Cultures of Trial and Error: When the Spell Breaks: Magic, Money, and the EPR Effect

JOTE x NanoBubbles present Cultures of Trial and Error: a peer reviewed blog series on error correction in science.

How nanomedicine conjured billions from thin air and what happened when the enchantment failed

“Nano scale [is] magic scale.” This was not a late-night laboratory joke. It was official U.S. government policy. The phrase appeared verbatim in early National Science Foundation solicitations for the National Nanotechnology Initiative (NNI). Nano-evangelist Mihail Roco coined it, and the very first NNI programme description placed it alongside technical aims (Mody 2004; Roco 2023). When President Clinton announced the program in 2000, he promised breakthroughs that were “only imagination away.” Magic sat next to milestones in official funding calls. The government had converted a charismatic promise into a bureaucratically sustained future.

What follows offers a case study in how weak scientific claims gain extraordinary performative power. It also shows how the entanglement of technoscience with financial capital can disable science’s self-correcting mechanisms. The protagonist is the Enhanced Permeability and Retention (EPR) Effect: the hypothesis that tumours’ leaky blood vessels would passively accumulate drug-loaded nanoparticles. For over two decades, EPR organised billions in research funding, launched dozens of biotech ventures, and became nanomedicine’s constitutive fiction. Its empirical basis was tenuous from the start. It worked reliably only in certain tumours transplanted into certain mice.

Drawing on our article in Social Studies of Science (Rasmussen & Kuch, 2026), this post uses the concept of enchantment to illuminate how EPR’s spell was cast, why it held for so long, and what its collapse reveals about contemporary technoscience. Max Weber (1991 [1919]: 139) called enchantments “mysterious incalculable forces” and denied their presence in modern science. Science and Technology Studies (STS) scholars have tracked them nonetheless, through concepts like promissory economies (Brown 2003), fictional expectations (Beckert 2013), and sociotechnical imaginaries (Jasanoff & Kim 2015).

The Claim That Launched a Thousand Startups

In 1986, Japanese pharmacologists Matsumura and Maeda observed that a polymer-modified cancer drug accumulated preferentially in mouse tumours. They proposed a mechanism. Tumours have chaotic, leaky blood vessels (enhanced permeability) and poor lymphatic drainage (enhanced retention). Macromolecules and particles of appropriate size should therefore concentrate in tumours passively, without molecular targeting (Matsumura & Maeda 1986). If true, this “EPR Effect” would turn cancer’s chaos into therapeutic advantage. Leaky vessels would become gateways. The drug would find the tumour on its own.

The empirical basis was tenuous from the outset. EPR worked reliably only in specific tumours transplanted into specific mice. Not all tumours showed enhanced permeability. Results varied considerably. As early as 1992, pharmacologist Len Seymour noted that discouraging results from liposome researchers had “suppressed passive drug targeting research in general” (Seymour 1992: 151). This early warning would be buried under subsequent enthusiasm.

Yet EPR did not fade. Through the 1990s, it circulated within a small drug delivery community. Researchers like Ruth Duncan and Karel Ulbrich developed polymer-drug conjugates designed to exploit the Effect. Duncan’s landmark 1999 review positioned EPR as the linchpin for a new generation of “polymer therapeutics” (Duncan 1999). At that point, none of these drug classes carried the “nano” prefix. That rebranding was about to arrive.

From Hypothesis to Holy Grail

The NNI’s launch in 2000 changed everything. Billions in federal funding and explicit invocations of “magic” transformed EPR’s prospects. Coordinated across agencies including NSF, NIH, DOE, and NASA, NNI spending climbed from roughly $500 million in 2001 to nearly $2 billion annually by 2010. By 2023, cumulative spending passed $40 billion (Roco 2023). Similar nano-roadmaps in France, the UK, Japan, China, and elsewhere reinforced a transnational funding ecology. Funders treated EPR as a ready-made portal to futuristic cancer cures, years before robust human data existed (Kearnes & Rip 2009).

Under these conditions, EPR became what sociologists call a “boundary object” (Star & Griesemer 1989). It was flexible enough to mean different things to different communities, yet coherent enough to coordinate collective action. Polymer chemists saw a design principle. Venture capitalists saw a platform technology. Federal agencies saw a tangible health benefit to justify massive investment. Universities saw patents and spin-offs. Each community found what it needed in the same three-letter acronym.

