Cervical cancer, the leading cause of female cancer mortality worldwide, disproportionately affects women in low- and middle-income countries (LMCs). Four-fifths of the nearly 275,000 annual cervical cancer–related deaths occur in LMCs where routine gynecological screening is minimal or absent1,2. Two new prophylactic vaccines, Gardasil from Merck (Whitehouse Station, New Jersey, USA) and Cervarix from GlaxoSmithKline (GSK; London) have proven effective in preventing human papilloma virus (HPV)-induced cervical lesions and some sequelae. Both vaccines are composed of HPV-L1 major capsid antigen virus-like particles (VLPs), and both prevent persistent infection from HPV-16 and HPV-18, which cause nearly 70% of cervical cancers3. Gardasil also contains L1 antigens from HPV strains 6 and 11, which are associated with genital warts. Costing at least $300 for the three-dose regimen, Gardasil is one of the most expensive vaccines introduced so far4. Its private market price exceeds $500 in several developed and developing countries5, which few can afford in most LMCs1,6,7.

Price discrimination by pharmaceutical companies could improve vaccine access in LMCs. Merck introduced Gardasil in India at about $171 for the three-dose regimen8. Although such pricing enables middle-class access in some emerging economies, the vaccine remains unaffordable for most low- and middle-income populations in LMCs. Prices must fall below $2 per dose to make broad access possible in low-income populations, especially in countries where gross domestic product per capita is below $1,000 (refs. 9,10). It is unlikely that Merck or GSK can reduce vaccine prices to match these affordability targets because of the high production costs associated with their vaccines11. Company donations can also improve access12,13. Merck donated about 130,000 doses to PATH for demonstration studies in India, Peru and Vietnam14. Through the Gardasil Access Program, Merck aims to extend its support to LMCs and has pledged to make 3 million doses of vaccine available to eligible countries15. However, reliance on pharmaceutical company donations alone is unsustainable. Alternatively, donor-aided vaccine purchase can significantly increase vaccine access by facilitating distribution of highly discounted vaccines in eligible LMCs16. The Global Alliance for Vaccines and Immunization (GAVI) recently prioritized HPV vaccines17. However, owing to a $4 billion deficit and existing financial commitments to other vaccines, it might be unable to finance HPV vaccines18. Donors often face tradeoffs, and high prices limit the quantities of vaccines or treatments they can subsidize19.

Vaccine manufacturing in LMCs can also reduce prices. Over the past decade, manufacturers in India, Cuba, China and Brazil have demonstrated their capacity to produce low-cost vaccines that meet international quality standards. They primarily serve low-income markets, supplying 64% of childhood vaccines procured by UNICEF and 43% of vaccines procured by GAVI20,21. Recombinant hepatitis B (HBV) vaccines illustrate the potential impact of such manufacturers in improving vaccine access in LMCs. When introduced in the early 1980s, the HBV vaccine was priced at $50–80 per dose, one of the most expensive prophylactic vaccines at that time21. However, developing country vaccine manufacturers (DCVMs), using alternate expression platforms suitable for low-cost development, successfully brought inexpensive HBV vaccines to market in the 1990s. The ensuing competition reduced market prices to less than $0.30 per dose21,22. Procurement costs for large vaccination programs consequently decreased, allowing wider access to the vaccine in LMCs21. DCVMs are therefore potentially important suppliers of low-cost HPV vaccines.

For successful production, DCVMs require access to relevant technology, which can be protected by intellectual property (IP) rights. Although DCVMs have faced few patent barriers so far, changes adopted by developing countries to comply with the World Trade Organization's Agreement on Trade Related Aspects of Intellectual Property Rights might create new obstacles for vaccine development23. Before the agreement, many LMCs did not award product patents for biopharmaceuticals, including vaccines. Now, however, DCVMs must consider international pharmaceutical companies' product patent rights on vaccines and related technologies. Patents granted in developing countries might constrain the ability of manufacturers to develop vaccines and/or to sell vaccines in local and international LMC markets (Fig. 1), thereby reducing their development incentives.

