Five of the most cited names in research integrity — John Ioannidis (2005), Amgen's Begley & Ellis (2012), Bayer's Prinz, Schlange & Asadullah (2011), Freedman, Cockburn & Simcoe (2015), and Schneider & Kilicoglu (HHS/ORI-funded, ongoing) — have independently documented the same structural failure in biomedical research.
Pharma companies cannot reproduce 65–89% of the published findings they buy programs on, the U.S. spends roughly $28 billion a year on preclinical work that cannot be repeated, and 25.4% of medical articles contain citation errors where the cited paper does not support what the citing paper claims.
The problem is documented at industry scale. The process that produces $100M-to-$10B biotech deals has not been fixed to address it.
| Finding | Source | |
|---|---|---|
| 89% | Amgen could not reproduce 47 of 53 landmark cancer biology papers, with cooperation from the original authors. | Begley & Ellis, Nature, 2012 |
| 65% | Bayer could not reproduce 65% of 67 published drug-target findings they used to source internal programs. | Prinz et al., Nat Rev Drug Discov, 2011 |
| 25.4% | Medical articles contain citation errors — the cited paper does not support what the citing paper claims. | Schneider & Kilicoglu, HHS/ORI (ongoing) |
| $28B/yr | U.S. spend on preclinical research that cannot be reproduced by anyone else. | Freedman et al., PLOS Biology, 2015 |
| 0 | Firms offering systematized, named citation-chain forensics as a service for biotech M&A diligence. | White space confirmed (04-2026) |
AbbVie paid $5.8 billion for Stemcentrx to get Rova-T, an antibody-drug conjugate targeting DLL3 on cancer stem cells. The deal rested on a foundational assumption: that the cancer stem cell model, functionally validated in acute myeloid leukemia and breast cancer, applied to small cell lung cancer.
It had never been functionally validated in SCLC. The standard assay (in vivo limiting dilution) was never published for that tumor type pre-acquisition. Two peer-reviewed papers (Hidalgo 2014, Ehm 2015) directly challenged the assumption and were available in the literature before close.
Phase III halted December 2018 — patients on Rova-T died sooner than on standard chemotherapy. Program terminated 2019. AbbVie took a $4 billion impairment charge.
A forensic audit tracing the citation chain backward from Bonnet & Dick (1997) to Saunders (2015) would have surfaced the missing validation in days. AbbVie could have demanded the assay before closing, restructured payments as milestones contingent on functional validation, or walked away. Without the flag, none of those conversations happen.
The same forensic method applies across the biotech M&A record. Four more citation-chain audits follow the same format: trace the evidence backward, flag where validation is missing, show what was available before close.
The twist case. The hypothesis was right. The Phase II data was real. Manufacturing scaled up for Phase III, phosphorothioate stereochemistry drifted, and Smad7 knockdown collapsed from 60–85% to 0–15%. A $10,000-per-batch head-to-head knockdown comparison — the same assay used in Phase II — would have caught the drift before the trial enrolled a single patient. Root cause not surfaced for six years.
Emraclidine is an M4-selective muscarinic agonist for schizophrenia. The confirmatory evidence cited at deal close (Karuna's KarXT Phase II data) was generated with xanomeline — an M1/M4 co-activator, not M4-selective. M1 receptor knock-out mouse data showing independent M1 contribution to antipsychotic-like effects was published before close. Ten days of forensics would have surfaced the gap. Phase II trial failed weeks after close.
Pfizer torcetrapib, Roche dalcetrapib, Merck anacetrapib. The same HDL-causation hypothesis run three times. The first failure (torcetrapib, 2006) was where the hypothesis broke. Voight et al. Mendelian randomization data — showing HDL levels are not causally linked to cardiovascular disease — was available four years before dal-OUTCOMES and well before REVEAL. Attempts two and three were avoidable.
Twenty years removing plaques the hypothesis said were causing the disease. DIAN familial-Alzheimer's studies showing tau pathology (not amyloid) better predicts cognitive decline, Jack et al. biomarker trajectory models, and longitudinal cohorts of amyloid-positive cognitively normal patients were all published before the major late-stage failures. The hypothesis predicted those cognitively normal patients should not exist. They did.
The forensic audit is not only a diligence tool for acquirers. Applied backward to an investor's portfolio, it answers: what did the science actually say at the time you entered, and what decision would that have enabled? Nine lookbacks from a single institutional portfolio, generated from public 13F filings and primary literature. → View the full portfolio audit dashboard
The drug worked. The thesis resolved correctly. But the forensic audit at entry would have found a specific gap — the superiority claim over tafamidis rested on pharmacodynamic extrapolation, not clinical evidence — and would have recommended a smaller starter position with a milestone-triggered add on Phase 3 data. The investor held through an 85% drawdown before the Phase 3 confirmed the thesis. The audit doesn't change the outcome. It changes the experience of getting there.
The positive control. Every link in the citation chain holds at HIGH confidence. COL7A1 causation is 30-year textbook biology. The GEM-3 Phase 3 was published in the NEJM two months before entry. FDA approval came four months later. The forensic audit at entry would have said: proceed at conviction size. The science is as clean as clinical-stage biotech gets. And it was.
