Hemostatic powder TC-325 as first-line treatment option for malignant gastrointestinal bleeding: a cost–utility analysis in the United Kingdom

Abstract

Background

Randomized controlled trials have shown that hemostatic powder (TC-325) results in greater immediate hemostasis and lower 30-day rebleeding rates than standard endoscopic therapy (SET) for management of malignant upper gastrointestinal bleeding (MUGIB). We explored whether TC-325 would be a cost-effective first-line option for patients with MUGIB compared with SET in the United Kingdom.

Methods

A decision tree was developed for patients with MUGIB, assessing initial therapy with TC-325 or SET over a 30-day period. Patients with failed initial hemostasis or a rebleed within 30 days underwent further endoscopic treatment, escalation to either transcatheter arterial embolization or surgery, or radiotherapy. Overall 30-day mortality was applied. Costs, in GBP, were based on the United Kingdom National Health Services costs for 2023/2024. Results were reported as incremental differences in cost, quality-adjusted life years (QALY), and net monetary benefit. Deterministic and probabilistic sensitivity analyses and scenario analyses were performed.

Results

The cost of treating MUGIB patients with TC-325 was £245.88 lower than treatment with SET, with an incremental increase of 0.001 QALYs. TC-325 remained a cost-saving approach in sensitivity and scenario analyses. Probabilistic sensitivity analysis revealed that TC-325 was more effective and cost saving in 80.1% of simulations (range 67.5%–98.63%).

Conclusions

Initial treatment of MUGIB with TC-325 compared with SET was more effective (higher primary hemostasis and lower 30-day rebleeding) and cost saving owing to the requirement for fewer interventions, readmissions, and length of stay. Additional studies are needed to address model uncertainties in the follow-up management of these complex patients.

Introduction

Acute upper gastrointestinal bleeding (UGIB) is common, with an incidence of 84–170 per 100 000 adults a year in the UK, resulting in approximately 70 000 annual admissions to UK hospitals [1] [2]. UGIBs incur a high financial and resource burden on the National Health Service (NHS), due to in-hospital costs, readmission rates, and post-discharge expenses; treatment of acute UGIB is estimated to cost over £155.5 million annually [3].

Malignant causes of gastrointestinal bleeding account for 4% of all UGIBs [4], and the prevalence is increasing due to advancements in the diagnosis and treatment of gastrointestinal cancers [5]. Endoscopic therapy for malignant bleeds can be technically challenging due to the large surface area of tissue requiring treatment, tissue friability, and possible underlying coagulopathy [6]. Data regarding the efficacy of standard endoscopic treatment (SET) for malignant UGIB (MUGIB) are variable, with primary hemostasis rates reported between 31% and 86%, and rebleeding rates between 28% and 80% [7] [8] [9] [10].

Topical hemostatic powders have been gaining popularity due to their ease of endoscopic application and the ability to apply them quickly and easily over a diffuse area while causing minimal tissue trauma [6]. A recent large-scale randomized controlled trial (RCT) demonstrated that use of the hemostatic powder TC-325 resulted in significantly greater immediate hemostasis (100%) and lower 30-day rebleeding rates (2%) compared with SET (68% immediate hemostasis and 21% 30-day rebleeding) in malignant gastrointestinal bleeds [11]. These results are consistent with trends noted in a prior pilot RCT on a different continent [12].

A potential limitation to the widespread use of TC-325 is the increased initial purchase costs compared with SET options. Even though the economic impact of using TC-325 as a first-line therapy in nonvariceal UGIB has been reported [13] and a cost analysis has been performed in MUGIBs [14], no formal cost-effectiveness analysis has specifically addressed MUGIB. We therefore performed a cost–utility analysis to determine whether using the hemostatic powder TC-325 would be a cost-effective first-line option for MUGIB compared with SET in the UK.

Methods

Model

A decision tree was developed in Microsoft Excel 2016 to estimate the overall costs and consequences of treating MUGIBs, with either SET or TC-325. SET for MUGIB includes the use of epinephrine, hemostatic clips, thermal coagulation, or argon plasma coagulation alone or in combination [4] [8] [9] [11] [15] [16] [17]. An NHS provider perspective was adopted with a time horizon of 30 days, under which no discount rate was applied as the time horizon was less than 12 months. The model begins with a hypothetical cohort of MUGIB patients (mean age 63.4 [SD 11] years, 60.4% male) [11], who are treated with either SET or TC-325.

