Learning from tragedy: safety and dosing in first-in-human trials

The Rennes clinical trial disaster in January last year highlights the dangers posed by first-in-human trials of high-risk drug candidates, with dosing strategies coming under the spotlight. What makes early-stage trials go wrong, and how can dosing protocols be improved to minimise health risks to trial participants?


Phase I first-in-human clinical trials, which test the safety profile of new experimental drugs on human volunteers for the first time, are an invaluable part of the drug development process. Data from these trials helps to reveal the pharmacokinetics (PK: how the body affects the drug) and pharmacodynamics (PD: how the drug affects the body) of a new treatment, informing the recommended dosage as the drug moves through the clinical trial process.

Despite the risks involved in testing new drugs in humans for the first time, for the most part these studies have an excellent safety record. According to the Association of the British Pharmaceutical Industry (ABPI) in its Phase I clinical trial guidelines, the overall incidence of serious adverse events related to investigational medicinal products is around 0.02%. A complex web of regulations and protocols guiding the preparation of pre-clinical data, risk assessment, dosing decisions and adverse events is in place to minimise the risk of harm to human volunteers.

First-in-human: laying out the dosage risks

Nevertheless, the rarity of such events means that the clinical research community’s attention is quickly drawn to cases where something does go awry. This is especially true when an early-stage trial is so disastrous that patients either die or are severely incapacitated as a result of the drugs that have been administered. These tragic cases are shocking enough to shake our faith in the safety standards governing first-in-human trials, and they tend to prompt detailed and long-running investigations into what went wrong.

One of the biggest sources of risk in early-stage human trials is working out the optimal safe starting dose of an experimental drug, and then planning the schedule for gradually increasing the dose to study its tolerability. If the starting dose is too high, the risk of unexpected adverse reactions rises, while escalating the dose too rapidly narrows the window of opportunity to identify emerging side effects before larger doses are administered to more volunteers.

"Assessing the risks in a Phase I study in which the risk pro­file is not well-established can be challenging."

“Assessing the risks in a Phase I study in which the risk pro­file is not well-established can be challenging,” wrote PPD senior regulatory affairs specialist Kirsten Messmer and Schulman IRB director of operations Julie Blasingim in an analysis of Phase I safety issues, published in the February 2017 edition of regulatory affairs association TOPRA’s journal, Regulatory Rapporteur.

“Reviews of Phase I research should consider study design, starting dose, rate of dose escalation, time interval between dose escalations, data assessable before dose escalation decisions, number of subjects dosed simultaneously, standard assessments of subject eligibility criteria, and monitoring provisions,” the analysis added.

This risk is especially pronounced during first-in-human trials of therapeutic agents with a higher potential for causing harm during first human exposure, such as biological molecules with novel mechanisms of action, new agents targeting the immune system and new agents with a high degree of species-specificity.

TGN1412: a 21st century turning point

Several high-profile incidents in the last decade have proven the difficulty involved in minimising dosing risk in first-in-human trials, and the shocking consequences that can follow when an unforeseen complication rears its head. In 2006 at London’s Northwick Park Hospital, a first-in-human trial of TeGenero Immuno Therapeutics’ TGN1412 – a monoclonal antibody targeting immunological diseases such as rheumatoid arthritis and leukaemia –caused devastating cytokine storms and systemic organ failure in all six of the first volunteers to receive the drug.

All of the affected volunteers survived, but not without life-changing long-term health consequences, including devastated immune systems that have left them highly susceptible to cancer and auto-immune diseases, and one subject who lost his fingers and toes as a result of the TGN1412 infusion.

Subsequent investigations into the disaster by the UK’s Medicines and Healthcare Products Regulatory Agency (MHRA) and an Expert Scientific Group (ESG) found that TeGenero and trial manager Parexel had broadly kept to trial and pre-clinical data protocols, and that the catastrophic adverse reactions were caused by an unpredicted biological action of the drug in humans. The candidate had been tested on non-human primates (macaques) without reaction and the initial starting dose for the trial was 500 times lower than a dose found to be safe in animals.   

