified constant variance for the assessments and constant covariance between the assessments over time, was used as the covariance structure for the GEE model. For the study’s primary objectives, P values were provided for testing the statistical significance of change from baseline to week 24 and week 78 assessments. For the week 24 analysis (treatment period), only measurements up to week 24 were included in the model. For the week 78 analysis (treatment plus extension periods), all measurements were included in the model. Based on the scheduling of assessments, the 12MWT measurements taken at weeks 18 and 60 were considered the primary time point for the week 24 and 78 analyses, respectively. For analyses with insufficient observations for a GEE model, a descriptive summary was provided, and analyses using observations at a single time point (e.g., paired t-test or other nonparametric methodologies) were considered.

After final database lock, several analyses were added, including annualized event rate and annualized event days for total MCEs (i.e., for the composite of rhabdomyolysis, hypoglycemia, and cardiomyopathy events) to evaluate the aggregate effect of triheptanoin on metabolic crises. Sensitivity analyses for handling missing data using a negative binomial regression model, to account for different follow-up times and address impact on MCEs of patients who discontinued the study prematurely, were also added. Furthermore, subgroup analyses of MCEs were performed for age group at initial treatment with triheptanoin, age of presentation, LC-FAOD subtype, and administration route of triheptanoin. These analyses were also added after database lock. Ad hoc sensitivity analyses were performed for 12MWT and cycle ergometry results after 1 patient in each test was noted to have performed very differently from others in the group; sensitivity analyses were performed after excluding these patients.

MCEs with completely missing dates were excluded from both the number and duration of event calculations during both treatment periods. MCEs with either an admission or start date were counted in the number of events but were excluded from the total duration for both treatment periods. For each patient, randomly missing assessments remained as missing data. When a change from baseline was assessed, only patients with a baseline and at least 1 post-baseline measurement were included in the analysis. As previously mentioned, to address the impact of missing data due to premature discontinuation, a negative binomial regression model was used to compare MCEs before and during triheptanoin treatment.

Study CL202

Approximately 100 patients meeting the eligibility criteria were planned to be enrolled in the study. The sample size was intended to provide information regarding the long-term safety of triheptanoin, as well as indicators of sustained efficacy and durability of response. Further details on recruitment or determination of this sample size were provided in the statistical plan. No formal interim analysis was planned for CL202; however, an interim analysis including both efficacy and long-term safety data was conducted to support a regulatory filing (data cut-off of June 1, 2018).

Statistical analyses were mainly descriptive. Statistical tests were performed for selected end points; all were 2-sided and tested at a statistical significance level of 0.05.

The annualized MCE event rate (combination of rhabdomyolysis, hypoglycemia, and cardiomyopathy events) during the CL202 study period (i.e., study period 1) was calculated for the full analysis set (FAS; refer to the Analysis Populations section for definition) in the primary efficacy analysis; the annualized MCE event rate was also calculated by 18-month intervals. For secondary efficacy analyses, observed values and corresponding changes from baseline by visit during the CL202 study period were summarized for echocardiographic variables (ventricle size, ejection fraction, and shortening fraction), and annualized MCE event rate and event days were summarized for the FAS, as described for the primary efficacy end point. Summary statistics were produced for all exploratory end points using all available assessments in the CL202 study period. No covariates were planned for use in the analysis of primary and secondary efficacy end points.

When sample size and number of observations were allowed, end points measuring change from baseline over time were analyzed using a GEE model that included time as the categorical variable and adjusted for the baseline measurement. Compound symmetry, which specifies constant variance for the assessments and constant covariance between the assessments over time, was used as the covariance structure for the GEE model. For analyses with insufficient observations for a GEE model, a descriptive summary was provided, and analyses using observations at a single time point (e.g., paired t-test or other nonparametric methodologies) could be considered.

For patients in the CL201 rollover cohort, the long-term effects of triheptanoin treatment were assessed on the basis of MCE rates and echocardiographic variables (ventricle size, ejection fraction, and shortening fraction) using an integrated analysis by combining data collected from the CL201 an CL202 study periods. Annualized event rate and annualized event days of MCE (total and by type) were calculated for 3 periods (18 months pre-triheptanoin, 18-month triheptanoin treatment in CL201, and triheptanoin treatment in CL202). Baseline measures from Study CL201 were used for echocardiographic variables (ventricle size, ejection fraction, and shortening fraction). Statistical significance and P values were assessed, comparing the frequency and duration of MCEs between the pre-triheptanoin period and the first 36 months of triheptanoin treatment. The 36-month treatment window was selected to mitigate variability that may have been introduced from different triheptanoin treatment durations and to confirm the primary findings from Study CL201 over an extended treatment period.