In the model, patients with failed initial hemostasis ([Fig. 1]; see also Fig. 1s in the online-only Supplementary material) are treated with rescue TC-325 or are escalated to either transcatheter angiographic embolization or surgery [11] [12] [17] [18]. The decision tree includes a possibility of one rebleed within 30 days, following which patients can be treated with repeat endoscopy matching the initial treatment allocation (i.e. no crossover at rebleed), surgery, transcatheter angiographic embolization, or radiotherapy. The model assumes that all secondary treatments resolve the initial bleed and patients are at risk of rebleed. Additionally, the model assumes all patients are admitted as emergencies. Overall, 30-day mortality was applied and, implementing the notion of half cycle correction, death is assumed to occur on Day 15, with all costs incurred before death and utilities calculated up to Day 15. The clinical validity of the model was reviewed by five experienced clinicians (A.B., B.N., N.D.H., S.H., and A.T.). Results are reported as incremental differences in cost, quality-adjusted life years (QALYs), and net monetary benefit (NMB).

Fig. 1 Decision tree for assessing the cost-effectiveness of TC-325 for malignant upper gastrointestinal bleeds. Failure to achieve immediate hemostasis (red cross) or a downstream rebleed (red droplet) within 30 days results in further treatment. Overall 30-day mortality is applied. MUGIB, malignant upper gastrointestinal bleed; RTx, radiotherapy; Sx, surgery; TAE, transcatheter angiographic embolization; green check, hemostasis achieved/maintained; red cross, hemostasis not achieved; red drop, rebleed.

Literature search

A comprehensive literature search was conducted in PubMed and Cochrane Library (February 2024), following the guidance of the Cochrane Handbook for Systematic Reviews of Interventions [19]. Inclusion/exclusion criteria and search strings are presented in Table 1s and Table 2s in the online-only Supplementary material. Screening was performed by D.M.C., B.N., and A.T., and is reported according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (Fig. 2s). All identified studies were extracted using standardized data tables after a consensus agreement was reached.

Three RCTs were identified comparing TC-325 with SET specifically in malignant bleeds [11] [12] [20]. The studies were heterogeneous; one study included 30.5% of patients without an active bleed [20] (active bleeding is required for TC-325 use), another was a pilot RCT [12], and the final RCT, which included only patients receiving nonpalliative treatment, was the only study sufficiently powered to inform significance [11]. There were some discrepancies in SET compared with UK practice in one of the studies [20]. However, although all three studies were performed outside the UK, the patient demographics of these base case studies were comparable with patients in a UK registry [21] [22]. To provide the most generalizable results despite the heterogeneity between studies, the three RCTs were pooled to inform the base case [11] [12] [20] [23] ([Table 1]). To overcome some of the heterogeneity issues, the impact of using only the large, powered RCT on the cost-effectiveness outcome was explored in scenario analysis [11].

Table 1 Pooled data from randomized controlled trials.

Study	Immediate hemostasis, n/N (%)	30-day rebleed, n/N (%)
NR, not reported; SET, standard endoscopic treatment; TC-325, hemostatic powder.
Pittayanon et al. 2023 [11] (N = 106)
TC-325	55/55 (100)	1/48 (2.1)
SET	35/51 (68.6)	10/47 (21.3)
Chen et al. 2020 [12] (N = 20)
TC-325	9/10 (90.0)	2/10 (20.0)
SET	4/10 (40.0)	6/10 (60.0)
Martins et al. 2022 [20] (N = 59)
TC-325	28/28 (100)	9/28 (32.1)
SET	NR	6/31 (19.4)
Pooled (N = 175)
TC-325	92/93 (98.9)	12/86 (14.0)
SET	39/61 (63.9)	22/88 (25.0)

Clinical inputs

The clinical parameters and sources utilized are reported in [Table 2]. In brief, primary hemostasis, 30-day rebleed, and mortality were used directly from the pooled RCT data ([Table 1]). Given the short 30-day time horizon, death was not adjusted for. The base case RCTs reported limited follow-up information regarding downstream hemostatic treatments. Therefore, data from wider papers, with longer follow-ups of malignant bleeds, were pooled to inform downstream treatments following the initial hemostatic failure [11] [12] [17] and rebleeding [8] [9] [15]. These data sources informed transition probabilities for surgery and repeat endoscopy, with transcatheter angiographic embolization probabilities being calculated by limiting total probabilities to 1. Downstream treatments were assumed to be the same for both arms.

Table 2 Base case, clinical parameters.

