Giving Developing Countries the Best Shot: An overview of vaccine access and R

8/7/2019 Giving Developing Countries the Best Shot: An overview of vaccine access and R

1/31

Giving developingcountries the best shot:An overview of vaccineaccess and R&D


2/31

Author: Paul Wilson

This paper draws on studies conducted byAndrew Jones for MSF and Paul Wilson for Oxfam

Cover photos: Nico Heijenberg/MSF, Oliver AsselinBack cover photos: Guillaume Ratel, Oliver Asselin

Cover design and layout: Daniel Jaquet

April 2010


3/31

An overview of vaccine access and R&D 1

Executive summary 2

Background 5

Overview of vaccines and immunisation

Vaccine markets

Vaccine suppliers and business models

Access to vaccines 7

The current situation

International nancing for vaccine purchase

GAVI

The situation of middle-income countries

The pneumococcal Advance Market Commitment

Vaccine prices

Tiered pricing of vaccines and the PAHO controversy

Manufacturing costs

Barriers to new entrants

Summary of access challenges

Vaccine research and development 15

Overview of vaccine development

Regulatory approval of vaccines

R&D costs

Alternative models for R&D

Push funding

The Malaria Vaccine Initiative (MVI)

The Meningitis Vaccine Project (MVP)

Pull funding

Advance Market Commitments

FDA Priority Review Vouchers

Prize fundsAccess to technology

Summary of R&D challenges

Conclusion and policy objectives 22

Abbreviations and Acronyms 24

Glossary 25

References 27

Table of Contents


4/31

Vaccines have made possible some of the greatestpublic health successes of the past century.Immunisation helps avert an estimated 2.5 millionchild deaths each year, as well as millions more boutsof illness and disability.1 Poor countries as well asrich have benefited, although developing countriesalmost always benefit only after long delays. Basicchildhood immunisation is one of the few healthinterventions to which most of the worlds poor haveaccess, free of charge and through the public sector.In fact, immunisation is one of the most equitablehealth interventions, protecting girls and boys alike,and reaching the poor within countries at higher ratesrelative to the wealthy than other services.2

Despite their impact, vaccines have generally receivedless attention than drugs. But the vaccine landscapeis shifting, and new opportunities, challenges, anddebates have pushed vaccines to the centre of globalhealth discussions. The issues are complex andthe experience gained in the struggle for access toHIV medications is an imperfect guide. The basicprinciplesequitable access and research anddevelopment (R&D) based on needsare the same,but vaccines differ from drugs in important ways.

There are multiple factors that make delivering

vaccines to children in developing countries difcult.These include among others high prices of newervaccines, the lack of R&D for better-adapted andneeded vaccines, as well as weak health systems withcorresponding health worker shortages. This paper willfocus on the rst two points, outlining some of themajor issues and exploring possible solutions.

Two fundamental challenges surround vaccineaccess and R&D: First, the newest vaccines are oftenprohibitively expensive, in part because of a lack ofadequate competition in the market, hindering theiruse in developing countries. Second, because there

is little incentive for pharmaceutical companies toconduct R&D for diseases that affect populations withlimited purchasing power, some diseases continue tobe unaddressed by vaccines altogether, while manyvaccines are not well-adapted for people in developingcountries.

Challenges in expanding accessto newer vaccines

Basic immunisation coverage has improved substantiallyin recent decades. According to WHO/UNICEF estimates,coverage of the third dose of the DTP (diphtheria-tetanus-pertussis) combination vaccine, a standardmeasure of immunisation coverage, reached about80% in 2007 a remarkable increase from only 20%in 1980.3 Nonetheless, 26 million children roughlyone in ve born each year remained unimmunisedglobally.4 Wider use of available vaccines could helpavert a further two million deaths annually in childrenunder ve years.5

The Global Alliance for Vaccines and Immunization(GAVI) was founded in 2000 to expand the scope andaccelerate the delivery of newer vaccines to childrenin the poorest countries. Initially, GAVI focused onadding two newer vaccines to the basic package: Hib(Haemophilus influenzae type b) and hepatitis B vaccines that were already in routine use in wealthycountries by the early to mid 1990s.6 By the end of2008, global coverage of the Hib vaccine was 28%,while Hep B was 69%.7 While this demonstratesimportant progress, it also illustrates the lag in vaccinedelivery between developed and developing countries.

GAVI is currently focusing on adding two newervaccines that have been introduced in wealthycountries over the last several years: vaccines toprevent the severe diarrhea caused by rotavirus andvaccines to prevent pneumococcal disease, whichtogether account for 1.3 million child deaths per year.8It is also looking to provide a new meningitis vaccine,the human papillomavirus (HPV) vaccine that helpsprevent cervical cancer, and several others. The goal isto shorten dramatically the delivery time lag betweenrich and poor countries.

Yet since these newer vaccines remain relativelyexpensive, and since GAVI is facing serious fundingshortfalls, ambitious introduction plans may bejeopardized. GAVIs ability to provide broad access tonewer vaccines will depend on bringing prices downdramatically, as well as on filling the multi-billiondollar funding gap. Of the U.S.$7 billion it will needfor the ve-year period until 2015, GAVI has securedonly 40%, leaving a $4.3 billion funding shortfall.9

Although an increasing number of developing countryemerging producers (from whom 53% of vaccines

funded by GAVI are purchased)10 have entered theglobal vaccine market, the new and most expensivevaccines continue to be produced by a handful ofmultinational pharmaceutical companies whoseoligopoly status allows them to charge high prices.

Executive summary

2 An overview of vaccine access and R&D


5/31


There has been considerable consolidation in thevaccine industry, with global vaccine revenue largely inthe hands of just ve companies: GSK, Merck, Novartis,Sano-Pasteur and Wyeth/Pzer. The limited numberof innovative firms along with patent protectionand the large economies of scale for vaccines resultin markets for new kinds of vaccines typically beingcontrolled by one or two companies for long periodsof time.

These companies do employ tiered pricing, andhave agreed to provide the new vaccines to GAVI atsignicant discounts for use in the poorest countries.However, these tiered prices are almost certainly higherthan those that could be achieved through competition.

And the problem for middle-income countries is evenmore acute, as they do not qualify for GAVI-negotiatedprices. Although middle-income prices are lower thanthose charged in wealthy markets, they often exceedsubstantially the prices offered to other developingcountries. Middle-income countries are thus faced withthe difcult decision of introducing these new vaccinesat the expense of other national health priorities,limiting introduction to only the highest-risk groups,or not introducing them at all.

The key to bringing prices down signicantly and

thereby enabling wider coverage of newer vaccines will be the competition generated by emergingsuppliers entering the market. However, emergingsuppliers face multiple barriers to swifter marketentry: the increasing complexity of newer vaccines;increased regulatory stringency; lack of technologicalcapacity; and intellectual property barriers. Unlikewith medicines, though, building capacity amongemerging vaccine suppliers requires intensivetechnology, or know-how, transfer, as it is not possibleto simply reverse-engineer a vaccine. There is, inessence, no such thing as a generic vaccine, as itis impossible to certify that vaccines produced by

different manufacturers are identical. Therefore, aswith biologic drugs, vaccines cannot be licensed basedon bioequivalence to already-licensed products.

One way to address the know-how challenge is with thehub model, whereby a public or non-prot institutionprovides a training platform for emerging suppliers,as opposed to relying on bilateral technology transferrelationships between individual producers.

Weak national regulatory authorities in emergingsuppliers countries also constitute a barrier to their

market entry, as approval by a domestic regulatorybody is a prerequisite to applying for quality assuranceby the World Health Organization (prequalication).WHO prequalication, in turn, is required for vaccinesto be procured by GAVI/UNICEF.

Challenges in vaccine researchand development

As with drugs, vaccine R&D is dominated by theparadigm of multinational pharmaceutical/vaccinecompanies charging high prices for products tailoredto wealthy markets. Companies argue that high pricesare needed to recoup R&D costs. This model distortsR&D priorities such that companies are not necessarilydeveloping products to tackle the greatest globalmedical needs and are not producing products thatare adapted to the particular needs of developingcountries. In addition, companies are not necessarilyproducing products in a manner that ensures thelowest cost, focusing instead primarily on bringing

vaccines to market as quickly as possible, as even afew months of advance over competitors can bring bigcommercial gains.

Most of the basic research upon which vaccinedevelopment depends continues to be carried out byuniversities and public laboratories, with multinationalsassuming the later stages of vaccine development andmarketing. With their large revenues, experience andexpertise, companies have been able to bring newvaccines through the expensive and time-consumingdevelopment process. However, the public sector has

historically played an important role in vaccine R&D inthe U.S. and Europe. For example, the U.S. Army ledthe development of several vaccines after World WarII. Re-engaging the public sector in vaccine R&D couldhelp close the R&D gap.