For patients in the triheptanoin-naive cohort, statistical comparisons for all MCE-related end points (i.e., annualized event and annualized event days of MCEs overall and by event type) were made by comparing the retrospective period (18 months pre-triheptanoin) with the first 18 months of triheptanoin treatment. The comparison between pre-triheptanoin and triheptanoin treatment in MCE rates was also performed using a negative binomial regression model to account for different follow-up times during either the pre-triheptanoin period (for infants with less than 18 months of retrospective data collection) or triheptanoin treatment (for patients who discontinued the study early, before reaching 18 months of participation). The ratio of the event rate during triheptanoin treatment to the pre-triheptanoin event rate was provided, along with the 2-sided 95% CI and P values. For non-MCE end points measuring changes from baseline, 2-sided 95% CIs were provided for the current interim analysis; statistical significance and P values are planned for the final analysis.

For MCE-related end points measured in both cohorts, pre-triheptanoin and triheptanoin treatment comparisons were made using the paired t-test. For distributions that violated the normality assumption due to heavily skewed underlying distributions (i.e., P value from the Shapiro-Wilk normality test was less than 0.05), the Wilcoxon signed rank test was used instead of the paired t-test.

Several changes were made to the planned analyses, most notably to the CL201 rollover cohort. This included adding a pre-triheptanoin and triheptanoin treatment comparison, as previously described (i.e., statistical significance of annualized event rates and duration of MCEs during the pre-triheptanoin period compared to first 36 months of treatment). A negative binomial regression model was also performed to account for different follow-up times. Furthermore, for both CL201 rollover and triheptanoin-naive cohorts, sensitivity analyses to remove outliers that were identified for annualized event rate and event days of MCEs and MCE hospitalizations were added. An outlier was defined as a value that lies outside the interval (Q1 − 1.5 × IQR, Q3 + 1.5 × IQR), where IQR is the interquartile range (i.e., Q3 – Q1 of the underlying distribution during either the pre-triheptanoin or triheptanoin treatment period).

Subgroup analyses of MCE end points (annualized event rate and annualized duration of overall MCEs and MCE hospitalizations) were performed for the following subgroups in the CL201 rollover and triheptanoin-naive cohorts: age group at initial treatment with triheptanoin, age of presentation, LC-FAOD diagnosis, and administration route of triheptanoin. For the CL201 rollover cohort, the analysis included the pre-triheptanoin period and then first 36 months of total triheptanoin treatment, whereas, for the triheptanoin-naive cohort, the analysis included the pre-triheptanoin period and the first 18 months of treatment. However, due to limited sample size within each group, statistical testing was not conducted.

Missing assessments for each patient remained as missing, unless otherwise specified. When a change from baseline was assessed, only patients with a baseline measurement and at least 1 post-baseline measurement were included in the analysis. Furthermore, only observed data (not imputed data) were presented.

Gillingham et al. (2017) Study

In the Gillingham et al. (2017) study, a sample size of 32 patients (16 in each treatment group) was planned for enrolment. The authors targeted 5 to 6 patients per diagnostic group (i.e., CPT II, VLCAD, or TFP and LCHAD deficiency) at each study centre. Sample size calculations were based on a previous preliminary trial in patients with LC-FAODs, in which 6 patients were randomized to a standard high-carbohydrate diet and 6 were randomized to a high-protein diet, and the change in adjusted TEE was compared between treatment groups. The adjusted TEE outcome used was a z score for TEE, which was measured via doubly labelled water, and normed based on age, height, weight, and sex. The mean adjusted TEE z score from the preliminary data on the 12 patients was −1.12 (SD = 3.12) for the group on a high-protein diet, compared to −5.22 (SD = 1.82) for patients on a standard diet, with the mean difference of 4.1 observed in the TEE z score between the 2 diets. The study investigators anticipated a similar difference in change from baseline between triheptanoin and MCT diets. Using these SDs and an assumption that the SDs at baseline and follow-up are the same within each treatment group, the investigators calculated that the sample sizes per group needed to detect a 4.1 mean difference in z score change for a range of correlations between 0 and 0.75. A simple 2-sample t-test for change was used for the calculations. Specifically, a sample size of 14 patients was required to detect a mean difference of 4.1 in TEE z score with 80% power at significance level 0.05. This assumed zero correlation between baseline and follow-up TEE z scores, with an estimated SD of TEE z score difference from baseline of 2.57 for MCT and 4.41 for triheptanoin. Thus, a sample size of 32 patients was deemed sufficient to achieve greater than 80% power, and enrolment of sufficient patients was achieved.

