Response to the Letter to the Editor concerning ‘Cost-effectiveness of velmanase alfa vs. bone marrow transplantation or no causal therapy in patients with mild to moderate alpha-mannosidosis’

Dear Editor,

We thank you for the opportunity to respond to the comments that the authors of the letter Stevenson et al. made about our work published in your journal [1,2]. We are also grateful to the authors of the letter, Stevenson et al. for carefully reading our paper and openly expressing their observations, criticisms, and suggestions; we believe that an open discussion in scientific journals regarding important topics, such as the cost-effectiveness (C/E) and availability of enzyme replacement therapy (ERT) for rare but severe diseases, is of great importance for reducing uncertainty that can improve health care. In the following text, we respond point by point to the most important objections in the letter mentioned above.

Stevenson et al. first pointed out that the model did not take into account the effects of ERT on the forced vital capacity of the lungs and the mobility of the patients, measured by the three-minute stair climb test (3MSCT) and the six-minute walk test (6MWT), particularly among pediatric patients, suggesting that this led to an underestimation of the clinical effects of ERT [1]. However, we deliberately did not take into account the mentioned effects because in the publicly available scientific literature, at the time of the model creation, there were no publications that would transpose the positive effects of the mentioned tests into concrete reductions in treatment costs or increases in quality of life. These tests are surrogate markers of actual drug effects and can only be quantitatively linked to hard outcomes (costs and quality-adjusted life years gained) once appropriate studies of sufficient duration are conducted. Precisely, in our study’s limitations, we stated that the long-term effectiveness of velmanase alfa (VA) is still unknown because most of the previously publicly available studies lasted up to 5–6 years. Faced with such a situation during model development, we decided not to introduce additional uncertainty into the model by considering the effects on the mentioned surrogate markers.

The second objection by Stevenson et al. was related to the lack of precise input data into the model concerning the time until the occurrence of a certain event, especially regarding the effect of velmanase alfa [1]. From the table (Table 1. ‘Model inputs for events onset’ from the reference 2) with input data of our model [2], it can be seen that for six events, we had precise data from clinical studies on the effects of VA [3,4]. For another six events, due to the lack of published data from clinical studies, we took the same extent of the effects for all three therapeutic options based on studies on the natural course of the disease without specified therapy in order not to unjustifiably favor any of the therapeutic options. For the remaining three events, we equated VA with the option ‘no causal therapy’ due to lack of published data, while for bone marrow transplantation, the existing published data were used. Please note that the 15 events used in the model were selected based on published studies describing the natural course of the disease.

The next objection of Stevenson et al. concerns the use of the average price of VA from the countries of the European Union because the drug was not authorized in Serbia and did not have an administratively established price [1]. Also, Stevenson et al. commented that the costs of certain events and their consequences are taken from other countries because they may not apply to the Serbian environment [1]. We have already stated in the limitations of our study that the drug does not have a marketing authorization and that we formed the price based on the average price in European countries because that is the rule in Serbia applicable to authorized drugs; in the case of unauthorized drugs, there is no strict rule on how the price of the drug is determined, but experience with other unauthorized drugs shows that manufacturers most often ask for and negotiate for a maximum drug price to the price level similar to prices in other European Union countries. In the absence of data from cost-of-illness studies in Serbia, one of the approaches to make the most accurate estimate is to take data from other countries and then adjust it according to the percentage of GDP (which we did), which is an established practice in pharmacoeconomic analysis. Please note that our study was purely a cost-effectiveness analysis without considering other aspects (budget impact, disease prevalence, market access, etc.). Also, we mentioned in the paper that when making a definitive decision on reimbursement, not only should the results of the cost-effectiveness analysis be considered, but other aspects as well.

We thank Stevenson et al. for pointing us to the positive results of the model submitted to the UK’s National Institute for Health and Care Excellence (NICE) for ‘treating alpha-mannosidosis in people under 18 years and in people who turn 18 while on treatment’ [1,5]. It is known that companies that prepare pharmacoeconomic models for Health Insurance Funds have access to numerous data that are not published or not publicly available, which reduces the uncertainty of their models (these data remain hidden/redacted even in the public reports of NICE or the other relevant health technology assessment agencies or bodies). At the time we built our model (first half of 2023), we used all the data we could find at the time, and we took into account mild to moderate patients of all ages without limiting our model to people under 18. Our model inevitably has more uncertainty in input data, but we tried to compensate it with a probabilistic and one-factor sensitivity analysis. Perhaps the most important factor that influenced the fact that VA is not cost-effective in Serbia is the significant difference in GDP between the UK and Serbia (GDP is about 4.5 times lower in Serbia), which is directly referred to in the cost-effectiveness threshold, which is also in Serbia several times lower than in the UK.

Finally, in our paper’s conclusions, we just acknowledged that VA is not cost-effective according to the standard C/E analysis, and we made no recommendation that this drug should not be financed and made available to patients in Serbia. On the contrary, we suggested that alternative cost-effectiveness assessment methods are needed to assess ERT cost-effectiveness, which could inform a final decision on whether this drug should be reimbursed in Serbia.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was funded by the Ministry of Science, Technological Development and Innovation, Republic of Serbia, through a Grant Agreement with the University of Belgrade – Faculty of Pharmacy No 451-03-65/2024-03/200161 and with the University of Kragujevac – Faculty of Medical Sciences No. 451-03- 65/2024-03/200111.