Because the pharmaceutical industry for the mostpart does not conduct R&D targeted to developingcountry health needs, alternative mechanisms tostimulate needs-based R&D must be employed.There are several models that aim to either pushR&D via upfront funding (e.g. product developmentpartnerships PDPs), or to pull R&D via incentives

that entice industry to invest in developing neededproducts (such as advance market commitments, prizefunds and GAVI itself).

An example of a successful push mechanism is theMeningitis Vaccine Project, a PDP established in 2001by PATH and WHO that has developed a neededvaccine to address a specic strain common in Africasmeningitis belt. Technology was licensed through theU.S. National Institutes of Health to the Serum Instituteof India, which agreed to provide the vaccine at anaffordable price in exchange for transfer of know-how

support for clinical trials in Africa and India, and theprospect of a GAVI-supported market. The total projectcost amounted to just $60 million, excluding plantcosts, and the vaccine is expected to be available foruse in Africa toward the end of 2010. While this project


6/31


represents a success story, and shows how emergingsuppliers can play an increasing role in vaccine R&D,the model will mainly be useful for adaptations ofexisting vaccines with known technologies, rather thanfor the development of entirely new and more complexvaccines, such as those for TB, malaria and AIDS.

An example of a successful pull mechanism isGAVI itself, which through its long-term purchasecommitments sends a signal to industry that developingcountries are a viable and long-term market. However,GAVIs pull has not been strong enough to stimulatethe development of new and sophisticated vaccines.

The advance market commitment (AMC) model is a

pull mechanism that ideally promotes needed vaccineR&D by guaranteeing a subsidized market for theresulting product if it meets certain specificationsand is purchased by countries/donors. But the AMClaunched by donors in 2009 to accelerate delivery ofpneumococcal vaccines was aimed at two vaccines thatwere already in nal stages of development and closeto gaining marketing approval, thus rendering it morea procurement mechanism than an R&D incentive.It is possible that these vaccines could have beenpurchased as or more cheaply through conventionalUNICEF tender procedures than with an AMC. Whether

or not the pneumococcal AMC accomplishes its goals,the question will remain whether this is an appropriatemechanism to stimulate the development of new

vaccines, as originally hoped. While well-designedAMCs could play a role in mid-stage developmentor for less complex vaccines as a complement topublic sector research funding, PDPs, and other pushmechanisms they are unlikely to be a practical wayto drive R&D for challenging early-stage vaccines thatface substantial scientic obstacles.

Prize funds are another innovative pull mechanismto stimulate health-needs-driven R&D while securingaccess to a product at an affordable price upfront.The prize awarded to a successful product developerwould be linked to the products public health impact,and producers may be asked to agree in advance tosell the product at an affordable price. Several prize

fund proposals are on the table, but none has yetbeen established for vaccine development.

Because several new vaccines with vital implicationsfor developing countries have come to market recently,every effort must be undertaken to ensure that childrenin these countries gain access to them. This will dependboth on raising additional funds and bringing pricesfor newer vaccines down by accelerating the entry ofemerging suppliers into these markets. Technology andknow-how transfer will be crucial, and will also helpemerging suppliers to assume a larger role in vaccine

R&D. And innovative models to stimulate R&D will becritical to meeting still-unaddressed vaccine needs.


7/31


Overview of vaccines and immunisation

In all countries, the basic package of vaccines comprisesat least the six vaccines that were included in theoriginal WHO Expanded Programme on Immunization(EPI): BCG (against tuberculosis), polio, measles,diphtheria, tetanus, and pertussis (the last three makeup the DTP combination). With the help of fundingfrom the Global Alliance for Vaccines and Immunization(GAVI), most countries also provide or are about tointroduce vaccines against Hepatitis B (Hep B) andHaemophilus inuenza type b (Hib). Vaccines againstrubella, mumps, and yellow fever are also widely usedin developing countries.

Three new vaccines important to developing countrieshave been developed in the last decade: thepneumococcal conjugate, rotavirus, and HPV vaccines.Pneumococcal disease causes over 800,000 under-vedeaths in developing countries and between one andfour million episodes of pneumococcal pneumonia inAfrica alone.11 Rotavirus causes over 500,000 deathsevery year.12 Thus, the new pneumo and rotavaccinescould have a major impact on child mortality. Humanpapillomavirus (HPV) is the cause of cervical cancer,which kills more than 160,000 women each year,more than 90% of them in developing countries. The

new HPV vaccines could theoretically prevent about70% of these deaths. These vaccines command largemarkets in the developed world, but they are far moreexpensive than older vaccines and are not yet widelyavailable in low- and middle-income countries. Muchof the policy debate over vaccines in the last few yearshas focused on how to make these important vaccinesavailable to poor countries.

Furthermore, several important new vaccines are indevelopment, including the first vaccines againstmalaria and dengue fever as well as improved orcheaper versions of vaccines against meningococcal

disease, Japanese encephalitis, cholera, typhoid andtuberculosis. Vaccines against other infectious diseasesand even against some non-communicable diseases,including cancer, are being pursued. HIV vaccineswould have an enormous impact, but are many yearsaway.

Finally, the 2009 inuenza scare and the controversyover allocation of a limited supply of vaccines brought about a belated recognition that developingcountries are at least as vulnerable to pandemic uand other unpredictable global epidemics and need

fair access to protective vaccines.

Vaccine markets

Vaccines make up a small (about 3%) but rapidlygrowing segment of the global pharmaceutical market.Total sales are expected to grow from about $20.5billion in 200813 to $34 billion by 201214. Sales in low-and middle-income countries were estimated at about$1.6 billion in 200815, or less than 10% of the total.UNICEF, which buys most vaccines for low-incomecountries and many for middle-income countries,procures 40% of global vaccine doses, which, however,account for only 5% of global market value.16 But low-and middle-income country markets are expected tocontribute substantially to future growth.

This is a remarkable turnaround for an industry thatwas widely perceived to be in decline just a fewyears ago.17 Reports of rms abandoning vaccines18have been replaced with giddy celebrations ofblockbuster sales. Much of the new enthusiasm ofindustry and investors for vaccines derives from theunprecedented commercial success of two products,Wyeths pneumococcal vaccine Prevnar and MercksHPV vaccine Gardasil, each of which brought in morethan $2.8 billion in sales in 2008. These huge revenueswere made possible by very high prices (over $300 fora three-dose course of Gardasil), which have shattered

the notion of vaccines as low-margin commodities.19

Vaccine suppliers and business models

Vaccine manufacturers are conventionally dividedinto two groups: the established multinational rmsbased in the U.S. and Europe, and the emergingsuppliers based in developing countries. Mergers andthe departure of many rms from the vaccine businessin recent decades have left the rst segment of thevaccine industry extraordinarily concentrated:20 the vetop multinational rms (GSK, Merck, Sano-Pasteur,

Wyeth - now part of Pzer -, and Novartis) accountedfor about 85% of global sales in 2008.21 Their share ofthe market is much lower in volume terms, however,as the emerging suppliers produce large volumes ofcheaper vaccines.

Traditionally, multinational vaccine companies,building on publicly funded basic research, have beenresponsible for most vaccine innovation, drawing ontheir greater revenues, experience, and expertise tobring new vaccines through the expensive and time-consuming development process. Their business model

depends on charging high prices for new vaccines inorder to recover R&D costs and return large prots totheir investors. Bringing vaccines to market as quicklyas possible is typically more important to them thanachieving the greatest efciencies in production.

Background


8/31


The multinational rms are primarily focused on high-price, relatively low-volume rich-world markets, whichprovide the great bulk of their revenues. But thereare important differences among the rms. GSK andSanofi have long experience in developing-countrymarkets, public and private, and both sell in largevolumes to UNICEF, together accounting for 52% ofUNICEFs purchases in 2006. These manufacturersderive real if modest prots from sales in developingcountries, which thus constitute an integral part oftheir business models. In contrast, Merck22 and Wyethhave only entered these markets recently and in manyways are still rening their approach. At least initially,they appeared to be motivated more by a sense ofcorporate responsibility (or by the fear of adverse

publicity) than by a belief that low-income marketsoffer meaningful commercial opportunities. Both rmshave explored donation strategies, but donation is nota sustainable business model and programmes arevulnerable to management turnover and changes incorporate strategy.