	Base value	Lower value	Upper value	Distribution	Source
LOS, length of stay; SET, standard endoscopic treatment; TAE, transcatheter angiographic embolization; TC-325, hemostatic powder. 1Length of stay is used only to inform utilities not costs.
Transition probabilities; TC-325
Primary hemostasis	0.9892	0.9607	0.9979	Beta	[11] [12] [20]
30-day rebleed rate	0.1395	0.0757	0.2198	Beta	[11] [12] [20]
Transition probabilities; SET
Primary hemostasis	0.6339	0.5160	0.7541	Beta	[11] [12] [20]
30-day rebleed rate	0.2500	0.1658	0.3450	Beta	[11] [12] [20]
Transition probabilities; TC-325 and SET
30-day mortality	0.1515	0.1011	0.2099	Beta	[11] [20]
Surgery post-failure	0.0345	0.0009	0.1234	Beta	[11] [12] [17]
Repeat endoscopy post-failure	0.9310	0.8165	0.9912	Beta	[11] [12] [17]
Surgery post-rebleed	0.1301	0.0769	0.1947	Beta	[8] [9] [15]
TAE post-rebleed	0.0650	0.0287	0.1146	Beta	[8] [9] [15]
Radiotherapy post-rebleed	0.0488	0.0183	0.0931	Beta	[8] [9] [15]
Rebleed post-TAE	0.1600	0.0564	0.3072	Beta	[8] [9] [15]
Rebleed post-surgery	0.0752	0.0040	0.1440	Beta	[10] [24]
Utilities and LOS
In-hospital utility	0.45	0.34	0.57	Beta	[25]
Discharge utility	0.78	0.70	0.85	Beta	[25]
LOS surgery1	12.88	6.70	21.30	Gamma	[26]
LOS TAE1	5.20	3.50	7.60	Gamma	[26]
LOS endoscopy	6.71	5.30	11.70	Gamma	[26]
LOS radiotherapy1	2.00	1.00	9.00	Gamma	[27]
Health resource use costs
Endoscopy, TC-325	£1072.00	£964.50	£1515.5	Gamma	[28]
Endoscopy, SET	£514.63	£355.21	£930.21	Gamma	[28]
Bed day cost endoscopy	£298.00	£268.20	£327.80	Gamma	[28]
TAE procedure	£5243.14	£4387.00	£5889.00	Gamma	[28]
Surgical procedure	£10 776.9	£5675.00	£13 851.00	Gamma	[28]
Radiotherapy	£1220.00	£870.00	£1700.00	Gamma	[28]

The primary effectiveness measure was the QALY, which incorporates health-related quality of life and mortality. Utilities were estimated from a UK-based analysis of patients with acute UGIB [29], utilizing the EuroQol EQ-5D instrument. This provided utility values for a patient at home (after discharge), and a patient in the hospital, with upper and lower limits [29]. No adjustments were made to the inpatient utility based on treatments. To quantify the in-hospital utility, the duration of hospital stay was calculated by summing the length of stay associated with the total pathway of procedures and capped at 30 days. Any difference from 30 days was assumed as the time the patient spent discharged and at home with the at-home utility applied. Length of stay for different procedures was derived from hospital-admitted patient care activity data reported in NHS Digital [25] for the general population based on procedural codes reported in Tables 3s–7s.

Costs

The healthcare resource use associated with the initial endoscopic treatment of the acute bleed was derived from the NHS England tariff for diagnostic endoscopy. The tariff includes healthcare professional costs, endoscopy suite time, and general consumables for an endoscopy, although it does not include treatment consumables or length of stay. Consumable cost for SET is the weighted use of epinephrine, hemostatic clips, thermal coagulation, argon plasma coagulation, and combinations of these treatments according to the reported use for malignant bleeds [4] [8] [9] [11] [15] [16] [17]. TC-325 prices were provided by Cook Medical (Limerick, Ireland). Length of stay cost for index endoscopy was based on the excess day’s trim point costs according to the tariff, and the mean length of stay for therapeutic endoscopic procedure codes was based on hospital episode statistic data [25]. Healthcare procedure costs for surgery and embolization were derived from weighted nonelective NHS tariff costs in 2023/24, based on the frequency of procedure codes from 2021/22 [3] [25] [26] and weighted according to comorbidities [28]. Radiotherapy costs were an average of tariff costs for the delivery of 10 fractions [30]. To prevent double counting of the length of stay, it was assumed the tariff sufficiently covered bed stay costs for surgery, embolization, or radiotherapy. Details of procedure codes and tariffs can be found in Tables 3s–7s.

