Key messages from the 2023 funding data

One area of R&D that has experienced a significant long-term decline is vaccine development for S. pneumoniae and N. meningitidis (‘pneumonia and meningitis’). Vaccine funding for both pathogens is down by more than 90% from their respective peaks, falling from a total of $136m in 2013 to less than $20m in 2023.

Broadly speaking, this shift makes sense. New conjugate vaccines for both pneumonia and meningitis have been launched over the last 15 years, which our modelling credits with averting more than 138 thousand disability-adjusted life years (DALYs) through to the end of 2024. These new vaccines remain vulnerable to the rise of resistant serotypes, leading to an arms race of multi-valent vaccines targeting an ever-changing list of dominant variants. As a result, there is still some ongoing development of whole-cell-based and non-conjugate-protein-based vaccines, which would limit the virus’s ability to select for resistant strains.

However, compared to the world before the new conjugate vaccines, the burden from both pneumonia and meningitis is substantially lower. Death rates from meningitis fell from 4.66 per hundred thousand people in 2010 when MenAfriVac began widespread distribution to 2.71 in 2021. It makes perfect sense that funding for a now much smaller problem has declined, shifting to ensuring access to those new products or to other areas with a relatively greater untreated burden.

Malaria vaccines show a similar, if less striking, response to the success of new products. While early-stage research continues on vaccines that could supersede R21 and RTS,S by delivering higher efficacy or disrupting transmission, malaria vaccine R&D has fallen sharply from the peak associated with those vaccines’ late-stage trials. This, too, seems like broadly good news – provided the displaced funding is allocated wisely. Production and distribution of R21 remains a barrier to reaching its full potential, with current plans accounting for only around 25 million doses, compared to the 200 million needed to maximise its impact.

Another area where funding is understandably down is onchocerciasis drug R&D, which peaked in 2016 with trials of the repurposed drug moxidectin before falling steeply once it was approved.

Figure 8: Product approvals versus R&D funding

Open to enlargeDownload

Less obviously explicable is the long-term decline in microbicide funding for HIV prevention, which has dropped by over 90% from its peak. Unlike other areas of long-term decline, however, this downward trend began well before the launch of any approved product; funding was already far below its peak by 2021, when the WHO recommended the dapivirine ring, still the only regulatory-approved microbicide.

Some of the missing microbicide funding has shifted into multipurpose prevention technologies (MPTs). These are products which act as contraceptives while also protecting from sexually transmitted infections, including HIV. This shift reflects an emerging consensus that MPTs deliver more impact than single-purpose microbicides, offering users better options for pre-exposure prophylaxis and reducing their need to interact with the health system. However, while some funders have explicitly shifted their microbicide budgets towards MPTs, there has been nothing like enough MPT funding to account for the more than $200m decline in annual microbicide spending. One culprit might be the failed 2022 Phase III trial of the MPT candidate EvoGuard, which may have dented funders’ enthusiasm – hopefully only temporarily.

Funding for HIV vaccines has fallen alongside microbicides, albeit much more slowly. Unlike pneumonia, meningitis, and malaria – where reductions follow successful product launches – this drop reflects multiple late-stage setbacks, forcing a move back to less costly early-stage R&D as part of an overall shift away from HIV vaccine research. Since 2018, HIV vaccine funding has fallen by nearly a quarter of a billion dollars, across both early- and late-stage research – reaching a record low in 2023. As with microbicides being replaced by MPTs, some of this dip may reflect the success of alternative product categories, including improved HIV diagnostics and the long-acting injectables which have significantly reduced the burden of HIV even without a new vaccine.

The net effect of all these shifts is that global vaccine funding has trended gradually down since its peak in 2009, with one major exception: tuberculosis.

Figure 9: Vaccine R&D funding, TB vs other diseases 2007-2023 (participation adjusted for 2023)

Open to enlargeDownload

While reported vaccine funding rose slightly in 2023, this reflects funding from new survey participants (which we excluded from Figure 9) and, more importantly, a second consecutive leap in funding for TB vaccines.