The emerging suppliers are a more diverse group thatincludes both traditional state-owned rms devotedto supplying national programmes with basic vaccinesand privately owned manufacturers. Several of theprivate firms, particularly in India, have expanded

rapidly and now supply a significant proportion ofbasic vaccines purchased by UNICEF. A few of thepublic manufacturers in developing countries havealso begun to export their products and considersales to other developing countries and public sectorbuyers such as PAHO and UNICEF. Among the moreimportant and forward-looking public sector firmsare Brazils BioManguinhos and Butantan, ChinasChengdu, and Indonesias Biofarma; private firmswith WHO-prequalied vaccines include Indias SerumInstitute, Panacea, Shanta, and Biological E. Bothpublic and private sector emerging manufacturersare represented by the Developing Countries Vaccine

Manufacturers Network.23

The emerging suppliers have traditionally sold older,less complex vaccines in high-volume, low-marginmarkets (emerging suppliers provide 86% of traditionalvaccines globally24). To thrive in these markets, theyhave focused on exploiting cost advantages ratherthan innovation. This is changing, however, and themore ambitious rms are increasing their investmentin R&D and looking to develop more complex vaccinesand, eventually, to enter high-income markets.Emerging manufacturers already produce Hepatitis B-and Hib-containing vaccines, included the pentavalentvaccines, and several rms are developing rotavirus,pneumococcal conjugate, and Japanese encephalitisvaccines, among others. Serum Institutes newMeningitis A conjugate vaccine has just been licensed

(see below).

However, the emerging manufacturers are still wellbehind the multinational rms in technology, know-how,and regulatory expertise. Many are quite procient atscaling up manufacturing processes for mass productionbut are still quite weak in earlier stages of R&D.

Building the innovative capacity of emerging supplierswill make markets more competitive and lessen, butnot eliminate, dependence on the multinationals fornew vaccines. For vaccines with both high-income

and developing-country markets, tiered pricing is oneway to facilitate access in low- and middle-incomecountries until competition can reduce prices for all,and generate vaccine versions tailored to developing-country needs. Yet many developing countries that donot qualify for the lowest price are dissatised withthe current tiered pricing schemes and have resistedimplementation of tiered prices for new vaccines.For vaccines with little prospect of rich-world sales,push funding, or perhaps large prizes, which de-linkthe price of a product from the cost of R&D, may bethe best way to cover R&D costs while keeping priceslow. These issues are discussed in more detail in the

following sections.


9/31


A majority of the worlds children now receive a setof basic childhood vaccines, but millions more remainunimmunised and several important newer vaccineshave yet to reach children in most developing countries.The major barriers to access are high vaccine prices,inadequate nancing for immunisation, and weaknessof national immunisation systems. This paper willnot address health system issues but will focus onprice and nancing. Since increasing the number ofsuppliers is a powerful way to reduce vaccine prices,this section will also discuss barriers to entry of newmanufacturers.

The current situation

The struggle to bring vaccines to the worlds poorhas shifted dramatically in recent years. Severaldevelopments have brought real improvements forchildren in the poorest countries.

First, basic immunisation coverage has improvedsubstantially, bringing, at the least, the six EPI vaccinesto about 80% of the worlds children, according toWHO/UNICEF estimates. These vaccines are very cheap,so weak health systems constitute the primary obstacleto expansion and maintenance of coverage.

Second, the creation of GAVI in 2000 has providednew resources for the purchase of more expensivevaccines on behalf of low-income and some lower-middle-income countries. (Eligible countries weredened at GAVIs inception as those with a GNI percapita below $1,000; in 2011, the threshold will beincreased to $1,500 GNI per capita).25 These fundshave enabled most eligible countries to introduceHep B and Hib vaccines, and should soon allow theintroduction of rotavirus and pneumococcal conjugatevaccines as well, if GAVIs current nancial challengescan be overcome.

Third, the major multinational vaccine firms haveaccepted in general the principle of providing theirproducts to the poorest countries at discountedprices through UNICEF. At the same time, emergingsuppliers have entered the UNICEF/GAVI market for allbut the newest vaccines, increasing competition andcontributing to signicant price reductions for somevaccines, especially Hep B. Prices for other vaccines,notably the pentavalent combination vaccines (GAVIfunds the pentavalent vaccine combining DTP, Hiband Hep B vaccines into one), have not yet fallen asrapidly as anticipated.

These developments have improved access to vaccinesin GAVI-eligible countries. But GAVIs own finances

are now under severe strain. The organisation willbe challenged to honor existing commitments andis unlikely to finance the large-scale purchase ofadditional vaccines for some time. The Bill and MelindaGates Foundation in January 2010 announced that itwill commit $10 billion to research, development anddelivery of vaccines over the next decade.26 To date,however, it remains unclear if any of these funds willbe allocated specically to GAVI.

Moreover, the situation in middle-income countries,which receive very little donor support for immunisation,

is increasingly contentious. The pharmaceutical industrysees these countries, especially the rapidly growingemerging economies such as Brazil, China, and India,as potentially lucrative markets, and is not willing toprovide new vaccines to these countries at the samelow prices it offers to UNICEF/GAVI. There is growingconcern that these countries, in particular those withincomes only slightly above the GAVI threshold, maynot be able to afford new vaccines or may be forced todivert funds from other health programmes to do so.As a result, middle-income countries, especially thosethat procure vaccines through PAHO, have become thenew battleground over vaccine prices.

Access to vaccines


10/31


International nancing for vaccine purchase

GAVIEven the poorest countries are generally able topurchase the six basic EPI vaccines from their ownhealth budgets, but many would not be able to affordthe newer vaccines without external assistance. GAVIwas created in 2000 to accelerate the adoption ofnew and underused vaccines in poor countries; itspent about $600 million in 2008, mostly to purchasevaccines through UNICEF for 72 eligible low- andlower-middle-income countries. GAVI has helped mostof these countries introduce Hep B and Hib vaccines,and it is poised to nance the introduction of rotavirusand pneumococcal vaccines.

There is no doubt that GAVI has done a great dealto facilitate access to vaccines among the poorestcountries. But it is currently facing a serious nancialcrisis: spending on pentavalent, rotavirus, andpneumococcal vaccines is expected to push totalexpenditures to $1.6 billion in 2013 while expectedresources fall from a peak of $1.0 billion or so in 2010(see Figure 1). (The expected fall in resources comesin large part from a big decline in income from theInternational Finance Facility for Immunisation after2010. The IFFIm was designed to frontload resources

for immunisation by issuing bonds in capital marketsthat are to be subsequently paid off through donorpledges of future support). Unless donor pledgesincrease dramatically or a new source of funding isfound, GAVI will have to make difcult choices in the

next few years. It may have to delay support for somevaccines, prioritise approved applications from eligiblecountries, or dramatically increase the share of vaccinecosts borne by countries (co-nancing rates).

In this more difcult scal environment, GAVIs goalof introducing new vaccines as soon as possible indeveloping countries will depend on accelerating adecline in prices, both by facilitating the entry of newsuppliers and by placing greater emphasis on price inprocurement. In vaccine procurement, however, pricemust be balanced against security of supply and theneed to keep the GAVI/UNICEF market attractive to arange of rms.

The situation of middle-income countriesEven if GAVI can sustain its current model, it will notbe able to expand its support to the bulk of middle-income countries for the foreseeable future. When theGAVI Boards recent decision on eligibility policy takeseffect in 2011, eligibility will be limited to countrieswith per capita GNI below $1,500 (taking ination intoconsideration, this is roughly equivalent to $1,000in 2000).28 This change will reduce the number ofeligible countries from the current 72 to about 58,although existing support for the graduating countrieswill continue at least through 2015. The Board also

increased the minimum level of immunisation coveragethat countries must achieve before they can introducenew vaccines with GAVI support; this provision willaffect several large countries, including India.29

2000

200

0

400

600

800

1000

1200

1400

1600

2001

Cash inflows Cash outflows

2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Figure 1: GAVIs projected resources and expenditures27


11/31


Middle-income countries face a double challenge inaffording new vaccines: they have almost no accessto international assistance to buy vaccines, but at thesame time they must pay signicantly higher pricesthan GAVI countries for many vaccines as a result ofindustrys practice of tiered pricing (see below). Asa result, many of these countries will have troubleintroducing the new pneumococcal and HPV vaccines.Approaches to ensuring access to new vaccines inmiddle-income countries are discussed below in thesection on tiered pricing.

The pneumococcal Advance Market CommitmentAn advance market commitment (AMC) is an innovativefinancing model that subsidizes pharmaceutical

companies for the development and production of newvaccines. The subsidy is meant to reduce the risk forpharmaceutical companies of investing in products fordeveloping country markets with limited purchasingability, and is only paid once a vaccine meeting certainspecifications is purchased by eligible developingcountries (or donors on their behalf ) at a pre-set price.The subsidy covers an agreed volume of vaccines,after which a predetermined and lower long-term price(also called tail price) is offered to countries. Thisaims to ensure the vaccines use is sustained beyondthe duration of the subsidy.

