Stop TB Working Group on New Vaccines: Task Force on Economics and Product Profiles Discussion Points London School of

2. OVERVIEW SLIDES

3. REVIEW OF EVIDENCE ON EPIDEMIOLOGY, SOCIOECONOMIC IMPACT OF TB AND COST-EFFECTIVENESS OF NEW TB VACCINES Introduction Measurement of progress against targets set The complexities of TB � latent infection, HIV and MDR- and XDR-TB Existing vaccination strategies � the BCG vaccine New vaccination strategies and the health impact of new TB vaccines Drugs, vaccines and diagnostics working together Cost-effectiveness of TB control and vaccination Limitations of cost-effectiveness analysis Socioeconomic analysis and cultural issues Summary of implications for policy making A very selective bibliography

4. BASIC MARKET AND REVENUE FRAMEWORKS Some introductory thinking Comparing market and revenue frameworks Replacement vaccine Boost vaccine Total portfolios

5. RATE OF ADOPTION AND MECHANISMS OF UPTAKE Introduction Hep B as a model for adoption? EPI levels Efficacy and coverage

6. PRICING STRUCTURES, TIERED PRICING, WILLINGNESS TO PAY The principles of tiered/Ramsey vaccine pricing Consultancy model pricing structures Incentives for rich market and �high risk� market uptake, and implications for tiered pricing strategies Scenarios for non-rich markets and implications for tiered pricing strategies Vaccine catch-up �Willingness to pay�

7. POLICY PROCESS METHODOLOGY TO IMPROVE ACCURACY OF PRICING AND DEMAND FIGURES AND SPEED OF VACCINE UPTAKE  Introduction Aim Consultancy models Vaccination and health policy literature Political science and public policy literature Next steps Vaccination and health policy literature for future review Political science and public policy literature for future review

8. COST ISSUES Cost to develop a new vaccine Attrition rates / probabilities of success Portfolio reasoning of �at least one� success Applied Strategies portfolio analysis Commentary: Attrition rates and costs of development Other cost of development issues including biomarkers, trial site limitations, time to licensure COGS and LDC sales brought forward Booster COGS Some sensitivity analysis Costs caused by plant size and capacity issues

9. TB VACCINE AMC OR SIMILAR INITIATIVES Relative role of such mechanisms as: Prize Product subsidy (including how best to structure it) Form of production capacity insurance Role of competition Other incentive issues Lessons from other cases (see case studies)

10. LESSONS FOR TB FROM A SELECTION OF OTHER VACCINES Hepatitis B Pneumococcal Human papillomavirus (HPV) Rotavirus Haemophilus influenzae type B (Hib)

11. TOPIC-BY-TOPIC SLIDES

12. Distribution of tuberculosis in the world in 2003

13. High incidence of active TB High incidence of active tuberculosis: High level of latent TB infection (around one-third of the world�s population) Relatively recent problems of coinfection with HIV/AIDS Increasing multi-drug resistance to treatment 9.2 million new cases of TB in 2007 1.6 million deaths (137 cases/100,000 population, 25 deaths per 100,000 population) 31% of which in the African Region 55% of which occurred in South-East Asia and the Pacific *Global tuberculosis control: surveillance, planning, financing: WHO report 2008. (WHO/HTM/TB/2008.393).

14. Following the peaking of epidemics in Africa and Europe, TB case rates are stable or falling

15. Trajectories of tuberculosis epidemic for nine epidemiologically different regions of the world

16. Current Targets Targets centre around expanding Directly Observed Treatment Strategy (DOTS) and on achieving reductions in TB incidence, prevalence and death rates The Millennium Development Goals set out the impact target of halting and reversing TB incidence by 2015 WHO�s Stop TB Strategy (launched in 2006) has set further targets: By 2005 - At least 70% of people with sputum smear-positive TB will be diagnosed (i.e. under the DOTS strategy), and at least 85% cured. These are targets set by the World Health Assembly of WHO (first set in 1991) By 2015 - The global burden of TB (per capita prevalence and death rates) will be reduced by 50% relative to 1990 levels By 2050 - The global incidence of active TB will be less than 1 case per million population per year The principal WHO measure of case detection is the rate of case detection for new smear-positive cases in DOTS programmes

