Cost-effectiveness analysis of telotristat ethyl for treatment of carcinoid syndrome diarrhea inadequately controlled with somatostatin analogs

Abstract

Aims: This study evaluated the cost-effectiveness of telotristat ethyl (TE) added to somatostatin analog octreotide (SSA + TE) compared to octreotide alone (SSA) in patients with carcinoid syndrome diarrhea (CSD) whose symptoms remain uncontrolled with SSA alone.

Materials and methods: A deterministic Markov model evaluated the costs and quality-adjusted life-years (QALY) gained with SSA + TE vs SSA per a third-party US payer perspective. The model reflected clinical practice and resource use estimates based on current standards of care, with utility estimates based on similar symptoms from ulcerative colitis. Treatment efficacy was based on the phase III clinical trial of SSA + TE vs SSA alone [TELESTAR, NCT01677910]. According to TELESTAR, 44% of SSA + TE and 20% of SSA patients responded to therapy after 12 weeks. At each 4-week assessment period, SSA patients not adequately controlled received increasing doses of SSA and SSA + TE patients discontinued TE and moved to SSA only. Drug costs for adequately and not adequately controlled patients were $4,291.75 and $5,890.57 for SSA, respectively, and $9,456.07 and $5,890.57 for SSA + TE, respectively.

Results: The base-case analysis demonstrated lifetime QALYs of 1.67 at a cost of $495,125 for the SSA cohort and 2.33 ($590,087) for SSA + TE with an incremental QALY for SSA + TE of 0.66 for an additional $94,962. The incremental cost per QALY gained was $142,545. Sensitivity analyses demonstrated high probability (>99%) of SSA + TE being cost-effective at thresholds for rare diseases and orphan drugs of $300,000–$450,000.

Limitations: The recent availability of TE precluded the incorporation of clinical and economic inputs based on real-world practice patterns. The scarcity of epidemiology and utility information for this rare condition required the use of some proxy estimates.

Conclusions: This analysis demonstrated TE is a cost-effective treatment option when used on top of standard of care in CSD patients.

Keywords:

• Carcinoid syndrome
• cost-effectiveness
• telotristat ethyl
• neuroendocrine tumors
• diarrhea
• somatostatin analogs

Introduction

Carcinoid syndrome (CS) is known to develop in patients with neuroendocrine tumors (NETs), caused by tumoral secretion of vasoactive amines and peptides, and is associated with diarrhea, abdominal pain, vasomotor symptoms, bronchoconstriction, and complications such as carcinoid heart disease¹. The prevalence of NETs in the US is ∼103,000², of whom 63,000 (61%) are estimated to have gastroenteropancreatic tumors³. Approximately 10–20% of these patients have been reported to have functioning tumors, often leading to CS^4,5. The US FDA defines rare disease as having a prevalence <200,000⁶. Despite the limited availability of precise epidemiological data for CS diarrhea (CSD), the extrapolated prevalence of CSD must be far smaller than this threshold. The age-adjusted incidence of carcinoid tumors is estimated to be 1.9 per 100,000 persons in the US, for both men and women⁷, representing 0.5% of newly-diagnosed tumors⁸. This places the incidence of CS well below 1.9 per 100,000 persons in the US, with specific complications such as CSD in an even smaller sub-set of the population. Patients with CSD require more hospitalizations, office visits, and higher healthcare costs than those without diarrhea, highlighting the deleterious effects of this rare condition⁹.

The National Comprehensive Cancer Network guidelines recommends the somatostatin analogs (SSA) octreotide and lanreotide for treatment of CSD¹⁰. Daily subcutaneous injections of octreotide (100–600 mcg, mean daily dosage 300 mcg) are approved for use in the US in 2–4 divided doses for 2 weeks, which may be followed by the long-acting release (LAR) suspension formulation (20 mg every 4 weeks)^11,12. The SSAs were designed to block the release of vasoactive peptides and amines, but patients may develop reduced responsiveness to SSA therapy and experience subsequent breakthrough symptoms^13,14.

