The politics and governance of drug production in public-private partnerships: Brazil’s response to hepatitis C

ABSTRACT

The local manufacture of advanced pharmaceutical products has been a long-standing objective of health and industry policy in many developing countries, including in Latin America. This strategy has been applied to fight epidemics such as HIV/AIDS, malaria, and the COVID-19 pandemic. However, we still know little about the politics and governance that enable such arrangements, especially when there is no consent from the originator company. This study focuses on the case of Brazil, a country that is well-known for its health-industry policy, which includes the local production of direct-acting antivirals (DAAs), a new treatment for hepatitis C. We seek to explain the factors that have contributed to Brazil’s successful production of generic versions of DAAs, and, later, to the decision by the Ministry of Health (MoH) to procure drugs from multinational pharmaceutical companies rather than from local laboratories. A lack of support for domestic production by important stakeholders, the patent holder’s attempt to block domestic production and the MoH’s adoption of more modern treatment guidelines under a different procurement logic all created an unfavourable environment for local production and procurement of DAAs. Our study draws implications for middle-income countries that wish to produce drugs domestically without voluntary license agreements.

KEYWORDS:

• Domestic production of medicines
• technology transfer
• hepatitis C
• Brazil
• sofosbuvir
• direct-acting antivirals

Introduction

Pharmaceutical production and the related technology transfer from originating companies to local producers in low – or middle-income countries have the potential to secure steady and affordable access to medications while enhancing technological capacities and innovation (Mackintosh et al., 2016; Russo & Banda, 2015; Shadlen & Fonseca, 2013). Indeed, the World Health Organization (WHO) has launched global initiatives to encourage links between local production and innovation, public health needs, and the development of local economies (WHO, 2021). The Health for All initiative revamps this agenda by calling for better integration between health and industrial goals (WHO, 2023). However, while a consensus regarding the need for local production exists, we still know little about the governance arrangements and politics that may promote or hinder these initiatives at local levels.

Brazil has historically been known for promoting policies that foster generic production and local pharmaceutical industry (Shadlen & Fonseca, 2013; Caliari & Ruiz, 2014). Most notably, its National AIDS Programme¹ has successfully promoted universal access to antiretroviral (ARV) drugs produced in state-owned laboratories (Flynn, 2015; Nunn, 2009). But it was not until the mid-2000s that Brazil began to encourage, more systematically, public-private initiatives in technology transfer. These were called ‘partnerships for product development’ (PDPs) of medications.

As part of this strategy, and in line with the WHO’s goal to eliminate hepatitis C as a public health concern globally, Brazil’s response to hepatitis C included locally producing a new treatment based on the direct-acting antivirals (DAAs) sofosbuvir and, later, daclatasvir. Launched by Gilead Sciences and Bristol-Myers Squibb, respectively, these DAAs revolutionised the treatment of hepatitis C by curing more than 95% of people infected by the hepatitis C virus (HCV virus) (Mayburry & Lee, 2019). Gilead established voluntary licensing agreements with a few firms, mostly in India, which were able to produce generic versions of DAAs and commercialise them in 101 countries (Douglass et al., 2018). Brazil, however, was not included in Gilead’s strategy.² Nonetheless, a local public-private partnership between the Brazilian companies Blanver and Microbiológica and the public laboratory Farmanguinhos (henceforth, BMK-FAR) was established, and these two private companies successfully produced a generic version of sofosbuvir (Cassier & Correa, 2018). Yet, such a strategy faced numerous challenges that limited BMK-FAR's possibility to supply these products.

Our study seeks to contribute to the debate on the domestic production of generics through technology transfer³ to address major public health challenges. Although many studies have contributed to this debate by exploring favourable conditions and challenges, such as established capabilities, infrastructure, political will, and procurement strategies (Chataway et al., 2016; Simonetti et al., 2016; Tibandebage et al., 2016; Russo and Oliveira 2016), little is known about the political dynamics and governance for local production of DAAs for fighting hepatitis C. This is particularly important, as technology transfer is contingent and shaped by context, i.e. it does not follow a general recipe that applies everywhere. We seek to explain the factors that contributed to Brazil’s successful production of generic versions of DAAs, and the later decision to procure drugs from multinational pharmaceuticals rather than the local consortium of laboratories. In doing so, we offer lessons from Brazil, which have implications for other middle-income countries that wish to produce drugs domestically without voluntary license agreements, as these need to be enabled politically.

Our study shows that, as a highly complex and risky project, technology transfer needs to be enabled politically and needs support from state and non-state actors, especially when there is no voluntary licensing of patents by the originating company. We also illustrate that having capabilities in place for domestic production is an important factor when negotiating prices for treatments with multinationals. This has further implications for middle-income countries (MICs) that wish to pursue domestic production of generic pharmaceuticals, since the existence of capabilities does not guarantee local production but may be a fundamental factor in negotiating favourable prices.