Figure 1: Impact of patents on manufacturing, sale, and/or export of 'bio-similar' HPV vaccines by developing country vaccine manufacturers.
figure 1

Patents existing in the country of the manufacturer that claim the 'composition of matter' of necessary antigens (e.g., nucleic acid and/or amino acid sequences of the L1 proteins) or the vaccine itself (e.g., VLPs made of L1 antigens) would prevent DCVMs from developing and selling vaccines in their country without a license from the patent owner. If, however, patents only claim specific methods or processes for producing the vaccine, manufacturers may have freedom to operate if they use alternate processes to 'work around' those patents. Patents granted in jurisdictions outside the manufacturing country—especially in potential export markets—might also affect DCVM vaccine development plans. Manufacturers exporting vaccines to these countries would infringe patents if their vaccines embodied the compositions or processes protected by such patents. However, if vaccines had different formulations or were developed by using alternate processes, they could still be sold in the importing country.

Two recent reports have suggested that IP might be a barrier for DCVMs who are interested in developing HPV vaccines24,25. However, publicly available information about the patenting and licensing of HPV vaccine technologies in developing countries is minimal. We therefore systematically investigated the extent to which patents are a barrier to producing HPV vaccines in LMCs, focusing on India for several reasons. India bears nearly 25% of the global cervical cancer burden26,27 and its growing middle class is a potentially large private market for HPV vaccines. Including this vaccine in national immunization programs that target low-income populations will further expand the market, creating strong incentives for local manufacturing of inexpensive alternatives. In addition, several Indian manufacturers who are interested in developing HPV vaccines have concerns about potential patent impediments to such efforts.

The HPV vaccine patent landscape

First-generation prophylactic vaccines. Technologies that enabled the development of L1-VLP-based vaccines originated at the US National Institutes of Health's (NIH) National Cancer Institute (NCI), University of Rochester (Rochester, New York, USA), Georgetown University (Washington, DC) and the University of Queensland (Brisbane, Queensland, Australia) (Table 1). MedImmune (Gaithersburg, Maryland, USA), Merck and GSK developed these technologies further and performed safety and efficacy clinical testing to bring the vaccines to market28. The IP landscape for HPV vaccines is complex, with 81 US patents granted so far, linked to 86 specific Patent Cooperation Treaty (PCT) applications. Eighteen entities—ten of which are nonprofit—own these US patents. Nonprofit organizations own 20 of the 81 US patents, for-profits own 55, and for-profit and nonprofit entities jointly own 6 (Supplementary Table 1). Merck owns the most patents (24), followed by GSK and the US Government (arising from the NCI), who own 8 patents each.

Table 1 Timeline of patenting and licensing of HPV L1-VLP–based prophylactic vaccines

As of December 2008, 19 of the 86 international PCT applications were filed in India (Table 2). The universities and NIH have not sought patent protection for technologies underlying L1-VLP vaccines in India. However, Merck and GSK have applied for patents on HPV vaccine compositions. GSK alone has filed 13 of these applications. The Indian Intellectual Property Office (IPO) has awarded six patents, four to GSK and one each to Wyeth Holdings (Wayne, New Jersey, USA) and the University of Cape Town (Cape Town, South Africa). Although the determination of patent scope is complicated and sometimes the subject of costly litigation, we offer our preliminary analysis of patent claims based on our understanding of these technologies and discussions with researchers who developed first-generation vaccines. Patent 203333, awarded to GSK, claims compositions of a prophylactic vaccine that contains VLPs composed of L1 antigens from HPV-16, 18, 31 and 45. This is detailed in Claim 1, the first independent claim that technically confers the broadest scope of protection and reads, “A vaccine composition comprising virus like particles containing L1 proteins or functional L1 protein derivatives from human papilloma virus 16, human papilloma virus 18, human papilloma virus 33 and human papilloma virus 45 genotypes wherein the antibody response generated by the vaccine is at a level similar to that for each human papilloma virus, virus like particle formulated alone.” Our analysis suggests that only a vaccine containing L1-VLPs from all four HPV strains mentioned in Claim 1 directly infringes the patent. Therefore, Indian manufacturers are probably free to develop a bivalent HPV vaccine containing L1-VLPs for HPV-16 and HPV-18 only or a quadrivalent vaccine containing any combination of three, two or one of these four strains in addition to other unclaimed oncogenic strains. Patent 209780, also awarded to GSK, claims a vaccine composition comprising L1-VLPs for HPV-16, HPV-18 and an adjuvant containing aluminum hydroxide and 3-O-desacyl-4'-monophosphoryl lipid A (3dMPL). Claim 1 specifically reads, “A vaccine comprising a human papillomavirus 16 L1 virus like particles, human papillomavirus 18 L1 virus like particle, aluminum hydroxide, and 3dMPL.” Furthermore, Claim 4 reads, “The vaccine consisting of an HPV 16 L1 VLP, an HPV 18 L1 VLP, aluminum hydroxide, and 3dMPL.” However, our analysis suggests that a bivalent (HPV-16, -18 L1-VLP) prophylactic vaccine developed by an Indian manufacturer would not infringe this patent if formulated with a different adjuvant. Additional patents awarded to GSK (Table 2) claim nucleotide sequences of HPV early antigens (214047) and compositions of combination vaccines containing HPV L1 antigens (202425) and other antigens, respectively. These too are unlikely to constrain Indian vaccine manufacturers developing Gardasil or Cervarix 'biosimilars'.