EPR-based cancer therapeutics were not a minor corner of nanomedicine. They were its centre of gravity. By the mid-2010s, cancer therapeutics accounted for roughly 40% of nanopharmaceutical value. Nanopharma itself represented about 90% of projected nanomedicine market value. EPR underpinned more than a third of nanomedicine’s overall worth (Loubaton 2012).

The Wizard and the Magic Bullet

Every enchantment needs a magician and a talisman. The EPR story found its charismatic figure in Robert Langer, the MIT scientist whose thousand-plus patents made him legendary in academic and investment circles alike. Media profiles called him a “wizard” of biotech and compared him to Thomas Edison (Prokesch 2017). His reputation operated as transferable charisma: symbolic capital that conveyed credibility through association. To have Langer’s name on your paper, his technology in your startup, his blessing on your grant: these conferred value.

Through strategically placed review articles in journals like Nature Nanotechnology, Langer positioned EPR-based nanocarriers as the natural heirs to Paul Ehrlich’s century-old “magic bullet” dream. Ehrlich, the early twentieth-century pharmacologist, had envisioned drugs that seek and destroy pathogens while sparing healthy tissue (Lenoir 1988; Bosch & Rosich 2008). In a prominent 2007 review, Langer and colleagues described the nanocarrier-based approach as “an important modality within therapeutic and diagnostic oncology,” with EPR as its core mechanism (Peer et al. 2007: 758).

Langer was candid about the logic. Publication in “highly selective and competitive journals validates that the idea may be a significant breakthrough,” he wrote (Langer 2013: 487). Peer review became a charisma-generating machine.

When Stories Become Securities

Here the political economy of promising becomes explicit. EPR did not merely organise research. It created financial value. Patents incorporated EPR assumptions as foundational claims. Investment prospectuses treated it as validated science. When Langer and colleague Omid Farokhzad founded BIND Therapeutics in 2006, they secured nearly $70 million in private capital by the end of 2007, plus billion-dollar milestone payments from Big Pharma partnerships (Maine & Thomas 2017). Many of BIND’s foundational patents depended explicitly on EPR. The Effect underwrote the value of the firm.

By the time BIND floated its IPO in 2013, its drug delivery platform was protected by 16 issued U.S. patents and 50 patent applications (Neuman & Chandhok 2016). The prospectus invoked EPR as the mechanism by which its Accurin® particles would reach tumours, asserting that “Accurins represent the next stage in the evolution of targeted therapies and nanomedicine” (BIND Therapeutics 2013). To many investors, BIND’s EPR-based “story” made for a compelling “value proposition,” according to one investment newsletter’s post-mortem (Nanalyze 2017).

Economic sociologist Jens Beckert (2013) calls such narratives “fictional expectations”: stories whose empirical truth matters less than their capacity to coordinate action under radical uncertainty. But fictionality alone cannot explain why some stories hold collective imagination for decades while others evaporate quickly. EPR’s empirical weakness was evident early on. Yet belief persisted. What Beckert’s framework captures in coordination, it underspecifies in effect.

This is where enchantment adds analytical depth. Drawing on Jane Bennett’s (2001) concept of secular enchantment and Josephson-Storm’s (2017) work on modernity’s persistent magical thinking, we use “enchantment” to describe the sticky, affective commitment that binds heterogeneous actors into shared belief. Scientists, investors, regulators, patients: EPR bound them through something that preceded and exceeded rational calculation. Mouse experiments became talismans. Langer’s reputation became portable charisma. The “magic bullet” metaphor promised transformation, not mere treatment. It promised to convert “leaky” from pathological liability into investment opportunity.

Why Self-Correction Failed

Here we confront a troubling question: why didn’t science’s self-correcting mechanisms identify and reject EPR’s weak empirical foundations?

Part of the answer lies in what we call “distributed complicity.” No single actor committed fraud or conscious deception. Yet the system collectively sustained a fiction. The concept draws on Mirowski’s (2011) analysis of the “commercialisation of epistemology” but extends it by emphasising the distributed, emergent quality of the problem. No conspiracy was needed. Each node in the network optimised rationally within local constraints. The aggregate effect was that a weak claim became structurally impossible to dislodge.

Scientists needed NNI funding and could access it by framing existing research as “nanomedicine.” Peer reviewers faced reputational costs in challenging papers from prestigious laboratories. Funding officers needed success narratives to justify continued investment to Congress. Universities needed patent royalties and spin-off companies. Venture capitalists needed compelling stories to deploy capital. Regulatory agencies lacked resources for aggressive pre-clinical scrutiny. Each actor’s behaviour was locally rational. The collective outcome was epistemic capture.