The primary analyses, comparing change in energy expenditure, cardiac function, and exercise tolerance between the 2 treatment groups, were performed using linear mixed models, with a separate model for each outcome.

All models treated patients as a random effect, while treatment group (triheptanoin versus trioctanoin), time (baseline versus 4-month follow-up), and the interaction between these dichotomous factors, were held as fixed effects. Randomization factors (i.e., medical diagnostic group and investigation site) were also included as fixed effects. The formal test for comparing change between groups was based on an interaction assessing whether change over time (baseline versus 4-month time point) differs by treatment and serves as the effect of interest. In instances where pronounced (positive) skewness or extreme ranges (i.e., > 3-fold separation between the maximum and minimum observation) occurred, responses were log-transformed before fitting the models. End points involving a series of measurements over time (e.g., treadmill test) had mixed-effects models fit to the observed change between follow-ups. All P values were 2-sided and tested at a statistical significance level of 0.05 for each end point considered. No details were provided on subgroup analyses or the handling of missing data.

Analysis Populations

In CL201, the FAS included all patients enrolled (e.g., who participated in the run-in period) in the study. The primary analysis set (PAS) consisted of patients in the FAS who completed the 4-week run-in period and received at least 1 dose of triheptanoin. In general, the PAS set was used for analyses of the efficacy data. The PAS for a specific assessment was defined as follows:

• cycle ergometry: the subset of patients in the PAS who had at least 1 cycle ergometry test performed with any duration

• 12MWT: the subset of patients in the PAS who had at least 1 12MWT performed with any distance walked

• SF-10: the subset of patients in the PAS who had at least 1 SF-10 test performed

• SF-12: the subset of patients in the PAS who had at least 1 SF-12 test performed

As the number of patients who were included in the PAS was the same as the FAS (n = 29), the 2 were considered interchangeable for the purposes of this CADTH review, and the PAS data have been reported and referred to as the FAS in the following tables. The per-protocol analysis set consisted of patients in the safety analysis set who completed at least 80% of triheptanoin doses and did not have a major protocol deviation recorded in the category of treatment compliance. A complete treatment dosage was defined as at least 25% of DCI from triheptanoin. The per-protocol analysis set was used for additional analyses of the key efficacy end points in cycle ergometry and 12MWT. The analysis of the MCEs was also assessed with the per-protocol analysis set. The safety analysis set included all patients who received at least 1 dose of triheptanoin and was used for all safety analyses.

In CL202, the FAS included all patients enrolled who had at least 1 post-baseline efficacy assessment. Unless noted otherwise, all efficacy analyses were based on the FAS. The FASs for specific assessments are defined as follows:

• SF-10: subset of patients in the FAS whose age at enrolment was 6 or older to less than 18 years and who had at least 1 SF-10 test performed while in the CL202 study

• SF-12v2: subset of patients in the FAS whose age at enrolment was 18 years or older and who had at least 1 SF-12v2 test performed while in the CL202 study

The safety analysis set included all patients who were enrolled and treated with at least 1 dose of triheptanoin while in the CL202 study.

In the Gillingham et al. (2017) study, the primary analysis, comparing change in energy expenditure, cardiac function, and exercise tolerance between triheptanoin and MCT supplementation, was performed on the intention-to-treat population, which included all randomized patients.