The share of global vaccine funding going to TB has risen from just 5% of the global total in 2018 to 12% in 2023, rising sharply as vaccine R&D for most other diseases has fallen. As shown below, these falls are headlined by the $250m (27%) reduction in HIV vaccine funding and the even larger proportional falls for malaria (down by 33%), diarrhoeal diseases (down by 43%) and especially pneumonia & meningitis (down by 89%), most of them reflecting one or more product launches.

Figure 10: Change in vaccine R&D 2018-2023 by disease

Open to enlargeDownload

The single biggest strategic shift in the funding landscape over the last two years has been the near doubling of spending on TB vaccines, with the vast majority of the new funding being directed to the Gates Medical Research Institute for late-stage trials of the M72 vaccine.

M72 is the most advanced of the 16 different TB vaccine candidates currently in the pipeline. In Phase IIb trials, M72 has demonstrated around 50% efficacy in preventing disease in children – a small but meaningful improvement on the existing BCG, which offers roughly 45% protection. The small number of actual infections observed during the trial, though, means there is significant overlap in the confidence intervals for M72 and BCG’s efficacies: we can’t yet be sure that M72 will improve on the existing standard of care.

The scale and epidemiology of TB infections are such that even a 5% increase in efficacy would avert millions of deaths and tens of millions of cases over the next 25 years: as many as two million deaths and almost 13 million cases in India alone.

M72 offers a significant opportunity to drastically reduce the global burden of TB and slow the rise of extensively drug-resistant strains. Its development reflects a big bet by its philanthropic backers, the Gates Foundation and Wellcome, who plan to invest a significant share of the neglected disease funding in hopes of a successful trial result. Practically, though, all this funding for M72 has come at the expense of other neglected diseases: both major funders have reduced their non-TB expenditures over the past two years. Evaluating whether this gamble is worthwhile – accounting for the roughly 30% chance that M72 might demonstrate lower efficacy than BCG – requires careful consideration of where the money is coming from.

The calculus for investing in TB R&D looks very different if M72’s efficacy proves to be high enough to make elimination of the disease a realistic goal. The prospect of eliminating TB, rather than merely controlling it, ought to push funders towards an ‘all of the above’ TB R&D strategy, including improved diagnostics and therapeutics to detect and treat the cases where the vaccine fails and improved public health measures to control its spread while the other measures take effect. These other approaches all remain valuable investments even if – as seems likely for a vaccine with roughly 50% efficacy – M72 can’t form part of a realistic elimination strategy; but they are much less valuable if elimination is off the table; and less valuable than they were in a world with no M72 and more cases of TB, despite the obvious medical synergies between vaccines, diagnostics and therapeutics. If M72 doesn’t reduce TB’s prevalence, and with it the likely impact of a novel TB drug, we should not be spending $500m testing it.

In a world where funding for neglected diseases is – mostly – fixed, big bets like M72 require big sacrifices somewhere else. Not everything can be prioritised.

Each year, we carefully catalogue the most significant changes in funding over the previous year and try to explain why, for example, private sector investment went up. Sometimes, that same trend reverses itself the following year, and we are left explaining why private sector funding went down again. This sort of variation can represent two real and opposing trends, both worthy of explanation, or it might just be random variations in companies’ year-to-year spending as their trials ramp up or wind down.

One way to avoid being drawn into building complex explanations for simple randomness is to restrict ourselves to long-term trends: the declines in vaccine or microbicide funding have been going on far too long to be written off as mere blips. The downside of only writing about long-term trends, though, is that we need to wait years for them to emerge from the data, and we miss big new shifts as they happen.