The audit reaches its own boundary. NicHealth entered after FDA approval, after ASPEN Phase 3 confirmation, after the NEJM publication. The science chain is not where the risk lives. The forensic audit confirms the mechanism is de-risked — and then names what the investor is actually betting on: commercial execution in an underdiagnosed indication. The audit's job is to clarify which chapter you're in. Science chapter closed. Commercial chapter open.
The paradigm bet. Prior KRAS inhibitors (sotorasib, adagrasib) target the GDP-bound inactive state of RAS — they catch it at rest. Revolution's RAS(ON) multi-selective inhibitor targets the GTP-bound active state: the oncogenic form that drives tumor proliferation. Phase 1/2 data shows ORR 38% in KRAS-mutant NSCLC and 47% in PDAC combination — compelling signals in the hardest solid tumor in oncology. The forensic audit confirms the mechanism is a genuine paradigm shift but names the open question: ORR does not equal OS, and PDAC has a graveyard of promising Phase 1 signals that died in Phase 3. RASolute 302 data expected 2026.
The mechanism audit. Obefazimod's claimed mechanism — a small molecule CBC binding site ligand that enhances lncRNA splicing to upregulate miR-124 25-fold in rectal tissue, suppressing inflammatory cytokines — is the most novel and least independently replicated in the portfolio. Phase 3 induction was positive (16.4% placebo-adjusted remission, p<0.0001). The forensic audit identifies two gaps: maintenance data was missing at entry (Q2 2026 readout), and the miR-124 anti-inflammatory literature base is predominantly neuronal, not intestinal. NicHealth's position size ($11.2M, smallest major position) is forensically consistent with the audit's verdict — sized for a maintenance binary event, not a confirmed thesis.
The cleanest chain in the portfolio. ROS1 as an oncogenic driver is one of the most definitively validated targets in oncology. Zidesamtinib is designed to solve the specific problem that prior ROS1 inhibitors created: the G2032R resistance mutation (dominant post-TKI mechanism, ~40% of patients) and CNS penetrance failure. ARROS-1 Phase 1/2 (n=117 TKI-pretreated): ORR 44%, DOR 78% at 12 months, CNS ORR 56%. NDA accepted, no AdCom required, PDUFA September 18, 2026. The forensic audit finds no material gaps — defined driver, defined resistance mutation, defined clinical solution. Proceed at conviction.
The class rehabilitation bet. XEN1101 is a Kv7.2/7.3 potassium channel opener — the same mechanistic class as ezogabine (Potiga), which was approved in 2011 and withdrawn in 2017 not for efficacy failure but for pigmentation toxicity caused by its melanin-binding aromatic ring system. The forensic audit confirms XEN1101 structurally avoids the responsible moiety and that preclinical pigmentation studies are negative. Phase 2b X-TOLE (n=325, JAMA Neurology 2023): 52.8% vs 18.2% median seizure reduction. The +123,891-share add — largest single-quarter position build in the portfolio — is a high-conviction Phase 3 bet. But the audit makes one point non-negotiable: pigmentation monitoring at longer Phase 3 exposure durations is the kill condition that overrides any efficacy outcome.
The audit boundary case for rare disease. Maralixibat is approved — FDA and EU — for pruritus in Alagille Syndrome and PFIC. The ASBT inhibitor mechanism is textbook hepatology. The science chapter is closed. What the audit identifies is the open chapter: the EXPAND Phase 3 in biliary atresia and additional rare cholestatic diseases. Biliary atresia is pathophysiologically distinct from ALGS and PFIC — it is an obstructive, structural disease, not a genetic transporter deficiency. The ASBT mechanism is plausible but not guaranteed to translate. EXPAND enrollment completed March 2026; top-line data expected 2027. The Q4 2025 trim (–3,677 shares) is forensically consistent: managing position size ahead of a multi-year data wait.
The audit that lands at the event. Auvelity is the first oral NMDA antagonist approved for MDD (August 2022) — dextromethorphan's NMDA antagonism plus sigma-1 agonism, sustained in plasma by bupropion's CYP2D6 inhibition. The MDD science is de-risked. NicHealth's +42,907-share Q4 2025 add is a specific bet on the Alzheimer's agitation NDA, PDUFA April 30, 2026 — three days from this audit date. ACCORD Phase 3 was positive (NEJM 2024). The forensic audit names the risks that remain: bupropion's seizure risk and DXM's falls risk are more consequential in an elderly Alzheimer's population than in working-age MDD patients, ACCORD (n=178) is not large enough to fully characterize them, and brexpiprazole (Rexulti) is already approved for the same indication. Approval resolves the binary. Commercial success is the chapter that opens after.
An independent forensic audit of a biotech target's scientific evidence chain — tracing every citation link behind the mechanism-of-action claim backward to primary sources, checking whether cited studies actually support the claims built on them, and flagging where standard validation is missing. The audit does not run wet-lab experiments. It produces an evidence-chain map the acquirer can use to restructure the deal before close, not after — or the investor can use to size the position correctly from the start.
→ Read the full product spec (methodology, process, case studies, pricing anchor)