The paper is structured as follows. The next section situates our study within the broader debate on technology transfer for generic production in the Global South. In the third section, we contextualise biomedical technology transfer in Brazil. The fourth section discusses Brazil’s production of DAAs as a means for expanding access to treatment in line with the WHO’s plan to eliminate hepatitis C as a public health concern by 2030, and its limitations. In the conclusion, we draw out some implications of our findings for middle-income countries wishing to pursue domestic production of medications.

Technology transfer for local production of medicine in the Global South

Much has been written about technology transfer to the Global South as a means for addressing major public health challenges while at the same time contributing to local economies (Dong & Mirza, 2016; Mackintosh et al., 2016; Mujinja et al., 2014; Russo & Banda, 2015; Shadlen & Fonseca, 2013; Sidibé et al., 2014). In Making Medicines in Africa, Mackintosh et al. (2016) describe countries’ reliance on technology transfer in fighting epidemics such as HIV/AIDS, malaria, and other diseases. Successful technology transfer initiatives in African countries have relied on established infrastructure and capabilities, political will, substantial financing and a ‘buy local’ procurement strategy (Chataway et al., 2016; Simonetti et al., 2016), whereas a lack of technological support and training, low capabilities, and small market size constituted the main challenges for countries aiming to produce generic medications (Tibandebage et al., 2016; Russo & Oliveira, 2016). India’s pharmaceutical sector has made impressive progress by internalising external technologies (Chaudhury, 2005; Kale & Little, 2007), but also by building on its existing capacities to encourage more endogenous R&D (Lander & Thorsteinsdóttir, 2011). India has successfully improved manufacturing capacity and technical skills in biotechnology, which has proved fruitful during the Covid-19 pandemic, when the country became one of the largest vaccine producers (Kumraj et al., 2022; Chaudhuri, 2022).

All these studies have contributed to understanding how technology transfer for the production of medications occurs in the global South. However, while some of the aforementioned studies’ main findings are corroborated by our study, none of them provide a detailed account or analysis of technology transfer in the absence of voluntary licensing – which is where the political dynamics and interactions between various stakeholders become even more crucial to the success or failure of domestic production.

In the case of the generic production of DAAs for hepatitis C treatment in the global South, there is limited literature exploring specific cases. Cassier and Correa (2018) examine the Brazilian production of sofosbuvir from 2014 to 2017. Their study is important for understanding the initial establishment of the national consortium BMK-FAR, Brazil’s political and economic context, and the industrial property situation. But the study could not provide a more thorough analysis, because much has happened since 2017. Our article mostly focuses on events and outcomes that happened after 2017.

Without analysing any specific technology transfer for DAA production in African countries, Assefa et al. (2017, p. 8) argue that universal access could only be achieved through ‘compulsory licensing and patent opposition; […] political commitment and increased government expenditure; […] global solidarity and global health financing mechanism; and […] advocacy by civil society and people living with HCV’. Douglass et al. (2018) briefly mention Egypt’s generic production of DAAs, enabled by its government’s rejection of patent applications for sofosbuvir. There are also records of the domestic production of DAAs in Latin America (DNDi, 2018), in Pakistan (The Express Tribune, 2016), and in India, the latter through a voluntary license agreement which would benefit 91 countries (Gilead, 2014). However, none of these studies explore how local production unfolded on the ground, i.e. the political dynamics and specific arrangements that enabled or hindered local production of DAAs.

In this article, we explore how politics and governance together contributed to both the initial successful production of generics, and later to the decision to procure the treatments mostly from Gilead. Understanding the dynamics of the generic production of DAAs in an middle-income country is important, considering the high price of these drugs and the barrier it represents for universal treatment. The case of Brazil is particularly relevant given its history of domestic ARV production as well as its adherence to the WHO’s strategy for eliminating hepatitis C as a public health concern by 2030.

This article is based on documentary analyses of resources from Brazil’s MoH, the Federal Accountability Office, non-governmental organisations, and state governments, as well as on more than 20 semi-structured interviews conducted between 2021 and 2023, with current and previous health experts from the MoH, non-governmental organisations (NGOs) advocates, members of the consortium, and personnel from the Oswaldo Cruz Foundation (Fiocruz), the public institution in charge of producing DAAs domestically. We also consulted local media reports and specialised literature.

Contextualising biomedical technology transfer in Brazil

Brazil is well-known for promoting biomedical technology transfer to meet the needs of its National Health Service (Sistema Único de Saúde, henceforth SUS) (Benchimol, 2017; Homma et al., 2005; Homma et al., 2013). SUS is one of the largest publicly funded health systems in the world. One of the notable aspects is its broad and generous provision of medicines and vaccines. In the mid-2000s the Brazilian Ministry of Health (MoH) created the PDP (Ministério da Saúde, 2010), public-private partnerships which transfer technology from a private institution into a public laboratory in order to produce generic drugs, which are then acquired through public procurement by the MoH (Camargos & Moreira, 2015). PDPs were meant to promote integration between what was seen as a dichotomy: the healthcare sector and the realm of industry (Gadelha, 2003; WHO, 2023). The rationale behind PDPs is to provide a stable supply of ‘strategic’ and affordable drugs, among other medical products, for SUS while simultaneously strengthening the ‘Economic Industrial Healthcare Complex’ (Gadelha, 2003).