Table 2 Patent landscape for HPV vaccines in India

The University of Cape Town patent claims methods to produce HPV-16 L1-VLPs in tobacco plants and their use in a vaccine composition. However, plant-based expression has so far been unsuccessful in yielding high amounts of purified HPV-16 VLPs11, thus limiting the commercial viability of this technology. Patent 220842, awarded to Wyeth, covers polypeptides of HPV early antigens E6 and E7, which are likely to be used in therapeutic cervical cancer vaccine compositions but are less relevant to L1-VLP–based prophylactic vaccines. We found no patents on HPV-16 and HPV-18 L1 nucleic acid sequences filed or awarded in India. However, Merck has four pending patent applications, claiming L1 nucleic acid sequences of HPV subtypes 31, 45, 52 and 58, optimized for expression in several yeast strains. Because the IPO can choose to significantly narrow the scope of or deny some claims during examination, it is difficult to assess whether Merck's applications will affect vaccine development in India.

Second-generation prophylactic vaccines. Currently, marketed first-generation vaccines are costly to produce as they use expensive expression systems to produce the L1 antigens. Moreover, both miss several oncogenic HPV strains that are present in India and other LMCs29, which is particularly problematic when countries lack the screening programs necessary to detect cancers not prevented by current vaccines. Researchers at the NCI and Johns Hopkins University (Baltimore, Maryland, USA) have developed an L2 (minor capsid antigen)-based vaccine. This approach would protect against infection by all oncogenic strains and would eliminate the costs of increasing the valency of L1-based vaccines11.

NCI and Johns Hopkins have partnered with Shantha Biotechnics (Hyderabad, India) to commercialize this candidate. They jointly filed Indian patent application 6219/DELNP/2007 (Table 2) with the explicit rationale of preserving freedom to operate and market exclusivity for Indian partners (J.T. Schiller & M. Schmilovich, NCI, NIH; personal communication). Shantha has signed a Cooperative Research and Development Agreement with the NIH, gaining access to biological materials such as codon-optimized plasmids as well as the expertise and training necessary to develop this vaccine. Shantha has nonexclusively licensed this technology from Johns Hopkins30. Using an Escherichia coli expression system to purify L2 antigenic peptides, Shantha hopes to lower development costs, thereby making it possible to reduce the price of vaccines significantly30 (A. Khar & R. Chaganti, Shantha Biotechnics; personal communication).

Indian patent application 131/CHENP/2007 also bears on second-generation vaccine development and is based on research performed at the University of Lausanne (Lausanne, Switzerland). Denise Nardelli-Haefliger and colleagues showed that recombinant clones of attenuated Salmonella enterica (serovar Typhi and Typhimurium) strains expressing HPV-16 and HPV-18 L1 antigens can induce a strong immune response11,31. This technology would allow oral or mucosal immunization against HPV-16 and HPV-18 infection. Lower development and implementation costs associated with this oral vaccine make it highly suitable for use in LMCs. To maximize the potential benefits of this technology to LMCs, Nardelli-Haefliger and colleagues assigned ownership of enabling IP to Indian Immunologicals (Hyderabad) (D. Nardelli-Haefliger, personal communication). Indian Immunologicals has a memorandum of understanding with Lausanne and has received biological materials, know-how and training. It has also filed international patent applications (Table 2) but will not seek patent protection in Organisation for Economic Co-operation and Development (OECD) markets (R. Sriraman, D. Thiagarajan & K. Kumar, Indian Immunologicals; personal communication). With assured access to essential patents and know-how, Indian Immunologicals has strong incentives to invest in the development of an oral HPV vaccine. Both the Shantha and Indian Immunologicals vaccines are currently in the preclinical phase. Shantha projects a 2015 market entry at an initial price of $15 per dose32. Both manufacturers believe, however, that prices will drop further as vaccine adoption increases, eventually reaching around $1–2 per dose, which would make broad access feasible.