Telotristat ethyl (Xermelo™, Lexicon Pharmaceuticals, Basking Ridge, NJ) is a novel oral small-molecule tryptophan hydroxylase (TPH) inhibitor with a high molecular weight and acidic moieties that keep it from crossing the blood–brain barrier that is approved in the US for the treatment of CSD in combination with SSA therapy in adults inadequately controlled by SSA therapy^15–17. The TELESTAR phase III randomized, controlled clinical trial (ClinicalTrials.gov number NCT01677910) demonstrated significant reductions in bowel movement (BM) frequency and urinary 5-hydroxyindole acetic acid (u5-HIAA) among patients with CSD not adequately controlled by SSA therapy alone¹⁸. Telotristat ethyl, in addition to SSA therapy, was shown to be generally safe and well-tolerated through an open-label extension study showing a sustained effect on bowel movement control¹⁸.

Advancements in the management of diseases and associated complications have seen closer discrimination and interpretation of value in therapeutic decisions. Leading oncology centers, in particular, have adopted value frameworks to support evidence-based coverage decisions. More than 70% of US-based managed care medical and pharmacy directors recently indicated the use of value frameworks to inform decision-making related to oncology therapeutic management¹⁹. Rigorous assessment of cost-effectiveness using such value frameworks is central to understanding the most efficient use of new health technologies. Rare diseases with orphan drug treatment options are known to carry special considerations in health technology assessment and cost-effectiveness analysis^20,21.

The National Institute for Health and Clinical Excellence (NICE) Highly Specialized Technologies (HST) program evaluates new technologies for rare diseases, and has adapted some of its cost-effectiveness calculation assumptions for orphan drug treatments²¹. The Institute for Clinical and Economic Review (ICER) recently proposed an assessment framework for “ultra-rare” conditions with ≤10,000 patients per year, or ∼3.0 per 100,000²². ICER maintains the evidence standard for ultra-rare conditions, but acknowledges the challenge of evidence generation in this population, and extends the upper bound of the willingness-to-pay (WTP) threshold to $500,000 per quality-adjusted life-year (QALY)²².

The objective of this study was to evaluate the cost-effectiveness of telotristat ethyl with SSA (SSA + TE) compared to SSA alone among patients with CSD whose symptoms remain uncontrolled with SSA therapy alone.

Methods

Model overview

A deterministic Markov model was constructed to evaluate the costs and QALY gains associated with SSA + TE vs SSA treatment in patients with CSD uncontrolled with initial SSA therapy over a lifetime time horizon according to a third-party US payer perspective.

The model included three health states: Adequate Control; Control Not Adequate; and Dead. The model simulation extrapolated costs and patient outcomes from health states observed in the TELESTAR trial. The model utilized the primary end-point definition of treatment response from the TELESTAR trial of ≥30% reduction in daily BM for ≥50% of the time over 12 weeks¹⁸.

Patients maintaining Adequate Control on SSA + TE throughout the simulation continued therapy until study end or death. Patients receiving SSA who did not achieve treatment response were assumed to receive increasing doses of octreotide until loss of response and treatment discontinuation or death. SSA + TE patients who did not achieve Adequate Control were considered to discontinue TE treatment and receive SSA therapy only with increased dosing, mirroring the increased dosing of non-responders in the SSA cohort. As a rule, drug therapy for symptomatic conditions such as carcinoid syndrome is continued in patients who continue to derive tangible benefit, and otherwise discontinued when no benefit is evident or there are toxicity concerns. This assumption is consistent with health economic analyses in immunology, and has been demonstrated in the case of ulcerative colitis²³. A schematic overview of the included health states and transitions is presented in Figure 1.

Figure 1. Markov model health states and patient transitions. Abbreviations. SSA, somatostatin analog; TE, telotristat ethyl.