A PDP agreement typically involves three organisations⁴: (i) a public institution responsible for internalising the technology and producing the drug, (ii) a public or private institution responsible for producing the active pharmaceutical ingredient (API), and (iii) a private institution responsible for developing the drug technology and transferring it to the public institution or laboratory (Varrichio, 2017). In the case of the BMK-FAR arrangement, Farmanguinhos represents the public institution that is supposed to internalise the technology to produce sofosbuvir, Blanver is the private laboratory responsible for developing and transferring the technology to Farmanguinhos, and Microbiológica is the API producer.

Two crucial pillars comprise the PDP arrangement. First and central is that the federal government is responsible for purchasing high-cost drugs. This ensures that private and public laboratories have greater incentives for engaging in technology transfer and for producing large quantities of drugs for the SUS through PDPs. PDPs are thus called ‘demand-driven policy tools’, as firms within a PDP take advantage of the economic potential of the SUS’s demand. In the case of sofosbuvir and daclatasvir purchases, for instance, when the MoH decided in 2017 to provide treatment for all, there were an estimated 632,000 HCV-positive patients in Brazil (Benzaken, 2019). Brazil’s National Elimination Plan (Ministério da Saúde, 2018) established that, by 2020, 50,000 patients per year would be treated (Coutinho et al., 2021).⁵

The second pillar is that public laboratories do not have to bid for government contracts, according to the Public Procurement Law (Law 8.666/1993), which requires that the public administration must acquire the cheapest product that meets certain quality standards. Additionally, a 2012 amendment to the Procurement Law (Law 12.715) established that the Public Administration can participate in strategic partnerships with public-private laboratories (without having to open public bidding) that promote technology transfer, development of the national industry, and technological autonomy (Varrichio, 2017). Public laboratories are thus crucial to the PDP arrangement; they allow government to hold the expertise of the technology being transferred but also, from a governance perspective, help the MoH maneuver Brazil’s strict procurement laws.

As of late 2022, there were 85 ongoing PDPs for medicines, vaccines, and hemoderivatives, with four of them involved in the production of DAAs.⁶ For daclatasvir, by 2022 the most advanced PDP was being developed by the BMK-FAR consortium but it was under suspension. For sofosbuvir, the PDP by the BMK-FAR consortium was also the most advanced; however, the technology had not been transferred to the public laboratory. With the new government inaugurated on 1 January 2023, the MoH is laying the groundwork for launching innovative changes to the programme that will shape its trajectory in the coming years, with a strong emphasis on the development of the health-industrial complex (Agência Brasil, 2023).

Expanding access to hepatitis C treatment through technology transfer

Brazil responded to the hepatitis C epidemic by scaling up the diagnosis and treatment of people living with HCV (Benzaken, 2019). The emergence of highly effective, interferon-free therapy in 2014 revolutionised treatment (Brennan & Shrank, 2014) and enabled a higher cure rate of approximately 95%.⁷ From 2014 to 2017, Brazil treated 70,000 patients with the new DAAs sofosbuvir and daclatasvir⁸ at the cost of (US) $1 billion (Benzaken, 2019).

In line with the WHO’s plan to eliminate hepatitis C as a public health concern by 2030 (WHO, 2016), treatment started with patients with severe liver fibrosis and coinfections and expanded in 2018 to encompass the whole patient population (Ministério da Saúde, 2018). Scaling up treatment with the new DAAs posed a significant economic challenge to the MoH. An estimated 1,091,000 people were infected with HCV in Brazil in 2016 (Benzaken, 2019). International prices for DAAs were at (US) $140,000 per treatment per patient in 2015.⁹ Against this background, the MoH engaged in negotiations with the pharmaceutical companies selling sofosbuvir, daclatasvir, and simeprevir, namely Gilead, Bristol-Meyers Squibb, and Janssen, aimed at significantly lowering their prices.¹⁰

In 2015, as a result of these negotiations, the MoH managed to purchase DAAs for less than 10% of international market prices – an unprecedented discount for a high-middle income country (Mesquita et al., 2016).¹¹ According to one of the main negotiators for DAA prices, Brazil’s policy of generic production alongside its experience in issuing compulsory licenses for the AIDS drug efavirenz were key to the pharmaceutical companies’ decision to grant the MoH a significant discount.¹² There was also the possibility of domestic production of DAAs, which we will explore below.

