Appendix 5. Assumptions behind Research and Development Cost Estimates
Appendix 5 explains the methodology used to estimate the global cost of malaria R&D through 2050. Specifically, the model evaluates the cost for malaria drugs, vaccines, vector control and diagnostics, including the early research, development and information needs after launch. R&D cost estimates were derived from interviews with experts in the malaria community, historical data and industry analysis.
Given the inherent uncertainty in predicting time and costs associated with technology development, this model is based on assumptions that should be continuously updated and cross referenced as new information comes available. A multiplier of 1.2 has been applied to the final cost estimates at this time to account for this uncertainty. As the research agenda for elimination and eradication becomes more defined, the model will need to be further refined.
Early Research. Preclinical research costs are included in the model estimates. In the 2004 Malaria R&D Alliance Report, basic research was estimated to be 16% of the total malaria R&D costs. For the purpose of this model, the percentage of basic research was assumed to double to 32% of the 2007 R&D costs given the increased efforts necessary to enable malaria eradication. As a result, since the global malaria R&D spend in 2007 was estimated to be ~US$ 422 million, the basic research was estimated to be 32% of this or US$ 133 million. This basic research cost was assumed to be constant going forward through 2050.
The annual research allocation of US$ 133 million was divided according to the breakdown outlined in the following table. The basic research need for diagnostics was developed separately and is described in the diagnostic section. The model assumes that basic research on vaccines is 50% of the total basic research costs since the technology is further behind relative to the other interventions. The remaining basic research costs are assumed to be split equally between drugs and vector control.
Table A.9: Allocation of basic research costs by intervention
|Intervention||% of basic research||Annual basic research cost (US$ millions)|
|Diagnostics||n/a||See section below|
Source: GMAP costing model, 2004 Malaria R&D Alliance Report and expert interviews
Information Needs R&D Costs. The information needs cost is comprised of post-launch and product integration studies. Specifically, this cost captures implementation research, effectiveness studies, and resistance monitoring. In 2004, these costs were estimated to be 17% of the total R&D cost presented in the 2004 Malaria R&D Alliance Report. For the purpose of this model, these costs were estimated as 20% of the total annual R&D costs. This estimate is based on the assumption that there is a necessity for greater information needs support given the heavy commitment to developing new tools and ensuring they are used effectively in future decades.
Drug R&D Costs. R&D costs for drugs are grouped into preventive and therapeutic.
Preventive Drugs. The priorities for future preventive drugs are aimed at filling gaps in the tool kit by developing IPT-specific drugs. It was assumed that 1 novel combination drug and 1 monotherapy will be launched in the next 10 years. Drug requirements assume non-artemisinin combinations and as a result, 2 active ingredients will be developed for novel preventive drugs in the next 10 years. Furthermore, 1 active ingredient would need to be developed in subsequent decades in order to have enough products to prevent resistance buildup. In addition to developing novel preventive drugs, it is estimated that 4 reformulations will be developed in 10 years and 2 reformulations will be developed each subsequent decade.
The cost of developing a novel active ingredient is estimated to be US$ 250 million and the development time is assumed to be 10 years. For modeling purposes, the cost was evenly spread out over the 10 year development cycle. The cost of developing a reformulation is estimated to be US$ 25 million or 10% of the cost of a new active ingredient. The development time is assumed to be 2-6 years. For modeling purposes the cost of 4 reformulations was spread out over 10 years (2008-2018) and the cost of 2 reformulations was linearly spread out over each subsequent decade.
Therapeutic Drugs. Current thinking in the R&D community indicates that the following types of therapeutic drugs are needed:
- Next generation ACT for P. falciparum
- Therapy targeting the hypnozoite of P. vivax in the liver
- Drugs blocking P. falciparum and P. vivax transmission (Gametocytocides/sporontocides)
- Drugs aimed at avoiding resistance
It is assumed that 4 novel combination drugs will be developed in the next 10 years. The first combination drug, a next generation ACT for P. falciparum, will require the development of 1 new active ingredient in addition to artemisinin. The second combination drug, a therapy targeting P. vivax hypnozoites in the liver, will require at least 1 new active ingredient in addition to a pre-existing active ingredient. It is assumed that 2 new active ingredients are needed for a third combination drug. This drug will block both P. falciparum and P. vivax transmission through the vector and also simultaneously treat the disease at the red blood cell stage. It may be challenging to design and develop this drug; therefore a more conservative estimate was used in the model. It was assumed that a third combination drug will block P. falciparum transmission and a separate, fourth combination drug (also requiring the development of 2 new active ingredients) will block P. vivax transmission. These drugs may or may not treat the disease at the red blood cell stage. If they did not treat the disease at the red blood cell stage, a separate, single combination therapy could be developed that accomplishes this for both P. falciparum and P. vivax. In total, the model assumes 6 new active ingredients will be developed in the next 10 years to yield 4 novel therapeutic combination drugs.