The structure of the model was designed to reflect current clinical practice for patients receiving SSA therapy and utilized SSA + TE treatment efficacy based on the TELESTAR trial. Patients with adequate response to initial SSA + TE treatment remained in this state through the simulation or until death, according to the durability of SSA + TE treatment response demonstrated in the open-label extension of the TELESTAR trial¹⁸. A proportion of patients receiving SSA with an initially adequate response to treatment experienced a loss of response to octreotide, as reflected in clinical experience such as that reported by Toumpanakis et al.²⁴. Patients experiencing inadequate control with SSA received dose escalations in line with standard clinical practice to maintain SSA therapy beyond recommended dosing levels, despite non-response²⁵. Approximately 40% of carcinoid patients are reported to receive SSA above the indicated dosing frequency or 30 mg maximum dose in efforts to manage uncontrolled symptoms, tumor progression, elevated u5-HIAA levels, or other factors²⁶.

Transition probabilities

The base case analysis placed 44% of SSA + TE and 20% of SSA patients in the Adequate Control state according to the 12-week TELESTAR responder analysis¹⁸. Patients in the SSA cohort were assumed to move into the Not Adequate Control state at a constant rate of 0.01 every 4 weeks, based on Toumpanakis et al.’s²⁴ report of 3-month probability of inadequate control with octreotide therapy.

The probability of death in this cohort was calculated using u5-HIAA as a predictor of mortality, based on an analysis in patients with mid-gut carcinoid tumors²⁷. This regression model was used to predict the probability of death for each cycle, health state, and comparator, assumed to estimate mortality rates for an individual with age and u5-HIAA level equivalent to the mean values reported in TELESTAR. Age-specific mortality rates were estimated and averaged according to the age distribution of TELESTAR patients. The 12-week TELESTAR u5-HIAA levels were applied to model cohorts where SSA patients were assigned a mean of 54.83 mmol/mol creatinine and SSA + TE patients a mean of 36.27 mmol/mol¹⁸. Since serious adverse events and related discontinuations were similar among TELESTAR treatment groups and were not expected to significantly impact quality-of-life or healthcare costs, adverse event-related health states and transitions were not included in this analysis.

Costs

All costs were based on an analysis of administrative claims from a large US health insurance plan including members with carcinoid syndrome from both commercial and Medicare Advantage plans²⁸. Evidence of SSA dose escalation was used to determine SSA treatment responsiveness and subsequent healthcare resource utilization according to the Burton and Lapuerta²⁸ analysis. Resource use estimates, excluding drug costs, were assumed to be identical for SSA + TE and SSA patients with adequate or inadequate response to treatment (Table 1).

Table 1. Base-case resource utilization and treatment input costs.

Intervention	Adequate control cost	Not adequate control cost
Resource utilization
Outpatient visit*	$650.39	$1,128.19
Outpatient pharmacy	$176.92	$597.25
Emergency department visit	$11.68	$13.85
Inpatient stay	$887.49	$1,464.32
Other medical encounters	$74.38	$156.95
Total resource use costs	$1,800.86	$3,360.56
Treatments
SSA, octreotide (24.32 mg average monthly dose)	$4,291.75	$5,890.57
Telotristat ethyl (21,000 mg average monthly dose)	$5,164.32	$5,164.32
SSA + TE	$9,456.07	$5,890.57 (discontinued TE, SSA only)

* Outpatient visit cost does not include SSA physician service and drug costs in order to avoid duplication of SSA costs in the analysis.

Drug costs for SSA (octreotide LAR) were based on the average monthly dose identified in the analysis conducted by Burton and Lapuerta²⁸ for adequately controlled (24.32 mg) and not adequately controlled (33.38 mg) patients. In the absence of information on the impact of telotristat ethyl on octreotide dosing, it was estimated that SSA dosing was equivalent for all adequately controlled SSA and SSA + TE patients. The recommended dose for telotristat ethyl is 250 mg, three times daily¹⁷. Patients continuing SSA + TE treatment were assumed to remain persistent and, thus, require the full cost of therapy in that cycle. The unit cost of octreotide was $176.47, based on the lowest calculated price per milligram, according to the published wholesale acquisition cost (Optum, EncoderPro, February 2007). The cost of telotristat ethyl was $61.48 per 250-mg unit, or $5,164 per 4-week cycle.