GAVI and UNICEF will procure the new pneumococcalconjugate vaccines through an AMC. Although AMCswere originally proposed as a way to stimulatedevelopment of new vaccines for neglected diseases,the two leading pneumococcal vaccines were already inadvanced development for high-income markets whenthe AMC was launched (GSKs 10-valent conjugate,marketed as Synflorix, and Wyeth/Pfizers 13-valentconjugate, marketed as Prevnar 13). The AMC mayaccelerate development of a pneumococcal vaccineby one or more emerging suppliers, but it is servingprimarily as a procurement mechanism rather than

an R&D incentive. The pneumococcal AMC has beencriticized as too expensive and too complicated, andas favoring the multinational firms over emergingsuppliers.30 It is possible that these vaccines couldhave been purchased as, or more, cheaply throughconventional UNICEF tender procedures, and the useof this complicated new mechanism to buy almost-licensed vaccines has certainly confused the discussion.Moreover, the agreement includes no provisions toencourage technology transfer to developing countrymanufacturers. But if the pneumococcal AMC works ashoped, it will bring a new generation of pneumococcal

vaccines to many of the worlds poorest countries atalmost the same time they are introduced in the richworld.

Pzer and GSK signed on to the AMC in March 2010,committing to supply 30 million doses of vaccine eachfor ten years,31 but it remains to be seen whetherdeveloping country producers will be able to meetAMC requirements and tap into allocated funds beforethese are exhausted. In addition, the implicationsfor the AMC of the new GAVI eligibility policies willhave to be worked out, in particular the higherimmunisation coverage requirement and the likelygraduation of several countries in 2011. Whether ornot the pneumococcal AMC accomplishes its goals,the question remains whether this is an appropriatemechanism for stimulating development of newvaccines, as originally hoped. This issue is addressedin the R&D section.

Vaccine prices

Vaccine prices, like those of other products, areshaped by the balance of supply and demand. Butseveral unusual features of vaccine markets have astrong influence on prices in developing countries.First, markets for new kinds of vaccines tend to becontrolled by one or at most two rms for extendedperiods, because of the very small number ofinnovative multinational rms, the large economies

of scale in vaccine production, and patent protection.This lack of competition gives originating firmssubstantial freedom to set prices during the rst phaseof a vaccines lifecycle, before additional suppliersenter the market. Second, this supplier marketpower is balanced in part by the market influenceof public sector purchasers, in particular the pooledprocurement mechanisms operated by UNICEF andPAHO. Third, tiered pricing, or market segmentationby national income, is increasingly standard (seebelow). Fourth, the large fixed costs and long leadtimes required to build new manufacturing plantsmean that predictability of demand is very important

to manufacturers, who will offer lower prices in returnfor long-term commitments.

Production costs are not in general an importantcomponent of the price of new vaccines in rich-worldmarkets, where rms are able to charge well abovecost. And while some margin is necessary to recoverR&D costs, actual prices are not determined in asimple way by, or justied by, R&D costs. But marginalcost of production does set a floor for the bottomtier of new vaccine prices (prices charged in GAVIcountries) and become an important determinant of

price in mature, competitive markets. Production costsin turn vary considerably among classes of vaccines,by production volume, and by site of production:emerging suppliers have signicant cost advantages insome but not all cases.


12/31


Thus, prices of vaccines in low- and middle-incomecountries for new vaccines tend to be determinedinitially by the pricing practices of the multinationaloriginator rms. Marginal cost of production sets atleast a theoretical minimum for the poorest countries.The speed at which new rms can enter the marketis the key factor over the longer run, however, asprices can be expected to fall in all markets withincreased competition and, in many cases, lowerproduction costs for emerging suppliers. But vaccinetechnology can constrain how far prices can fall evenwith competition and production efciencies.

Tiered pricing of vaccines and the PAHO controversyThe multinational vaccine firms all support and

practice, to the extent they are free to do so tieredpricing: a policy of charging high prices in rich countries,low prices in GAVI countries, and intermediate prices inmiddle-income countries. Tiered pricing is less centralto the business models of the emerging manufacturers,who typically sell older vaccines in more competitivemarkets.

For rms, tiered pricing is a prot-maximizing strategy:by charging different prices to classes of customerswith different willingness and ability to pay, they areable to maximize profits in high-income countries

while earning money in countries that cannot affordrich-country prices. Tiered pricing also helps firmsprevent or defuse criticism over high prices. Fordeveloping countries as a whole, tiered pricing isclearly preferable to a situation in which all countriespaid the same high price. With respect to vaccines,low-income countries benefit the most from tieredpricing, receiving the lowest prices, although the pricesthey pay under this system are not necessarily as lowas they would pay in fully-competitive markets. Thepractice of charging higher prices in middle-incomecountries than in the poorest countries has beencontentious, however. Firms argue that middle-income

countries, especially better-off upper-middle-incomecountries such as Brazil, have substantially greatercapacity to pay for vaccines than do GAVI countries.Middle-income countries argue that their populationsinclude many poor people and that the prices they areasked to pay are in any case too high and very high inrelation to what least-developed countries pay.

The controversy over tiered pricing came to a headin 2009 in a dispute between firms and the PAHORevolving Fund, a pooled procurement mechanismused by most countries of the Latin American region.

The Revolving Fund, which has helped membercountries to strengthen immunisation systems andintroduce new vaccines, has until recently been ableto buy most vaccines at prices very similar to thosepaid by UNICEF on behalf of GAVI (see Table 1). But

the suppliers of the new pneumococcal conjugate andHPV vaccines insisted that PAHO must pay more thanGAVI; PAHO, in turn, insisted on a so-called most-favored nation (MFN) clause in its contracts withsuppliers that requires that PAHO receive the lowestprices available to any purchaser, including UNICEF.

PAHOs decision not to procure the new GSK andWyeth vaccines, at least in the short run, removed theobstacle that the dispute over the MFN clause posedto the launch of the AMC. However, this conict mayre-emerge later in 2010 when UNICEF begins to procurerotavirus vaccines on behalf of GAVI.

Since the current state of GAVIs nances will delay

purchase of HPV vaccines, the most immediatethreat posed by the PAHO conict has been to GAVIprocurement of pneumococcal vaccines. PAHO iscurrently paying $65 to vaccinate a child ($21.75 perdose) for three doses of the Wyeth 7-valent vaccine,while the AMC will pay $21 to vaccinate a child ($7.00per dose) (and eventually $10.50 per child ($3.50 perdose)) for three doses of the superior and moreexpensive to produce 10- and 13-valent vaccines.

It is difficult to know how much middle-incomecountries outside the PAHO region pay for vaccines,

since these prices are not in general made public.Anecdotal evidence suggests that these prices vary agreat deal, with some lower-middle-income countriesthat procure through UNICEF receiving prices close tothose paid by GAVI, while others pay much more.

Tiered pricing can help developing countries (andGAVI) to afford new vaccines in the initial periodbefore the entry of additional suppliers makes marketsmore competitive. But these prices must address theneeds of all developing countries. While many middle-income countries have somewhat greater capacity topay for vaccines than the poorest countries, higher

prices mean fewer resources for other health priorities.PAHO should continue using its collective bargainingpower to ensure that the mark-up charged by rms ismodest. Other regions would probably benet fromcreating similar pooled procurement mechanisms,which can strengthen countries bargaining powervis--vis rms.

The problem of vaccine affordability is particularlyacute for lower-middle-income countries, as some ofthese countries will soon become ineligible for GAVIsubsidies. One measure that would help is to allow

these countries to continue to procure through UNICEFat GAVI prices. At the same time, PAHO and GAVIshould avoid being played off one another by therms and develop joint strategies that further theirshared interests.


13/31


Over the longer run, more efcient solutions should be

explored that would cover the cost of developing newvaccines appropriate for developing country populationswhile allowing prices to fall closer to marginal cost(see R&D section). Middle-income countries can andshould contribute to these solutions. In the absence ofnew models of funding R&D, measures that acceleratethe entry of new suppliers would help to bring pricesof new vaccines down more rapidly (see table 1).

Manufacturing costsManufacturing costs do not in general directly determinevaccine prices, which are set by prot-seeking rms in

markets that are more or less competitive. But theydo set a oor below which prices cannot fall even intiered pricing or highly competitive situations. Vaccineproduction costs include variable costs associated witheach dose of vaccine (cost of goods), which include

the cost of the vaccine components, vials, and so on;

semi-fixed costs associated with each productionbatch, including quality tests; and the fixed costs ofplant and equipment. Fixed and semi-xed costs makeup the bulk of total cost, typically contributing 60%and 25% respectively; this gives vaccine manufacturelarge economies of scale, since average costs fall withincreasing volume over a large range.33

The main factors that determine manufacturing cost arevaccine and production technology, presentation, andmanufacturing scale. Location of production matters,as some costs are lower in developing countries.

Design choices are also critical, and market incentivesin rich countries may drive the development of morecomplexand thus expensivevaccines than arerequired for public health impact.