Cost-effectiveness analysis

The results of the cost–utility analysis are reported as incremental costs (2023/24 cost year), QALYs, and NMB. Incremental costs and QALYs are the difference in costs and QALYs for the TC-325 arm compared with the SET arm. The incremental cost-effectiveness ratio (ICER) is the incremental costs divided by the difference in QALYs. A negative incremental cost indicates a cost saving. The ICER is reported against the National Institute for Health and Care Excellence recommended willingness-to-pay threshold of £20 000 per QALY [27]. A cost-effectiveness acceptability curve demonstrates uncertainty around cost-effectiveness at varying willingness-to-pay thresholds. NMB, the difference in net monetary benefit between a new intervention and the standard interventions, was calculated at specific willingness-to-pay thresholds by multiplying the incremental difference in QALY by the willingness-to-pay threshold and subtracting the incremental difference in costs. A positive NMB indicates cost-effectiveness [31].

Sensitivity analysis

Deterministic sensitivity analysis and probabilistic sensitivity analysis were performed to assess the model’s robustness. Deterministic sensitivity analysis was conducted by varying input parameters within plausible bounds, and the impact of these changes on the total incremental cost and NMB are presented as tornado plots. Probabilistic sensitivity analysis, using a Monte Carlo simulation, was conducted to assess the simultaneous impact of uncertainty around key parameters. All cost, probability, mortality, and utility variables were included. Transition probabilities of upper and lower boundaries were calculated as 95%CIs ([Table 2]). The lower cost of the endoscopy procedure was the current diagnostic tariff minus 10%, and the upper cost was the emergency procedure rate for this tariff. The lowest and highest identified costs for SET consumable costs informed upper and lower bounds for SET, while TC-325 device costs varied by ±10%. The estimated length of stay for the procedures was varied by the upper and lower limits reported by NHS England [26] and the bed stay cost varied by ± 10% of the base case value. The costs of surgery, transcatheter angiographic embolization, and radiotherapy were the lowest and highest identified tariffs, respectively [26]. The probabilistic sensitivity analysis was run for 1000 iterations, and incremental costs in GBP for 2023 were plotted against incremental QALYs.

Scenario analysis

Multiple scenarios were investigated to explore uncertainties around base case assumptions including, 1) use of the large, powered RCT of patients undergoing nonpalliative treatment (Eastern Cooperative Oncology Group performance status 0–2 only), 2) varying costs by upper or lower gastrointestinal location, 3) using standard tariff cost in place of emergency cost, 4) mortality determined by downstream interventions, 5) use of median length of stay for index endoscopy, and 6) using therapeutic endoscopy tariff in place of diagnostic tariff. Further details are available in Tables 8s–18s.

Validation

In addition to the scenario analyses exploring the uncertainty around the cost assumptions, the overall per-patient cost predicted by this model was validated compared to the estimated UK bleed costs reported previously [3], inflated to 2022 prices using the NHS inflation index [32].

Results

Cost-effectiveness

Total costs for treating MUGIBs over a 30-day time horizon were lower for TC-325 than for SET ([Table 3]). The TC-325 pathway was £245.88 less costly than SET per patient (5.4% reduction). The base case analysis indicates a gain of quality of life of 0.001 QALYs when using TC-325 compared with SET. As the model predicted TC-325 to be cost saving and QALY gaining, the estimated ICER predicted TC-325 as dominant over SET ([Table 3]). Based on these costs and consequences, the NMB was estimated at £265.63, at a maximum willingness-to-pay threshold of £20 000 per QALY.

Table 3 Costs and quality-adjusted life years; base case and scenario analyses.

	Cost, £	QALY	ICER	NMB, £	Probability of being cost-effective at £0 WTP, %
GI, gastrointestinal; ICER, incremental cost-effectiveness ratio; NMB, net monetary benefit; QALY, quality-adjusted life year; SET, standard endoscopic treatment; TC-325, hemostatic powder; WTP, willingness to pay.
Base case
TC-325	4324.78	0.0503	Dominant	265.63	80.1
SET	4570.66	0.0493	–	–	–
Increment	–245.88	+0.0010	–	–	–
Scenario 1 – Nonpalliative care patients
TC-325	3779.51	0.0516	Dominant	617.29	98.3
SET	4365.14	0.0496	–	–	–
Increment	–585.63	+0.0016	–	–	–
Scenario 2 – Upper and lower GI bleeds
TC-325	3709.81	0.0521	Dominant	182.31	71.2
SET	3874.25	0.0512	–	–	–
Increment	–164.44	+0.0009	–	–	–
Scenario 3 – Elective care costs
TC-325	4254.63	0.0503	Dominant	148.80	67.5
SET	4383.68	0.0493	–	–	–
Increment	–129.05	+0.0010	–	–	–
Scenario 4 – Downstream procedure mortality
TC-325	4324.78	0.0525	Dominant	278.17	79.5
SET	4570.66	0.0509	–	–	–
Increment	–245.88	+0.0016	–	–	–
Scenario 5 – Median LOS for index endoscopy
TC-325	2696.43	0.0552	Dominant	139.95	68.4
SET	2822.93	0.0545	–	–	–
Increment	–126.51	+0.0007	–	–	–
Scenario 6 – Therapeutic endoscopy tariff costs
TC-325	2626.70	0.0503	Dominant	140.72	71.7
SET	2748.19	0.0493	–	–	–
Increment	–121.50	+0.0010	–	–	–