Utility

Overall, the frequency of CSD-related BM and number of flushing episodes have been associated with lower health-related quality-of-life among patients with NETs²⁹; however, utility estimates based on controlled vs uncontrolled symptoms have not been published. This analysis used ulcerative colitis as a proxy for estimates of active disease and remission states with controlled or uncontrolled CSD. Utility estimates for these states were based on a time trade-off study of ulcerative colitis patients in the US yielding proxy values of 0.32 and 0.79 for uncontrolled and controlled carcinoid syndrome, respectively³⁰. Similar to that of CSD, the symptomatology of ulcerative colitis includes diarrhea, abdominal pain, urgency, weight loss, fever, and pain, and treatment goals are to relieve symptom severity in the short-term and avoid symptom flares in the long-term³¹. The US Food and Drug Administration’s guidance on clinical trial endpoints for ulcerative colitis emphasizes the value of patient-reported daily bowel movement frequency³¹, as was the primary efficacy endpoint of the TELESTAR trial. Untreated patients in the TELESTAR trial had 5.2–6.1 mean daily bowel movements at baseline, similar to stool frequency reported for untreated ulcerative colitis patients of 4–5 per day³², or where >90% have had one or more bowel movements than normal per day³³. The severity of ulcerative colitis related to bowel movement frequency includes a similar range, where mild disease may be indicated by <4/day, 4–6/day moderate disease, and >6/day severe disease, based on the presence of other symptoms³⁴. The similarity of symptomatology and expected response to treatment in the form of reduced bowel movement frequency between ulcerative colitis and CSD offered the closest proxy condition for this analysis.

Cost-effectiveness analysis

This analysis was conducted for patients using a lifetime time horizon to evaluate LY gains, lifetime costs, QALY gains, and incremental cost effectiveness ratios based on costs per QALY gained. A 3% discount rate was applied to costs and outcomes occurring beyond 1 year, based on common health economic analysis practices in the US³⁵. A WTP threshold of $150,000 per QALY gained was used as a benchmark for cost-effectiveness, based on the most recent ICER Value Assessment Framework cost-effectiveness threshold range³⁶. Probabilistic sensitivity analyses (PSA) were conducted with WTP thresholds more closely aligned with value frameworks used with orphan drugs for rare and ultra-rare conditions^21,22. Regarding the distributional assumptions of variables that entered the PSA, the gamma distribution was used for all cost-related inputs that tended to be right-skewed. The average SSA doses in the SSA and SSA + TE arms were assumed to follow a normal distribution. The beta distribution was used for those model inputs whose values were limited to a range of 0–1, such as probability and utility inputs. These analyses considered thresholds 2–3-times higher than the conventional value, in the range of $300,000–$450,000 per QALY gained. To assess the impact of base-case model assumptions, further sensitivity analyses were conducted related to resource utilization costs, treatment dosing and costs, and utility estimates (Table 2).

Table 2. Sensitivity analysis parameters for the cost-effectiveness analysis.