Table 1: UNICEF, PAHO and U.S. public sector prices in 201032

(U.S.$; 10 dose vials unless otherwise indicated)

Vaccine

BCG (PAHO 20 dose vial)

DTPw

MMR (Zagreb strain forUNICEF & PAHO)

Yellow Fever

HepB (1 dose vial)

Hib (lyophilized)

DTP-HepB-Hib (pentaval-ent; 1 dose vial, liquid)

Rotavirus

Pneumococcal (7-valent

for PAHO and U.S., 10-or 13-valent for GAVI)

1 Weighted average prices per dose.

Minimum calendar (plus booster) for children less than 5 years old.

Full vaccination package (including booster) included in cost of vaccination.

*The booster doses for Hep B and Hib are not ofcially recommended in the WHO guidelines, but they are listed as an option ifgiven in combination vaccine. Prices are based on including the 4th dose (booster dose).

Booster not nanced by GAVI (only for 1-11 month old children).

** Not yet procured by UNICEF/GAVI.

*** Weighted average prices per dose from 2009.

UNICEF/GAVI1

0.11

0.18

0.93

0.90

0.27***

3.40 (1dose vial)

2.94

**

7.00

(via AMC)

PAHO1

0.10

0.15

0.92

0.65 BrazilOrigin

1.15 France

Origin

0.28

2.25 (1dose vial)

3.20

5.15Rotateq

7.50Rotarix

20.00

U.S.publicsector

--

--

18.64

--

10.25

8.66 (10doses vial)

--

59.18Rotateq83.75Rotarix

91.75

No. of dosesas per WHOrecommend-

ations

1

3+1

2

1

3+1*

3+1*

3+1

3Rotateq

2Rotarix

3

Cost ofvaccinationUNICEF/GAVI/

per child

0.11

0.72

1.86

0.90

1.08

13.60

11.76

**

21

Cost ofvaccination

PAHO/per child

0.10

0.60

1.84

0.65-1.15

1.12

9.00

12.80

15.45Rotateq15.00

Rotarix

60.00

Cost ofvaccination

U.S./per child

--

--

37.28

--

41.00

34.64

--

177.54Rotateq167.50Rotarix

275.25


14/31


Technology: Vaccines of different types can havedramatically different costs of production (see Figure 2).At one end of the spectrum, simple live-attenuatedvaccines such as oral polio or measles can be producedin very large batches from very inexpensive inputs, andcost no more than a few pennies per dose. At the otherextreme, the new pneumococcal conjugate vaccinesrequire the manufacture or purchase of puried bacterialpolysaccharides, conjugation of these polysaccharidesto a protein carrier, mixing and carefully balancing ofcomponent vaccines against as many as 13 bacterialstrains or serotypes, and a very large number of qualitycontrol steps. These vaccines cost $1-3 per dose toproduce not counting the cost of the plant itself.

Production technology also determines batch size,which is itself an important determinant of cost, sincesome expenses are incurred on a per-batch basis.

Access to technology and superior know-how maygive the multinational rms a cost advantage in somecases. On the other hand, emerging suppliers, forwhom reducing cost is a greater concern, may be moremotivated to realize efciencies.

The cost of producing technologically sophisticatedvaccines can be expected to come down over time

as rms accumulate experience and as more efcientproduction technologies are developed. But the natureof some vaccines makes them intrinsically more costlyto manufacture than others. For this reason, choicesmade during the earliest stages of vaccine developmenthave critical consequences for later manufacturingcost. Once a technology has been chosen, producersare largely locked in. For example, it will probably benecessary to switch to an entirely different vaccine

concept to achieve a low-cost pneumococcal vaccinewith broad serotype coverage.

Presentation: The form in which vaccines are packagedand the number of doses per unit can also contributesignicantly to cost: pre-lled syringes are more expensivethan single dose vials (and less suitable for developingcountries); 10-dose vials are cheaper per dose than1-dose vials. The relative contribution of presentationto total cost is greater for cheaper vaccines. With moreexpensive vaccines, the risk of wastage can outweigh thecost benets of multi-dose presentations.34

Scale of operations and capacity utilization: Becauseof the importance of xed costs, a plant producing 100

million doses per year will in general have lower costsper dose than a plant producing 50 million doses peryear. As a rule, emerging suppliers serving high-volume,low-margin markets tend to build larger plants. This canhave a signicant impact on vaccine production costs.

These gains from scale are realized only if plantsproduce at full capacity: per dose costs will be higherto the extent that capacity is underutilized.

Location of operation: Lower costs of brick and mortarconstruction generally make vaccine plants cheaper

to build in developing countries, even if much of theequipment has to be imported. The labor costs ofoperating a plant are also signicantly lower in Indiaor China than in Europe or the U.S. 35 In addition,the EMEA/FDA regulatory environment imposes someadditional quality control and operations costs. Withincreasing harmonization in good manufacturingpractices (GMP) standards and increasing labor costshowever, this difference is shrinking.

Figure 2: Relative complexity of different types of vaccines

Simpler More complex

PolioCost of goods difference of over 20X

Up to 100 M doses per batch

< 100 QC steps

Up to 1.5 M doses per batch

Up to 100 QC steps

JE

Rotarix

HPV

Rotateq Pneumo


15/31


Determining the cost of a specific manufacturingplant is difcult, mainly because manufacturers oftenbuild plants that produce multiple vaccines, makingthe attribution of the portion of the plant that dealswith a single vaccine difficult. Further, the numberof doses produced by a plant is often kept vague.However, a review of facilities constructed in the pastdecade36 suggests that costs for multinational rmsrange from $200 to $400 million, with additions toexisting facilities costing less than $200 million andmultiple vaccine plants up to $700 million. Emergingmanufacturers generally have much lower costs, whichare usually less than $100 million and can range fromless than $50 million up to $150 million.

Barriers to new entrants

Facilitating the entry of new suppliers is the mostpowerful way to accelerate the decline in prices ofnew vaccines in developing countries. But would-beproducers of versions of new vaccines, especiallyemerging suppliers, face a more complicated set ofobstacles than generic drug producers.

Increasing complexity of, and regulatory stringencytowards, new vaccines work together with know-how

and intellectual property (IP) obstacles to delay theentrance of competitors and preserve markets for thelargest companies.

Technological sophistication and access to know-how: Although patent barriers are becoming moreimportant (see below), the primary obstacle to thedevelopment and production of newer, more complexvaccines by emerging manufacturers has been lack oftechnological capacity. While the leading manufacturersin India, China, and Brazil are investing more in R&Dand closing the technology gap, their capability stilllags well behind that of the multinational companies.

This slows or prevents the development of versionsof existing vaccines as well as entirely new vaccines.Unlike small-molecule drugs, vaccines are not easilyreverse-engineered, as the greatest challenges oftenlie in details of production processes that cannot beinferred from the nal product. Thus proprietary know-how is often the greatest impediment to the entry ofnew suppliers into markets for sophisticated vaccines.

Lack of a generic pathway: Vaccines, like biologic drugs,cannot be licensed on the basis of bioequivalenceto already licensed products, essentially because it

is in general impossible to certify that vaccines madeby different manufacturers are identical. Even whena new vaccine is closely modelled on an existingone, its safety and efficacy must be independentlydemonstrated in clinical trials. Thus, there is technically

no such thing as a generic vaccine. Although the WHOprequalication process works quite well to endorsefollow-on versions of vaccines, in some cases withabbreviated requirements for trials, this remains animportant distinction between drugs and vaccines.

Regulatory requirements: The most ambitiousemerging suppliers, who hope to gain access to high-income markets, are preparing to meet increasinglystiff regulatory requirements. Moreover, there is sometendency for the standards set by national regulatoryauthorities in developing countries and by the WHOin its prequalification process to converge withthose of the FDA and EMEA, especially in the area ofgood manufacturing practices (GMP). Meeting these

standards increases costs. But the WHO prequalicationprocess has been highly successful, and it remains apractical way for developing country rms to sell theirvaccines to UN agencies and enter markets in otherlow- and middle-income countries. The weaknessof national regulatory authorities (NRAs) remainsa problem, however, both because approval by afunctioning regulator in the manufacturing country is arequirement for WHO prequalication (which is in turna requirement for all vaccines procured by UNICEF/GAVI), and because many vaccines purchased fromdomestic manufacturers are not prequalied.

Intellectual property: Historically, patents were notconsidered as important a barrier to follow-on vaccinesuppliers as access to technology and proprietary know-how. But patents are apparently an impediment in thecase of HPV vaccines, for example, and IP barriers, aswell as data exclusivity provisions, are likely to grow inimportance. Developing country manufacturers state thatcircumventing these barriers may in some cases delaytheir efforts to introduce competing vaccines by years.As is often the case, the new technologies that underliethe HPV vaccines were developed in publicly-fundedacademic labs, and the adoption by universities of more

open licensing policies that facilitated production bymultiple suppliers in developing-country markets wouldbe an important step. Governments need to examinethe impact of patents on vaccine availability and makeuse of exibilities enshrined in the TRIPS agreementto limit the negative impact of patents. Manufacturersin developing countries will also have to build theircapacity to manage IP.