Sensitivity analysis

The results for the top 15 parameters that impact the deterministic sensitivity analysis are presented in a tornado diagram in [Fig. 2], where the central line indicates base case incremental costs, and in Fig 3s for NMB. The parameters exerting the most influence were the probability of 30-day rebleeding for both SET (incremental cost –£631 to £95) and TC-325 (incremental cost –£540 to £121). All other parameter variations continued to return an incremental cost saving and a positive NMB for TC-325.

Fig. 2 Tornado plot showing the influence of increasing or decreasing the top 15 key variables on incremental costs. TAE, transcatheter angiographic embolization; LOS, length of stay.

The results of 1000 Monte Carlo iterations for the probabilistic sensitivity analysis are presented in a cloud diagram ([Fig. 3]). At a willingness-to-pay threshold of £20 000 per QALY, approximately 82.0% of the simulations are within this threshold. The average probabilistic estimate also predicted a dominant ICER in the southeast quadrant of the cost-effectiveness plane. At a willingness-to-pay of £0, TC-325 had an 80.1% probability of being cost effective, i.e. cost saving.

Fig. 3 Cost-effectiveness plane, demonstrating 1000 Montecarlo simulations (dots), deterministic result (square), probabilistic result (triangle), and willingness-to-pay threshold (line).

Scenario analysis

Scenario analyses explored the use of different source data: immediate hemostasis and 30-day rebleeding, procedure costs, mortality assumptions, length of stay for endoscopic procedures, and cost assumptions for endoscopy. Each scenario continued to predict a QALY gain with a cost saving (£121.50 to £585.63) for TC-325 compared with SET, and TC-325 continued to be dominant in all scenarios, with probabilities of being cost-effective at £0 willingness-to-pay ranging from 67.5% to 98.3% ([Table 3]).

Overall cost validation.

The overall estimated per-patient cost for the 30 days following an acute MUGIB, ranged from £2626.70 to £4324.78 for TC-325 and from £2748.19 to £4570.66 for SET, with a mean overall estimated cost of £3789. The estimated UK bleed costs reported previously [3], inflated to 2022 prices using the NHS inflation index [32], estimated the average cost of any acute bleed, including bleeds not requiring any treatment, at £2855.58.

Discussion

The increasing body of clinical evidence supporting the use of TC-325 as initial monotherapy for MUGIBs [11] [12] [22] is raising timely questions about the cost-effectiveness of such a treatment approach. Given the increased purchase cost of TC-325 compared with SET, we sought to explore, through formal cost–utility analysis, whether the reported improved immediate hemostasis and reduced 30-day rebleed would make TC-325 a cost-effective first-line option for treating patients with MUGIB in the UK. These findings indicate that, compared with SET, initial treatment of MUGIB with TC-325 is both less costly and increases the overall quality of life for patients with a malignant bleed. This model predicts a dominant ICER for TC-325 across a wide range of plausible willingness-to-pay thresholds, including a threshold of zero, indicating that TC-325 is not only cost effective but also cost saving. With reported significant improvement in immediate hemostasis and 30-day rebleeding [11] [12], it is not surprising that TC-325 is cost effective given that the costs associated with prolonged admission and readmission in this patient population are high compared with the purchase cost of TC-325 [3] [33].

Currently, only one other formal cost-effectiveness analysis of TC-325 exists, exploring the position of TC-325 in the treatment pathway for acute nonvariceal upper gastrointestinal bleeds in the USA [13]. The authors reported that adding TC-325 to traditional endoscopic treatment was less costly than traditional endoscopic treatment alone in these patients [13]. Although the authors included a subgroup for MUGIBs in the model, they did not specifically report on the impact of treating patients with malignant bleeds. Furthermore, the authors acknowledged that a limitation of their study was that the data were sourced from limited single-arm studies [13]. Recently, Shah and Law explored the cost of rebleeding in a cost analysis for MUGIB in the USA [14]. The authors explored bleed based on location; however, they did not incorporate the impact of failed immediate hemostasis, initial treatment costs, or the impact on patients’ quality of life. Our work adds significantly to the body of evidence in that it is the first cost–utility analysis in MUGIB reporting on ICER per QALY from the UK NHS perspective. This utilizes recent RCT efficacy data, valid cost sources, validated cost estimates, and multiple sensitivity and scenario analyses, thus providing robust conclusions.