Meanwhile, as soon as Gilead launched sofosbuvir in the US market under the brand name sovaldi, several key actors in Brazil began collaborating to produce the drug domestically. Two private laboratories, Microbiológica and Blanver, and a private consultant, Karin Bruening, formed a consortium to produce generic sofosbuvir and, later, daclatasvir. Both laboratories had experience producing antiretroviral drugs for the MoH, and Bruening had worked for Farmanguinhos in the early 2000s, when the public laboratory became the pillar and producer of generic antiretrovirals for the National AIDS Programme.¹³ Importantly, Microbiológica founder Jaime Rabi had previously collaborated with Pharmasset, the pharmaceutical company that discovered sofosbuvir and was afterwards acquired by Gilead. Thus, Microbiológica had knowledge of the research and development processes related to sofosbuvir (Cassier & Correa, 2018).¹⁴ In sum, the arrangement was as follows: Microbiológica would produce the molecule (the API), Bruening would develop the pharmaceutical product, and Blanver would produce and sell the medication (Cassier & Correa, 2018).

At the same time, Farmanguinhos also started mobilising actors and resources in order to produce sofosbuvir. Health experts from Farmanguinhos argued that the new drug could be produced for a significantly cheaper price than sovaldi.¹⁵ Furthermore, sofosbuvir was considerably more efficient (with a cure rate exceeding 90%) than interferon-based therapies (with a 50% cure rate), which Fiocruz already produced; sofosbuvir also offered a better adherence to treatment, as it was an oral drug with fewer side effects than the injectable interferon.¹⁶ It is important to note that the rationale behind the domestic production of sofosbuvir also mentioned the strength, or rather weakness, of the sovaldi patent, which was still pending in Brazil. According to Farmanguinhos’ analysis, the sovaldi patent was weak and unmerited, since it was based on old science.¹⁷ While Gilead had already applied for local patent rights, Farmanguinhos bet that the Brazil’s National Institute of Industrial Property (INPI) would not approve it.¹⁸ Essentially, health experts at Farmanguinhos believed Brazil would recreate the successful experience of producing HIV medicines domestically (Cassier & Correa, 2018).

Before BMK-FAR applied for a PDP, Microbiológica and Blanver had already mastered the technology for the production of sofosbuvir. According to one stakeholder, ‘we were ahead of everyone […] there was no competition for us’,¹⁹ The BMK consortium was fully equipped to produce and supply the sofosbuvir. In fact, BMK alone had already produced the first batch of sofosbuvir, even before winning a contract in 2019 to supply the SUS through a PDP.²⁰

One remarkable feature of the sofosbuvir production case was the lack of consensual support by non-state actors. Historically, Brazil has counted on robust mobilisation within civil society to support universal health treatment as a matter of human rights and democracy (Fleury, 2011). Particularly in the case of HIV/AIDS, the federal government’s tough stance on efavirenz producer Merck would not have succeeded without organised pressure and support on the part of its civil society (Eimer & Lutz, 2010). By contrast, there has been no consensus in the case of hepatitis C. NGOs have remained divided on, or indifferent to, the issue of domestic production of sofosbuvir.²¹ Important NGOs supported differing stances despite counting on Brazil's leading HIV/AIDS research and advocacy organisation, ABIA,²² and despite a public challenge from Brazil’s Working Group on Intellectual Property to Gilead’s patent application, as well as pressure for domestic production of sofosbuvir (GTPI, 2016). On the one hand, one of the main national organisations, the Brazilian Movement of Fight Against Viral Hepatitis, was publicly supportive of Brazil’s efforts to produce sofosbuvir as part of the PDP rationale (MBHV, 2018). On the other hand, the equally active organisation Independent Alliance of Supporting Groups, represented by Grupo Otimismo, was publicly against the generic production of sofosbuvir, supportive of patent protection, and even requested a public audit of the BMK-FAR PDP from the federal accountability office (Agência de Notícias das Hepatites, 2018).

But what has characterised pro-hepatitis C patients’ organisations, according to informants from the MoH and civil society, was their indifference to whether a medication is locally or internationally procured: ‘having medications available at the SUS (rather than advocating for generic production) was their main concern, and once they achieved it, they became less active’.²³ For pro-patient organisations, ‘[i]t doesn’t matter the price […] the patient wants medication’.²⁴ Having a civil society mobilised for generic production requires raising awareness about the importance of domestic production, which is not an easy task.²⁵ Considering that patients may have free access to medicines and vaccines through SUS (once the Health Technology Evaluation Commission approves them), their concern was less about generic production than access.