Parameter	Base case value	Low value	High value	Probabilistic sensitivity analysis
				Distribution	SD of the distribution
Costs
Ambulatory visit costs (AC)	$650.39	$520.31	$780.47	Gamma	$44.45
Emergency department visits (AC)	$11.68	$9.34	$14.02	Gamma	$3.38
Inpatient stays (AC)	$887.49	$709.99	$1,064.99	Gamma	$130.12
Other medical (AC)	$74.38	$59.50	$89.26	Gamma	$25.80
Outpatient pharmacy (AC)	$176.92	$141.54	$212.30	Gamma	$36.82
Ambulatory visit costs (CNA)	$1,128.19	$902.55	$1,353.83	Gamma	$47.76
Emergency department visits (CNA)	$13.85	$11.08	$16.62	Gamma	$2.11
Inpatient stays (CNA)	$1,464.32	$1,171.46	$1,757.18	Gamma	$296.66
Other medical (CNA)	$156.95	$125.56	$188.34	Gamma	$63.70
Outpatient pharmacy (CNA)	$597.25	$477.80	$716.70	Gamma	$73.26
SSA, octreotide LAR depot	$176.47	$141.18	$211.76	Gamma	$17.65
Average SSA dose
SSA only group (AC)	24.32	19.5	29.2	Normal	2.4
SSA only group (CNA)	33.38	26.7	40.1	Normal	3.3
SSA + TE group (AC)	24.32	19.5	29.2	Normal	2.4
SSA + TE group (CNA)	33.38	26.7	40.1	Normal	3.3
Treatment efficacy (12 weeks)
SSA patients with AC at end of trial	20.0%	16.0%	24.0%	Beta	0.02
SSA + TE patients with AC at end of trial	44.0%	35.2%	52.8%	Beta	0.04
Transition probability (4 weeks) from AC to CNA	1.0%	0.8%	1.2%	Beta	0.1%
Long-Term probability of mortality (every 4 weeks)	0.013	0.010	0.015	–	–
Utility
Utility (AC)	0.79	0.63	0.95	Beta	0.08
Utility (CNA)	0.32	0.24	0.36	Beta	0.03

Abbreviations. AC, adequate control; CNA, control not adequate; SSA, octreotide; SSA + TE, octreotide and telotristat ethyl.

Results

Cost-effectiveness analysis

The base case analysis according to a conventional disease-intervention value framework provided lifetime QALYs of 1.67 at a cost of $495,125 for the SSA cohort and 2.33 ($590,087) for SSA + TE, with an incremental QALY for SSA + TE of 0.66 at an additional cost of $94,962. The incremental cost per QALY gained was $142,545, making the telotristat ethyl-containing regimen cost-effective within the conventional value framework WTP threshold of $150,000 per QALY.

Sensitivity analyses

When considering WTP thresholds more appropriate to a rare or ultra-rare disease value framework, the probability of SSA + TE being cost-effective was high (Figure 2)^21,22. At a 2-fold WTP threshold of $300,000, the probability of SSA + TE being cost-effective was 99.1%, and at a 3-fold threshold of $450,000 it was 99.9%, compared to 57.2% at the conventional disease-intervention threshold of $150,000. One-way sensitivity analyses, which rank model inputs by their impact on the incremental cost-effectiveness ratio, showed the model was most sensitive to the utility value of patients achieving adequate control (Figure 3) and, to a lesser degree, to the utility value assigned to inadequate control. The incremental cost-effectiveness ratio was also sensitive to the number of SSA units administered for patients not achieving adequate control, SSA + TE units administered, and patient responsiveness to treatment (probability of achieving adequate control for each regimen).

Figure 2. Cost-effectiveness acceptability curve. Abbreviations. SSA, somatostatin analog; TE, telotristat ethyl; WTP, willingness to pay.

Figure 3. One-way sensitivity analysis results. Abbreviations. AC, adequate control; CNA, control not adequate; SSA, somatostatin analog; TE, telotristat ethyl.

Discussion

The TPH inhibitor telotristat ethyl has demonstrated clinically meaningful reductions in BM frequency and u5-HIAA in patients with CSD not achieving adequate response to SSA therapy alone¹⁸. The efficacy of telotristat ethyl in reducing daily BM was sustained with no new safety events in an open-label extension trial¹⁸. In the reality of limited resources and high demand for quality care with meaningful health outcomes, healthcare decision-makers must contemplate the efficiency of health technology innovations in addition to their clinical benefits and risks. The prevalence and impact of disease are central to such assessments. This analysis demonstrated the cost-effectiveness of telotristat ethyl for patients with carcinoid syndrome not adequately controlled with SSA therapy alone according to both conventional and rare disease value assessment frameworks. The cost-effectiveness of the SSA + TE regimen was largely driven by improved symptom reduction and quality-of-life demonstrated in the phase III clinical trial program.