17. Progress National estimates suggest 77 countries met the 70% target for case detection rates by the end of 2006* Although some countries have met the target, progress has been much slower in the African, Eastern Mediterranean and European regions Nevertheless, the fall in incidence has been slow and projected incidence will still be 100 times the elimination threshold in 2050 Observe that as an epidemiological measure of success, incidence usually changes more slowly than prevalence or deaths in response to control efforts *Global tuberculosis control : surveillance, planning, financing : WHO report 2008

18. Projected incidence/million/year to 2050

19. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Surveillance and data: Sound cost-effectiveness analysis needs sound epidemiological analysis Poor-quality data and potential bias in incidence estimates (for example, case detection uses total number of smear positive cases as the denominator) Not controlled experiments - difficult to do inference See Table 1.1, p11 of Part 1 of Scoping Study Discussion Points: To what degree does the Task Force take such surveillance problems as given and work analysis according to these constraints? To what degree does the Task Force engage in discussions about the surveillance systems and the data generated from such systems with an eye to improving data available to its tasks? What are the advantages and disadvantages of various approaches to measuring tuberculosis incidence, prevalence and mortality? Data �ownership� issues

20. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Latency: The latent nature of TB means that it will take a long time to achieve a low rate of infection Latent stage TB infection (LTBI) can be detected via a tuberculosis skin test (TST), and thence treated to decrease the chance of progression to active TB. However, the interpretation of TST can be complicated if a person has had a vaccine Discussion points: What are the full complications and implications of the latent nature of TB for vaccine markets? How does this affect the decision of countries to use vaccines as part of a control strategy and hence cost effectiveness analysis of such strategies? For example, in countries where exposure to active TB is rare, treating latent TB is more likely to be the strategy of choice How might this incentive change over time for countries that start off using vaccines?

21. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Cost effectiveness methodologies � quantity and distribution: See p22-31 of Part 1 of Scoping Study Used to determine the relative cost of an intervention, such as vaccination, and the outcome or effect compared to no intervention (or another alternative) In global health, cost of a health intervention per unit of outcome e.g. healthy life year averted (usually presented in Disability or Quality Adjusted Life Years - DALYs or QALYs) Cost-effectiveness used to identify of knowledge gaps and the need for further research Few cost effectiveness analyses of control strategies for TB, let alone of TB vaccines

22. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Floyd (2003), 1982-2002, 66 studies, 28% were in high-burden countries Few related to vaccination strategies Expensive, and time-to-build long Evidence from other vaccine launches: especially useful to design such studies around vaccine trial sites on the basis of a targeted early launch strategy that creates momentum for success Baltussen* et al: Most economic studies of tuberculosis in high burden countries do not assess the impact of interventions on transmission Most use indicators of effectiveness that are specific to tuberculosis control, rather than generic measures This can prevent the results on cost-effectiveness of tuberculosis control being compared with that of interventions for other diseases This limits the usefulness of these studies in decision-making at the policy level *Baltussen R, K Floyd and C Dye �Cost effectiveness analysis of strategies for tuberculosis control in developing countries� British Medical Journal (2005) doi:10.1136/bmj.38645.660093.68 (published 10 November 2005)

23. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Potential benefits to different institutions of an improved vaccine or vaccines:

24. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Discussion points: See p24-p27 of scoping study for some results Update? Take the level and range of cost effectiveness studies as given or expand the level and range of such studies? Current distribution of such studies? Priorities for the future distribution if such studies are to be used as part of an evidence base for vaccine launch? Priorities at different horizons? Optimal combination of cost effectiveness studies by region or country? Multivariate cost effectiveness sensitivity analysis of authors such as Zif et al?

25. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Breadth of coverage of measures and Common protocols: See p 28 for limitations and issues Goal to make studies comparable Often studies are found at a later date to be non-comparable or not as comparable as they could have been with a bit more prior thought and coordination Weakens the ability to use such analysis for advocacy and to influence policy decision process Most studies use indicators of effectiveness specific to tuberculosis control, rather than generic measures, making it difficult to compare with the cost effectiveness of interventions for other diseases Further limits ability to use such evidence in the policy-making process Discussion point: Should the Task FOrce be more involved in standardizing the protocols of such studies so at to make results more comparable (e.g. hospitalization costs, productivity impacts, treatment of exchange rates, appropriate sensitivity analysis, end points, etc.)?

26. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines

27. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Discussion point: How can DOTS expansion complement the launch of a new vaccine or vaccines and vice versa? How to use DOTS to identify gaps? What are the implications at different horizons for infrastructure and costs? How should analysis be adapted to take account of new drugs being developed?* Are the intermediate goals and metrics of success different for different interventions, even if the final goals are the same? What are full benefits and limitations of integration? *In many of the analyses, the public markets for vaccines in poor countries are valued relative to costs (and efficacy) of non-vaccine options, and this is usually presumed to be the current state of play vis a vis DOTS. There is a presumption underlying some investment case figures (those of BVGH/BCGH) that there will be no new drugs between now and 2030

28. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Pre- and postexposure vaccine strategies: Debate about the best strategy to improve immunisation against TB (pre- or postexposure, combination of vaccines and boost strategies) Ziv et al, effectiveness of a pre- or postexposure vaccination strategy in developing countries with high incidence and prevalence of disease At first, postexposure vaccines have a greater impact in reducing the number of cases of disease This effectiveness declines over time (even with continuous vaccination campaigns) effectiveness of preexposure vaccines rises Pre vaccination ultimately nearly twice as effective in reducing the number of new infections Over a 20-30 year horizon cases prevented under post- or preexposure vaccination campaigns roughly equal Either way a widely deployed and highly effective vaccine, whether pre- or postexposure, the number of cases of TB in high-incidence countries is likely to remain high Need for several vaccines or combination * Ziv et al (2004), extending the work Lietman T, Blower SM. The potential impact of tuberculosis vaccines as epidemic control agents. Clin Infect Dis. 2000;30:S316�22. Lietman T, Blower SM. Tuberculosis vaccines. Science.1999;286:1300�1

29. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines 70%-80% reduced infection rate does not translate into 70%-80% reduced disease rate Incidence of TB disease is driven by: Susceptible persons who become infected and quickly progress to disease Preexposure vaccines, given to uninfected individual, to reduce this route, but has little impact on the second route Latently infected persons who slowly progress to disease often with a long time lag. but with little impact on the second Postexposure vaccines, given to latently infected persons, act mainly on the second route with little impact on the first

30. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines

31. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines

32. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines

33. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines

34. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Discussion points: See p19-p20 of report Current consensus regarding pre- and postexposure vaccines? How important in determining product profile, market, etc., is the mechanism by which a preexposure vaccine works? Investment case literature: Are there nuances that that literature misses and ways that that literature could be improved by a more sophisticated approach to choice between vaccine strategies?

35. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Multi-drug resistance, discussion points: How important is multi-drug resistance as a driving force for uptake of vaccines? How is multi-drug resistance captured in models involving combinations of prevention and treatment measures? HIV: It is unclear what the effect of co-infection with HIV will be on TB vaccine effectiveness About 40% adults with TB (15-49 age) are co-infected with HIV Discussion point: Does the TF need to know more about the long-term role of the HIV epidemic or does it take the work of others as given? Impact on: Incidence of TB vaccine effectiveness cost effectiveness in developing countries control programs Implications for product profiles and market sizing?

36. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Duration of protection of new vaccines: Product profiles in willingness to pay questionnaires usually specify ten years Discussion points: Significance of finding that duration of immunity has a much greater impact on cost effectiveness if it is greater than ten years? Implications for cost-effectiveness analysis of ten versus twenty versus life-long protection (delivery costs, long-term consequences for epidemiology of TB, etc.)? Transmission dynamics: Ideally estimates of averted DALYs should incorporate any transmission effects of the disease How disease incidence may change in the future A more substantiated claim on the effectiveness of a new intervention Discussion point: State of such analysis in the case of TB?

37. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Cohort studies: Discussion point: What is the situation regarding cohort studies for informing disease burden and existing vaccination and treatment coverage and impact?

38. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Other effects of a vaccine � discussion points: When assessing a new vaccine candidate, how important is the interrelation with BCG and both direct and indirect effects of the new vaccine candidate? How significant are non-specific survival benefits from vaccination? Active pulmonary TB: Active pulmonary TB is the leading cause of death in HIV/AIDS patients in the developing world and has spiraled in the last decade or so Discussion points: What is the state of play of modeling and evidence on this? No modeling relating to this in the investment case analyses

39. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Socio-cultural limitations: See p28-30 for issues There are many socio-cultural factors associated with TB, which can have severe implications for accessibility and acceptability of treatment and prevention strategies E.g. care-seeking strategies of the poor E.g. Chinese cost recovery systems E.g. social stigma Such factors have a direct implication for the validity of any assumptions about coverage made in a standard cost-effectiveness analysis Other costs include early exit from school by children on account of their parents� TB, the impact of social stigma, etc. Discussion point: To what degree does the Task Force regard these as an area the TF should investigate, or should this be left to others?

40. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Health systems: Concern about DOTS Wide variation in the selection and implementation of control strategies within and among countries DOTS programmes have mostly recruited patients who would have been detected and treated anyway in the public health systems DOTS has failed in some countries to reach deeply into the private sector, and in others to provide access to patients living in areas with inadequate health services Rates of improvement limited by quality of health system Bloom et al, reaching populations gets harder and harder after the first wave of a new intervention Discussion point: If vaccines are to optimally �integrate�, what are the implications for modeling vaccination strategies and cost effectiveness of new TB vaccines � especially boost and prime boost � of the health systems issues?

41. Epidemiology, socioeconomic impact of TB and cost-effectiveness of new TB vaccines Discussion points: What are realistic success metrics for vaccines over such horizons presuming a vaccine is launched, and presuming that vaccines are combined with other interventions? How do such horizons impact: investment decisions product profile decisions the nature of follow-on generations of vaccines sponsor funding, etc.? Epidemiologically-driven targets: TB targets, as encapsulated in the MDGs, are not driven by an epidemiological framework Discussion Points: What would epidemiologically-driven TB targets look like, and what are the implications for vaccines? How can the Task Froce influence an improved epidemiological basis for such targets?

42. Basic market and revenue frameworks t = the time of the cash flow n = the total time of the project r = the discount rate Ct = the net cash flow (the amount of cash) at time t. C0 = the capital outlay at the beginning of the investment time (t = 0) So long as NPV is positive, risk neutral* investors should invest *Perhaps because they can diversify such as to be in effect risk-neutral on this investment. The above is a little too simplistic. If there is any potential variability of input values, uncertainty, even disagreement, the NPV derived will vary � potentially very considerably

43. Basic market and revenue frameworks Current market and investment case analyses See p35-36 for basic comparisons See p36-p42 for a case See p44 on for boost and global XDR See p48-p50 for total portfolio figures Discussion points: How satisfactory are the global market and investment case analyses so far created? What is the balance between the need to generate absolute market numbers (the focus of current market and investment case analysis) and the need to improve the functioning of the market for vaccines, and more effective launch strategies? Strategic thinking discussion point: Because of the concern to get numbers, much of the market analysis is remarkably devoid of strategic thinking. For example, over simple issues like: the levers that can be used to improve adoption of new technology; the relationship between efficacy and ease of uptake; catch-up strategies. Would it be too difficult to have investment case reports adopting a more strategic approach?

44. Basic market and revenue frameworks Public and private sectors: The treatments of the division between public and private sectors in the various market/investment cases so far have been relatively rudimentary Discussion points: How is the division between public and private sectors best modelled? Should this division be based on country-level data pertaining to other vaccines or past vaccine launches? Does it depend on price? Is it useful having more granular details about this, or will the effort just complicate the process of deriving a market figure while adding little value? Conversely, is knowing the division useful for knowing how to adapt launch strategies?