Serum Institute of India (Pune) and Bharat Biotech (Hyderabad) are both also developing L1 VLP-based vaccines. Serum's candidate will probably be a bivalent HPV-16 and HPV-18 vaccine. The company will seek a nonexclusive license from the NIH for cell lines optimized for high expression of L1 antigens and will nonexclusively license the Hansenula polymorpha expression platform from Rhein Biotech (Düsseldof, Germany). Serum anticipates a market entry of three to four years after project initiation (S. Singh, Serum Institute; personal communication).

In addition to the L1-VLP vaccine, Bharat scientists are exploring a chimeric L2-HPV vaccine. They plan to express an L2-HBV small surface antigen fusion protein in Picchea pastoris to produce VLPs containing HPV-L2 antigens at high density. Because the HBV surface antigen spontaneously assembles into VLPs, Bharat hopes to reduce the price of the vaccine by circumventing the high costs of purifying and assembling VLPs. Bharat filed a provisional patent application for this vaccine in India last year. Despite developing this technology in-house, Bharat might seek a nonexclusive license from the NIH for the cell lines used in neutralizing assays (S. Kandaswamy, Bharat Biotech; personal communication).

Freedom to operate

Despite considerable patenting activity, our analysis suggests that IP will not preclude manufacturing of first-generation L1-VLP–based vaccines unless they are identical in formulation or strain coverage to those compositions claimed in granted Indian patents. We cannot, however, make this claim definitively owing to uncertainties in interpreting claims and the fate of pending applications. The claims analysis we present is therefore not legal advice but rather a starting point for independent freedom to operate (FTO) analyses by interested parties. Although there are several patents and pending applications on promising second-generation technologies (e.g., L2-based vaccines or oral L1 vaccines), they so far appear to preserve freedom for DCVMs.

IP transparency

Recent studies suggest that the lack of IP transparency could be an important impediment for DCVMs exploring new vaccine candidates33. Lack of patent claim information in publicly available Indian patent databases made our own research slow and expensive. Our experience mirrored those of Serum (Y. Dalvi, Serum Institute; personal communication) and Bharat. Indeed, Bharat's R&D was delayed owing to uncertainty about the status of patent protection for HPV antigens in India (S. Kandaswamy, Bharat Biotech; personal communication). Moreover, many countries in Africa, Latin America and Southeast Asia—potential markets for HPV vaccines—lack online patent databases, making it very difficult to determine which LMCs have pending or granted patents. More importantly, LMC companies generally lack the substantial financial and human resources necessary to perform FTO analyses using proprietary databases available in developed countries.

Shantha, Indian Immunologicals and Bharat often rely on researchers to conduct in-house patent searches (R. Chaganti, A. Khar, Shantha; R. Sriraman, D. Thiagarajan, K. Kumar, Indian Immunologicals; S. Kandaswamy, Bharat; personal communication). The WHO Initiative for Vaccine Research or other agencies could help coordinate FTO and patent landscaping services to advise regional manufacturers on potential IP barriers for HPV vaccine development. The creation of resources to map and update the IP landscape for novel HPV vaccines could facilitate regional manufacturing efforts. This could be developed in partnership with the DCVM network.

Potential roles for universities and funders

Universities and nonprofit research institutions exploring new HPV vaccines can expedite access to technology in LMCs. The IP management practices of academic institutions, who are the primary generators and gatekeepers of IP for vaccine technologies, will greatly affect regional vaccine manufacturing. The Lausanne-Indian Immunologicals partnership, for example, harnesses the capacity of a DCVM to commercialize a vaccine candidate with potentially high public health impact in LMCs despite little commercial interest in OECD countries. The NCI-Johns Hopkins-Shantha partnership to commercialize L2-based vaccine technology further illustrates how IP management can create a pathway for product access in low-income markets.