Despite the lack of consensus among NGOs, as soon as it became clear that producing sofosbuvir for Brazil’s SUS was feasible BMK-FAR applied to the MoH for three PDPs, also including daclatasvir and simeprevir in their application (Cassier & Correa, 2018). In 2017, the MoH approved BMK-FAR’s PDP proposals to produce sofosbuvir and daclatasvir (Cassier & Correa, 2018). The consortium signaled a willingness to provide sofosbuvir at (US) $8.50 per pill, a quarter of Gilead’s price (Fonseca et al., 2019). Given this newfound competition, Gilead reduced the price of sofosbuvir from $75 per pill in 2015 to $53 per pill in 2016 and 2017 (see graph 1). As the consortium advanced in the production of sofosbuvir, Gilead was granted a patent and attempted to block domestic production through lawsuits and patent disputes.

In September 2018, the INPI finally analysed Gilead’s request, but granted a patent related only to an intermediate molecule that did not affect BMK’s production.²⁶ A decision that was weeks later suspended in court following a request from a presidential candidate and her running mate (Estado de SP, 2018).

Gilead launched a lawsuit against BMK-FAR’s PDP for sofosbuvir, claiming that Farmanguinhos had not opened a bidding process prior to hiring the private laboratories. Gilead’s arguments were based on The Public Procurement Law (law 8.666).²⁷ In 2018, Brazil’s federal accountability office (TCU) suspended the PDP pending further analysis by the court.²⁸ This was not the first time Brazil’s public authorities (either the TCU or the judicial system) accepted a lawsuit against a PDP based on law 8.666 (Sundfeld & Souza, 2013), which illustrates that PDP regulations were not sufficient to override public procurement law and demonstrates the lack of coordination amongst the different actors involved in enabling PDPs.

Although the TCU later decided in favour of BMK-FAR, its initial decision stopped national production through a PDP, which in turn sowed doubts as to whether the BMK-FAR PDP would in fact materialise. Furthermore, without national production of DAAs through PDPs, Brazil’s MoH had to negotiate with multinational companies, who ended up supplying most of the needed DAAs to the SUS, though for a significantly reduced price. However, BMK lost an important element of the PDP programme, namely, guaranteed public procurement.

Gilead’s strategy involved more than legally challenging the BMK-FAR arrangement. In 2015, Gilead had signed a memorandum of understanding with Bahiafarma, a public laboratory belonging to the State of Bahia, for technology transfer in order to produce sofosbuvir (Governo da Bahia, 2015). Another agreement was later signed with Lafepe, a public laboratory belonging to the state of Pernambuco (Governo de Pernambuco, 2018), also to produce sofosbuvir. The idea was to submit a proposal for a PDP in case the INPI did not approve the patent (Cassier & Correa, 2018). Some BMK-FAR stakeholders argued that the Bahiafarma-Gilead and Lafepe-Gilead PDPs were a ‘trap’ or ‘publicity stunt’, a strategy Gilead was using to win the right to be a supplier of DAAs in Brazil but without having any intention to transfer the technology into Bahiafarma or Lafepe.²⁹ This could be a valid claim, considering that both agreements have never materialised.

Public procurement was a major challenge BMK faced in its efforts to supply sofosbuvir to the MoH. While TCU suspended the BMK-FAR PDP, the MoH opened emergency public auctions. Between July 2018 and January 2019, three auctions were held, in which Gilead, Blanver (separately from the BMK-FAR arrangement), and Bristol, the producer of daclatasvir, tried to win contracts with the MoH. The first auction took place in July 2018, the second in November 2018, and the third in January 2019. These three auctions, according to stakeholders from the consortium and the MoH, constituted a blow to BMK-FAR’s ambitions to be the main supplier of hepatitis C treatments.³⁰

In July 2018, the MoH sent price requests for 50,000 treatments to DAA producers.³¹ In the view of BMK stakeholders, however, this constituted a regular public bidding, where Gilead, Blanver, and Bristol made bids to win contracts under the Public Procurement Law. It is important to reiterate that the treatment for hepatitis C consisted of either single pills produced by Gilead or of both daclatasvir, which the BMK consortium did not produce, and sofosbuvir. Thus, to win the contract, Blanver also depended on Bristol, the holder of the daclatasvir patent,³² to lower its prices so that the two pills would be cheaper than the single pills produced by Gilead. At the time, Blanver offered ‘very attractive royalties’ as part of a voluntary license for Bristol’s daclatasvir, so that Blanver could supply generic sofosbuvir alongside Bristol’s daclatasvir – but Bristol refused the offer and made their own bid.³³

Despite offering the lowest prices alongside Bristol, Blanver never supplied the 50,000-treatment sofosbuvir batch. A Blanver representative reported, ‘We had prepared ourselves to produce 50,000 treatments […] we ordered the raw material, made changes to our plant, […] but we never received the order [from the MoH]’.³⁴ According to BMK, there was no justification for the MoH’s decision to not consummate the deal, especially given that the 50,000 treatments would have cost the same as 15,000 treatments.³⁵ Informants from the MoH, however, said the meeting was not a public bidding but rather a ‘price request’ that would guide them in future procurements.³⁶ Furthermore, the INPI’s initial granting of a patent to Gilead in September 2018 (which soon after was invalidated by a federal court) delayed the procurement process and forced the MoH to resume negotiations with the pharmaceutical industry.³⁷

In another public bidding later in 2018, BMK won a contract to supply approximately 15,000 sofosbuvir treatments, though at a lower price than they had expected.³⁸ A dramatic situation unfolded around Christmas 2018 when Gilead resorted to the courts to attempt to block the delivery of BMK’s sofosbuvir (Folha de São Paulo, 2018b). Trucks loaded with sofosbuvir were prevented from entering the MoH facility due to the court injunction. A situation that was later reversed.