45. Basic market and revenue frameworks Global MDR-TB and XDR-TB: Analysis of impact of global MDR-TB and XDR-TB seems rudimentary in much market analysis In theory governments and individuals should be prepared to pay to avoid MDR-TB and, even more so, XDR-TB because of the high costs of treatment One of the market analyses (Applied Strategies) relies heavily on global XDR-TB to generate large revenue figures and NPVs from booster vaccine take-up (3X the NPV without) Not clear how global XDR-TB feeds through to dictate prices and country decision-making processes Given such heavy weight on global XDR-TB to drive large NPVs in some analysis, there seems little epidemiological and cost-effectiveness modelling of the impact of it

46. Basic market and revenue frameworks Discussion points: What is the scientific basis of any probability of MDR-TB/XDR-TB How should it be used in market analysis? Why is Global XDR-TB modelled as having little impact on NPV of replacement vaccines, though it does for booster vaccines? It seems that Global XDR-TB drives the biggest differences in levels of tiered pricing � but on what basis?

47. Basic market and revenue frameworks High-income high-risk groups: We find repeatedly that the role of the high-income country and high-income high-risk groups still needs clarifying Potentially, such markets could be important for tiered pricing strategies Middle-income and private sectors discussion points: What is the role of middle-income countries and the private sector in general? Replacement vaccine discussion point: What are the strengths and limitations in presuming that replacement vaccine will replace current BCG vaccine?

48. Basic market and revenue frameworks Cases of large NPV: In some of the boost vaccine and global XDR-TB scenarios, NPVs are very large Discussion points: Is this realistic? What happens if there are other interventions (drugs and diagnostics) �competing� at the same time? If the NPVs are truly that large, why does private investment not flood in and drive NPVs down? If the problem is scientific uncertainty, and if the NPVs truly are that large, if a vaccine candidate gets through to stage III, should we not expect private investors to take it all the way to licensure? Is something missing in this reasoning?

49. Basic market and revenue frameworks Efficacy and duration of protection of BCG vaccine discussion point: Given that efficacy levels presumed in the various investment case materials are defined relative to BCG vaccine (and phrased that way in �willingness to pay� questionnaires) how do we view these in the light of there being no consensus regarding the effectiveness or duration of protection of BCG vaccine, and the natural bias and the role of other factors in assessing those vaccinated and non-vaccinated individuals?

50. Rate of adoption and mechanisms of uptake Hep B is used in one market analysis as a template:

51. Rate of adoption and mechanisms of uptake

52. Rate of adoption and mechanisms of uptake

53. Rate of adoption and mechanisms of uptake

54. Rate of adoption and mechanisms of uptake Discussion points: Are the assumptions regarding rate of adoption accurate in any of the analyses? (see p51-53 for some presumptions about uptake) How might a methodology/system be created to gather and process the requisite information to improve the accuracy of this information? Is Hepatitis B a good template for modelling rate of TB vaccine adoption, in particular of booster and prime-boost vaccines? In some of the analysis there is no connection between efficacy and rate of adoption. What specific role does efficacy play in dictating the shape of the adoption curve?

55. Rate of adoption and mechanisms of uptake China and India: In none of the models is this particularly well defined, yet we know that the emerging and middle income markets in such countries will be important in ten to twenty years time, particularly with respect to global tiered pricing strategies Discussion points: What are sensible assumptions to make about speed of vaccine uptake in China and India? At horizons of ten or twenty years? What are the sources of information? What are the policy processes in such countries?

56. Rate of adoption and mechanisms of uptake Catch-up strategies: Catch-up is treated inconsistently across the investment case literature At best there is (apparently) a limited catch-up phase modeled in many of these analyses Catch-up is a way of generating lots of early sales and hence high NPV for every sale made Price is also usually modelled as higher earlier, before competitive events lock in, further reinforcing the impact on NPV of early sales Therefore catch-up strategies are likely to heavily impact the value of R&D investments Discussion points: What are likely catch-up strategies, particularly of booster vaccines? How will catch-up strategies likely vary across countries? How can this be worked into the policy-making literature?