Protocol Amendments

Several amendments were made to the protocols of all 3 studies. The protocol for Study CL201 was amended 4 times. Notable amendments included an increase in number of patients planned for enrolment from 20 to 30, a change in the age criteria to change the lower limit to 6 months and remove the upper limit (original protocol was 6 to 25 years). The definition of severe LC-FAODs was also expanded in the inclusion criteria to capture more patients with severe disease. Also, TFP deficiency was added to the inclusion criteria. Due to unexpected performance issues, electronic diaries were discontinued and replaced by paper versions. Interim analysis at week 48 was removed. All protocol amendments were made before the first patient signing consent; only protocol amendment 4 (i.e., removal of the week 48 interim analysis and electronic diaries) occurred during patient enrolment.

The protocol for Study CL202 was amended 5 times. Notable amendments include expansion of study eligibility criteria by allowing patients who had failed conventional therapy and who had documented severe unmet need, and extending the study duration to 5 years (previously 3 years). The 12MWT, PEDI-CAT, and patient diaries for patient-reported fatigue, exercise tolerance, muscle pain, and activity level were removed, as they were deemed unlikely to provide informative data in a single-arm, open-label study setting (although fatigue and muscle pain were still considered AEs). Also, primary, secondary, and exploratory end points were clarified based on the study objectives. Exploratory end points were identified to characterize the potential effect of triheptanoin treatment on functional disability and cognitive development, clinical biomarkers, and growth measurements. All protocol amendments were made during patient enrolment.

The protocol for the Gillingham et al. (2017) RCT was amended 4 times. Notable amendments include the following. The age criteria were initially restricted to 7 to 45 years due to concerns about patients completing the exercise treadmill test, but the upper age limit was later removed, with no scientific justification given. The exclusion criterion of history of myocardial infarction was added. The number of patients recruited at each site was adjusted owing to institutional limitations; as a result, 20 patients were recruited at the Oregon Health and Science University and 12 patients were recruited at the University of Pittsburgh (total remained unchanged). Cardiac MRI was eliminated and was replaced by standard echocardiogram. Also, the end point measuring phosphocreatine depletion and recovery was added. All protocol amendments were made during patient enrolment.

Several analyses added after database lock for studies CL201 and CL202 are discussed in the Statistical Analysis section.

Results

Patient Disposition

Patient disposition is summarized in Table 12 for Study CL201 and Study CL202, and Table 13 for Gillingham et al. (2017).

Study CL201

In Study CL201, 30 patients were screened, 29 of whom were enrolled in the study as part of the FAS as well as the safety analysis set. Twenty-two patients were included in the per-protocol analysis, which reflected patients who were compliant with treatment (i.e., consumed at least 80% of triheptanoin doses and had no major protocol deviations). A significantly smaller group of patients relative to the enrolled population were included in the assessments of muscle function or exercise tolerance and HRQoL. By the data cut-off date, 5 patients (17.2%) had discontinued from the study, 25 patients (86.2%) had completed 24 weeks of triheptanoin treatment, and 24 patients (82.8%) had completed 78 weeks of treatment.

Study CL202

In Study CL202, a total of 75 patients were enrolled by the data cut-off date. Of these 75 patients, 24 had completed Study CL201 (CL201 rollover cohort), 20 were enrolled into the triheptanoin-naive cohort, and 31 had received previous triheptanoin treatment through ISTs or other mechanisms and were enrolled in the IST or other cohort. The mean duration of treatment was 25.92 months overall. The mean duration for each treatment cohort was as follows: 23.01 months for the CL201 rollover cohort (excludes CL201 study duration), 15.68 months for triheptanoin-naive cohort, and 34.77 months for the IST or other cohort.

One patient in the triheptanoin-naive cohort had been enrolled in the CL201 study but had discontinued the study after taking triheptanoin for 3 days, and subsequently did not receive triheptanoin for 2 years before enrolling in CL202. As a result, this patient was considered triheptanoin-naive. In the IST or other cohort, most patients (n = 27) had received triheptanoin under a compassionate use program. Seven patients who were previously enrolled in the Gillingham et al. (2017) trial were also included in the IST or other cohort, including 4 who also had received triheptanoin through the compassionate use program before enrolling in CL202.

All patients enrolled in CL202 were included in the FAS as well as the safety analysis set in their respective cohorts. Data for a per-protocol analysis set were not presented for this study. A significantly smaller group of patients relative to the enrolled population were included in the assessments of HRQoL. Ten patients (13.3%) had discontinued from the study; as CL202 is ongoing, no patients had completed the study at the time of the data cut-off.

