Abstract
Background Currently, only dexamethasone, tocilizumab and sarilumab have conclusively been shown to reduce mortality of COVID-19. No drug for prevention or treatment in earlier stages of COVID-19 are yet found; although several new candidates including molnupiravir, ivermectin, baricitinib, budesonide and fluvoxamine are being studied with some early promising results. Safe and effective treatments will need to be both affordable and widely available globally to be used alongside vaccination programmes. This analysis will estimate and compare potential generic production costs of a selection of COVID-19 drug candidates with available international list prices.
Methods Costs of production for new and potential COVID-19 drugs (dexamethasone, ivermectin, fluvoxamine, budesonide, baricitinib, tocilizumab, sarilumab and molnupiravir) were estimated using active pharmaceutical ingredients (API) data extracted from global shipping records. This was compared with national pricing data from a range of low, medium, and high-income countries. Annual API export volumes from India were used to estimate the current availability of each drug.
Results Repurposed therapies can be generically manufactured at very low per-course costs, ranging from $4.16 for fluvoxamine and $2.58 for IV dexamethasone (or $0.19 orally) to $0.55 for ivermectin. No export price data was available for baricitinib, tocilizumab or sarilumab. When compared against international list prices, we found wide variations between countries.
Conclusions Successful management of COVID-19 will require equitable access to treatment for all populations, not just those able to pay high prices. Analysed drugs are widely available and affordable, whilst IV treatment courses are more expensive.
Key Points
Re-purposed drugs under evaluation for COVID-19 must be affordable worldwide to compliment vaccine programmes.
Estimated costs/course were: ivermectin ($0.55), budesonide ($4.34), baricitnib ($6.67), molnupiravir ($255.57), dexamethasone ($0.22), tocilizumab ($410.59), sarilumab (875.70) and fluvoxamine ($4.16).
High drug prices may limit access.
Introduction
18- months from the initial COVID-19 outbreak, drug repurposing with intention to treat and prevent is still an important strategy alongside vaccine rollout. The uneven global vaccine deployment, “vaccine nationalism” [1], difficulties caused by vaccine hesitancy and emerging new variants of concern means that an excessively vaccine-dependent strategy may not be fully effective. Even in countries with high vaccine coverage, infected patients will still need treatment. For many developing countries, mass vaccination required to suppress transmission and achieve herd immunity may remain unattainable for several years.
Treatments, when they emerge from the vast number of clinical trials currently underway, will need to overcome the same barriers of efficacy, safety, affordability, and availability as vaccines. In our 2020 analysis of potential costs of generic production for drug candidates then under active investigation [2], we proposed four key tests:
Efficacy – drugs should demonstrate therapeutic clinical efficacy and/or prophylactic properties.
Safety – drugs should be safe and well tolerated in all populations with minimal side effects or drug interactions.
Affordability – drugs should be widely accessible as low-cost generics so that all populations can equitably afford the cost of treatment.
Availability – drugs must be readily available or have the potential for rapid up-scaled production to meet global demand.
Many previously investigated drugs have now been proven ineffective, or even harmful, for example hydroxychloroquine [3]. Large-scale studies such as SOLIDARITY and RECOVERY have found previously leading candidates such as lopinavir/ritonavir [4, 5], interferon beta-1a [4] and azithromycin [6] to have no clinical benefit, and despite FDA approval, remdesivir demonstrated no clinical benefit except for shorter duration of hospitalisation [4]. Colchicine is currently studied in PRINCIPLE [7], following preliminary evidence of efficacy from the published COLCORONA trial [8], although no efficacy was found recently by RECOVERY [9].
However, questions of affordability and availability should be equally important as efficacy and safety. Treatments are of limited value if their supply is limited and the cost prohibitive, as in the case of tocilizumab and sarilumab. These monoclonal antibody interleukin-6 (IL-6) receptor inhibitors were found to improve outcomes in critical care patients [10], but cost several hundred dollars per dose, and require complex bioreactors to produce. Conversely, dexamethasone, the first treatment found to increase survival and is widely available as a low-cost generic, is only applicable in critical care settings [11] - an effective anti-viral agent for asymptomatic, mild, or moderate patients that limits progression and treats COVID-19 is still absent.
We present an updated analysis to reflect recent developments in the clinical trial pipeline, with a view to pre-empt any efficacy and safety conclusions of drug candidates we deemed to be most promising based on current literature. Only drugs with early promising results to either lead to preventions of infection, or reduce mortality, with a pipeline to be tested in over 4000 patients were reviewed (Appendix 1).