Deterministic sensitivity analysis and probabilistic sensitivity analysis confirmed the robustness of the initial findings. The deterministic sensitivity analysis demonstrated that the incremental costs and NMB are largely unaffected by changes in the unit cost of the TC-325. The univariate deterministic sensitivity analysis demonstrated that cost-effectiveness is sensitive to several input parameters. Given that the base case studies have variable rebleeding rates, it is not surprising that cost-effectiveness is most responsive to the probability of rebleeding in both the TC-325 and SET arms. However, despite the results of deterministic sensitivity analysis, the probabilistic sensitivity analysis indicated that the probability of TC-325 being cost effective is 82.0% at the UK willingness-to-pay threshold of £20 000 per QALY gained; furthermore, at a willingness-to-pay threshold of £0, the incremental costs of TC-325 continue to have a probability of falling below the threshold of £0 (80.1%), implying a high likelihood of TC-325 being cost saving compared with SET.

Multiple scenarios utilizing different input values and sources were explored. Utilizing data from the large, powered RCT in patients receiving nonpalliative treatment [11] reduced the heterogeneity of the patient population and variances in practice, and resulted in a higher incremental cost saving, indicating an upper saving value whereby TC-325 has very low rebleeding rates. Scenarios exploring other cost input parameters yielded lower cost savings; however, all scenarios continued to report a cost saving and QALY gain with TC-325 treatment, further validating the robustness of this model.

Costs of an acute gastrointestinal bleed in the UK reported by Campbell et al. [3] and inflated to 2022 costs, enabled external validation of the model. Our model predicted patients’ costs to be between £2626 and £4570, and a mean cost of £3789. The inflated estimate from Campbell et al. (£2855) is within the bounds of our current model. The inflated estimated per-patient cost is likely to be an underestimate compared with this model, as the real-world data in the Campbell et al. study included the cost for patients where 14% of patients received no treatment, 57% underwent diagnostic endoscopy, and only 29% underwent a therapeutic endoscopy [3]. The patients in the current model all underwent a therapeutic endoscopy; hence, it is not surprising that the per-patient cost predicted in this model exceeds that of the inflated value by Campbell et al. [3]. In addition, the current model was built specifically for malignant bleeds, which occurred in only 3% of cases in the study by Campbell et al. Additional costs would be expected in the current model due to the increased length of stay associated with this particularly complex and comorbid group of patients. Indeed, Campbell et al. reported an average length of stay of 5.34 days compared with 8–11 days for MUGIBs in the Pittayanon study [3] [11]. Nonetheless, the fact that the scenarios in this model encompass the values predicted supports the methods used here, and the base case estimates are not too far outside the Campbell et al. estimate, further validating our model.

Economic dominance, overwhelming our findings across broad scenario analyses, makes the take-home message likely correct, but this work has some limitations. This conservative model means costs could be underestimated. For example, the model assumes hemostasis following surgery or embolization, which will underestimate any costs due to the failure of these practices [10] [24]. Similarly, the model cannot incorporate costs due to readmissions associated with a rebleed, thus underestimating these costs, and presenting a conservative estimate of cost savings.

A 30-day time horizon was used in this model, which, while suitable to answer the short-term impact, does not address any possible long-term implications of rebleeding on patients’ quality of life or quantity of life. Two long-term follow-up studies of patients with MUGIBs both report a significant increase in median overall survival in patients who did not rebleed compared with patients who did rebleed [8] [15]. None of the randomized studies have reported longer-term mortality data; thus, in the absence of reliable data, it was not feasible to extend this model to a 2-year time horizon.

The model was informed by RCTs performed outside the UK [11] [12] [20]. Results from contemporary observational studies in the UK support a high immediate hemostasis rate and low 30-day rebleeding rate with TC-325 used as a monotherapy [21] [22] adding credibility to this model. Unfortunately, these observational studies did not report on SET, preventing their use in this comparative analysis.