A third auction took place at the beginning of 2019, and the MoH required a number of treatments that BMK could not produce within the stipulated time frame. According to members of the consortium, they withdrew from the bidding because it was impossible for them to produce ‘tons of sofosbuvir in just 30 days’.³⁹ According to MoH health experts, two problems had arisen: first, Bristol had increased its price for daclatasvir; second, it was clear that the domestic industry did not have sufficient scale to meet MoH’s demand at a ‘much lower price’.⁴⁰ Thus, Gilead was successful in this procurement call.

Finally, starting in 2019, the MoH adopted a new procurement logic, namely, ‘cost minimization’ (Ministério da Saúde, 2019). In line with the government’s austerity measures that sought to freeze social spending for 20 years (Souza Lima, 2019), the MoH began prioritising alternative treatments in order to reduce impacts on the MoH budget while at the same time guaranteeing access to therapy for HCV-positive patients (Minist'erio da Saúde, 2019). To further complicate matters, Bristol later decided to discontinue the distribution of daklinza (their brand name for daclatasvir) worldwide,⁴¹ making it impossible for BMK to become the main provider of hepatitis C treatments at that moment.⁴² At the same time, the MoH hepatitis C guidelines for pangenotypic treatment prescribed the most modern treatment, i.e. the sofosbuvir plus velpatasvir, which offered the promise of a single pill, with efficacy supported by scientific evidence (Conitec, 2018).

Notwithstanding the impact of Gilead’s lawsuits against the BMK-FAR PDP, the MoH’s support for domestic production by BMK was not consensual. Informants agreed that, initially, local production was effective in reducing the price; however, with new products launched and the cost minimisation logic, the PDP became impracticable. In 2019, the BMK-FAR PDP was later suspended pending a judicial decision (Agência Fiocruz, 2019).

Nonetheless, the real possibility of sofosbuvir production by a domestic consortium significantly increased the MoH’s bargaining power in relation to Gilead. In 2015, Gilead’s sofosbuvir cost the MoH approximately US $72.94 per pill. With the establishment of the BMK-FAR PDP in 2016, the MoH obtained a 38% discount per 400 mg sofosbuvir pill by Gilead. In 2018, the procurement of generic sofosbuvir meant the MoH saved 83.5% in comparison to the 2017 contract. Considering the purchase volume, the MoH was able to save US $49.5 million with the acquisition of the generic version of sofosbuvir.

Table 1. Timeline of important events

Between 2019 and 2023, Gilead supplied a treatment consisting of sofosbuvir and ledipasvir combined into a single pill (Harvoni)⁴³ taken once a day for 12 weeks.⁴⁴ Other treatments involved the combination of two distinct pills containing sofosbuvir or daclatasvir (Epclusa), produced by Gilead and Bristol, respectively. In 2022, for the full pangentoypic treatment of sofosbuvir plus velpatasvir, the acquired treatment cost approximately US $18.3, which is 64% cheaper than the initial price Gilead charged for sofosbuvir alone – that is, for half the treatment.⁴⁵

Final remarks and practical implications

In this study, we found an unfavourable environment for local production and procurement of DAAs has been created by each of the following: a lack of support for – and consensus about – domestic production by important stakeholders, such as civil society organisations and the Ministry of Health; patent holder attempts to block domestic production through patent applications, legal challenges, and the establishment of competing consortia; the MoH’s adoption of new, more modern treatment guidelines under a novel procurement logic in line with austerity measures implemented by the federal government.

Despite the numerous setbacks for the BMK-FAR PDP and the MoH itself, Brazil has achieved considerable success in treating hepatitis C, thus demonstrating progress towards the WHO plan. By 2015, the possibility of domestic possession of the technologies for producing a generic version of sofosbuvir meant big pharmaceuticals such as Gilead had to lower their prices considerably to compete with the national version. In addition, building on this experience, in 2023, Fiocruz partnered with the Drugs for Neglected Diseases Initiative, a collaborative, non-profit pharmaceutical research organisation, to develop ravidasvir, which received approval from the regulatory agency for use in combination with sofosbuvir. Whether there will be new disputes over patents remains to be seen (Santos, 2023).