In some cases, the formation of patent thickets canimpede the development of new classes of vaccines aswell as the entry of competitors.37

Capital: Interestingly, private-sector emergingmanufacturers do not describe access to capital asa barrier to entry, although it is an issue for manypublicly-owned manufacturers.


16/31


Summary of access challenges

GAVI has greatly improved access to vaccines suchas Hep B and Hib in low-income countries, buta dramatic funding shortfall is endangering itscommitment to introduce the pneumococcal androtavirus vaccines, as well as its plans to supportpurchase of HPV and other important vaccines. Theimpact of the economic crisis on donors and thehigh prices of these new vaccines, which are stillsupplied only by multinationals, are the main causesof GAVIs financing troubles. A truly competitivemarket for pentavalent, another important GAVIvaccine, has also been slower to emerge thanexpected, contributing to persistently high prices.

Most middle-income countries are not eligible forGAVI support, and growing market segmentation hasled to a conict between GAVI and PAHO over pricingof new vaccines. Any efforts to promote access andaffordability must keep middle-income countries inmind. A more comprehensive approach could includenew pooled procurement mechanisms (modeled onPAHOs Revolving Fund), access to GAVI prices forsome countries that might not qualify for subsidies,regional exports by government-owned producerssuch as Brazils BioManguinhos, and forms of tiered

pricing acceptable to middle-income countries.

Shortening the time it takes for competitive productsto reach market is probably the most promisingstrategy to make vaccines more affordable forcountries and donors. Strategies that could help toreduce barriers to entry include:- Mechanisms for facilitating technology transfer;- Mechanisms for preventing or removing patentbarriers, including open licensing policies on thepart of universities and government research bodiesand the use of TRIPS exibilities when appropriate;and

- Procurement policies that support competitionand, at a minimum, do not inadvertently reinforcethe dominance of the handful of establishedmultinational suppliers.


17/31


Although ensuring access to the many useful vaccineswe already have is an immediate priority, new vaccinesare also urgently needed. There are no licensedvaccines against HIV, malaria, dengue, or any of theother tropical parasitic diseases, and only a grosslyinadequate vaccine against tuberculosis. Improvedvaccines are needed against cholera and typhoid fever,as well as new vaccine formulations and presentationsmore suited to use in low-income settings.

This section begins by reviewing the current system ofvaccine R&D. It then outlines some of the obstacles todevelopment of needed vaccines for low- and middle-income countries and a range of possible solutions.

Because the entry of new suppliers into vaccinemarkets is one of the most important ways to drivedown prices and ensure access, and because newsuppliers of vaccines (unlike generic drug producers)must carry out substantial development and testing oftheir products, many of the issues raised in this chapterare relevant to access as well as to the developmentof entirely new vaccines.

Overview of vaccine development

Vaccines are very different from drugs: vaccines arehighly complex macromolecules or whole organismsdesigned to stimulate the immune system to fightdisease pathogens, while most drugs are smallmolecules that directly inhibit disease processes.But vaccine research and development, like drug

development, involves progression through a seriesof increasingly expensive phases, from exploratoryresearch in the laboratory to large-scale clinicaltrials and development of manufacturing processes.Candidate vaccines can fail at any stage, making R&Da very risky undertaking. Perhaps a quarter of productsthat enter clinical trials reach market. Figure 3 providesa snapshot of the research and development processfor vaccines.

Although the later stages of vaccine development arenow carried out largely by industry, and in particularby a handful of multinational rms, the public sectorplays a very important role. Most basic research isdone at universities and public laboratories, which

also contribute to discovery in many cases. Biotechcompanies are also increasingly important in earlystage R&D. The R&D capacity of emerging suppliersin developing countries is growing rapidly, and publicsector manufacturers remain important in severalimportant middle-income countries, including Brazil,India, China, Indonesia, and Mexico. Historically, thepublic sector played a more important role in vaccinedevelopment in the U.S. and Europe, as well the U.S.Army in particular developed many important vaccines.Rebuilding public sector capacity to develop and testneeded vaccines may be an important element of a

public health strategy.

Regulatory approval of vaccinesLike drugs, vaccines are licensed by national regulatoryauthorities (NRAs), which require evidence of efcacy(in this case prevention of infection or disease), safety,

Vaccine research and development

Figure 3: Stages of vaccine research and development

DiscoveryTranslation of knowledgeinto candidate vaccines

Basic ScienceUnderstanding of disease organisms,disease pathology, and immune system

Typically done at academicinstitutions with public funds

Biotechs and universitiesplay important roles

Open-ended and risky:most ideas fail

Perhaps 1 in 10pre-clinical productsmake it to market

Small trials (


18/31


and quality, including proof of good manufacturingpractices (GMP). New vaccines developed by the bigmultinational rms are typically licensed rst by theU.S. FDA or the EUs EMEA, while vaccines producedby developing country rms are usually approved rstby their own NRAs.

Manufacturers from either developed or developingcountries hoping to supply developing countries (otherthan their home market) must also be prequalifiedby the WHO. Prequalification, which is required forpurchase by UN agencies, including UNICEF andPAHO, is intended to ensure that vaccines used innational immunisation services are safe and effectivefor the target populations and that packaging and

presentation are appropriate. The WHO standards arerigorous but distinct from those set by the FDA andEMEA, and the prequalication process is thus a vitalalternative to approval by a rich-world regulator.

For a vaccine to be considered for prequalication,the national regulatory authority in the country ofmanufacture must be considered functional by WHO.Strengthening regulatory agencies, especially in thekey exporting countries, is thus crucial to a strategythat relies in large part on the emerging suppliers.

A few points are worth noting:1) Demonstration of efcacy for an entirely new vaccineor class of vaccines requires a large placebo-controlledtrial, in which rates of infection or disease in peoplewho receive the vaccine are compared to rates in theplacebo group. Developers of subsequent vaccinesagainst the same disease and relying on the samemechanism of protection may not have to demonstratereduced incidence. If study of previous trials hasestablished a so-called correlate of protection (alevel of immune response to the vaccine above whichindividuals are protected), regulators will acceptevidence that the new vaccine consistently produces

this protective response in lieu of incidence data. Thiscan sometimes make much smaller trials possible.

2) Safety standards for vaccines have become almostabsurdly stringent in recent decades, especially inthe U.S., in part because the public in high-incomecountries is no longer particularly concerned aboutthe diseases vaccines are intended to prevent. As aresult, there is no longer the sense that safety risksshould be balanced against benefits. While from apublic health perspective this may be unfortunatefor the U.S. and Europe, it is a much more serious

problem for developing countries, which might set thisbalance differently but nd it technically and politicallychallenging to set a different safety standard than isused in the U.S. and Europe.

The most well-known case of regulatory stringency isthat of Wyeths rotavirus vaccine RotaShield, whichwas approved in the U.S. in 1998. The vaccine wastaken off the market one year later after 15 childrendeveloped intussusception, a serious side effect. As aresult, developing countries, where rotavirus infectionis estimated to kill more than half a million childrenevery year, were also unwilling to use RotaShield. Arecent study by the CDC showed that the number oflives saved in high-burden countries by a vaccine likeRotaShield would greatly outnumber new cases offatal intussusception.38

But the cost of excessive regulatory stringency is notonly that a particular vaccine will not come to market

but also that all vaccines become more expensive.The developers of the next generation of rotavirusvaccines had to conduct trials of unprecedented sizein order to rule out this very rare side effect (GSK dida trial involving 61,000 children), and developers ofall vaccines are aware that side effects too rare to bedetected in trials of ordinary size could result in theirvaccines being taken off the market.

3) As with drugs, there is no completely satisfactoryregulatory pathway for new vaccines that have nomarket in the U.S. or Europe. Although the WHO

prequalication programme works well for follow-onvaccines, most developing country regulators still lackthe capacity to rigorously evaluate wholly new vaccines.One option is the EMEAs new Article 58 procedure,under which the agency will offer an opinion on avaccine that will not be marketed in Europe. This isthe route that GSK and the Malaria Vaccine Initiative(MVI) will take with the RTS,S malaria vaccine. The FDAhas announced a similar programme. In the long run,the best solution is to build the capacity of regulatoryagencies in the countries where new vaccines willbe used, and in countries such as India, Brazil, andChina that can produce and increasingly develop

new vaccines for use in developing countries. Onepromising route to this end is regional regulatoryharmonization or even consolidation. An example isthe use of the South African Medicines Control Councilas a regional regulatory reference in southern Africa.


19/31


R&D costsThe costs of each of the stages of vaccine developmentvary greatly, accordingly to the scientific difficulty,technology, trial sizes, and the type of productdeveloper. Table 2 gives ranges for each stage.