Another way to avoid talking about randomness is to understand the processes that generate funding totals: we know, for example, that Japanese public funding to GHIT follows a two-year cycle, and that Australian PDP funding runs for five years. When we see these anticipated cyclical changes, we avoid getting too worried. But, despite our best efforts, we often don’t have any ability to determine when funders have changed their strategy – bad news about future commitments to neglected disease is unlikely to make it into a press release.

As a result, we are trying to develop an understanding of how random neglected disease R&D funding has been in the past so that we can separate the genuinely surprising shifts from the sort of transitory changes we observe all the time.

Figure 11 represents an attempt to use ‘autoregressive integrated moving average’ (ARIMA) modelling to explain past changes in overall funding and to forecast its future direction. This is a ‘dumb’ model: it doesn’t know anything about global funding, including the specific cyclical shifts in GHIT and PDP funding identified above. It treats funding as a sequence of numbers over time and tries to build a model to fit the observed data.

Figure 11: Observed and forecasted funding, totals

Open to enlargeDownload

What the model suggests is that overall global funding follows a comprehensible trajectory – the model is able to produce values that fit closely with the real ones. It also suggests that the slight drop in funding we observed in 2023 was a little unexpected. Based on its past behaviour, the model ‘expected’ funding to rise a little in 2023 and it forecasts that funding will continue to rise over the next two years. This suggests that the drop in funding we observed is perhaps more significant a departure from previous trends than the rest of the report acknowledges, but also that the observed level of funding is not hugely different from what an uninformed observer, armed only with trend data, would have predicted.

Looking instead at funding specifically for vaccines, we see a slightly different picture:

Figure 12: Observed and forecasted funding, vaccines

Open to enlargeDownload

The larger gap between actual and predicted funding implies that vaccine funding is harder for the model to explain than overall funding, which aggregates more individual decisions and disbursements, balancing out some of its random variation.

Second, the observed trend to date suggests (to the model) that vaccine funding will continue to fall; the model is ‘surprised’ by the slight uptick in vaccine funding we saw in 2023. We know, though the model doesn’t, that the very small rebound in reported vaccine funding reflects new survey participants, and secondly, that funding would have fallen sharply were it not for the big increase in TB vaccine R&D. So, the model is telling us that these events do represent a break from the existing trend – they are worth explaining and probably won’t disappear in next year’s data.

Funding from the US government dropped sharply for the second year in a row, with big cuts from almost every major funder: the NIH, the DOD, the CDC and USAID.

Overall US public funding is now closer to its level prior to the 2009 pre-American Recovery & Reinvestment Act than to its extended peak between 2018 and 2021, when it averaged more than $2.2bn a year.

The broader decline in US funding for neglected disease R&D highlights the importance of making a compelling case for continued US engagement. As outlined in last year’s Doing Well by Doing Good report, emphasising the domestic as well as global benefits of such investments is critical.

At the same time, diversifying funding sources is essential to reduce reliance on the shifting priorities of US defence and basic research funding, which often reflects the interests of America’s soldiers and scientists, rather than the burdens faced by LMICs.

Alongside the direct health impact of rising global temperatures and their accompanying extreme weather events, climate change will increase the range of many tropical diseases and provide new habitats for the insects and other vectors that carry them.

A key part of adapting to the impacts of climate change will be preparing naïve populations and inexperienced health systems to deal with the threat from diseases previously confined to countries closer to the equator. A 2021 systematic review of links between climate change and tropical disease found that several neglected diseases had already spread to new areas and predicted significant further increases in both range and incidence as warming continues. This is particularly true of vector-borne diseases like dengue, which it predicted would spread to coast cities in Eastern China and Japan, alongside a 40-fold projected increase in cases in existing endemic areas like Dhaka, Bangladesh.

These predictions would prove prescient. In 2024 incidence of dengue exploded across the globe, leading to an estimated 3,000 deaths, including several major outbreaks across parts of Pakistan previously considered outside its range. In this context, both the surging private sector investment in dengue drug R&D – which leapt to a second consecutive record high in 2023 – and their earlier focus on vaccines (see Figure 13) is both easy to understand and likely too little, too late to address the ongoing rise of dengue infections.