Gillingham et al. (2017) Study

In the Gillingham et al. (2017) study, 39 patients were screened and 32 were randomized, with 16 in each treatment group. All patients had completed the study by the data cut-off date. However, patients were excluded from some planned analyses due to uninterpretable or missing data. A smaller group of patients relative to the enrolled population were included in the exercise protocol for assessment of phosphocreatine recovery. Also, echocardiogram evaluations were not available for 11 patients due to a logistic change in the protocol from cardiac MRI to echocardiogram after the first 7 patients were enrolled and technical difficulties leading to uninterpretable echocardiograms in 4 patients.

Exposure to Study Treatments

A summary of treatment exposure and concomitant therapies can be found in Table 14.

Study CL201

In Study CL201, the mean duration of treatment was 15.86 months (SD = 6.07). On average, patients were prescribed a mean triheptanoin dosage of 66.8 mL (SD = 33.85), which corresponded to a mean DCI percentage of 31.20% (SD = 8.88%). The mean dosage of triheptanoin consumed was 27.5% (SD = 4.58) of DCI. During the study, patients received the prescribed breakdown of macronutrients consistent with dietary guidelines for management of patients with LC-FAODs, and macronutrient breakdown (percentage of DCI) was relatively stable between the pre-triheptanoin and triheptanoin treatment periods. The most notable change was a 10% DCI increase (from average 17.4% to 27.5%) in the amount of medium-chain fat consumed compared to the pre-triheptanoin period. Various other dietary management was prescribed during the study; of note, 69.0% of patients received carnitine supplementation.

Study CL202

As of the data cut-off date, the mean duration of treatment was 25.92 months (SD = 11.080). Patients in the IST or other cohort had the longest mean duration of treatment (34.77 months; SD = 5.628) and triheptanoin-naive patients had the shortest duration (15.68 months; SD = 11.530). Patients in the CL201 rollover cohort received a mean duration of triheptanoin of 23.01 months (SD = 6.186) while enrolled in CL202, in addition to approximately 18 months of triheptanoin treatment while enrolled in CL201 (mean duration of study participation for combined CL201 and CL202 of 41.4 months). The mean dosage of triheptanoin prescribed was 64.35 mL (SD = 24.610), which corresponded to a mean DCI percentage of 26.95% (SD = 7.48); the mean prescribed dosage and percentage of DCI were overall similar among the 3 cohorts. The mean triheptanoin dosage (percentage of DCI) actually consumed was not reported, although most patients consumed more than 90% of their prescribed dosage. According to the Clinical Study Report, patients received the prescribed breakdown of macronutrients consistent with dietary guidelines for the study and management of patients with LC-FAODs. However, a detailed breakdown of macronutrients received (percentage of DCI) averaged over visits throughout the study was not reported for each cohort separately. Various other dietary management was also received by patients in the CL201 rollover and triheptanoin-naive cohorts; most notably, 66.7% and 80.0% of patients in the CL201 rollover and triheptanoin-naive cohorts, respectively, received carnitine supplementation during the CL202 study. Details of concomitant treatment in the IST or other cohort were not available.

Gillingham et al. (2017) Study

Based on analysis of the 3-day diet records in the Gillingham et al. (2017) study, patients consumed 16.62% (SD = 2.66) and 14.83% (SD = 3.40) of DCI from triheptanoin and trioctanoin, respectively. During the 4 months of study treatment, based on measures of compliance (i.e., amount of oil dispensed minus amount returned), patients had consumed 73.94% (SD = 22.59) and 75.59% (SD = 19.04) of the prescribed triheptanoin and trioctanoin, respectively. A detailed breakdown of the actual percentage of DCI of macronutrients received by the patients during the study was not reported, nor were concomitant treatments taken during the study.

Efficacy

Only efficacy outcomes and analyses of subgroups identified in the review protocol are reported as follows. Refer to Appendix 3 for detailed efficacy data.

No data were available for the following key efficacy outcomes identified in the CADTH review protocol: survival, symptom relief, concomitant medication, and productivity. Also, for efficacy outcomes with available data, many were not evaluated across all 3 studies.