University licensing terms are generally not publicly available, except when parties choose to disclose them voluntarily. This has precluded a definitive analysis of whether Rochester, Queensland and Georgetown preserved freedom to use these technologies or subsequent improvements in LMCs when negotiating exclusive licenses with Merck and MedImmune. However, the licensing of vaccine technologies underlying Gardasil and Cervarix does not conform to recent university technology transfer practice guidelines to maximize benefit for the global poor34,35. This is understandable because the licenses in question were crafted in the 1980s, before these guidelines were developed. In addition, limited recombinant vaccine production capacity in LMCs rendered humanitarian licensing largely unnecessary at the time. The inaccessibility of HPV vaccines, however, illustrates why recently recommended practice guidelines deserve attention, especially as new technologies for prophylactic or therapeutic vaccines emerge.

Moving forward, universities and other nonprofit research institutions should adopt IP management strategies that preserve options for DCVMs. Preferred practices include default nonexclusive licensing, exclusive licenses with geographic fields of restriction (to ensure LMC companies have FTO), retaining rights to sublicense to regional manufacturers, nonprofit organizations and/or public-private partnerships, humanitarian-use clauses for patented technologies and products, and 'White Knight' clauses to ensure vaccine affordability36,37.

Universities can also help to promote IP transparency. Secrecy surrounding licensing exacerbates uncertainties in FTO analyses. Publicly available licensing information can prevent regional manufacturers from wasting time and money on technologies that are blocked by patents and licenses. More importantly, illuminating unblocked pathways can create incentives to commercialize vaccines that are of little interest to OECD manufacturers. Universities owning upstream technologies can promote and compel disclosure of licensing terms as part of licensure to improve transparency. Alternatively, sponsors of university research can make transparency a condition of funding by stipulating (i) disclosure of what geographic regions and fields of use exclusive licenses cover or (ii) publication of licensing contracts. Regional manufacturers can easily identify technologies that are available for licensing and potential partners for vaccine development if a central portal or electronic clearinghouse of all HPV vaccine technologies is created.

Technology transfer: beyond patents

Although this study focuses on patents, researchers and DCVMs affirm that additional know-how is also crucial for developing new vaccines23. Even when technologies are in the public domain or are available for licensing, vaccine development requires considerable expertise38. Universities and other non-profit entities can address this need by creating collaborative technology transfer partnerships modeled, for example, on the NIH Rotavirus Technology Transfer program39,40. Transfer of three second-generation HPV vaccine technologies to Indian companies potentially increases the likelihood of producing a vaccine better suited for LMCs than current vaccines. Oral, needle-free delivery of HPV vaccines, for example, might reduce the risk of infection of other sexually transmitted agents such as HIV and HSV, eliminate multiple healthcare visits, and increase patient compliance in resource-poor regions if doses can be administered at home as reconstituted oral drops. Collaborative partnerships with DCVMs might also produce vaccines designed from the onset to meet specific implementation characteristics of resource-poor regions, such as heat-stable formulations and single-dose or combination vaccines. Finally, market competition between one or more second-generation vaccines will probably reduce prices.

Conclusions

Experiences with introducing new vaccines suggest that 20 years could pass before women in LMCs gain access to HPV vaccines41. Meanwhile, every five-year delay in vaccine introduction could result in nearly 1.5 to 2 million more HPV-related deaths1. The prevention of these fatalities will require vaccines that entail fewer doses, minimize interactions with healthcare professionals, and are suitable for delivery in a resource-poor setting. Although these are tremendous challenges independent of price, improved access to HPV vaccines will depend on a reduction in vaccine prices. Regional manufacturers can accomplish this by lowering production costs and by developing vaccines tailored for resource-poor settings. Furthermore, increased DCVM competition can lower prices. Our patent landscape suggests that patents on first-generation vaccines do not seriously inhibit the development efforts of DCVMs. Regional manufacturers, national governments and international agencies should consider this an opportunity and take the necessary steps to make low-cost vaccine production a possibility.

Academic research institutions, from which most HPV vaccine technologies emerged, can play an important role in supporting regional manufacturing. Their technology transfer practices can promote new channels for regional manufacturing while ensuring that licensing does not block pathways to low-cost regional manufacturing of existing vaccines. Improving access to know-how and creating IP transparency can further facilitate regional manufacturing. By participating in technology transfer partnerships and adopting favorable IP management practices, universities can expedite access to new generations of life-saving HPV vaccines and increase the public health impact of these vaccines in LMCs.