Cost-effectiveness of ambroxol in the treatment of Gaucher disease type 2

Abstract Objective Our aim was to compare the costs and efficacy of ambroxol in combination with imiglucerase with the costs and efficacy of imiglucerase only in the treatment of Gaucher disease type 2 (GD2) in the socio-economic settings of the Republic of Serbia, an upper-middle-income European economy. Methods The perspective of the Serbian Republic Health Insurance Fund was chosen for this study, and the time horizon was 6 years. The main outcomes of the study were quality-adjusted life years gained with ambroxol + imiglucerase and comparator, and direct costs of treatment. The study was conducted through the generation and simulation of the Markov chain model. The model results were obtained after Monte Carlo microsimulation of a sample with 1,000 virtual patients. Results Treatment with ambroxol in combination with imiglucerase was cost-effective when compared with imiglucerase only and was associated with positive values of net monetary benefit regardless of the onset of the disease. Such beneficial result for ambroxol and imiglucerase combination is primarily driven by the low cost of ambroxol and its considerable clinical effectiveness in slowing the progression of neural complications of GD2. Conclusion If ambroxol and imiglucerase are used in combination for the treatment of GD2, it is more cost-effective than using imiglucerase alone.


Introduction
Gaucher disease is caused by the accumulation of glucocerebroside in lysosomes due to the lack of the glucocerebrosidase enzyme; it is inherited autosomal recessively.The disease has three phenotypic forms, designated as type 1, type 2, and type 3 [1].In type 1, which is much more common (about 94% of all patients with Gaucher disease have this type of disease), accumulation of glucocerebroside predominantly occurs outside the central nervous system, in organs rich in reticuloendothelial cellsliver, spleen, bone marrow, and lungs.Type 1 is also called the non-neuronal form of the disease, because the nervous system is not affected by pathological changes.In types 2 and 3, which are called by one name the neuronal form of Gaucher disease, the primary damage is to the central nervous system, while the parenchymatous organs are affected little or not at all [2].
When it comes to types 2 and 3, i.e., the neuronal form of the disease, they differ from each other in terms of the type of genetic disorder and the severity of the clinical picture.Type 2 is also called the acute form, because neurological disorders develop quickly, so patients rarely live longer than 5 years.Type 3 is also called the chronic form, and with it, the disease progresses much more slowly, so patients live longer [3].With type 2, we additionally distinguish two subgroups, formed on the basis of the onset of the disease.The first group consists of patients in whom the disease begins with signs and symptoms already visible at birth, and in the second group, the disease begins in infancy, that is, during the first year of life.While for types 1 and 3, there is a causal therapy that has shown effectiveness (enzyme replacement therapy, imiglucerase, velaglucerase, or taliglucerase), for type 2, it is mostly ineffective [4,5].A few years ago, in one pilot study [6], and then in several case reports, a positive effect of ambroxol on neurological disorders in type 2 was shown, when it was used together with enzyme replacement therapy (which prevents the progression of pathological changes in parenchymatous organs).Ambroxol is the so-called pharmacological chaperone, which protects damaged glucocerebrosidase molecules from degradation and thus prolongs their effect.Additional clinical studies are needed to determine the true effect of ambroxol, and it is necessary to look at the cost-effectiveness ratio of the drug.
Our aim was to compare the costs and efficacy of ambroxol in combination with imiglucerase with the costs and efficacy of imiglucerase only in the treatment of Gaucher disease type 2 (GD2) in the socio-economic settings of the Republic of Serbia, an upper-middle-income European economy.