Survival

Survival was not an efficacy end point in any of the studies included in this review. However, deaths that occurred during CL201 and CL202 are summarized in the Harms section.

Major Clinical Events

In studies CL201 and CL202, the majority of MCEs, both before and during triheptanoin treatment, were supported by clinical and/or laboratory assessments. Rhabdomyolysis events were associated with elevated CK levels. Hypoglycemia events were defined by clinical symptoms (e.g., altered mental status, fatigue, pallor, palpitations, slurred speech, excessive sweating, shakiness, and/or light-headedness) or by glucose level readings below the lower limits of the reference range used by the institution. Cardiomyopathy events showed reduced LVEF by echocardiography.

Study CL201

Key efficacy end points for CL201 included the annualized event rate and event days (i.e., duration rate) of MCEs, which was a measure of hospitalization, ED or acute care visits, or emergency interventions due to pre-specified clinical manifestations. Pre-specified key efficacy end points were MCEs due to rhabdomyolysis, hypoglycemia, and cardiomyopathy events individually; a composite end point including all 3 clinical manifestations to evaluate the aggregate effect was added after final database lock. Analyses was performed at week 78, with data compared between the pre-triheptanoin and triheptanoin treatment periods.

In total, 70 MCEs occurred in 22 patients during the pre-triheptanoin period, and 39 events occurred in 16 patients through week 78 of the triheptanoin treatment period (Table 36). Of note, 4 events that had occurred during the pre-triheptanoin period were excluded from analyses due to missing dates. The majority of MCEs were due to rhabdomyolysis. A reduction in annualized event rates and event days occurred across all 3 clinical manifestations with triheptanoin treatment but was most favourable for the aggregate measure including all event types. For total MCEs, including all event subtypes, the difference in the mean annualized event rate was 0.813 events per year, and the difference in mean annualized event days was 2.997 in favour of triheptanoin (Table 15). Figure 7 and Figure 8 in Appendix 3 present individual-level data for MCEs as well as annualized events and duration rates.

Study CL202

The primary efficacy end point for CL202 was the annualized MCE rate, including skeletal myopathy (rhabdomyolysis), hepatic (hypoglycemia), and cardiomyopathy events. Secondary efficacy end points for CL202 included annualized event days of all MCEs, as well as annualized event rates and annualized event days of each MCE type (i.e., rhabdomyolysis, hypoglycemia, and cardiomyopathy). The definition of MCE was the same as for CL201 and included any visit to the ED or acute care, hospitalization, emergency intervention, or any similar event due to or complicated by LC-FAODs. The focus of the interim analysis was on the effects of triheptanoin on patients in the CL201 rollover and triheptanoin-naive cohorts. Statistical comparisons were performed after 36 months of treatment in the CL201 rollover group (including CL201 and CL202 studies) and after 18 months of treatment in the triheptanoin-naive cohort, with data compared between the pre-triheptanoin and triheptanoin treatment periods.

As of the data cut-off date of June 1, 2018, there were 259 MCEs experienced by 47 patients in the overall population. The majority of MCEs were due to rhabdomyolysis. Of the 3 cohorts, patients enrolled in the IST or other cohort had the highest number of MCEs (as well as the longest duration of participation). However, comparative analyses were not performed for the IST or other cohort. Annualized MCE event rate and event days during the CL202 study in all 3 cohorts, plus the overall population, are presented in Table 37.

In the CL201 rollover cohort, 22 of the 24 patients (91.7%) had at least 36 months of study participation. In total, 60 MCEs occurred in 18 patients during the pre-triheptanoin period, and 67 MCEs occurred in 17 patients through the first 36 months of the triheptanoin treatment period (Table 38). The majority of MCEs were due to rhabdomyolysis events. The most notable improvement with triheptanoin was in the annualized event rate of total MCEs. For this primary efficacy end point, the difference in the mean annualized event rate of total MCEs, including all event subtypes, was 0.8 events per year in favour of triheptanoin (Table 16). For the remaining annualized event rates and event days (secondary efficacy end points), reductions was generally occurred favour of triheptanoin across all comparisons, but none were significant. The exception was major rhabdomyolysis events, for which the mean annualized event days appeared to increase with treatment, although median days decreased. This may be due to the highly skewed distribution of annualized event days observed in this cohort; thus, the Clinical Study Report used median values to describe annualized event days.