For more than four years (2019-2023), the possibility of relying on the domestic industry to supply the treatment was frozen. It was only after the inauguration of a new federal government in January 2023 that the strategy for the national development of the industrial health complex was revitalised. The BMK-FAR PDP slowly resumed manufacturing, supplying its first batch to the MoH in November of that year (Fiocruz, 2023). Whether Gilead will litigate this supply remains to be seen. Future studies should investigate the processes of technology transfer and absorption, the capability gains to FAR, and whether the consortium has become sustainable over time.

Our study has practical implications for policymakers and governments that wish to encourage domestic production without voluntary licensing. First, as the literature has shown, relevant capabilities and policies, such as PDPs, are essential (Chataway et al., 2016; Simonetti et al., 2016; Tibandebage et al., 2016). However, without political support, consensus, and coordination amongst the various actors, any such initiative becomes unfeasible. In that sense, we corroborate the findings of Fonseca et al. (2023). While our study echoes many of the findings in the literature on technology transfer for domestic production, we provide further insights and a dynamic view of the ways in which domestic and international stakeholders’ actions, as well as governance arrangements such as public procurement, may affect domestic production. Thus, global health scholars and practitioners would benefit from further in-depth case studies exploring governance arrangements and political dynamics between the various stakeholders, and how these vary according to countries’ political institutions and economic contexts.

The second main practical implication is that such consensus and support from state and non-state actors become even more important in the absence of voluntary patent licensing. In that context, domestic production will almost invariably face legal challenges, patent applications, and other measures available to pharmaceutical companies – such as the establishment of competing consortia – that require coordinated opposition amongst state and non-state actors. Another pathway suggested by the literature is the issuing of compulsory licensing through the use of the Trade-Related Aspects of Intellectual Property Rights flexibilities (Douglass et al., 2018). Brazil has had a successful experience with compulsory licensing for ARVs (Nunn et al., 2007; Ford et al., 2007), yet it is important to recognise that the political, social, and economic context of such successes cannot be easily replicated (Cassier & Correa, 2018). Therefore, while finding loopholes or devising strategies to navigate patent protection, as demonstrated by BMK-FAR, may offer an alternative route for generic producers, we acknowledge that this may be a less viable option.

Finally, our study also illustrates that having capabilities in place for domestic production is an important factor when negotiating treatment prices with multinationals. This has further implications for middle-income countries that wish to pursue domestic production of generic pharmaceuticals, as the possession of the relevant manufacturing capabilities does not guarantee local production, though it can be a key factor in negotiating favourable prices.

Acknowledgments: This study was funded by the Newton Fund Institutional Links grant 2019, a partnership between Fiocruz, the LSE, and FGV, project ‘Brazil's Fight Against Hepatitis C: Surveillance, Control, and Care’. HMA and EMF are funded by the Sao Paulo State Research Foundation grants # 2022/07849-0 and #2021/06202-0, respectively.

We would like to thank all the respondents who kindly agreed to participate in this research, and the editor and anonymous reviewers of Global Public Health for their invaluable feedback.

Disclosure statement

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

Additional information

Funding

This study was funded by the Newton Fund Institutional Links grant 2019, a partnership between Fiocruz, the LSE, and FGV, project “Brazil's Fight Against Hepatitis C: Surveillance, Control, and Care.” HMA and EMF are funded by the Sao Paulo State Research Foundation grants # 2022/07849-0 and #2021/06202-0, respectively.

Notes

1 The National AIDS programme, as it was known when implemented in 1985, came to an end in 2019, through a decree by far-right president Jair Bolsonaro. However, since 2023, the new government has restructured the department/policy, currently named the Department of HIV/AIDS, Tuberculosis, Viral Hepatitis, and Sexually Transmitted Infections.

2 Gilead’s agreement included 91 “developing countries” loosely defined as resource-limited settings, which did not include Brazil. See the guidelines here: https://www.gilead.com/~/media/files/pdfs/other/hcv%20generic%20agreement%20fast%20facts%203215.pdf (accessed March 10, 2024)

3 It is important to note that as most studies (including those cited in this paper) approach technology transfer as the transfer of the manufacturing process for a drug from a transnational corporation to a domestic producer. Our case study, however, refers to technology transfer between a local private company that developed the drug through reverse engineering to a public laboratory, as will be explored.

4 See the current list of PDPs for medications: https://www.gov.br/saude/pt-br/composicao/sectics/pdp/medicamentos-vacinas-e-hemoderivados/medicamento-vacina-e-hemoderivados-parcerias-vigentes/view (accessed Mach 9, 2024).

5 In comparison, in 2019 and 2020, 36,658 and 16,874 patients, respectively, were treated. The drop in treatments from 2019 to 2020 is explained by the impacts of the COVID-19 pandemic (see Coutinho et al., 2021).