Methods
The health economic analysis plan that the authors developed before the study's start was followed throughout the investigation.A full description of the study plan is available at: https://drive.google.com/drive/folders/1emX2mamfMqmrKd6c0AKdvRJiq8JKuwGA?usp=share_link.
The study setting was the healthcare system of the Republic of Serbia, which is an upper-middle-income European economy according to the World Bank [7].Patients with GD2 of both sexes, with the disease beginning at birth or during infancy, represented our study population.We examined the cost-effectiveness of two therapeutic options for the treatment of GD2: (1) experimental, off-label combination of ambroxol and imiglucerase and (2) monotherapy of imiglucerase, the only approved causal therapy of GD2 until now [8].The analysis was conducted from the perspective of the Republic Fund for Health Insurance (RHIF) of the Republic of Serbia [9].We have defined the horizon as a period of 6 years, with cycles of one month each, according to the natural course of GD2 and the fact that almost none of the patients survive more than 5 years [10].Table 1 shows the model inputs.
Our major outcomes were quality-adjusted life years (QALYs) gained with ambroxol + imiglucerase and imiglucerase only, and direct costs of treatment.Within direct costs, we have included the costs of drugs, diagnostic and therapeutic healthcare services, materials for healthcare, and costs of diagnosing and treating adverse drug reactions.Given that we conducted the analysis from the perspective of RHIF, as an institution that is the dominant insurance carrier, we took into account only direct costs, while we did not count indirect and intangible costs.We used published cost-effectiveness studies or clinical practice guidelines for the treatment of GD2 for resource utilization.For the determination of unit prices, we used the following sources: the Tariff book of the RHIF (for prices of health services in the Republic of Serbia) [35] and "List of Reimbursable Drugs" of the RHIF [36] or from the "Decision on Maximum Prices of Medicines of the Government of the Republic of Serbia" [33] (for unit prices of drugs).We expressed all costs in the Republic of Serbia Dinars (RSD), which is the official currency of the Republic of Serbia.The estimated resource quantities and unit costs related to year 2022.When converting costs from foreign currency to RSD, the National Bank of Serbia's middle exchange rate was utilized, which was in effect on the day the primary data were published or collected.To implement the results of our analysis in the future, we introduced a 5% annual discount rate for all costs and QALYs starting from the second year until the end of the horizon [37].The costeffectiveness of the therapeutic options was compared by calculating incremental cost-effectiveness ratio (ICER) and comparing it with threshold values, shown as lambda lines at the incremental cost-effectiveness plane.The threshold values were set at 1, 3, and 9 gross domestic products (GDPs) per capita per QALY saved; the lambda 3 line representing the threshold of 9 GDPs/QALY is the upper limit of acceptance by health insurance funds granted to orphan drugs only, taking into account difficulties of developing orphan drugs and limited return of investment due to small size of their market.
The study was conducted through the generation and simulation of the Markov chain model which describes course of the GD2 through a series of health states [38].The Markov chain model assumes that a virtual patient may spend any of the monthly cycles in one of the possible clinical states; at the end of each cycle, the patient may stay in the same state, or transit to one of the other available states.After assigning probabilities to each of the possible transitions, the model calculates the probability of being in each state in each cycle.Healthcare utilities and costs of the model states are multiplied by probabilities and then summed up for whole model horizon, giving as output total costs and QALYs for each virtual patient.Monte Carlo simulation uses random number generation to place every virtual patient in a certain state at each model cycle following previously calculated probability distribution [39].We developed our model in Microsoft Excel, version 2019, while we used Virtual Basic for the development of special Initial direct costs of state without complications 3931.00 ± 820,00 RSD Inverse gamma distribution (α = 16, β = 245.68)[25] Annual direct costs of tracheostomy state: 108236.92± 12800.00RSD Inverse gamma distribution (α = 16, β = 6764.81)[25] Initial costs of tracheostomy state: 30901.00± 3400.00RSD [25] (Continued) Cost-effectiveness of ambroxol for GD2  3 macros for the execution of model simulations.A graphic representation of the model is shown in Figure 1.
To test the model's robustness, we performed both probabilistic and one-way sensitivity analyses.The study population's heterogeneity was considered by conducting independent analyses on two patient subpopulations, as the rate of progression is directly influenced by whether GD2 was first detected at birth or during infancy.Gross domestic product per capita Gross domestic product per capita in Serbia 2021 917441.90RSD n.a.[34] *SD: standard deviation.On the other hand, for patients treated by monotherapy of imiglucerase: (1) average cost per patient was 14672000.42± 197051.56RSD (99% CI) and (2) average number of QALYs gained 1.34 ± 0.01.

Onset of the disease at birth
In comparison with imiglucerase as monotherapy, the use of a combination of ambroxol and imiglucerase was followed by a positive net monetary benefit (1622511.89± 247121.93RSD [99% CI]), while the value of the ICER per one more QALY gained was 6069649.10 ± 20628071.68RSD (99% CI).The ICER is displayed separately for each virtual patient in Figure 2a.The combination of ambroxol and imiglucerase was a cost/effective therapeutic option of GD2 evident at birth in comparison to imiglucerase alone.

Acceptability curve
The acceptability curve reveals the changes in the percentage of patients who fall below the current willingness to pay line in the ICER diagram under the condition that the willingness of RHIF to pay one more QALY gained increases from 200000.00 to 250000000.00RSD (Figure 3a).

One-way sensitivity analysis
The input variable values were varied by ±25% each, within the scope of a one-way sensitivity analysis.The net monetary benefit was then computed for each of the varied values.On the Tornado diagrams, we have shown the results for only the six most influential variables to make the graphics as clear as possible (Figure 4a).The results of the one-way sensitivity analysis showed that the effect of ambroxol on slowing the progression of neurological complications of the disease has the greatest impact on its cost-utility and net monetary benefit.Considering other variables, even if they take extreme values in both directions, ambroxol remains cost-effective, since the net monetary benefit stays positive.