In the triheptanoin-naive cohort, 7 of the 20 patients (35.0%) had reached 18 months of study participation or MCE data collection as of the data cut-off date. In total, 84 MCEs occurred in 16 patients during the pre-triheptanoin period, and 27 MCEs occurred in 12 patients during the 18 months of triheptanoin treatment (Table 39). The majority of MCEs were due to rhabdomyolysis events. Due to heavily skewed distributions observed in this cohort, median values were used to describe annualized event rates and event days in the Clinical Study Report. The mean values for total MCEs and for rhabdomyolysis events appeared to increase with treatment; however, median values decreased. The mean annualized event rates and event days for hypoglycemia and cardiomyopathy decreased, although no change occurred in the median values due to the small number of events. None of the changes were significant (Table 17).

In Study CL202, outliers skewed the distribution for the total MCE end points. Specifically, in the CL201 rollover cohort, outliers were identified for annualized event rate (n = 1) and event days (n = 4); in the triheptanoin-naive cohort, outliers were identified for identified for annualized event rate (n = 3) and event days (n = 5). Ad hoc sensitivity analyses were performed to remove outlier values. In the CL201 rollover cohort, results of this sensitivity analysis were consistent with the original dataset of this cohort, whereas, in the triheptanoin-naive patients, results were consistent with the CL201 study, in which a notable difference in favour of triheptanoin occurred in the annualized event rate and event days.

Gillingham et al. (2017) Study

MCEs were not measured as part of the efficacy analyses in the Gillingham et al. (2017) study.

Subgroup Analyses

Ad hoc analyses of the following 2 relevant subgroups, as identified in the CADTH systematic review protocol, were performed in CL201 and CL202: age at triheptanoin initiation and LC-FAOD diagnosis subtype.

Age at Treatment Initiation

To evaluate the effect of triheptanoin on MCEs by age at treatment initiation, annualized event rates and event days were examined according to the following age categories: younger than 6 years, 6 to less than 18 years, and 18 years or older. In Study CL201, the youngest age group (children < 6 years) experienced higher annualized event rates and annualized event days, both before and during triheptanoin treatment, compared with older children and adults. Overall, results were generally consistent with the overall patient population. Across all age subgroups, the mean annualized event rates and event days decreased during triheptanoin treatment compared with the pre-triheptanoin period. The most notable reduction was in children younger than 6 years of age (Table 40).

In the CL201 rollover cohort of Study CL202, the mean annualized event rates decreased during triheptanoin treatment compared with the pre-triheptanoin period across all age groups, although the median increased in the 2 older age groups. The mean annualized event days decreased in young children (< 6 years) but showed an increase in older children and adults; the median annualized event days also decreased in the pediatric age groups but increased in adults (Table 41).

In the triheptanoin-naive cohort of Study CL202, the median annualized event rates decreased during triheptanoin treatment compared with the pre-triheptanoin period across all age categories. The median annualized event days decreased in the pediatric age groups but showed an increase in the adult age group. Change in mean values were variable across the age groups. Of note, 2 of the 4 adult patients were identified as outliers and skewed the annualized event rate and event days calculations, limiting the interpretability of the results (Table 42).

LC-FAOD Diagnosis Subtype

To evaluate the effect of triheptanoin on MCEs in LC-FAOD subtypes, annualized event rates and days were examined according to the following LC-FAOD subtypes: LCHAD, VLCAD, CPT II, and TFP deficiency.

In Study CL201, patients with the more common LC-FAOD subtypes (LCHAD and VLCAD deficiencies) experienced reductions in annualized event rate and event days with triheptanoin treatment compared to the pre-triheptanoin period. The 4 patients with CPT II deficiency also experienced a reduction in annualized event and event days during triheptanoin treatment. Only 3 patients with TFP were enrolled in the study; a reduction in annualized event rate occurred, but not in annualized event duration (Table 43).

In the CL201 rollover cohort of Study CL202, patients with the more common LC-FAOD subtypes (LCAD and VLCAD deficiencies) also experienced reductions in annualized event rates and event days with triheptanoin treatment compared to the pre-triheptanoin period. The 3 patients with CPT II deficiency also experienced a reduction in annualized event rate and event days during triheptanoin treatment. Similar to CL201, only 3 patients with TFP deficiency were enrolled in the study, and they experienced an increase in annualized event rate and annualized event days. However, 2 of the 3 patients were identified as outliers in the study, which limits the interpretability of the results (Table 44).