Although actual expenditure on a particular licensedvaccine can be determined ex post by adding up thecosts of each stage of research and development,such an approach underestimates the true cost ofvaccine R&D (and the likely cost of developing a newvaccine) because it leaves out the cost of failure:expenditures on vaccine candidates that didnt makeit to market. Taking the risk of failure at each stageinto account gives the following formula for average

total expenditure on clinical development per licensedproduct:C=(C1 + C2*P1 + C3*P1P2)/(P1*P2*P3), where C1, C2,C3 and P1, P2 and P3 are the costs and probabilitiesof success of the different phases of clinical trials.The formula is easily extended to include discovery,preclinical and licensure phases.

Using this approach and the probabilities and costranges presented in Table 2 gives an estimate oftotal risk-adjusted R&D cost of $135-$350 million,not including basic research or the cost of building a

manufacturing plant.

These estimates are at best gross averages. Verychallenging vaccines such as malaria and TB (not tomention HIV) can be expected to cost more, whileeasy vaccines or those that are substantially basedon already licensed vaccines, will cost much less todevelop, as trials may be smaller and chances ofsuccess at each stage much higher. Discovery costs areparticularly variable, as this phase can be quick for newversions of existing vaccines but essentially open-endedfor research into entirely new classes of vaccines.

A final issue in considering the cost of R&D is thecost of capital. Private sector developers, who mustraise money in capital markets to finance vaccinedevelopment, take into account the cost of this money

(essentially an interest rate) in calculating total R&Dcosts. Governments and foundations do not bearthese expenses in the same way, although a rigorousaccounting of the true public sector cost should atleast include the cost of public borrowing.

Alternative models for R&D

The current profit-driven system of pharmaceuticalR&D has worked fairly well to deliver new, albeitincreasingly expensive vaccines with large markets inthe developed world. But the system does not work togenerate vaccines for diseases that affect predominantlylow- and middle-income countries, whose markets do

not offer sufficient revenues to motivate industryinvestment in R&D on the necessary scale. Thisproblem is well recognized; a variety of solutions havebeen proposed and some are being tried. Even in thecase of diseases that affect both high-income anddeveloping countries, such as pneumococcal diseaseor HPV, the current system has important deciencies.Because vaccines against these diseases are developedprimarily for rich-world markets, they may not be wellsuited to low- and middle-income countries, where thedistribution of disease serotypes may be different anddifferent presentations may be needed.41 Perhaps most

importantly, the new vaccines may be very expensiveto manufacture, even though an R&D strategy thatconsidered cost from the start might have led to farless expensive, but still useful, vaccines.

Ideally, then, alternative models of vaccine R&D wouldresult in new vaccines against neglected diseases ofdeveloping countries as well as versions of existingvaccines more suited to developing countries needsand circumstances. Some proposed approachesare designed to fill specific gaps in the currentpharmaceutical R&D system, while others promise amore sweeping transformation that would align R&D

investment with public health need and eliminatemany of the distortions of the current system. Thissection provides a brief overview of some of thesemechanisms.

Stage Discovery40 Phases 1 & 2 Phase 3 Licensure Total

& PreclinicalCost 5 15 4 - 10 50 - 120 2 - 3 60 - 145

Chance of success 40% 33% 75% N/A 10%

Risk-adjusted cost 135-350

Table 2: Estimated costs of research and development (U.S.$ millions)39


20/31


Most solutions to the lack of neglected disease R&Dinvolve public or philanthropic funding to compensatefor the lack of commercial investment, but thisfunding can come in a variety of forms. Direct grantsto universities, public researchers, or rms to carryout R&D are often called push funding: funding ofthis type pays directly for R&D or reduces the costsor risks to commercial product developers. Pullmechanisms, in contrast, seek to increase the rewardto successful development of a new drug or vaccinein the hope of motivating developers, typically privatesector firms, to invest their own resources in R&D.(Rich-world markets can themselves be considered apowerful pull mechanism, especially when the rewardto innovators is increased by the temporary market

exclusivity conferred by patent protection.) While pushfunding remains the mainstay of neglected diseaseR&D funding, several innovative pull mechanismshave been implemented or proposed in recent years.The following sections will consider several promisingnancing models.

While funding is essential to any strategy for neglecteddisease R&D, access to technology is also crucial. Someideas in this area are outlined in the last section.

Push fundingIn addition to public funding of university research,the leading approach by which governments andfoundations currently try to drive development ofvaccines for the developing world is the productdevelopment partnership or PDP. Vaccine PDPs,which include the Malaria Vaccine Initiative, the AerasGlobal TB Vaccine Foundation, and the InternationalAIDS Vaccine Initiative, channel funding from donorgovernments and foundations, especially the GatesFoundation, toward a portfolio of discovery andclinical development partnerships with universities,biotechs, and pharma companies. PDPs are popular

with donors and with industry, but it is too early toknow if they will prove successful in the long run: nolicensed vaccines have yet resulted from their efforts.The Malaria Vaccine Initiative can serve as an exampleof a vaccine PDP.

The Malaria Vaccine Initiative (MVI)MVI, established by PATH with funding from the GatesFoundation, seeks to accelerate the development ofmalaria vaccines and ensure their availability andaccessibility. MVI manages a series of partnershipswith universities, biotechs, the U.S. military, and big

pharma. Its most clinically advanced candidate isRTS,S, a product originally developed by the U.S.military in the 1980s and now owned by GSK. Afterseveral phase 2b trials, it is now being tested in alarge phase 3 trial in several African countries. MVI

is sharing the costs of these trials with GSK, usinga grant of just under $200 million to MVI from theGates Foundation. In exchange, GSK has agreed to anundisclosed set of volume-dependent price ceilings.It is difcult to know if these prices are substantiallydifferent from the prices GSK would have asked if ithad developed the vaccine without help from MVI.

Trials conducted to date suggest that RTS,S will beat best an imperfect vaccine, with an efficacy ofapproximately 50% against clinical disease and relativeshort duration of protection. It could nonethelesshave a signicant impact, given the enormous toll ofmalaria in sub-Saharan Africa.

MVIs role in the development of RTS,S has beenprimarily to subsidize clinical development by GSK.Although it is impossible to know for sure what paththe vaccine would have taken without this subsidy, it islikely that development would have been abandonedor greatly slowed: GSK had apparently terminated workon RTS,S until MVI began to share costs.42 Malariavaccines in general have at best modest rich-worldmarkets among travelers and the military; RTS,S willhave no market at all outside Africa. GSK will derivean important indirect commercial benet from testingRTS,S, however, because the vaccine includes a new

adjuvant that could be used in other vaccines with bigmarkets.

MVI became involved with RTS,S late in the game andplayed little or no role in the design of the candidateor in supplying needed technologies. It was thus notin a position to choose a development partner or toinuence production costs. But given the difculty ofdeveloping a malaria vaccine a challenge that hasdefeated researchers for decades working with GSKon RTS,S was the only option available to MVI likelyto result in a licensed vaccine in the next decade. MVIis working with a number of partners on other R&D

projects, but all of these are at quite early stages andnone are assured of success.

The Meningitis Vaccine Project (MVP)The Meningitis Vaccine Project (MVP) is an intriguingmodel of vaccine development for developing countries,in which a vaccine with specic characteristics tailoredto a particular population is developed at a modestcost and provisions to ensure sustainable access arebuilt in from the start. Although similar to the largervaccine PDPs, its approach has differed in severalways.

The MVP was established in 2001 by PATH and WHOto develop an effective and affordable vaccine tocombat the epidemics of bacterial meningitis, causedmostly by group A strains ofNeisseriameningitidis


21/31


(meningococcus), which plague the so-calledmeningitis belt in sub-Saharan Africa. 43 Whilepolysaccharide vaccines have been available for manyyears, they have limited duration of protection andare not effective in very young children.44 So-calledconjugate vaccines, like the one produced by Sano,in theory overcome these disadvantages. But theSano conjugate vaccine, an ACYW-strain tetravalent,is more expensive to produce (four conjugate vaccinesinstead of just one). Even at a heavily discountedprice, it would likely remain far more expensive than amonovalent vaccine.

After identifying and licensing an appropriateconjugation technology from the U.S. FDA and

negotiating with several possible industrial partners,the MVP reached agreement with Serum Instituteof India (SIIL) to produce the new monovalentmeningitis A conjugate vaccine at an affordable pricefor the African market. In exchange for price andsupply commitments, SIIL beneted from transfer oftechnology and know-how. PATH funded the clinicaltrials.

Significant progress has been made. In December2009 market authorization for the Men A conjugatevaccine was granted by the Drugs Controller General

of India and the WHO prequalification process hasbegun. Plans are being made for introduction at publichealth scale in Burkina Faso in late 2010.

The total research and development costs for thisvaccine are estimated to be about $60 million, notincluding the cost of the manufacturing plant; SIIL hascommitted about $15 million to the project.