Probabilistic sensitivity analysis (PSA)
For the PSA, values of the input variables were replaced with distributions, with the beta distribution being used for rate and utility variables and the gamma distribution for cost variables.After the Monte Carlo microsimulation, similarly dispersed values of output variables were recorded (Figure 5a), and their means with 99% confidence intervals are presented in Table 2.With consistently below-threshold values of ICER and positive values of net monetary benefit, the PSA confirmed that ambroxol in combination with imiglucerase is a cost-effective option for the treatment of GD2 when compared with monotherapy with imiglucerase.

Onset of the disease in infancy 3.2.1 Base case
The base case Monte Carlo microsimulation for 1,000 virtual patients treated by ambroxol + imiglucerase gave the In comparison with imiglucerase as monotherapy, the use of a combination of ambroxol and imiglucerase was followed by a positive net monetary benefit (1888149.51± 325318.82RSD [99% CI]), while the value of the ICER per one more QALY gained was −30202486.52± 63597201.79RSD (99% CI).The ICER is displayed separately for each virtual patient in Figure 2b.The combination of ambroxol and imiglucerase was a cost/effective therapeutic option for GD2 emerging in infancy in comparison to imiglucerase alone.

Acceptability curve
The acceptability curve reveals the changes in the percentage of patients who fall below the current willingness to pay line in the ICER diagram under the condition that the willingness of RHIF to pay one more QALY gained increases from 200000.00 to 250000000.00RSD (Figure 3b).

One-way sensitivity analysis
The input variable values were varied by ±25% each, within the scope of a one-way sensitivity analysis.The net monetary benefit was then computed for each of the varied values.On the Tornado diagrams, we have shown the results for only the six most influential variables to make the graphics as clear as possible (Figure 4b).The  results of the one-way sensitivity analysis showed that the effect of ambroxol on slowing the progression of neurological complications of the disease has the greatest impact on its cost-utility and net monetary benefit.Considering other variables, even if they take extreme values in both directions, ambroxol remains cost-effective, since the net monetary benefit stays positive.

Probabilistic sensitivity analysis (PSA)
For the PSA, values of the input variables were replaced with distributions, with the beta distribution being used for rate and utility variables and the gamma distribution for cost variables.After the Monte Carlo microsimulation, similarly dispersed values of output variables were recorded (Figure 5b), and their means with 99% confidence intervals are presented in Table 2.With consistently below-threshold values of ICER and positive values of net monetary benefit, the PSA confirmed that ambroxol in combination with imiglucerase is a cost-effective option for the treatment of GD2 when compared with monotherapy with imiglucerase.

Discussion
Our research revealed that, for the treatment of GD2 in the socioeconomic context of the Republic of Serbia, ambroxol in combination with imiglucerase was more cost-effective than imiglucerase monotherapy.Treatment of GD2 with ambroxol and imiglucerase in combination was clearly more effective and less costly than that with imiglucerase only.Such beneficial result for ambroxol and imiglucerase combination is primarily driven by the low cost of ambroxol  Cost-effectiveness of ambroxol for GD2  7 and its considerable clinical effectiveness in slowing the progression of neural complications of GD2.Ambroxol is not officially approved for the treatment of GD2 [8,40].It was identified in an experimental study from 2009 as a credible contender for a pharmacological chaperone for mutant glucocerebrosidase [41].In 2016, a pilot study investigating the efficacy and safety of ambroxol combined with imiglucerase in patients with neurological manifestations of Gaucher disease (GD2 and GD3) was published [6].The results of this study showed that the use of high doses of ambroxol in combination with enzyme replacement therapy was accompanied by satisfactory safety and clinical efficacy in patients suffering from neurological types of Gaucher's disease, since there was a significant improvement in myoclonus, a reduction in the frequency of epileptic seizures, and an improvement in other neurological manifestations [6].
Nonetheless, despite the encouraging preliminary results, it seems that pharmaceutical companies are not very interested in looking into whether ambroxol may be used to treat patients with neurological forms of GD [42].It is believed that pharmaceutical companies are not particularly interested in continuing clinical trials examining the efficacy of amborxol in patients with GD2 and GD3 due to amborxol's low costs and limited potential for revenue [32].This kind of approach by pharmaceutical companies has far-reaching consequences.First of all, this puts in jeopardy the fundamental humanitarian and egalitarian ideals that state that every patient has a right to the best medical care.[43].In addition, pharmaceutical companies that develop orphan drugs enjoy various tax and other incentives, which are missed in this case [44].
Meanwhile, ambroxol is reported to be used off-label in combination with enzyme replacement therapy for the treatment of children with GD2 and GD3 [14].Particularly, promising results came from a girl with GD2 who got ambroxol monotherapy at a daily dose of 25 mg/kg beginning at the end of her third month of life [8].The achieved effect of this therapy was impressive, since at the end of her first year of life; this girl achieved the neurocognitive and motor development that is expected for her age [8].Imiglucerase was introduced into therapy only at the 15th month of life, even 11 months after the administration of ambroxol [8].The combined use of ambroxol and imiglucerase had a long-lasting effect because the girl's good neurocognitive growth and satisfactory quality of life persisted even at age 3 [8].The results of our study, which showed the financial justification of using ambroxol in combination with imiglucerase in GD patients, should be a major impetus for parents of affected children as well as associations and foundations for rare diseases to exert the necessary pressure and raise the issue of the continuation of clinical trials.
Our study was limited by the fact that ambroxol is available over the counter and is not covered by the Republic Health Insurance Fund; as a result, we were forced to use the rates set by local vendors because Serbia did not yet have an official maximum price for this medication.