In the absence of utility data for patients with a malignant bleed, the utility data informing this model was from UK patients hospitalized with an acute UGIB; hence, the utility for both inpatients and those at home may be over- or underestimated, and there could be a more significant QALY gain than reported here due to the impact of hospitalization on such a fragile cohort of patients. Finally, there exists a lot of variability in managing patients with MGUIB, and the model could not capture all management schemes. Despite varying many parameters and obtaining consensus on model structure from experts, the model remains a pragmatic representation. Additional RCTs with better characterization of utilities and clarity on downstream interventions would help clarify these uncertainties and build future models.

Conclusions

The literature reports that initial treatment of patients presenting with MUGIB with TC-325 is more effective than treatment with SET, with higher primary hemostasis [22] and lower 30-day rebleeding [11]. The current work has demonstrated that using TC-325 as first-line treatment for MUGIB is likely to result in cost savings in the UK because fewer interventions are needed compared with SET. This work provides insights into the cost-effectiveness of TC-325 in the UK, and it would now be beneficial to assess the conclusions in other jurisdictions where cost structure and point estimates of health resource expenditures differ. Given the increasing body of evidence supporting the clinical efficacy [11] [12] [22], and now the cost-effectiveness, of TC-325, it would be prudent to consider this hemostatic powder as a first-line treatment for the management of MUGIB.

Conflict of Interest

D.M. Cooper is a salaried employee of Cook Medical, a Cook Group Company. R. Haidry has received educational grants to support research infrastructure from Medtronic Ltd., Cook Endoscopy (fellowship support), Pentax Europe, C2 Therapeutics, Beamline Diagnostic, and Fractyl Ltd. A. Barkun is a paid consultant for Cook Inc. B. Norton, N.D. Hawkes, S. Hebbar, A. Telese, and J. Morris declare that they have no conflicts of interest.