As of July 2021, this includes ivermectin, an anti-parasitic and anti-inflammatory [12–14] being tested in over 102 studies in 46,974 patients [15], inhaled budesonide being tested in 11 studies in 3,569 patients [16, 17], baricitinib, a janus kinase (JAK) inhibitor being tested in 14 studies amongst 56,168 patients and mostly given together with remdesivir [18], molnupiravir, an experimental antiviral [19] being tested in 10 studies in 5,742 patients, and fluvoxamine, an oral selective serotonin receptor antagonist with multiple proposed possible mechanisms of action have been suggested regarding its anti-inflammatory and anti-platelet effect [20, 21] which is being tested in eight studies in 5,112 patients,
We analysed their costs of production based on cost of active pharmaceutical ingredients (API), and offer a snapshot of global API availability, and current list prices in a range of countries where data was available. The aim is to contribute to the evidence base around their affordability and availability so that should any drug be proven effective, it may be quickly approved and adopted for widespread use.
Methods
Drug selection
Clinical trial registries such as clinicaltrials.gov and covid-nma.com were screened alongside current published literature to identify promising repurposed drug candidates. Only those being tested in >4000 patients and shown to have potential efficacy at either reducing mortality or preventing infection were selected. The Panjiva database [22] and national drug price sources were also screened to exclude drugs without sufficient available data (Appendix 1).
Oral medications
Methodologies for estimating minimum costs of oral medication production were previously published [2]. Costs of Indian API exports were analysed (JW and JL) and independently reviewed (LE) using the online tracking database Panjiva, which showed API shipment details including quantities and costs per kilogram. All available API data from 2016 to June 2021 were included. Except for molnupiravir (which had only five shipment datapoints) for all other oral drugs we excluded shipments <1kg in volume and excluded the 15% highest and lowest priced shipments were removed to exclude outliers. For oral drugs, 5% API loss during tableting process, and a conversion cost of $0.01 per tablet was factored into calculations, alongside a multiplier based on API mass to account for the expense of excipients (additional substances needed to convert API into the finished pharmaceutical product).
Profit margins of 10% and Indian taxation of 27% on profit were added. Indian production was assumed given the country’s leading role in global generic medicine manufacturing [23].
IV dexamethasone
For injectables, an adjusted published methodology was followed to account for the cost of ampules, additional transport costs, and projected 20% API loss during formulation and filling [2, 24].
Inhaled budesonide
In the STOIC and PRINCIPLE studies, patients were asked to use two puffs of a 400mcg inhaler twice daily, (800mcg BD) for 14 or 28 days, with median duration being seven days [17, 18]. Cost per kilogram of budesonide dry powder API was estimated using the same methodology above, before accounting for the cost of inhaler devices ($3), additional transport costs, and a projected 20% API loss during formulation and filling.
Tocilizumab, sarilumab, baricitinib
No data for the cost or volume of monoclonal antibodies production were available, as this depends heavily on the efficiency and size of bioreactors used [25]. Also, no API data was available for baricitinib. Therefore, list prices in a range of countries, in particular developing economies, were tracked as a proxy, as it is assumed that the generic medicine could be reasonably produced at this level or less. There are news reports of tocilizumab and sarilumab being sold in Indian black markets at around $658.50 per 600mg dose [26].
International list price comparison
Estimated minimum production costs were compared with published list prices in a range of countries across the economic development spectrum to give a representative sample of prices globally. Data sources by country are detailed in Appendix 2. Data was collected and reviewed independently by JL and LE, intended to offer a snapshot at time of writing.
For consistency, a single data source per country was used for all searches of drug prices, based on the organisation of data and perceived reliability. Not all drugs analysed were available in every country, and in some countries without official databases, online pharmacy sites were used instead. Where several prices were available in the same database, the lowest was selected.
API availability
Panjiva data was used to analyse API export volumes per drug from India and estimate the maximum number of treatment courses potentially producible over a 12-month period.
Key assumptions
Some drugs, such as dexamethasone and tocilizumab, which are dosed on a per kilogram basis we used an average adult body mass of 70kg.
The costs of regulatory filings and approvals are often significant add-ons to the initial use of drugs in any specific country. All drugs analysed in this study have been approved for treatment for some indications already. We have not included the cost or timing associated with regulatory approvals for the use of these drugs. We assume that the WHO and other influential regulatory agencies will cooperate to define a pathway for use of these drugs which does not include additional financial outlays or filing for marketing approvals.
Results
The results are summarised in Table 1.
Summary of list prices, estimated production costs, and current availability of potential COVID-19 drugs selected for analysis. OD = Once daily, BD = twice per day, EUA = Emergency Use Authorisation (only to be given with remdesivir)
*In most recent 12-month period.