Conclusion
With the current price, ambroxol in combination with enzyme replacement therapy is a cost-effective option in comparison to the therapy of GD2 with only enzyme replacement.Considering the encouraging results of a growing number of individual patient cases that indicate satisfactory efficacy and safety of ambroxol as an additional therapy, the health insurance fund should reimburse ambroxol for this indication to all patients with this disease once the drug is officially approved for this indication.The findings of our study might encourage regulatory agencies and organizations for rare diseases to persuade pharmaceutical companies to continue clinical trials of ambroxol in patients with GD2.Author contributions: MNM: conceptualization, methodology, formal analysis, software, roles/writingoriginal draft; MG: conceptualization, methodology, formal analysis, roles/writingoriginal draft; VJ: conceptualization, methodology, formal analysis, writingreview and editing; SV: conceptualization, methodology, roles/writingoriginal draft; MDJ: conceptualization, methodology, roles/writingoriginal draft; RI: conceptualization, methodology, roles/ writingoriginal draft; JM: conceptualization, methodology, roles/writingoriginal draft; SJ: conceptualization, formal analysis, methodology, software, supervision, writingreview and editing.
Conflict of interest: MNM serves as Editor for Open Medicine, but it did not affected peer-review process.There is no other conflict of interest.
Data availability statement: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Figure 2 :
Figure 2:The base case ICERs for virtual patients (aonset of the disease at birth; bonset of the disease at infancy).The x-axis: difference in QALYs gained (combination of ambroxol and imiglucerase vs imiglucerase only); the y-axis: difference in costs (combination of ambroxol and imiglucerase vs imiglucerase only).The line lambda 1the RHIF's willingness to pay one GDP per capita for one more QALY gained with a combination of ambroxol and imiglucerase vs imiglucerase only.The line lambda 2the RHIF's willingness to pay three GDPs per capita for one more QALY gained with a combination of ambroxol and imiglucerase vs imiglucerase only.The line lambda 3the RHIF's willingness to pay nine GDPs per capita for one more QALY gained with a combination of ambroxol and imiglucerase vs imiglucerase only.

Figure 3 :
Figure 3: Acceptability curve (aonset of the disease at birth; bonset of the disease at infancy).

Figure 4 :
Figure 4: Tornado diagram (aonset of the disease at birth; bonset of the disease at infancy).

Figure 5 :
Figure5: The probability sensitivity analysis ICERs for virtual patients (aonset of the disease at birth; bonset of the disease at infancy).The x-axis: difference in QALYs gained (combination of ambroxol and imiglucerase vs imiglucerase only); the y-axis: difference in costs (combination of ambroxol and imiglucerase vs imiglucerase only).The line lambda 1the RHIF's willingness to pay one GDP per capita for one more QALY gained with a combination of ambroxol and imiglucerase vs imiglucerase only.The line lambda 2the RHIF's willingness to pay three GDPs per capita for one more QALY gained with a combination of ambroxol and imiglucerase vs imiglucerase only.The line lambda 3the RHIF's willingness to pay nine GDPs per capita for one more QALY gained with a combination of ambroxol and imiglucerase vs imiglucerase only.

Funding information :
The study was financially supported by the Serbian Ministry of Science, Technological Development and Innovations, Grant No 175007, contract No 451-03-66/2024-03/200111.

Table 1 :
The model inputs

Table 2 :
Values of main output variables before and after the probabilistic sensitivity analysis (mean ± 99% confidence interval) *ICER: incremental cost-effectiveness ratio.