The classic Mertonian norms of organised scepticism presume disinterestedness. Financialisation erodes that disinterestedness. When everyone from graduate students to university presidents has material stakes in a narrative, correction requires more than better evidence. It requires collective willingness to write off massive sunk costs in careers, patents, institutional identities, and investment portfolios. Enchantment makes those costs prohibitive by distributing complicity so widely that no individual actor can afford to break ranks.

Even ambiguous evidence was systematically reinterpreted to sustain the spell. When BIND’s Phase I trials showed modest results (two of three patients exhibiting some response based on surrogate measures), this became “promising” data justifying three Phase II trials and a jump in company valuation (Fonseca et al. 2014). The field had developed sophisticated interpretive practices for extracting hope from equivocal findings.

What Finally Broke the Spell

The spell did not break through peer review, replication failures, or internal scientific debate. It broke through a market correction.

BIND’s Phase II trials showed limited benefit. In January 2016, the company dropped its prostate cancer indication. In April, results for lung cancer and a multi-tumour trial showed marginal or no efficacy. Layoffs came the same day. Within a month, BIND had filed for bankruptcy. Two months later, Pfizer acquired the remaining assets for $40 million, down from a peak valuation exceeding $230 million.

This was not unusual for cancer drugs. Phase II trials are not designed to prove efficacy, and many promising candidates fail at this stage. But BIND had come to symbolise EPR’s promise so conspicuously that its collapse carried disproportionate weight.

Drug delivery scientist Fabienne Danhier (2016: 109) concluded: “The verdict has been handed down: the EPR effect works in rodents but not in humans. What is the future of nanomedicine?” The question confirmed the point. EPR-based cancer drugs had been nanomedicine’s most valuable fruit. Some senior scientists went further, arguing that the EPR fiasco reflected a political problem with the field’s distribution of resources (Park 2017).

The timing reveals something important. The spell broke from outside the scientific community, through bankruptcy, rather than from within. Financial failure is not epistemic correction. But it can trigger epistemic correction by making continued belief too costly to sustain. In this sense, the market compensated for science’s structural failure. Scientists still revised their views. The impetus came from outside.

The COVID Sequel: Re-Enchantment

Enchantments rarely stay dispelled. By late 2019, nanomedicine had reached the “trough of disillusionment.” The U.S. National Cancer Institute had discontinued funding for its flagship Centres of Cancer Nanotechnology Excellence. In October 2019, French nanomedicine researcher Patrick Couvreur blamed overinvestment in EPR for the field’s parlous state (Couvreur 2019: 319–320).

Barely a month later, COVID-19 changed everything. Within months, Moderna’s and Pfizer-BioNTech’s lipid-nanoparticle mRNA vaccines proved spectacularly effective. Nanomedicine’s leaders seized the opportunity. A March 2020 editorial called for deploying “nanotechnology advances as frontline tools” against the coronavirus (Chan 2020). A later Nature Nanotechnology editorial framed the impending vaccine approvals as “a milestone for nanomedicine,” capable of restoring faith lost in the EPR debacle (Anon. 2020).

Robert Langer illustrates how charisma migrates under financialised technoscience. He had co-founded Moderna in 2010, six years before BIND’s collapse. The lipid-nanoparticle technology was technically distinct from EPR-based polymer platforms. The vaccines worked by a different mechanism and targeted a different disease. What transferred was not science but reputational capital: the aura of the “wizard” whose track record made investors confident. At its 2018 IPO, Moderna was valued at nearly $8 billion before it had a single approved product.

This rapid re-enchantment demonstrates what Fortun (2008) calls “speculative realism”: promissory narratives can survive empirical failure by migrating to new domains. Under financial capitalism, charisma and narrative transfer across applications faster than evidence accumulates. The COVID vaccines were genuine achievements. But the speed with which they rehabilitated nanomedicine as a whole owed much to enchantment’s fungibility.

Beyond Nanomedicine: Recognising the Pattern

The EPR story is not unique. Similar dynamics appear wherever promissory narratives, charismatic figures, and venture capital intertwine (Borup et al. 2006; Brown & Michael 2003).

Consider the current AI landscape. “Scaling laws” and “emergent capabilities” function much as EPR once did: they organise research programs, justify massive capital allocation, and resist falsification because too many actors have too much invested (Galanos 2023; Kotliar 2025). The charismatic figures are familiar types. The gap between laboratory benchmarks and real-world utility echoes the gap between mouse tumours and human patients. The interpretive practices for extracting hope from ambiguous evidence mirror the endpoint drift that sustained EPR (Schaefer et al. 2024).