In order to avoid resistance, it is estimated that 2 new active ingredients will be needed to develop new combination therapies every subsequent decade. One active ingredient may be needed for the therapeutic combination targeting P. vivax in the liver stage and one active ingredient for the combination used to block P. falciparum and/or P. vivax transmission.
In addition to developing novel therapeutic drugs, it is estimated that 10 reformulations will be developed in 10 years and 6 reformulations will be developed each subsequent decade. Specifically, given the 4 new combination therapies being developed in 10 years as discussed above, each therapy requires a reformulation for various populations: adults (accounted for), pregnant women (4), children (4), infants (1), and intravenous (IV) formulation for severe cases of malaria (1). As a result, 10 reformulations are needed in 10 years. Given the target of developing 2 therapies in subsequent decades to combat resistance, 6 reformulations will be developed every 10 years starting in 2018: pregnant women (2), children (2), infants (1), and IV reformulation for severe malaria (1).
As with preventive drugs, the cost of developing a novel active ingredient for therapeutic drugs is estimated to be US$ 250 million and the development time is assumed to be 10 years. For modeling purposes, the cost was evenly spread out over the 10 year development cycle. The cost of developing a reformulation is estimated to be US$ 25 million or 10% of the cost of a new active ingredient. The development time is assumed to be 2-6 years. For modeling purposes the cost of 8 reformulations was spread out over 10 years (2008-2018) and the cost of 6 reformulations was linearly spread out over each subsequent decade.
Table A.10: Estimated cost of research and development for drugs
|Drug R&D||Timeframe||Total cost|
|2 active ingredients||2008-2018||$500|
|2 active ingredients||Subsequent decades||$500|
|6 active ingredients||2008-2018||$1,500|
|2 active ingredients||Subsequent decades||$500|
Source(s): GMAP costing model, Medicines for Malaria Venture (MMV) and expert interviews
Vaccine R&D Costs. Many experts consider vaccine development to be a key activity for malaria elimination and eradication. Given the lack of success in moving a malaria vaccine through phase III clinical trials to date, predicting the cost and timing of future vaccine launches is highly uncertain. As a result, assumptions made in generating the vaccine R&D cost numbers will have to continuously be updated as technological progress is made.
Efficacious vaccines are needed for both P. falciparum and P. vivax. RTS,S, which targets P. falciparum, is the most advanced malaria vaccine and is currently in phase III clinical trials. Even if RTS,S launches in 2013-14, a more efficacious P. falciparum vaccine is likely necessary for malaria elimination. Based on current vaccine priorities, it is assumed that a next generation P. falciparum vaccine would have to exceed 80% efficacy in order to justify the cost of late stage development. In addition to overcoming this hurdle, efficacy comparison studies between both generations of vaccines would have to be conducted. As a result, it is estimated that after the launch of RTS,S 10 years are needed for the deployment of a second generation P. falciparum vaccine.
In addition to the vaccines for P. falciparum, an efficacious vaccine for P. vivax will be necessary for malaria eradication. Several other vaccines would also be tremendous assets for the malaria community: a vaccine that targets both P. falciparum and P. vivax, a transmission blocking vaccine, and a vaccine for pregnant women. For the purposes of this R&D costing effort, it was assumed that four vaccines could be developed by 2028. Furthermore, one subsequent vaccine would be developed every decade after 2028.