Figure 13: In-scope private sector dengue R&D funding, by product*

Open to enlargeDownload

_{* Dengue vaccine funding was excluded from the G-FINDER survey post 2012 following strong commercial funding growth.}

The rush to create viable treatments for dengue should serve as a reminder of all the other climate-sensitive G-FINDER neglected diseases – such as malaria, helminths, diarrhoeal diseases, leishmaniasis, and Chagas – that will spread across a warming world.

Neglecting these kinds of diseases was never a good policy, but climate change will transform funding from a question of providing neighbours with development assistance to one of health security at home.

Impact Global Health has consistently called for increased funding for neglected disease R&D. The Impact of Global Health R&D report we published early in 2024 demonstrates that funding for neglected disease R&D delivers a massive return on investment – more than 400 to 1 – and argues that funders of all kinds should be spending more money on securing even more of these gains.

We reiterate those sentiments here: funding for neglected disease R&D is a good investment, and the world should be doing more of it.

But after nearly two decades of urging funders to increase support, it’s clear that simply calling for “more money” rarely prompts a drastic shift in priorities.

To meaningfully expand the pool of funding for neglected disease R&D, advocates will actually have to make the case that it represents a good use of funders’ limited resources.

The good news is that we have strong arguments on our side:

• The first of these is the sheer scale of R&D’s global impact, mostly measured in the value to society of lives saved and DALYs averted.
• The second is that conducting R&D delivers tangible economic benefits to the nations where it takes place, including more than 15,000 jobs and €30bn in economic activity in Europe alone.
• The third is that neglected disease R&D can and should be viewed through the lens of climate abatement and health security. A stockpile of products to deal with a distant disease can rapidly transform from a tool for vaccine diplomacy to a means of keeping one’s population safe. From the point of view of the private sector, diseases can go from rare and unprofitable to endemic and lucrative much faster than candidates can advance through the pipeline.

No matter how persuasive our advocacy, there will inevitably be more unmet needs than funding to address them. While working to increase the size of the funding pie, funders and developers must also ask two questions: how to turn a fixed budget into as much R&D as possible, and how to maximize the impact of that R&D.

In an ideal world, funders would estimate the full cost of building a pipeline designed to meet every unmet need, incorporating the risks of failure and evolving disease burden. In practice, we have something in the order of $4bn to work with each year, and the question funders face is not ‘what is the optimal number of late-stage trials to guarantee a vaccine by 2030?’ but instead ‘could the cost of a late-stage vaccine trial deliver more health impact if spent on other things?’. Assessing ‘what we need’ in light of 'what we actually have' means funders need to know not only what a given outcome is likely to cost, but also develop a triage strategy to ensure their limited resources deliver the greatest possible impact.

As we discuss elsewhere in the context of EID product development, trial protocols and – especially – benchmarks of statistical certainty designed for peacetime use in rich country health systems often fail to strike the appropriate balance between risk and cost in poorer nations struggling with an ongoing health crisis.

Critics of the FDA argue that the costs and delays involved with a gold standard review are hard to justify, even in the case of expensive drugs for mild problems in rich countries. They are much harder to accept for something like a repurposed therapeutic with a known side effect profile being used against a communicable or potentially fatal disease.

Clinical trials account for an estimated 68% of the cost of bringing a new product to market, peaking with the late-stage trials designed to demonstrate the product delivers a meaningful benefit. Given this distribution of costs, the decision by the governments of Ghana, Burkina Faso and Nigeria to begin using the R21 malaria vaccine based on its promising Phase IIb trial results is almost certainly the right one.

This approach could form the basis of a more general trend towards approving products once they have been proven safe, based on suggestive but inconclusive evidence of efficacy, and then substituting real-world data collection for late-stage clinical trials.