Dexamethasone
The widely used corticosteroid, dexamethasone has an economy of scale which supports low-cost production. Current national guidelines for dexamethasone treatment in COVID-19 most commonly recommend a 7-10-day courses of 6mg oral or IV [27, 28]. Shorter courses are used if patients improve or are discharged early, and sometimes longer courses and higher doses are used in patients with ongoing fibrosis and hyper-inflammation confirmed by high-resolution CT in critical care.
This analysis used a 10-day, 6mg/day course. The weighted-mean API cost between 2016-2020 was estimated to be $1116/kg; after factoring in the costs of excipients, formulation, tax, and profit, a 10-day IV course could be reasonably producible for around $2.58 and around $0.19 as tablets (Figs.1a,1b). The higher cost of injectables reflect the potential for greater loss during production, cost of vials and increased cost of transport. Furthermore, significant non-drug costs associated with IV administration, including staff time, expendables required such as sterile water/syringes for reconstitution, and saline/IV lines for infusion were not included.
Flowcharts showing cost estimations for dexamethasone oral (a) and IV (b), ivermectin (c), molnupiravir (d) fluvoxamine (e) and budesonide (f)
Comparatively, current list prices for oral dexamethasone were found to range from around $30.79 and $20.69 per course in UK and Israel respectively, to only $0.26 and $0.22 in China and Peru. For IV dexamethasone, course costs ranged between $26.47 (UK) and $0.98 (Vietnam) (Figs.2a, 2b). Sufficient API export shipments were identified in the most recent 12-month period to form 7.4 million courses, a substantial decrease on the four-year average between 2016-2019 at a theoretical 34.5 million 10-day courses per annum.
Ivermectin
A variety of doses and durations were used in clinical trials [12]. We assumed a five-day course of 0.4mg/kg/day oral treatment in a 70kg adult, equivalent to a total API dose of 140mg. If all API shipments from Panjiva were included, then the weighted mean API cost would be $381/kg, leading to an estimated treatment cost of $0.12 per patient.
However, this analysis received expert advice that many shipments may be of insufficient quality to reach medicinal-use grade; and the higher-quality API was estimated to cost instead around $3000/kg [29], meaning that a course could therefore be reasonably produced for $0.55 after accounting for costs of excipients, formulation, tax, and profit (Fig.1c). List prices per course of ivermectin treatment were found to range between $840.14 (Germany) and $519.61 (Norway) to only $2.78 (Bangladesh) and $1.66 (India) (Fig.2c).
Graphs of national treatment course prices compared to generic estimates.
Indian ivermectin API shipments identified between 2017-2020 showed an increasing trend annually. API exported in 2020 was equivalent to 3.9 million courses, compared to approximated 1.2 million courses in 2019. Overall, volume remains low, as ivermectin is mainly used in regions with endemic parasite infections.
Tocilizumab and sarilumab
Dosages for both monoclonal antibodies are based on current UK [30]. Both are given as single dose IV infusions: sarilumab at 400mg, and tocilizumab at 8mg/kg (800mg maximum). In rare circumstances, a second tocilizumab dose can be given 12-24 hours later on a case-by-case basis, depending on response to first dose. This analysis is based a single infusion of sarilumab 400mg, and tocilizumab 600mg which are guideline dose-banding suggestions for bodyweights 66kg-80kg inclusive [30].
International list prices per single dose of sarilumab ranged between $4850.90 (USA) to $875.70 (France); whilst for tocilizumab, single dose costs ranged between $3,625.14 (USA) to $410.59 (Australia) (Figs.2e, 2f).
Molnupiravir
In randomised clinical trials, molnupiravir has been tested at 200mg BD, 400mg BD and 800mg BD for five-days [19]. We calculated the cost a total of 8g per course (five-day 800mg BD) to be $255.57 (Fig.1d). No national list prices were available for molnupiravir as it has not yet been authorised for use in any disease. It has been suggested that a key bottleneck in molnupiravir production may be at the enzymatic conversion of cytidine-2 to an intermediate called N-hydroxycytidine-7, however, newer research suggests potentially this could be circumvented [31].
Fluvoxamine
Fluvoxamine has been tested with a cumulative dose of 4.05g (50mg OD day 1, 100mg BD days 2-3, 100mg TDS days 4-15). Some studies used lower doses (50mg OD for 14 days), but this study based the dose per course (4.05g) at that used in the RCT [32]. Fluvoxamine courses can be generically produced and sold profitably for $4.12 (Fig.1e), and found national prices ranged from $30.12 (USA) to $1.45 (Kenya) (Fig.2g).