57. Rate of adoption and mechanisms of uptake Efficacies in durability in product profiles: The product efficacies in the investment cases so far are fairly unsophisticated. It is not clear how they are derived as �targets� (see p56) The investment case literature regards this as still to be formulated There is no proper �sensitivity� analysis done of this in the investment case analyses Investment analysis has little to say about impact of durability of response on price and demand Discussion points: What is the mechanism for integrating discussion about such efficacies into analysis of all other parts of the analysis Such as rate of uptake, pricing, and indeed R&D costs if higher R&D expenditure leads to higher expected efficacy)? How sensitive are all the demand figures to percent efficacy? Including, e.g., willingness to expand delivery mechanisms, especially amongst adult populations Willingness to change organizational practice)? What is the impact of durability of response on price and demand? How do we capture the impact of the durability of response which is not captured in the efficacy measure? What is the evidence for the impact of rBCG vs prime-booster from a durability point of view?

58. Rate of adoption and mechanisms of uptake EPI improvement initiatives, discussion points: Can many recent ongoing initiatives to improve EPI performance (BMGF, WHO, PATH all have recent initiatives to improve EPI performance) be integrated into thinking about TB vaccine launch strategies, especially of booster vaccines? Delivery systems and organizational change, discussion points: As a function of efficacy� What is the evidence of willingness to expand delivery mechanisms Especially amongst adult populations? To change organizational practice E.g. to provide vaccine to LTC and the homeless in the US?

59. Pricing structures, tiered pricing, and willingness to pay �I cannot believe that anyone seriously believes that America should manufacture vaccines for the world, sell them cheaper in foreign countries, and immunise fewer kids as a percentage of the population than any nation in this hemisphere but Bolivia and Haiti� With apologies to Bill Clinton. This is an oft-cited quote of in the vaccine literature, taken from a press conference on the Childhood Immunisation Initiative in 1993. Public papers of the president of the United States. William J Clinton. Book I. January 20 to July 31, 1993. Washington, DC: Office of the Federal Register, National Archives and Records Administration, 1994. These days the Clinton Foundation plays its part in getting product prices down for the developing world, usually well below rich-world prices. Others point out that this is as much a criticism of US immunization policies at the time as a comment on global tiered pricing.

60. Pricing structures, tiered pricing, and willingness to pay

61. Pricing structures, tiered pricing, and willingness to pay Tiered pricing: Tiered pricing will not work with all vaccine cases More difficult if the product used in rich markets deviates from that used in poor markets or if there is simply no demand in higher-income markets It seems that TB vaccines will more likely be universally applicable? Depends on efficacy considerations and ability to integrate with other control strategies in richer countries Unlike case of pneumococcal, appropriate serotypes covered, etc. Discussion points: A strategy of tiered pricing (and communicating this well in advance)? Combined with emphasis on achieving �affordable� production costs? What is the situation with regard to possible tiered pricing strategies for different TB vaccines? If not the same product in all markets, what effort could be made to ensure that in advance the product is much more similar, such that tiered pricing can be used?

62. Pricing structures, tiered pricing, and willingness to pay Overview of current studies, discussion point: What does the Task Force feel about the pricing structures used in the various TB market analyses? See p60-63 for pricing structures used Discussion points: What is the role of market research in setting prices? Are there useful market research lessons from other products (both vaccine and non-vaccine)? What mechanisms should be put in place to derive pricing information in future, and updating it, and at what horizons? For example, the further from product launch, the more unreliable one would imagine the information would be

63. Pricing structures, tiered pricing, and willingness to pay

64. Pricing structures, tiered pricing, and willingness to pay Low-income market prices: Farlow figures show the power of poorer and middle income markets, even at lowish prices, to boost NPV if uptake is quicker Mainly the issue is sheer quantity of sales and whether COGS can be low enough Even just a $1-$2 mark-up has big impact if the number of doses is large and sold early enough (given the discounting of the value of sales) A $7 per-dose sale in 5 years is worth about $3 now at 15% rate of discounting Discussion points: What are sensible low-income market prices? What about prices for China and India over likely horizons?

