Startups are the new axis of pharmaceutical innovation, but Brazil hasn't realized it yet

How the global shift in health innovation redefines the role of industrial policies and exposes the backwardness of Brazil's strategy

The pharmaceutical industry is experiencing a historic shift. The model that, for half a century, concentrated drug discovery in large, integrated corporations is fragmenting. Vertical integration, a symbol of strength in the 20th century, has given way to a decentralized network of small biotechs , academic laboratories, and deep tech startups that are taking the lead in biomedical innovation. What seemed disorganized has become method. The new innovation paradigm is distributed, collaborative, and anchored in shared risk .

This transformation, accurately analyzed by Charles H. Reynolds in The New Paradigm in Pharma R&D (ACS Med. Chem. Lett., 2025), reveals how the traditional model collapsed under the weight of its own scale. Large companies became portfolio managers, not knowledge generators. The capacity to innovate migrated to small companies. In 2024, more than 60% of new drugs approved by the FDA originated in biotechs, and the most innovative compounds, first-in-class therapies, came almost entirely from startups.

In Brazil, however, the debate on innovation still moves in a different timeframe. Policies like the New Industry Brazil program focus incentives on large companies , while the true engine of technological competitiveness—startups—remains marginalized among government priorities.

As we've pointed out in previous IBIS Insights articles, such as "The Impact of Brazil’s New Foreign Exchange Framework on the Country’s Technology Balance" and "Boosting Brazilian Deep Tech: A Bayh-Dole-Inspired Mechanism for University Spin-offs and National Innovation" insisting on this model perpetuates dependence. The new frontier in healthcare and biotechnology will be conquered by those who understand that tomorrow's scale is built with today's risks.

The new paradigm of drug discovery

Reynolds's study explains how the productivity crisis at Big Pharma in the 2000s led to their reconfiguration. The average cost per molecule jumped from $800 million to over $2.5 billion , while the number of new drugs approved fell by almost half. The old, hierarchical, and siloed internal R&D model could no longer sustain the speed of science.

The answer was to outsource, collaborate, and buy innovation. Companies began forming distributed research networks, involving CROs, universities, and startups specializing in synthetic biology, advanced therapies, or AI platforms. This more flexible architecture, according to Reynolds, began to produce faster and more cost-effective results.

In 1998, only 20% of innovative medicines came from small companies. By 2024, that number had exceeded 60%, according to Evaluate Pharma. This phenomenon is so consistent that large companies have begun to design their portfolios based on future acquisitions, what the author calls the "swimming innovation model" Science has moved from being inside corporations to orbiting them.

This movement also decentralized the geography of innovation. Medium-sized cities with strong universities and regional investment funds, such as Cambridge (UK), Basel, Tel Aviv, and Toronto, replaced the former monopolies of New York and Zurich as centers of biomedical discovery. The logic is simple: networks replaced hierarchies . And in networks, small companies can lead.

Anatomy of the Small Advantage in Pharmaceutical Innovation

Small businesses no longer innovate by accident. They have structures and cultures that foster discovery. Reynolds lists three core elements: cost, focus, and alignment.

Cost. Startups operate with a much lower proportion of overhead. While large pharmaceutical companies allocate more than 50% of their R&D budget to internal administrative and regulatory processes, small biotechs concentrate 80% on scientific experimentation. This allows them to iterate more, adjust more, and fail faster .

Focus. Large companies, pressured by shareholders and multiple therapeutic fronts, dilute their efforts. Startups do the opposite: they concentrate everything on a single hypothesis. This intensity generates radical discoveries and also failures. But, on average, it produces more relevant innovation per dollar invested .

Alignment. In a 10- or 20-person startup, the CEO and chief scientist share the same goal and the same lab. Communication is direct, decisions are made quickly, and encouragement is shared. This model eliminates what Reynolds calls the “organizational iceberg”: the gap between what happens on the bench and what reaches management.

There's also a fourth factor: the culture of interdependence . Today's innovation ecosystem is made up of transient ties and dynamic partnerships between startups, universities, and large companies. This flexibility makes the system more resilient, adaptable, and creative. Innovation becomes a flow, not a fiefdom.