Indian fluvoxamine API shipments identified between 2016-2020 showed an increasing trend. API exported in the most recent 12-month period was equivalent to 3.5 million courses, compared to the five-year average of 2.65 million courses equivalent per annum.
Baricitinib
Baricitinib is an anti-rheumatic drug which in clinical trials [18] is dosed at 4mg/day for 14 days orally. No API data is currently available. International list prices for a 14-day treatment course were found to range between $2,326 (USA) to $6.67 (Bangladesh) (Fig.2f).
Budesonide
The STOIC study used a protocol of 800mcg BD up to 28 days, with a median duration of seven days [16, 17]. Clinically, patients generally receive individual inhaler devices containing 60-200 doses of 400mcg/dose, 200mcg/dose, or 100mcg/dose which is more than sufficient for the proposed treatment duration.
Therefore, production cost of dry power budesonide inhalers containing 20mg of API was estimated to be $4.34 (100 doses of 200mcg puffs) or $4.48 if containing 40mg of API (100 doses of 400mcg puffs), based on a weighted-mean API cost of $4325/kg. This was after accounting for the cost of the inhaler device ($3), a 20% potential API loss during production, transport costs, profit, and tax (Fig.1f). International list prices vary by country depending on the availability of the 200mcg/dose or 400mcg/dose inhaler devices and are shown in Fig.2h.
API export from India in the most recent 12-month period was estimated to be sufficient for 80 million inhalers, whilst the five-year average was estimated to be around 88 million inhaler equivalents, should the entire API production volume be used for inhalers.
Discussion
Repurposed drugs with available API data could all be produced generically at very low cost. Oral tablets such as dexamethasone, fluvoxamine, ivermectin and baricitinib can potentially all be sold for from $7 to less than $1 per course, whilst inhaled budesonide could cost <$4. Intravenous drugs were more expensive, ranging from $1/course of IV dexamethasone to $411 for tocilizumab and $875 for sarilumab. Indeed, list prices in developing economies such as Bangladesh, Kenya and India are already sometimes less than our estimated costs. API costs were previously found to be a key determinant of generic medicine prices - in a 2015 study of sofosbuvir/daclatasvir, an antiviral combination for hepatitis-C, API costs accounted for 65-95% of final drug prices in competitive markets [33].
Analysis of list prices, although only a snapshot, demonstrated large variations between countries, especially for monoclonal antibodies and highlights again the importance of international global health coordination to ensure accessibility. Public-private partnerships such as the Global Fund showed that it is possible to offer drug access at near cost-price through voluntary mechanisms. Strategies such as pooled procurement, differential pricing and voluntary licenses should be utilised more widely, as over time these have become more exclusionary and available only to a shrinking list of low- and middle-income countries (LMICs). Studies of differential or tiered pricing have shown that they can result in higher prices for MICs due to arbitrary market segmentation [34].
Other tools pertinent to expanding access to medicines in the current pandemic includes the use of TRIPS flexibilities by national governments and civil society, particularly of compulsory licensing and patent oppositions. On multiple occasions this proved to be effective in addressing high prices on medicines protected by intellectual property (IP) monopolies and triggering radical price reductions [35]. Given the pandemic situation, the so-called TRIPS waiver proposal at the WTO would be a crucial initiative for removing IP barriers to access generic and/or biosimilar medicines based on cost of manufacturing [36].
In terms of availability, dexamethasone, ivermectin, budesonide and fluvoxamine and are already available as generic medicines with a large volume of API being produced already. If promising results are demonstrated for any of these drugs in large studies, they could become an option earlier on in the disease process and production must be rapidly upscaled to meet demand. Regulators and decision makers should plan for this possibility and achieve accelerated deployment especially given that the safety profiles of these drugs are already established. An important aspect will be to protect supplies for existing patients treated for indicated conditions, for example parasitic infections and rheumatoid arthritis.
Neither steroids, nor IL-6 receptor antagonists are a silver bullet against COVID-19. Dexamethasone has shown a small improvement in survival for treatment of oxygen-dependent, hospitalised patients (18%), with a larger benefit seen for the treatment of mechanically ventilated patients (36%) [11]. A recent meta-analysis found a small, improved survival benefit for tocilizumab in the treatment of hospitalised patients in high income countries [37]. Conclusive benefits for other outcomes were inconclusive. It is unclear whether the same survival benefit would be seen in LMIC settings where infection control can be more challenging. It is also unclear whether sufficient supplies of tocilizumab could be manufactured to meet worldwide demand. Similarly, significant uncertainties persist over the efficacy of molnupiravir following the failure of Merck’s MOVE-IN trial for hospitalised patients [38]. Even if the ongoing MOVE-OUT outpatient study shows efficacy, the high production costs could be a barrier to adoption especially in LMICs.