- 1 RTS,S vaccine for P. falciparum (launch 2013-14)
- 1 next generation vaccine P. falciparum (launch 2024)
- 1 vaccine for P. vivax (launch 2024)
- 1 other vaccine (launch 2028): a vaccine that targets both P. falciparum and P. vivax, and/or a transmission blocking vaccine, and/or a vaccine for pregnant women
As of 2007, it was assumed the remaining cost to develop RTS,S was US$ 220 million. For modeling purposes, these costs were linearly spread out through 2013. In general, the baseline cost of the other vaccines was assumed to be US$ 800 million and the development timeline was assumed to be 13 years. Furthermore, 75% of the cost is spread over 10 years (pre-phase III) and 25% of the cost was spread over 3 years (phase III through launch). However, for the second generation P. falciparum vaccine (#2), the development timeline was lengthened to 17 years and the cost increased to ~US$ 1 billion for two reasons. First, the 80% efficacy hurdle makes the probability of success more challenging, thus the pre-phase III timeline was lengthened three years at an annual cost equal to the other pre-phase III years. Second, the Phase III efficacy comparison studies would take considerably longer and are more expensive than the Phase III studies for a first generation vaccine. Thus, one year of additional costs and time was added to the phase III portion of the model for this vaccine.
The US$ 800 million cost and 13 year vaccine development time estimates were derived from a compilation of historical malaria vaccine performance, comparable vaccine development data, and expert discussions. Specifically, pre-erythrocytic combinations and transmission blocking vaccines or combinations cost on average ~US$ 116 million and take ~13 years. Given an attrition rate that ranges from 0.6-2.4%, the total investment, including cost of failures, ranges from US$ 550 million - 1.5 billion. Similarly, blood stage vaccines cost on average ~US$ 114 million and take ~13 years. However, the success rate for blood stage vaccines ranges from 2.2-6.8%, therefore the total investment, including cost of failures, ranges from US$ 350 – 640 million. An average of these vaccine development costs (based on pipeline mix) yields ~US$ 800 million which is the estimate for vaccine cost used in the model. The average development timelines for each type of vaccine is ~13 years.
Table A.11: Estimated cost of research and development for vaccines
|Vaccine R&D||Timeframe||Total cost|
|RTS,S for P. falciparum||2008-2013||$189|
|Vaccine for P. falciparum||2008-2024||~$1,000|
|Vaccine for P. vivax||2012 -2024||~$800|
|Future vaccines||Post-2028||~$800 / vaccine|
Source(s): GMAP costing model, Malaria Vaccine Initiative (MVI), Bill and Melinda Gates Foundation, expert interviews
Vector Control R&D Costs. Vector control R&D is aimed at developing new active ingredients, new formulations and new paradigms for killing vectors. Specifically, it was assumed that 3 novel active ingredient classes, 15 reformulations, and 3 new paradigms such as larviciding, consumer products, etc. would need to be developed in the next 10-12 years to achieve control and elimination objectives. The new active ingredients are needed to develop safer, longer lasting, less expensive pesticides and chemicals for new paradigms that emerge. It is estimated that 1 novel active ingredient, 10 reformulations, and 1 new paradigm is needed to prevent resistance in each subsequent decade.
The cost of developing a novel active ingredient for vector control is estimated to be ~US$ 200 million and the development time is assumed to be 12 years. For modeling purposes, the cost was evenly spread over a 12 year development cycle. The cost of developing a reformulation is estimated to be between US$ 1 - 5 million, so on average ~US$ 3 million. The development time is assumed to be 2-6 years. For modeling purposes, the cost of 15 reformulations was spread evenly over 10 years (2008-2018) and the cost of 10 reformulations was linearly spread over each subsequent decade. The cost of establishing a new paradigm is ~US$ 4 million and there is a 50% failure rate associated with this effort. The development time to validate the utility of a new paradigm through experimentation is ~5 years. The US$ 24 million cost associated with developing 3 new paradigms was spread linearly over 10 years as was the US$ 8 million cost for 1 paradigm each decade thereafter.
Table A.12: Estimated cost of research and development for vector control
|Vector control R&D||Timeframe||Total cost (US$ millions)|
|3 active ingredients||2008-2020||$600|
|1 active ingredient||Subsequent decades||$200|
Source(s): GMAP costing model, Innovative Vector Control Consortium (IVCC) and expert interviews
Diagnostic R&D Costs. While the current diagnostic methods, microscopy and rapid diagnostic test technologies (RDTs), can confirm clinical diagnoses and provide treatment information, there are still several R&D opportunities in this field. Microscopy can identify which of multiple parasite species are in circulation, and determine the parasite density (quantitation). However, the experience of the technician and the quality of the equipment determines the sensitivity of microscopic diagnosis, and it is therefore limited to larger clinics and inappropriate for most village-based situations. Development of RDTs for malaria offers the potential to extend accurate malaria diagnosis to remote areas without microscopy services.See RDT Info
Click for source However, RDTs are not without challenges either: inconsistent quality within and across batches leads to a perception of unreliability in some circumstances. Priorities and assumptions described below were developed with these key issues in mind.