• References

• 1 Chatten K, Purssell H, Banerjee AK. et al. Glasgow Blatchford Score and risk stratifications in acute upper gastrointestinal bleed: can we extend this to 2 for urgent outpatient management?. Clin Med (Lond) 2018; 18: 118-122
• 2 National Institute for Health and Care Excellence. Acute upper gastrointestinal bleeding in over 16s: management. Clinical Guideline CG141. London: National Institute for Health and Care Excellence; 2012
• 3 Campbell HE, Stokes EA, Bargo D. et al. Costs and quality of life associated with acute upper gastrointestinal bleeding in the UK: cohort analysis of patients in a cluster randomised trial. BMJ Open 2015; 5: e007230
• 4 Savides TJ, Jensen DM, Cohen J. et al. Severe upper gastrointestinal tumor bleeding: endoscopic findings, treatment, and outcome. Endoscopy 1996; 28: 244-248
• 5 Minhem MA, Nakshabandi A, Mirza R. et al. Gastrointestinal hemorrhage in the setting of gastrointestinal cancer: anatomical prevalence, predictors, and interventions. World J Gastrointest Endosc 2021; 13: 391-406
• 6 Karna R, Deliwala S, Ramgopal B. et al. Efficacy of topical hemostatic agents in malignancy-related GI bleeding: a systematic review and meta-analysis. Gastrointest Endosc 2023; 97: 202-208.e208
• 7 Loftus EV, Alexander GL, Ahlquist DA. et al. Endoscopic treatment of major bleeding from advanced gastroduodenal malignant lesions. Mayo Clin Proc 1994; 69: 736-740
• 8 Song IJ, Kim HJ, Lee JA. et al. Clinical outcomes of endoscopic hemostasis for bleeding in patients with unresectable advanced gastric cancer. J Gastric Cancer 2017; 17: 374-383
• 9 Sheibani S, Kim JJ, Chen B. et al. Natural history of acute upper GI bleeding due to tumours: short-term success and long-term recurrence with or without endoscopic therapy. Aliment Pharmacol Ther 2013; 38: 144-150
• 10 Koh KH, Kim K, Kwon DH. et al. The successful endoscopic hemostasis factors in bleeding from advanced gastric cancer. Gastric Cancer 2013; 16: 397-403
• 11 Pittayanon R, Khongka W, Linlawan S. et al. Hemostatic powder vs standard endoscopic treatment for gastrointestinal tumor bleeding: a multicenter randomized trial. Gastroenterology 2023; 165: 762-772.e762
• 12 Chen YI, Wyse J, Lu Y. et al. TC-325 hemostatic powder versus current standard of care in managing malignant GI bleeding: a pilot randomized clinical trial. Gastrointest Endosc 2020; 91: 321-328.e321
• 13 Barkun AN, Adam V, Lu Y. et al. Using hemospray improves the cost-effectiveness ratio in the management of upper gastrointestinal nonvariceal bleeding. J Clin Gastroenterol 2018; 52: 36-44
• 14 Shah ED, Law R. Valuing innovative endoscopic techniques: hemostatic powder for the treatment of GI tumor bleeding. Gastrointest Endosc 2024; 100: 49-54
• 15 Kim YI, Choi IJ, Cho SJ. et al. Outcome of endoscopic therapy for cancer bleeding in patients with unresectable gastric cancer. J Gastroenterol Hepatol 2013; 28: 1489-1495
• 16 Schatz RA, Rockey DC. Gastrointestinal bleeding due to gastrointestinal tract malignancy: natural history, management, and outcomes. Dig Dis Sci 2017; 62: 491-501
• 17 Paoluzi OA, Cardamone C, Aucello A. et al. Efficacy of hemostatic powders as monotherapy or rescue therapy in gastrointestinal bleeding related to neoplastic or non-neoplastic lesions. Scand J Gastroenterol 2021; 56: 1506-1513
• 18 Gralnek IM, Stanley AJ, Morris AJ. et al. Endoscopic diagnosis and management of nonvariceal upper gastrointestinal hemorrhage (NVUGIH): European Society of Gastrointestinal Endoscopy (ESGE) Guideline – update 2021. Endoscopy 2021; 53: 300-332
• 19 Higgins JPT, Thomas J, Chandler J. et al. Cochrane Handbook for Systematic Reviews of Interventions. 2nd edn. Chichester: John Wiley & Sons; 2019
• 20 Martins BC, Abnader Machado A, Scomparin RC. et al. TC-325 hemostatic powder in the management of upper gastrointestinal malignant bleeding: a randomized controlled trial. Endosc Int Open 2022; 10: E1350-E1357
• 21 Hussein M, Alzoubaidi D, O’Donnell M. et al. Hemostatic powder TC-325 treatment of malignancy-related upper gastrointestinal bleeds: international registry outcomes. J Gastroenterol Hepatol 2021; 36: 3027-3032
• 22 Papaefthymiou A, Aslam N, Hussein M. et al. Hemospray (hemostatic powder TC-325) as monotherapy for acute gastrointestinal bleeding: a multicenter prospective study. Ann Gastroenterol 2024; 37: 418-426
• 23 Drummond MF, Sculpher MJ, Claxton K. et al. Methods for the economic evaluation of health care programmes. 4th edn. Oxford: Oxford University Press; 2015
• 24 Beggs AD, Dilworth MP, Powell SL. et al. A systematic review of transarterial embolization versus emergency surgery in treatment of major nonvariceal upper gastrointestinal bleeding. Clin Exp Gastroenterol 2014; 7: 93-104
• 25 NHS England. Hospital admitted patient care activity. 2022 Accessed December 09, 2024 at: https://digital.nhs.uk/data-and-information/publications/statistical/hospital-admitted-patient-care-activity/2022–23
• 26 NHS England. National tariff. 2023 Accessed December 09, 2024 at: https://www.england.nhs.uk/pay-syst/national-tariff/
• 27 National Institute for Health and Care Excellence. NICE process and methods [PMG9]: Guide to the methods of technology appraisal 2013. London: National Institute for Health and Care Excellence; 2013
• 28 Balic M, Hilbe W, Gusel S. et al. Prevalence of comorbidity in cancer patients scheduled for systemic anticancer treatment in Austria. memo – Magazine of European Medical Oncology 2019; 12: 290-296
• 29 Leontiadis GI, Sreedharan A, Dorward S. et al. Systematic reviews of the clinical effectiveness and cost-effectiveness of proton pump inhibitors in acute upper gastrointestinal bleeding. Health Technol Assess 2007; 11: iii-iv
• 30 Kondoh C, Shitara K, Nomura M. et al. Efficacy of palliative radiotherapy for gastric bleeding in patients with unresectable advanced gastric cancer: a retrospective cohort study. BMC Palliat Care 2015; 14: 37
• 31 Reed SD. Statistical considerations in economic evaluations: a guide for cardiologists. Eur Heart J 2014; 35: 1652-1656
• 32 Jones K, Weatherly H, Birch S. et al. Unit costs of health and social care 2022 manual. Technical report. Personal Social Services Research Unit (University of Kent) and Centre for Health Economics (University of York). 2023
• 33 Roy A, Kim M, Hawes R. et al. The clinical and cost implications of failed endoscopic hemostasis in gastroduodenal ulcer bleeding. United European Gastroenterol J 2017; 5: 359-364

Publication History

Received: 29 August 2024

Accepted after revision: 28 November 2024

Accepted Manuscript online:
03 December 2024

Article published online:
15 January 2025