What Brazil still hasn't learned

Industrial policies need to place startups at the center of their strategy . There is no technological sovereignty without a national policy for deep tech , especially in the areas of healthcare, biotechnology, and AI.

And there's data that reinforces this urgency. According to Emerge's Deep Tech Radar 2025 , Brazil is home to 952 deep tech startups , of which 36% operate in health and wellness and 27% in agriculture and food, two sectors that account for almost two-thirds of the ecosystem. Biotechnology is the main technology base , followed by AI and advanced materials, reflecting the confluence of biodiversity, life sciences, and digital transformation.

Around 55% of these companies are academic spin-offs , which shows the potential of Brazilian science, but 47% have never received private investment and depend mainly on public funding resources, such as FAPESP, Finep and Sebrae.

What's missing, therefore, is not knowledge, but an environment that translates this scientific depth into economic value and social impact . And this will come not from more generic public notices, but from an industrial policy capable of understanding the risk logic that drives deep tech innovation.

A policy for the 21st century

A national policy for deep tech in healthcare needs to be built on three principles: risk, coordination, and learning .

Risk, because science-based startups require time and uncertainty, and the government's role is precisely to cushion the cost of this uncertainty. Coordination, because science, business, and government still live in silos, and innovation flourishes where there are bridges. Learning, because innovation policies aren't designed all at once; they iterate, like science itself.

This entails creating sovereign wealth funds focused on biotechnology and AI in healthcare, with technical governance and technology maturity targets (TRL). It also entails reforming regulation, creating sandboxes for advanced therapies, medical algorithms, and new clinical trial models. And, above all, it entails supporting translational innovation: connecting universities, hospitals, and startups in proof-of-concept programs that transform discoveries into validated products.

Successful models already exist. The Danish BioInnovation Institute and the British Innovate UK Health Catalyst combine public-private investment, co-innovation with major pharmaceutical companies, and integrated regulatory support. These are experiences that Brazil could tropicalize, leveraging its universities, the Unified Health System (SUS), and its biodiversity as competitive advantages.

The risks of ignoring the turnaround

The cost of inertia is well known. Brazil lost the race for industrial biotechnology in the 2000s and for pharmacogenomics in the 2010s. Now it risks repeating the mistake with cutting-edge therapies. Technological dependence is deepening, and the balance of payments in intellectual property remains in deficit .

But ignoring the shift is not just an economic mistake, it's a strategic one. Without strong healthcare startups, the country will depend on imports even for essential medicines like insulin and RNA vaccines. Furthermore, without risk-based support mechanisms, Brazilian scientific talent will continue to migrate to more receptive ecosystems.

Of course, we must avoid the other extreme: euphoria over unsustainable startups. The global success rate in biomedical R&D remains below 8%. Smart public policies should mitigate risks, but with clear metrics, a focus on social impact, and commitments to reinvest in local research . Frontier innovation isn't built on promises, but on persistence.

A new role for large companies

The rise of startups doesn't eliminate the importance of big pharma; it redefines it. Big pharma remains essential for scaling, manufacturing, and certification. The challenge is to create synergies instead of hierarchies.

Tax incentives for co-innovation, corporate venture funds, public procurement of innovation, and co-development programs can align the interests of large companies with the potential of small ones. Brazil has already had successful trials in this regard, such as COINFAR , which brought companies and universities closer together in the early 2000s. However, they lacked continuity.

Today, a COINFAR 2.0 would make more sense than ever: a network that brings together startups, public institutions, and the pharmaceutical industry around strategic technology missions, immunobiologicals, clinical AI, and Amazonian bioactives. The country has the actors; what's missing is the roadmap.

From consumers to producers of the future

The new paradigm of healthcare innovation forces us to make a choice. Brazil can remain a consumer of foreign technologies, dependent on imports and royalties, or it can position itself as a producer of knowledge.

Startups are currently the most effective instrument for transforming science into sovereignty. But this requires policies that accept risk as part of the process and treat uncertainty as an asset . The future of the pharmaceutical industry will not be defined by the size of manufacturing plants, but by the speed of learning and the courage to experiment.

The country that understands this first will not only manufacture medicines but will set the course of science .

by Marcio de Paula

Brazilian Health Innovation Institute - IBIS