The prevalence of NETs is ∼100,000 in the US². Approximately 19% of patients with NETs have CS, and a small fraction of CS patients have associated diarrhea⁵. Although rare, the occurrence of CS continues to increase^1,2,5. The occurrence of diarrhea, as measured by daily bowel movements for patients with carcinoid syndrome, is prominent and costly, as patients tend to be of working age (mean = 51.5 years) with comorbidities⁹. CS patients with diarrhea have a nearly 50% greater odds of any hospitalization and 112% greater odds of carcinoid syndrome-related hospitalization, incurring $27,334 more in medical costs than matched counterparts without diarrhea⁹. Unfortunately, nearly half (40%) of the patients given the historical standard of care for CSD do not achieve adequate response, leaving significant medical and health economic unmet needs for this rare condition²⁸.

The interpretation of value in rare conditions with frequently occurring, expensive complications requires consideration of the impact on patients’ quality-of-life and survival, in addition to the efficiency of such meaningful improvements. The base-case incremental cost per QALY gained with telotristat ethyl of $142,545 was within the ICER threshold of $150,000 for conventional disease assessments, with a 57% probability of cost-effectiveness at this threshold. This analysis further demonstrated a very high probability (>99%) of cost-effectiveness with telotristat ethyl treatment within WTP thresholds more suited to value frameworks associated with rare and ultra-rare conditions of $300,000–$450,000. Since the model was most sensitive to the utility of patients achieving adequate control, which was derived from the published literature for ulcerative colitis, we varied this estimate in the sensitivity analyses and in threshold analyses. This allowed us to determine the minimum incremental utility required for TE to be a cost-effective treatment option, in addition to standard of care SSA therapy. The incremental utility values of >0.44, > 0.21, or >0.14 were required for the SSA + TE to be cost-effective at $150,000, $300,000, and $450,000 thresholds.

This cost-effectiveness analysis must be interpreted with certain considerations. It should be noted that the recent availability of telotristat ethyl has precluded a longer historical calculation of real-world practice patterns. The exact impact on healthcare resource utilization of telotristat ethyl, which may differ from that of SSA, cannot yet be quantified. The lack of published utility estimates in this rare condition required the use of proxy estimates from the case of patients with ulcerative colitis, which held notable influence on the sensitivity of results. Even though the disease symptomatology, burden, and treatment expectations are similar, further research on the utility associated with CSD specifically is needed. The treatment response threshold of ≥30% reduction in daily BM frequency was determined to be clinically significant from interviews with TELESTAR participants^37,38; however, we were not able to test other response thresholds in this analysis, which may be considered in future research. Furthermore, the present analysis is US-specific and, given the variability in practice patterns, SSA dosing, and drug acquisition costs, the findings should not be generalized to healthcare settings in other countries. Local adaptations of this model may be an interesting topic for future research.

This cost-effectiveness analysis was based on the results of the first and only randomized controlled double-blind clinical trial for CSD, in which telotristat ethyl demonstrated clinically meaningful improvements in key markers of disease. Response to therapy for CSD can be subjective and defined in several ways; however, this analysis applied the pre-specified definition of disease response used in the telotristat ethyl clinical trial program.

Conclusions

This analysis demonstrated that the addition of telotristat ethyl is cost-effective when given to patients with CSD not adequately controlled with somatostatin analogs.

Transparency

Declaration of funding

This study was sponsored by Lexicon Pharmaceuticals, manufacturer of telotristat ethyl, which is approved for the treatment of carcinoid syndrome diarrhea.

Declaration of financial/other interests

VNJ and PL are employees of Lexicon Pharmaceuticals. JF provided writing support, which was funded by Lexicon Pharmaceuticals. MK and MM are employees of Optum and were hired as consultants by Lexicon Pharmaceuticals.

Acknowledgments

The authors wish to acknowledge Jeff Frimpter, MPH, and Communications Symmetry for medical writing support, which was funded by Lexicon Pharmaceuticals. The authors also wish to acknowledge Michele Kohli and Michael Maschio for their technical input on this economic model.