6 https://www.gov.br/saude/pt-br/media/pdf/2021/novembro/03/pdp/medicamento-vacina-e-hemoderivados-parcerias-vigentes-parcerias-vigentes.pdf (accessed March 10, 2024)

7 Interview 1, São Paulo, 2021.

8 Although a third DAA, simeprevir, was initially part of the protocol, it was later removed from the protocol as it was less efficient and more difficult to take (Interview 1).

9 Interview 1, São Paulo, 2021.

10 Interview 1, São Paulo, 2021.

11 According to a study comparing nominal prices of sofosbuvir across 30 countries, including members of the Organisation for Economic Cooperation and Development and selected low—and middle-income countries, Brazil paid US$6,875, Egypt US$932, Mongolia US$900, and India the lowest at US$539—these were up to 120 times lower than the nominal price in the US (US$64,680) (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4886962/pdf/pmed.1002032.pdf). Egypt, Mongolia, and India had voluntary licensing agreements with Gilead, while Brazil did not (https://www.gilead.com/-/media/files/pdfs/other/hcv-generic-agreement-fast-facts-11-15-17.pdf) (accessed March 10, 2024).

12 Interview 1, São Paulo, 2021.

13 Interview 3, São Paulo, 2021.

14 It is worth noting that given this expertise, Rabi tried to establish a contract agreement with Gilead to produce the sofosbuvir in Brazil, but as the executives declined the offer, Microbiológica engaged in the consortium instead.

15 Interviews 5 and 11, São Paulo, 2021.

16 Interviews 1 and 5, São Paulo, 2021.

17 Inventiveness is one of the criteria for a patent to be granted. See similar arguments put forward by other countries in the case of Sovaldi (Pollack, 2015).

18 Interview 5, São Paulo, 2021.

19 Interview 3

20 Interviews 3, 4 and 8, São Paulo, 2021.

21 Interviews 6 and 7, São Paulo, 2021.

22 ABIA stands for “Brazilian Interdisciplinary Association for AIDS.”

23 Interview 7, São Paulo, 2021. It is important to note that a more in-depth study of these civil society organizations is necessary to understand the motivations behind certain patient organizations. We know, for example, that one of the largest organizations that offer support to hepatitis C patients receives funding from Gilead (Folha de São Paulo, 2018a).

24 Interview 6, São Paulo, 2021.

25 Idem.

26 Interview 8, São Paulo, 2021.

27 See Gilead’s precautionary measure and Brazil’s Federal Accountability Office’s preliminary response at: https://pesquisa.apps.tcu.gov.br/documento/acordao-completo/*/NUMACORDAO%253A735%2520ANOACORDAO%253A2020%2520COLEGIADO%253A%2522Plen%25C3%25A1rio%2522/DTRELEVANCIA%2520desc%252C%2520NUMACORDAOINT%2520desc/0 (accessed March 11, 2024)

28 https://pesquisa.apps.tcu.gov.br/documento/acordao-completo/*/NUMACORDAO%253A735%2520ANOACORDAO%253A2020%2520COLEGIADO%253A%2522Plen%25C3%25A1rio%2522/DTRELEVANCIA%2520desc%252C%2520NUMACORDAOINT%2520desc/0.

29 Interviews 3, 4 and 14, São Paulo, 2021. At the time, ABIA/GTPI released a manifest criticizing the Bahiafarma deal for lack of transparency and implying that it might interfere with BMK-FAR national production, which was already in advanced stages (ABIA, 2015). They suspected Gilead would purchase DAA generics from India and China indefinitely to supply the NHS without actually capacitating the national laboratories (personal communication).

30 Interviews 3, 4, 8, São Paulo, 2021. Interview 12, Rio de Janeiro, 2019.

31 Interviews 3, 10, São Paulo, 2021.

32 INPI had granted the patent to Bristol, and BMK did not produce a generic version of daclatasvir (Interview 4, São Paulo, 2021).

33 Interviews Interviews 3, 4, 8, São Paulo, 2021. Interview 12, Rio de Janeiro, 2019.

34 Interview 12, Rio de Janeiro, 2019.

35 Interviews 3 and 4, São Paulo, 2021.

36 Interview 10, São Paulo, 2021.

37 Interview 2, São Paulo, 2021.

38 Interviews 3, 4, 8, São Paulo, 2021.

39 Interview 8, São Paulo, 2021.

40 Interviews 7 and 13, São Paulo, 2021.

41 https://www.bms.com/assets/bms/brazil/documents/INFORMATIVO%20DAKLINZA%20final.pdf (accessed Feb 17, 2022).

42 The Consortium later was able to produce the generic version of daclatasvir (Interview 4).

43 Harnovi is for hepatitis C virus genotypes 1-6. Later, Gilead started offering a pangenotypic combination of sofosbuvir plus velpatasvir in a single pill (Epclusa).

44 Interview 2, São Paulo, 2021.

45 Ministry of Health information, obtained through the Law on Access to Public Information.