65. Pricing structures, tiered pricing, and willingness to pay Rich-world vaccines: Rich-market sales are potentially (since it is not clear yet) key to positive NPV and to tiered pricing strategies But it is not clear how new TB vaccines potentially fit into current prevention and treatment strategies according to efficacy in such high-income (and in the case of countries like the US, high-risk) settings See p64-75 Discussion points: How are high-income countries to be treated in market analysis? How do we find out exactly what policy makers in those countries are likely to do according to product profile? How should the problem they face be better modelled? Under what circumstance would the US and many EU countries that do not currently use BCG vaccine start to use new vaccines?

66. Pricing structures, tiered pricing, and willingness to pay Private market pricing (especially in middle income countries): At the moment the treatment of division of public and private market in low and middle income countries is very rudimentary in investment analysis In some of the analyses, private sector pricing is a small multiple of an already low price (2x, say, 10 cents) Discussion questions: Given the importance put on the private sector market, how should the division between public and private sector, including this division according to country, be accessed? How can the division be made more accurate?

67. Pricing structures, tiered pricing, and willingness to pay Production capacity and tiered pricing, discussion points: Role of production cost decisions in achieving tiered pricing Including technology used in the US (such as the Aeras production facilities), and manufacture in some of the most affected countries? Do other vaccine stories suggest lessons applicable to TB (Hep B, Meningitis, pneumococcal, HPV, etc.)? Questionnaires, discussion points: What sort of questions should questionnaires about pricing contain? Who should such questionnaires be directed to? What is the best timing for using such questionnaires? How should efficacy be communicated in such questionnaires? Competition:, discussion points: What is the impact on �willingness to pay� if there are competing (non-TB) vaccines and other control interventions?

68. Policy process methodology to improve accuracy of pricing and demand figures and speed of vaccine uptake  Booster v. replacement: The various investment case analyses count booster vaccine �in� or �out� but once �in� do not treat them as particularly problematic Discussion points: Are booster vaccines more problematic for policymakers to deal with than replacement vaccines? If so, what are the likely specific challenges? Empirical evidence about decision making: Boston Consulting Group�s and Applied Strategies� models do not have an underlying analytical structure for the decision making component Discussion points: What is the current level of use of models and empirical evidence of decision-making processes in informing market-building and procurement strategies? Is it worthwhile improving this, or too expensive and out of the remit of a Task Force on economics and product profile?

69. Policy process methodology to improve accuracy of pricing and demand figures and speed of vaccine uptake  Frameworks: Section 5 of the scoping study details a range of frameworks from the political science and public policy literature. Discussion point: Does any of this look promising? Mapping vaccine adoption stories, discussion point: Is it worth mapping out previous vaccine adoption stories for selected countries using any of these frameworks and creating country-specific models? Do we stick with the current consultancy models and improve them, or are there other approaches to taking advice? Vaccination and health policy literature, discussion point: Are there any specific comments on any of the vaccination and health policy literature or the political science and public policy literature?

70. Cost issues Portfolio analysis and attrition/success probabilities: Some analysis uses portfolio analysis This can look very sophisticated even if the underlying information is limited or even simply wrong The best way to control for this is by having openness about the figures used in calculations and an ability to do sensitivity analysis For example better stress-testing of probabilities Discussion points: Who will do this portfolio analysis in the future? Should it be a consultancy company under instruction, or some other entity? Are the success probabilities used in investment cases plausible? Do �current industry benchmarks� of success probabilities apply to the case of TB vaccine candidates? How do we make attrition/success probabilities for new vaccines more robust? What if we are being too optimistic?

71. Cost issues Optimal portfolio: The optimal portfolio seems a function of policy makers risk aversion about how high a probability of getting no vaccine they are prepared to tolerate, and the costs needed to achieve particular probabilities of getting �at least one vaccine�. Intuitively: As the pool of vaccine candidates increases in size, at first new additions to the pool have a big and initially growing positive impact on the probability of �at least one success� At some point, the addition of new candidates, while increasing the percent chance of success, does so at an ever-decreasing rate Assuring 95% chance of �at least one success� can be very expensive At that probability the act of raising the chance of �at least one success� by a few more percent is getting ever harder (and more expensive) to do The bigger the portfolio the higher the probabilities of getting variety of products. In this way decisions about portfolio size and shape interact with decisions about Target Product Profile

72. Cost issues Discussion points: How do we think about �optimal TB candidate portfolio� size? How does this interact with thinking about TPP? What is the portfolio formula for working this out? How do we turn portfolio measures into measures of impact? Can we, for example, model a link between a portfolio at a particular point in time to its expected impact, especially for candidates that have reached phase 3?

