The Pharma Papers: the buried trial that reached 40 countries

The documents

The archive began with a single encrypted drive handed to a reporter in a café: some 40,000 internal files from Aldermann Pharma, one of the world's larger generic-drug manufacturers, spanning study reports, regulatory submissions, pharmacovigilance exports and years of internal email. Over five months, a former manager in the company's drug-safety unit answered questions the documents could not, on the record and at professional risk. The files and her testimony point to the same conclusion: that a safety signal seen in the company's own data was softened, reclassified and, in places, left out of the dossiers sent to regulators.

The drug at the centre is olodipine, a calcium-channel blocker used to control high blood pressure and marketed under more than a dozen brand names. It is cheap, off-patent and, for tens of millions of patients, a daily tablet. That ubiquity is what makes the findings matter: a small distortion in a safety file, multiplied across 40 national markets, is not a paperwork error but a public-health event.

Authenticating the cache took the reporting team months. Study codes were checked against trial registries; internal email headers were verified; and the reclassified case reports were matched, line by line, against the versions that surfaced in national submissions. Independent clinical pharmacologists reviewed the key documents for the IAOIJ and agreed that the data contained a signal a diligent reviewer should have escalated. Nothing in the files suggests the drug is ineffective — only that its safety picture was managed.

What the trials showed

Aldermann's own safety study, coded AP-204, is where the signal first appears. In the raw trial data reviewed by the IAOIJ, roughly one participant in 140 taking olodipine developed laboratory markers of drug-induced liver injury, and a smaller cluster showed episodes of cardiac arrhythmia. In the summary that reached regulators, those cases were folded into a much larger pooled-safety table and described as "consistent with background population rates" — a phrase that, the internal email shows, was drafted by the company's regulatory-affairs team over the objection of two of its own reviewers.

The reclassification went further in individual cases. One trial participant who died during the study was recorded by the site investigator as a "possible" treatment-related cardiac event; by the time the case reached the submission, it had been reclassified as "unrelated" and attributed to pre-existing disease. The documents show no new medical evidence between those two versions — only a change of category, and of hands.

A dossier copied across borders

What turned a company's problem into a global one is the way modern medicines are registered. Rather than run fresh trials in every market, a manufacturer assembles one master dossier and files near-identical copies to national regulators, often through "reliance" pathways in which a smaller agency accepts the assessment of a larger reference regulator or of the World Health Organization's prequalification scheme. Efficient in theory, the system means a single flawed safety section can propagate, unaltered, into dozens of countries at once.

That is what the IAOIJ traced. The same summary of AP-204 — with the same softened language and the same reclassified death — appears in submissions the team obtained or reconstructed across 40 countries, from large European agencies to under-resourced regulators in Africa, South-East Asia and Latin America that approved olodipine largely on the strength of the reference decision. A contract research organisation that helped compile the master file appears in the correspondence of at least eleven of those submissions.

The reliance trap

Reliance was designed as a public good. It lets a health ministry with few specialist assessors approve a life-saving medicine without repeating million-dollar trials, leaning instead on the judgement of a well-resourced reference agency. But it assumes the reference file is complete and honest. The Pharma Papers show what happens when it is not: the same blind spot is inherited, uncorrected, by every regulator downstream — and the countries with the least capacity to catch it are precisely the ones that rely on it most.

Regulators in the dark

Few of the regulators involved ever saw the raw data. Under reliance and mutual-recognition frameworks, most received only the company's summaries and a reference decision to lean on. Several assessors did ask sharper questions: correspondence obtained by the IAOIJ shows at least four national agencies querying the liver-injury numbers. In each case they were sent the same pooled table and a reassurance, and in each case the query was closed. None had the raw case reports that would have let them check.

We were told the medicine was safe. The regulators trusted a file, and the file trusted the company — and somewhere in that chain, the patients disappeared. — Dr Naledi Khumalo, who leads a coalition of patient-safety groups in 14 countries

The patients

Olodipine is dispensed more than nine million times a year across the markets the IAOIJ examined. Establishing individual harm from a single drug is notoriously hard, and the reporting does not claim a body count. But in national adverse-event registries the team found hundreds of reports of unexplained liver injury and arrhythmia in olodipine patients — reports that, had the original signal been flagged, might have prompted earlier monitoring, clearer warnings on the label, or a simple blood test at the start of treatment. Among them is a 54-year-old teacher in one partner country whose liver failure was attributed to "unknown causes" a year before her family learned the drug carried a signal its maker had seen in trials.

Because olodipine is prescribed for a lifelong condition, patients rarely connect a new symptom to a tablet they have taken for years. That is exactly why an early safety signal matters: it is the difference between a doctor ordering a routine liver test and one searching, months later, for a cause that is never found. Clinicians the IAOIJ spoke to in six countries said they had never been told the drug carried a documented signal; several said a single line on the label would have changed how closely they monitored their patients.

Patient advocates say the deeper harm is to trust. Generics are the backbone of affordable healthcare, and the vast majority are safe and rigorously made. The danger of a case like this is that it hands ammunition to those who would discredit cheap medicines wholesale — when the real fault lies not in the chemistry but in the paperwork, and in the incentives that reward a flattering file.

Response

Aldermann Pharma, in a detailed statement, "categorically rejects any suggestion that patient safety was compromised", says AP-204 met all applicable standards, and attributes the reclassifications to routine clinical judgement. It declined to make its chief medical officer available. Since the first stories, two national regulators have suspended new approvals of the product pending review, three have ordered stronger label warnings, and the WHO has said it is re-examining the drug's prequalification. The IAOIJ is publishing the reclassified case reports, with patient identifiers removed, so that regulators and clinicians can judge the evidence for themselves.

Regulators, too, are now examining themselves. Two agencies told the IAOIJ they are reviewing how reliance decisions are checked, and whether assessors should have access to raw case reports rather than company summaries by default. Patient groups in nine countries have called for an independent audit of every submission that drew on AP-204. What began as one company's flattering file has become a test of whether the global system for approving affordable medicines can catch a problem hiding in plain sight.