A key barrier to regulatory approval has been evidence quality concerns – demonstrated by the latest WHO guidelines on ivermectin [14] despite the large volume of studies and participants, summarised in a recent meta-analysis which found improved outcomes amongst mild-moderate outpatients [12]. Funders and regulators should take a greater co-productive role with researchers to ensure clinical trials are well designed, adequately powered, and acceptable to the regulatory approval process when they are completed.
Moreover, overpriced treatments may harm already stretched healthcare systems. The USA has spent over $1billion pre-ordering around 1.7 million courses of molnupiravir, costing an estimated $1200 per course [39]. The price of a 5-day remdesivir course in the USA is approximately $2,335.50 despite offering no mortality reduction.
Limitations
This study presumed Indian generic production [23] with lower associated costs and no capital investment in new, dedicated facilities; however, following supply chain disruptions as seen earlier on in the pandemic, and currently with vaccines, countries may accept additional upfront costs and overheads for greater supply security in establishing domestic production. Fluctuating exchange rates and confidential in-country negotiated discounts for public healthcare providers means the prices we report are subject to change and are a snapshot of publicly available information at the time of writing [40]. The re-purposed drugs landscape is also a rapidly evolving pipeline, and many of the drugs we initially analysed were subsequently shown to be ineffective.
API production costs are subject to changes in supply and demand and optimisation of manufacturing processes. This study only examined data reliant on Indian custom records from Panjiva; data in other countries were unavailable so the real volume of global API production may be higher. Conversely, API could also be used in other formulations, for example veterinary or topical use, therefore actual API volumes available for oral or injectable formulations may in fact be less than estimated. Finally, regulatory approval costs were not considered, assuming that regulatory agencies such as WHO and FDA would coordinate rapid, emergency deployment of drugs without significant burden [41–43].
Our estimates do not include research and development costs. However, a majority of investment for drug repurposing in COVID-19 has originated from public sources, with significant amounts of research funded directly by governments, universities, and public institutions such as the NIH, Gates Foundation and the WHO. A recent estimate suggested that governments have spent almost $100 billion in the last year solely on vaccine research [44].
Aspects for further research
Historically, combination therapy proved most effective against HIV/AIDS, hepatitis-C and TB - a similar approach might be needed for COVID-19 [45]. Combinations such as antivirals given concurrently with anti-inflammatory drugs or at different disease stages may prove more effective than monotherapy and warrants further investigation. Occasionally, generic combination therapies can cost less than the sum of their parts due to increased efficiency in formulation, tableting and packaging.
There are good global accessibility and availability for steroids, ivermectin and fluvoxamine. High prices of JAK inhibitors and monoclonal antibodies are barriers for patient access in healthcare systems requiring out-of-pocket co-payments or LMIC settings. The entry of biosimilars can lead to significant price reductions, even if not on the scale of chemical medications [46, 47]. One reason for higher biosimilar costs could be the overly cumbersome regulatory pathway [48] that many have argued for reform. If greater emphasis is placed on encouraging biosimilar entry, prices for tocilizumab and sarilumab could be much lower and allow for greater availability.
Conclusion
In conclusion, repurposed drugs can be manufactured at very low costs if shown to be clinically effective. In the short term until vaccination access is more widely available, repurposed drugs could potentially offer affordable and accessible treatment options against COVID-19 patients at all stages of the disease, from pre-exposure prophylaxis to asymptotic and mild infections, through to critical care.
Data Availability
Data used for this study is available upon request to the authors.
Funding
This work was supported by the International Treatment Preparedness Coalition, Make Medicines Affordable Campaign.
Conflicts of interest
None to declare.
Acknowledgements
The authors would like to thank members of the International COVID-19 Research Group, Kajal Bhardwaj and Sergey Kondratyuk for their support of this study and their contributions towards the development of this manuscript including insights, feedback, and proof-reading.
The authors would also like to express their gratitude to Professor Joe Fortunak of Howard University, Washington for his expert advice and support in the estimation of the costs of API for several drugs analysed in this study.
Appendix
flow chart showing list of drugs screened for available data
Drug price sources
Footnotes
Methodology: clarified strategy and rationale for drug selection. Results: revised estimates for ivermectin, removed colchicine and dutasteride, added fluvoxamine, molnupiravir, inhaled budesonide. Discussion: updated.