The main controversy over the trial was its dosing regimen, as according to one participant who received a placebo, the six volunteers who received the drug were all infused within just 20 minutes, allowing no time to study its immediate effects and limit the damage.

The recommendations compiled by the ESG after the Northwick Park incident included some prominent points on dosing, noting that doses should be “administered sequentially to subjects with an appropriate period of observation between dosing”, that starting doses for high-risk drugs should “be calculated to err on the side of caution”, and that “careful consideration should be given to the route and the rate of administration of the first doses in first-in-man trials, with careful monitoring for an adverse or exaggerated response”.   

Rennes trial disaster: what went wrong?

More recently, a first-in-human clinical trial run by French contract research organisation Biotrial in January last year at a facility in Rennes claimed the life of one participant and hospitalised five others. The trial was assessing the safety of Portuguese pharma company Bial’s experimental fatty acid amide hydrolase inhibitor (BIA 10-2474), intended for the treatment of a range of conditions, from Parkinson’s disease to anxiety disorders and chronic pain.

Questions have been raised by scientific peers about the design of the trial since the incident. Again, dose escalation was at the forefront of concerns. At a December conference in London, hosted by the British Pharmacological Society (BPS), Bial and Biotrial were criticised after it emerged that researchers had not consulted pharmacodynamic data before increasing the dosage of the drug in volunteers, which is what caused the reaction in the lead subjects. Without PD data, the critics speculated, the companies would have been unable to tell when the copies of the enzyme in the brain were saturated, which is the point when increased doses can have unpredictable effects on other enzymes.

"Bial and Biotrial were criticised after it emerged that researchers had not consulted pharmacodynamic data before increasing the dosage of the drug."

“That was a revelation as far as we are concerned,” BPS president Professor David Webb told the Nature journal’s website in December. “Without the PD data, they were flying blind. That’s when accidents happen. I think that was negligent.”

The Bial trial bears more resemblances to the TeGenero case, in that the company involved had adhered to protocol and the trial had been approved by the medical regulator – in this case France’s National Agency for the Safety of Medicines and Health Products (ANSM). Bial was not legally obliged to incorporate PD data into its decision-making (although it is considered a best practice), and the company argues that no pre-clinical data could have explained the outcome.

“Towards the report conclusion regarding the escalating doses, including the passage from 20mg to 50mg, BIAL notes that the safety and tolerability profile of BIA 10-2474 was favourable up to 20mg,” the company said in a press statement in May. “There were no alerts or signals in any of the safety parameters collected from any of the previous cohorts that could have anticipated the tragic accident.”

Dosing and risk management

These recent incidents highlight the worst-case scenario risks involved in first-in-human trials. The Rennes tragedy, as with the Northwick Park incident, has prompted six new recommendations for future clinical trials, along with updated guidelines from the European Medicines Agency (EMA). All guidance stipulates extreme caution in the study of pre-clinical data, dose-ranging studies and decisions on initial dosing and dose escalation.

Moving forward, alternative initial dosing techniques could help reduce the risks of first-in-human trials. The no adverse effect level (NOAEL), which is often used for measuring safe initial human doses, is increasingly being recognised as sub-optimal for high-risk investigational drugs, and agencies have instead been recommending use of the minimum anticipated biological effect level (MABEL) for such drugs.

Micro-dosing, meanwhile, is a Phase 0 trial technique, by which subjects are exposed to tiny, non-pharmacologically active doses of a drug, providing an extra layer of safety as it allows for a drug’s pharmacokinetic profile to be defined with minimal risk to the subject. A 2008 study published in the Indian Journal of Pharmacology described micro-dosing as “a new viable concept in the ‘toolbox’ of the drug development activity” with potential to “complement standard animal-to-human scaling, redefining the existing concept of Phase I clinical research”.

It’s unlikely that any system of protocols and regulations could absolutely guarantee subject safety in first-in-human clinical trials, especially for biologics and other drugs that bear a heightened risk of unexpected reactions. Nevertheless, it is vital that best-in-class safety standards are mandated and adhered to for high-risk trials to avoid a repeat of the tragedies in London and Rennes.