Quantum computing tells a similar story. Repeated promises of imminent “quantum advantage” have organised funding for two decades, yet practical applications remain elusive (Verhoeven 2021). Foundational assumptions are rarely questioned in funding contexts. Longevity biotech follows the same grammar: massive capital, promissory narratives, wizard founders, and a gap between cellular demonstrations and clinical reality.

This is not to say these fields are fraudulent. EPR was not fraudulent either. The point is structural. When technoscience and finance entangle under these conditions, certain patterns recur. Recognising them is the first step toward better epistemic hygiene.

Navigating Enchantment: Notes for Early-Career Researchers

For PhD students and postdocs, the EPR story may feel uncomfortably familiar. You may be building your research program on foundational assumptions you have inherited rather than tested. You may suspect a paradigm is shaky but lack the career security to say so.

A few practical observations.

Learn to recognise enchantment from the inside. Ask yourself: What would falsify my core assumption? Has anyone looked? If the assumption were wrong, who would lose, and how much? If the answers are “nothing would,” “no,” and “everyone, enormously,” you may be inside an enchantment.

Understand the costs of scepticism. Questioning foundational assumptions is professionally dangerous. The costs fall hardest on those with least security: graduate students, postdocs, researchers without tenure. This asymmetry is how enchantments sustain themselves.

Diversify your epistemic portfolio. If your entire research program depends on a single foundational assumption, you are vulnerable. Where possible, develop expertise that would retain value even if the paradigm shifted.

Find your sceptics. Every field has people who harbour doubts but stay quiet for career reasons. Find them. Build relationships. Create spaces where concerns can be voiced without professional penalty.

Remember that enchantments end. EPR organised billions and careers for two decades, then collapsed in months. The researchers who adapted fastest had privately maintained critical distance. The researchers who suffered most had fully identified with the paradigm.

What Can We Learn?

Spectacular failures create privileged observation windows. When spells break, previously invisible infrastructure surfaces. The vacuum left by a collapsed enchantment reveals the coordination work it had been doing: review articles positioning EPR as inevitable, patents assuming it into existence, investment narratives making it bankable.

Metaphors perform work. When technologies promise to turn chaos into control, ask what the metaphor accomplishes. “Magic bullet” is not innocent description. It is strategic enchantment that canalises capital, labour, and imagination.

Scepticism has material costs. Questioning enchantments requires material security, not just intellectual courage. This is why EPR was broken by bankers rather than bench scientists. The costs of questioning fell hardest on those inside the scientific community.

Correction comes from unexpected places. For highly financialised technoscience, we may need to rethink where epistemic accountability operates. Bankruptcy triggered what peer review could not.

Which Enchantments, For Whom?

Max Weber argued that modern rationality means “there are no mysterious incalculable forces” and that “the world is disenchanted.” He was wrong. The EPR saga shows why. The boundary between rational calculation and magical thinking is more porous than we admit, especially when technoscience and finance intertwine.

The question is not whether science should dispel enchantment. Enchantment may be constitutive of the collective action that ambitious research requires. The question is which enchantments serve which communities and how we can make their operation visible before they collapse under their own weight.

A closing invitation. What enchantments structure your own field? What foundational assumptions have you inherited but never tested? What would it cost you to voice scepticism? And if those costs feel prohibitive, what does that tell you about the epistemic health of your discipline?

These are not comfortable questions. But they are the ones most worth asking.

This post draws on Nicolas Rasmussen and Declan Kuch’s article “‘Nano Scale [is] Magic Scale’: On EPR, Unicorns, and Enchantment in Nanomedicine,” published in Social Studies of Science. DK gratefully acknowledges support for some of this research by the ARC Centre of Excellence in Convergent Bio-Nano Science and Technology (CE140100036).

This blog post series has been financially supported by 'NanoBubbles: how, when and why does science fail to correct itself', a project that has received Synergy grant funding from the European Research Council (ERC), within the European Union’s Horizon 2020 programme, grant agreement no. 951393.

References

Anon. (2020). Nanomedicine and the COVID-19 vaccines. Nature Nanotechnology, 15, 963.

Beckert, J. (2013). Imagined futures: Fictional expectations in the economy. Theory and Society, 42(3), 219–240.

Bennett, J. (2001). The Enchantment of Modern Life. Princeton University Press.

BIND Therapeutics. (2013). Form S-1, United States Securities and Exchange Commission, filed 12 August 2013.