In the triheptanoin-naive cohort of Study CL202, patients with the more common LC-FAOD subtypes (LCHAD and VLCAD deficiencies) experienced reductions in median annualized event rate and annualized event days with triheptanoin treatment compared to the pre-triheptanoin period. The 2 patients with CPT II deficiency both also experienced a reduction in annualized event rate and annualized event days during triheptanoin treatment. Only 3 patients with TFP deficiency were enrolled, and no changes in the median annualized event rates or event days were noted. Changes in mean values were variable across the diagnosis subtypes. Two CACT patients and 1 CPT patient were excluded from the subgroup analysis due to the small sample size in the respective subgroup (Appendix 3, Table 19).

Hospitalizations

Hospitalizations were captured as part of the MCEs in CL201 and CL202.

Study CL201

Most MCEs during the pre-triheptanoin and triheptanoin treatment periods were hospitalizations. In total, there were 57 hospitalizations (81.4% of the 70 total MCEs) in 21 patients during the pre-triheptanoin period and 29 hospitalizations (74.4% of the 39 total MCEs) in 13 patients during triheptanoin treatment (Table 46). The majority of hospitalizations were due to rhabdomyolysis. Although few events due to cardiomyopathy occurred during the study, all led to hospitalization due to the serious nature of the event. A reduction in annualized hospitalization rates and hospitalization days occurred across all 3 clinical manifestations with triheptanoin treatment but was most favourable for the aggregate measure including all event types. For hospitalizations due to total MCEs, including all event subtypes, the difference in the mean annualized event rate was 0.739 hospitalizations per year, and the difference in mean annualized event days was 2.923 in favour of triheptanoin (Table 18).

Study CL202

As of the data cut-off date of June 1, 2018, there were 188 hospitalizations for MCEs experienced by 43 patients in the overall population. The majority of hospitalizations were due to rhabdomyolysis. Of the 3 cohorts, patients enrolled in the IST or other cohort had the highest number of hospitalizations. However, comparative analyses were not performed for the IST or other cohort. Annualized hospitalization rate and duration during the CL202 study in all 3 cohorts, plus the overall population, are shown in Table 48.

In the CL201 rollover cohort, there were 48 hospitalizations (80.0% of 60 total MCEs) in 17 patients during the pre-triheptanoin period and 53 hospitalizations (79.1% of 67 total MCEs) in 15 patients during the first 36 months of the triheptanoin treatment period (Table 49). The majority of MCEs were due to rhabdomyolysis. The greatest improvement with triheptanoin treatment occurred in the annualized hospitalization rate of total MCEs. The difference in the mean annualized hospitalization rate of total MCEs, including all event subtypes, was 0.67 events per year in favour of triheptanoin (Table 19). For the remaining annualized hospitalization rates and hospitalization days, reductions were generally in favour of triheptanoin across all comparisons, but none were significant. The exception was hospitalization for major rhabdomyolysis events, for which the mean annualized event days appeared to increase with treatment, although median days decreased. This may be due to the highly skewed distribution of annualized event days observed in this cohort; thus, the Clinical Study Report used median values to describe annualized event days.

In the triheptanoin-naive cohort, there were 64 hospitalizations (76.2% of 84 total MCEs) in 16 patients during the pre-triheptanoin period and 23 hospitalizations (85.2% of 27 MCEs) in 10 patients during the 18 months of triheptanoin treatment (Appendix 3, Table 50). The majority of MCEs were due to rhabdomyolysis. Due to heavily skewed distributions observed in this cohort, median values were used to describe annualized event rates and event days in the Clinical Study Report. The mean values for hospitalization due to total MCEs and rhabdomyolysis appeared to increase with treatment; however, median values decreased. The mean annualized event rates and event days for hypoglycemia and cardiomyopathy decreased, although there was no change in the median values due to the small number of events. None of the changes observed were significant (Table 20).

Gillingham et al. (2017) Study

In the Gillingham et al. (2017) study, 7 hospitalizations for acute rhabdomyolysis were reported in each treatment group. There was no difference in length of hospital stay.