73. Cost issues Average cost per new vaccine, Trials and phase III: For each percent chance of at least one success, a portfolio will produce a range of possible numbers of new vaccines Discussion points: What will be the average number of successful products, and the average cost per successful product, for any given probability of at least one successful product? What is the average cost per product based on the 95% reasoning? How many products are logically possible in phase 3 anyway? What are the potential trial site bottlenecks? What is the optimal sizing of the number, distribution, and sizing of trial sites given the lumpy nature of trials needs?

74. Cost issues Sponsor return, discussion point: What do sponsors get as their payout (in terms of contract deals on quantities and prices in exchange for their investments)? COGS, discussion point: What is the importance of COGS, especially of booster vaccines? How might lower booster manufacturing costs stimulate uptake? How does that feed in to NPV? Competitive event, discussion point: What is the �competitive event�? What is the social value of the �competitive event�? Plant size/capacity, discussion point: What are the various plant size and capacity issues?

75. Cost issues Investment externalities: The greater the investment and the earlier the success via drugs in tackling MDR-TB and XDR-TB, the lower will be the value of vaccine-based investments Similarly, a combination of new pre- and postexposure vaccines could substantially reduce the risks of emergence of MDR-TB and XDR-TB and reduce the value of the market for drugs Better diagnostics enable more precise targeting of booster vaccines, with impacts on booster vaccine sales (probably able to be lower) and prices (probably able to be higher) Better clinical biomarkers will shorten trial lengths, sizes and costs, and yet there will be poor incentive to develop and supply such biomarkers if they are costly to develop and if the benefits can�t be internalized Discussion points: What is the combination of negative and positive investment externalities across different possible areas of R&D? How does it impact investor incentives? How should any trade-offs be modelled? How should investment case analysis adapt to this?

76. AMC and similar mechanisms The amount of R&D to pay for and the sizing exercise, discussion points: In setting any prize-like mechanism to incentivise R&D, what amount of R&D would be targeted? What would be the methodology for setting the size for TB vaccines? How would the following be factored in: Complexity of underlying science R&D costs Epidemiological issues Already spent private and sponsor funding Production costs Would an �auction� be used as originally proposed, followed by a monitoring mechanism (to check firms are investing �enough�)? Follow-on incentives discussion points: What dynamic follow-on incentives are envisaged that TB vaccine development will need? How will this be best managed?

77. AMC and similar mechanisms PDP issues: How would this or any other mechanism interact with vaccine/drug/diagnostic PDPs? Given that it is supposed to work by pulling in private finance? Affordable prices: How is pressure kept up on affordability, especially in the light of the recent pneumococcal experiences? How does a �prize� stop becoming an inefficient pre-sunk subsidy? The pneumococcal lesson: The pneumococcal AMC case is providing a lesson; what are the salient features so far of that case? Given the heavy monitoring and evaluation being proposed (the recently-released report of the �monitoring and evaluability study� is 110 pages long), this suggests that by the time TB becomes a potential target, a great deal more will be known. Is there some sense in waiting?

78. AMC and similar mechanisms Alternatives/complements, discussion points: If an AMC is a prize-based R&D mechanism, what alternatives/complements are there in terms of: Procurement policy Other predictable funding streams Demand forecasting Tiered pricing Delivery mechanisms Production capacity insurance, etc.? How does this feed back on the original prize-based mechanism? At what point, does the AMC nomenclature stop referring to a �prize� and start referring to something else? Funding available and tradeoffs, discussion points: The current AMC counts on balance sheets earlier than originally thought. How does this affect the tradeoff of funding mechanisms? What is the probable budget constraint? How much funding is going to be available probably?