Borup, M., Brown, N., Konrad, K. & van Lente, H. (2006). The sociology of expectations in science and technology. Technology Analysis & Strategic Management, 18(3–4), 285–298.

Bosch, F. & Rosich, L. (2008). The contributions of Paul Ehrlich to pharmacology. Pharmacology, 82(3), 171–179.

Brown, N. (2003). Hope against hype. Science & Technology Studies, 16(2), 3–21.

Brown, N. & Michael, M. (2003). A sociology of expectations: Retrospecting prospects and prospecting retrospects. Technology Analysis and Strategic Management, 15(1), 3–18.

Chan, W.C.W. (2020). Nano research for COVID-19. ACS Nano, 14(4), 3719–3720.

Couvreur, P. (2019). Nanomedicine: From where are we coming and where are we going? Journal of Controlled Release, 311, 319–321.

Danhier, F. (2016). To exploit the tumor microenvironment. Journal of Controlled Release, 244, 108–121.

Duncan, R. (1999). Polymer conjugates for tumour targeting. Pharmaceutical Science and Technology Today, 2(11), 441–449.

Fonseca, N.A. et al. (2014). Bridging cancer biology and the patients’ needs. Cancer Treatment Reviews, 40(5), 626–635.

Fortun, M. (2008). Promising Genomics. University of California Press.

Galanos, V. (2023). A Historical Sociology of Expectations and Expertise in AI. PhD thesis, University of Edinburgh.

Jasanoff, S. & Kim, S.H. (2015). Dreamscapes of Modernity. University of Chicago Press.

Josephson-Storm, J.A. (2017). The Myth of Disenchantment. University of Chicago Press.

Kearnes, M.B. & Rip, A. (2009). The emerging governance landscape of nanotechnology. In S. Gammel et al. (Eds.), Jenseits von Regulierung. AKA.

Kotliar, D.M. (2025). Can’t stop the hype: Scrutinizing AI’s realities. Information, Communication & Society. https://doi.org/10.1080/1369118X.2025.2531165

Langer, R. (2013). A personal account of translating discoveries. Nature Biotechnology, 31(6), 487–489.

Lenoir, T. (1988). A magic bullet. Minerva, 26(1), 66–88.

Loubaton, B. (2012). Exploring the quantitative dimensions of the economic impact of nanomedicine. OECD & US NNI International Symposium, Washington DC.

Maine, E. & Thomas, V.J. (2017). Raising financing through strategic timing. Nature Nanotechnology, 12(2), 93–98.

Matsumura, Y. & Maeda, H. (1986). A new concept for macromolecular therapeutics. Cancer Research, 46(12.1), 6387–6392.

Mirowski, P. (2011). Science-Mart. Harvard University Press.

Mody, C.M. (2004). Small, but determined. Hyle, 10, 99–128.

Nanalyze. (2017). WTF Happened to BIND Therapeutics? Nanalyze, 1 August 2017.

Neuman, D. & Chandhok, J.N. (2016). Nanomedicine patents highlight importance of production methods. Nature Reviews Drug Discovery, 15(7), 448–450.

Park, K. (2017). The drug delivery field at the inflection point. Journal of Controlled Release, 267, 2–14.

Peer, D. et al. (2007). Nanocarriers as an emerging platform for cancer therapy. Nature Nanotechnology, 2, 751–760.

Prokesch, S. (2017). The Edison of medicine. Harvard Business Review, March–April.

Rasmussen, N. & Kuch, D. (2026). ‘Nano Scale [is] Magic Scale’: On EPR, Unicorns, and Enchantment in Nanomedicine. Social Studies of Science. https://doi.org/10.1177/03063127261424493

Roco, M.C. (2023). National Nanotechnology Initiative at 20 years. Journal of Nanoparticle Research, 25, 197–224.

Schaefer, M. et al. (2024). Misrepresented technological solutions in imagined futures: The origins and dangers of AI hype in the research community. arXiv:2408.15244.

Seymour, L.W. (1992). Passive tumor targeting. Critical Reviews in Therapeutic Drug Carrier Systems, 9(2), 135–187.

Star, S.L. & Griesemer, J.R. (1989). Institutional ecology, translations, and boundary objects. Social Studies of Science, 19(3), 387–420.

Verhoeven, P. (2021). On the social shaping of quantum technologies: An analysis of emerging expectations through grant proposals from 2002–2020. Minerva, 59, 421–440.

Weber, M. (1991 [1919]). Science as a vocation. In From Max Weber: Essays in Sociology (pp. 129–156). Routledge.