Microscopy. Microscopy has been the reference standard of diagnostic equipment since it enables direct parasite determination. Giemsa microscopy has been the primary microscopy technique used in the past. Newer microscopy techniques are being evaluated which offer greater detection capabilities. Technologies such as incident light fluorescence microscopy are growing rapidly in importance as investigational tools in the fields of medical and biological research, and may have a place in improving accuracy and reliability of malaria microscopy. It is assumed that annual R&D investment in microscopy of at least ~US$ 2 million will be needed through 2050 in order to continuously improve microscopy technologies.
RDTs. The immediate goal is to improve existing technology to get higher quality RDTs. In the medium- to long- term, one could try to develop new monoclonal antibodies, advanced polymerase chain reaction (PCR) technology, or other broader diagnostic technologies. Due to the uncertainty surrounding the time and cost of realizing each of these R&D options, it was assumed that the 5 R&D strategy scenarios shown below will be conducted in parallel:
- Current technology is improved thereby enhancing product quality and reproducibility.
- New monoclonal antibodies are developed to increase diagnostic sensitivity and test stability.
- PCR technology is developed for mass screening.
- Broader diagnostic technologies are developed: e.g. platform that performs differential diagnoses for a range of infectious diseases (currently in development at Claros Diagnostics), remote diagnosis via telemedicine, etc.
- Non-invasive tests are developed.
For the purpose of this costing model, it was assumed that improvements to current technology costing US$ 200,000 per year per research laboratory would be conducted through 2050. Assuming five research laboratories would be conducting R&D on the top diagnostic technologies yields an annual cost of US$ 1 million. Even if in the next few years significant improvements to current diagnostic technology is seen and higher quality tools were deployed in the field, other types of diagnostic R&D (specifically, scenario #2, #3, and #4 above) would persist.
Research on new monoclonal antibodies and new PCR technology would continue too in an attempt to improve diagnostic sensitivity. This increased sensitivity would enable improved diagnosis of malaria in pregnancy, in asymptomatic patients during population screening, and in other situations where parasite densities are low. Research to develop tests for specific markers of severe disease, such as cerebral malaria, is also expected to proceed. Furthermore, development of new monoclonal antibodies could help improve the thermal stability and shelf life of current diagnostics.
Developing a new monoclonal antibody is estimated to take a lab ~US$ 750,000 and 4-5 years. Given a 33% probability of success it would cost ~US$ 2.5 million over 4-5 years to identify a new monoclonal antibody. In addition, multiple field trials would be conducted in various endemic settings, each costing ~US$ 1 million. For the purpose of this model, the assumption was made that every 5 years through 2025, one new monoclonal antibody would be developed and 3 field studies would be performed resulting in an improved diagnostic at the cost of US$ 5.5 million. These costs were spread linearly over the 5 year development timeline. After 2025, it is reasonable to assume that countries will be more focused on elimination and thus advances in either PCR technology or broader diagnostic technology will be the dominant technology deployed in the field.
Development of new PCR technology is expected to take ~8-10 years. The model assumes that the development of this technology (scenario #3) or broader technologies (scenario #4) will go on in parallel through 2050 at an annual cost that is double the cost for developing a new monoclonal antibody. As a result, the annual cost to explore both PCR and broader diagnostic technologies is estimated to be ~US$ 4.4 million.
Development of non-invasive tests, which may use existing technologies to detect markers available from other means of sampling (e.g. saliva, urine), or new technologies such as photoabsorption are estimated to be ~US$ 1 million and US$ 2 million per year, respectively.
Overall, the peak diagnostic costs are expected to be US$ 11.5 million per year. The table below outlines the cost breakdown described above.
Table A.13: Estimated cost of research and development for diagnostics
|Diagnostic R&D||Timeframe||Total cost (US$ millions)|
|Improving current technology||2008-2050+||$1|
|New monoclonal technology||2008-2025||$1.1|
|Broader diagnostic technology||2008-2050+||$2.2|
Source(s): GMAP costing model, Foundation for Innovative Diagnostics (FIND), WHO and expert interviews