The MVPs approach differs in some respects from thatof the larger PDPs: Focus on low cost: Consultations at the start of the

project established that to ensure long-term access

a vaccine should cost no more than $0.50/dose.MVP emphasized this requirement in discussionswith vaccine manufacturers. The agreement betweenSIIL and MVP is also notable for its transparency:although all PDPs include access provisions in theiragreements with industrial partners, these provisionsare generally not made public.

Focus on a single candidate and single supplier:Unlike most product development partnershipsthat maintain a portfolio of vaccine candidates,MVP focused on a single candidate and on SIIL asthe sole supplier. The relatively high probability of

success in making a monovalent conjugate vaccinelowered the risk when compared to a malaria or TBvaccine, where the risks of R&D failure are high.

Partnership with an emerging supplier: The choiceof an emerging supplier was critical to achieving

the goal of low cost, in part because the access totechnology and visibility associated with the projectmade the partnership more attractive to SIIL thanto a multinational rm. In addition, producing for ahigh-volume, low-margin developing country marketwas consistent with SIILs established businessmodel.

Public sector technology transfer: The transfer ofa non-exclusive license for an efcient conjugationtechnology through the NIH Office of TechnologyTransfer, coupled with active transfer of know-how,was central to the success of the project.

To a great extent, these choices were made possiblebecause several meningococcal conjugate vaccines

had already been developed, the technologies werewell understood, and correlates of protection forMeningitis A were known. These data made focusingon one candidate a reasonable bet, allowed emergingsuppliers to be considered as manufacturing partners,and kept costs relatively low. The MVP model istherefore particularly suited to the development ofadapted versions of vaccines based on establishedtechnologies, such as for rotavirus, pneumococcal orHPV vaccines. On the other hand, the development ofAIDS, TB and malaria vaccines is far more challengingand probably requires different models.

Other forms of push funding for vaccine developmentare under discussion, including new funds that wouldpool donor resources and allocate them among PDPsand possibly rms, proposals to use bond markets toraise long-term funding for PDPs, as well as currencytransaction levies or nancial transaction taxes.

An alternative to the PDP model of publicly subsidizedcollaboration with the private sector would be torebuild capacity in the public sector to develop andtest vaccines. The public sector, notably the U.S.Walter Reed Army Institute of Research, played the

leading role in the development of several importantvaccines, including vaccines against inuenza, mumps,and meningitis, especially in the decades followingWorld War II.45

Pull fundingA simple example of pull funding is the existenceof an organisation like GAVI, which, by obtainingcommitments of several billion dollars from donors,served to signal to industry that the poorest countriescould be a viable market. However, the existence of this

subsidised but still low-margin market is not sufcientto drive the development of new and sophisticatedvaccines, although it has almost certainly acceleratedthe introduction of follow-on vaccines from emergingsuppliers.


22/31


Advance Market CommitmentsThe pilot AMC for pneumococcal vaccines, discussedabove, is primarily a procurement mechanism. However,AMCs were originally proposed as a way to stimulatethe development of new vaccines for developingcountries by creating an articial market sufcient tomotivate private investment in R&D. If a second AMCis created and discussions among GAVI, the WorldBank, and certain donors have begun it is likely thatit would seek to test the value of this mechanism forearlier-stage vaccines. The daunting scientic obstaclesfacing HIV vaccines make them an unappealing choice;this leaves malaria and tuberculosis vaccines asthe front-runners, although other diseases are alsopossible (see below).46 In fact, the original expert

committee that chose pneumococcal vaccines for thepilot suggested that malaria should be chosen if asecond AMC were funded.

At least in theory, AMCs are an elegant solution tothe problem of lack of private-sector investment indeveloping-country vaccines without large markets.But an early-stage AMC would face several seriouschallenges that did not have to be considered indesigning the pneumococcal AMC. First, the terms ofan early-stage AMC, especially its size, are difcult toset, since costs, risks, and potential returns cannot

be estimated with any precision, and neither theidentities nor the number of likely competitors isknown. Second, it is not clear that the university labsor biotech companies whose participation in R&D ismost needed at early stages can be easily reachedby the distant pull of an AMC. Third, an AMC, as aninherently competitive mechanism might interfere withthe networks of collaboration and information-sharingthat have been built by PDPs and other neglecteddisease R&D funders. Finally, and perhaps mostimportantly, an AMC of this kind would almost certainlyhave to be very large to attract substantial investment potentially so large as to be politically infeasible.

The very high discount rates used by industry, coupledwith the long time to market and the very high risks,mean that only the prospect of very large returnscould drive decisions to invest on purely commercialgrounds.47

Given these difculties, AMCs by themselves are probablynot a practical way to drive R&D for challenging early-stage vaccines or those facing substantial scientificobstacles. They could have a role, however, as acomplement to public sector research funding, PDPs,and other push mechanisms. Some assurance that

donors would pay a respectable price, at least enoughto cover the costs of plant and manufacture andphase 3 trials might help to ease the hand-over frompublic sector researchers or PDPs to a rm capableof bringing a new vaccine to market, once the major

obstacles had been overcome. There may be a rolefor well-designed AMCs in mid-stage or less difcultvaccines, those that have demonstrated some promisein preliminary efcacy trials or for which similarity toexisting vaccines offers a likely path to success.

FDA Priority Review VouchersAccording to this new U.S. programme, launchedin 2008, any organisation that wins FDA approvalfor a new drug or vaccine against a defined list ofneglected diseases is eligible for a priority reviewvoucher (PRV) entitling the holder to expedited FDAreview of another new drug application. The voucheris transferrable: it can be sold to and used by anotherorganisation. Since priority review allows a new

product to be marketed sooner and thus extendsthe period of market exclusivitythe voucher couldsubstantially increase returns from a blockbuster drug:one study estimated that it could be worth as muchas $300 million.48 In theory, then, the prospect ofwinning a PRV could be a powerful incentive for rmsto pursue neglected disease R&D. Like an AMC or aprize, the PRV is a pull mechanism focused primarilyon motivating neglected disease R&D by the for-protpharmaceutical industry. One PRV has been awardedso far, for Novartis malaria drug combination Coartem.Since Coartem has been on the market for several

years outside the U.S., in this case the PRV did notreward new neglected disease R&D.

It is too early to say whether PRVs will prove aneffective incentive for neglected disease vaccinedevelopment. An important flaw in the legislationis that it includes no access provisions: firms arenot required to manufacture or distribute a winningproduct, let alone make it available in developingcountries at an affordable price, or, as an alternative,to permit generic supply.

Prize fundsThe current pharmaceutical R&D system dependsprimarily on high prices in rich-world markets,buttressed by patent-protected market exclusivity, tocover R&D costs. Although this system has provideda strong incentive for the development of someimportant vaccines, it distorts both R&D priorities(toward products with large markets) and prices (whichare kept well above marginal cost of production).Some advocates have proposed a system of largeprizes for new drugs and vaccines as a way to solveboth problems. Incentives would be aligned withsocial benet by making the amount of the payoffs

proportional to public health benefit, while access(and economic efficiency) would be maximized byfacilitating or requiring prices close to marginal cost.Prizes are therefore a pull mechanism that allowsthe cost of R&D to be uncoupled from the price of


23/31


the product. The prizes would come from a large andregularly replenished public fund. These ideas wererst developed by Knowledge Ecology International(KEI);49 a more recent proposal called the HealthImpact Fund is derived from the KEI proposal butdiffers from it in several ways.50

If prize funds could be implemented and managedsuccessfully, they would offer comprehensive reformof the pharmaceutical R&D system, aligning R&Dinvestment with public health need and drivingdown prices of new medicines. These ideas faceformidable political, nancial, and practical challengesto implementation on a large scale. They are socompelling, however, that they should be tested on a

smaller scale.

Prize mechanisms that rely on open licensing andgeneric competition to ensure low prices face additionalchallenges, especially for vaccines. First, removingpatent barriers will not in many cases be sufcient toallow new manufacturers to produce versions of newvaccinesa mechanism for transferring know-how aswell as an abbreviated regulatory pathway would alsobe needed. Second, markets for some new vaccinesmight be too small to attract multiple suppliers even ifa patent pool and transfer of know-how substantially

reduced R&D costs for follow-on manufacturers. Ifopen licensing and competitive generic supply are notfeasible, a prize would need to include mechanismsto ensure low prices and supply, such as assurancesfrom manufacturers, backed by nancial guarantees,that the product would be manufactured in sufcientquantities and provided at an affordable cost.

Access to technologyFunding for R&D, either through grants or pullmechanisms like AMCs, or prizes, is necessary tomake up for the lack of market returns, especially forchallenging vaccines needed primarily in low-income

countries. But greater access to technology wouldgreatly expedite the entry of new suppliers intomarkets for existing kinds of vaccines and increase thecapacity of rms in developing countries to contributeto the development of new classes of vaccines.The te

Documents

Giving Developing Countries the Best Shot: An overview of vaccine access and R