This would require developers (and the regulators who control their market access) to weigh the trade-off between distributing a product earlier, and at perhaps half the R&D cost, against the costs and risks of approving a product which ultimately proves to be harmless, but useless. Proceeding on the basis of incomplete data may not often be the right decision, but sometimes it will be.

The costs of clinical development can also be brought down by building trial capacity in LMICs. Something like the UK’s COVID-era RECOVERY trial, which tested ten different interventions across 48 thousand patients at a cost of just $20m, is obviously partly an artefact of the pandemic, and of the UK’s National Health Service and its investment in clinical research facilities. But the concept of lower cost platform trials drawing on a pre-identified patient population ought not to be inherently more expensive when performed in, say, Nigeria instead of the UK.

Finally, and more generally, conventions regarding "certainty" and "significance" in clinical trials are not immutable; they are norms grown up around historical accident – including, notably, the calculations of an Irish beer scientist. When a clinical trial concludes that a medicine is worthless because the data show a 6% chance it has no effect, this reflects a deliberate choice. Evaluating whether this choice is the right one requires consideration, not just of the probability of being wrong (in this case, 94%) but also of the potential consequences of abandoning a working product.

It is vanishingly unlikely that a statistical threshold originally designed to optimise the purity of Irish beer would be perfectly suited to determining the approval of a vaccine in Nigeria.

Above, we consider the evidence that funders are rationally responding to their successes and failures when deciding where to spend their money.

Two new malaria vaccines don’t obviate the need for different and better vaccines in the future, but it makes them less of a priority than a decade ago. A still-unproven late-stage TB vaccine doesn’t mean we should abandon early-stage research on other candidates, but, at the margin, it probably means we should be spending less. The second approved therapeutic for onchocerciasis will probably provide less impact than the first.

Are funders striking the right balance between maintaining long-term commitments and responding to emerging priorities? While we certainly hope so, the reality is that we don’t truly know – and we strongly suspect that funders themselves may not know either. Deciding how much funding to allocate to TB vaccines versus Chagas diagnostics, for example, is an incredibly complex optimisation challenge. This is further compounded by the fact that individual funders often make these decisions without a clear understanding of how much other funders are investing in each area.

We like to think that G-FINDER provides funders with part of that picture by helping them monitor what others have spent, are spending, and will spend on a particular area; and by helping them identify other candidates and where they sit in the pipeline. However, knowing how much is being spent is only a starting point. Leading funders like Wellcome and the Gates Foundation refine their decisions by weighing each candidate’s probability of success, likely impact on clinical practice, and alignment with existing pipelines. Peer review and expert panels also shape funders’ priorities, though repeatedly relying on the same voices can introduce biases.

Estimating how many lives a yet-to-be-developed vaccine might save is daunting, but focusing on impact rather than just regulatory approval helps to keep the interests of patients at the heart of R&D decisions.

As with the M72 TB vaccine, looking at a product’s role in actual communities can guide not only how much funding it deserves, but how well it complements or replaces other tools, ensuring resources genuinely improve lives where they’re needed most.

Prioritisation is a particularly fraught debate in global health because almost everything we could be researching is, in some sense, incredibly worthwhile. When the average return on R&D funding is more than 400 to 1, even our least promising ideas will deliver health improvements worth ten times what we spend on them. Inevitably, though, some ideas will be more promising than others.

Funders need robust data – not wishful thinking or guesswork – to identify the most promising opportunities and ensure each investment has the strongest possible impact.

Figure 14: Projected DALYs averted 2024-2040, by product

Open to enlargeDownload

As funders strive to build complementary portfolios that maximise impact, it is vital that they regularly revisit past allocations with the benefit of new evidence. In that light, it is encouraging to see funders actively redirecting their efforts in response to changes in burden, and in the product landscape, to where past successes and failures point them next, making smart choices to drive future breakthroughs in global health.