By repeatedly comparing a (new) treatment with a placebo or standard care in a series of individual patients, the added value of that treatment can be clarified. Statistician Dr. ir. Joanna in ’t Hout, Team Leader of Innovative Trial Designs for Rare Diseases at Radboudumc in Nijmegen, explains why this approach can be particularly meaningful for rare diseases and drug repurposing.

“An N=1 trial always has an experimental design,” says in ’t Hout. “It is not an observational study, but an attempt to demonstrate causality. You want to be as certain as possible that any observed change is the result of the treatment being administered, rather than due to coincidental external factors such as seasonal variation.”

Individual

In its simplest form, an N=1 trial is a way to explore, within the context of patient care, whether a treatment provides added value for an individual patient. The treatment is compared with a placebo or standard care in a double-blind crossover design. The patient receives one treatment for a given period and subsequently the alternative for an equivalent period, potentially separated by a washout phase to ensure that all effects of the first treatment have dissipated. Outcomes are measured at predefined time points. The administered treatment and control appear identical in both periods, ensuring that neither the clinician, the patient, nor their environment knows which intervention is the new treatment (double-blind). Through the crossover design, the patient effectively serves as their own control.

This exploratory use of an N=1 trial in an individual patient is clinically motivated. It can be applied in rare diseases, but also in more common conditions. For example, the approach is sometimes used to assess whether methylphenidate provides added value over placebo in individuals with suspected attention deficit disorder. In some cases, a single crossover is sufficient; in others, multiple measurement periods (“blocks”) are required to increase certainty regarding causality.

Series

Beyond its exploratory use in individual patients, a series of N=1 trials can be employed to investigate the effectiveness of a treatment in rare diseases. In ’t Hout explains: “This is an area where we, from the Nijmegen Therapy Accelerator for Rare Diseases, have both practical experience and methodological involvement. In rare diseases, there are usually too few patients to conduct a classical clinical trial with two separate treatment groups, such as a randomised controlled trial (RCT). Moreover, there is often substantial heterogeneity in disease presentation, making it even more difficult to draw conclusions from comparisons between treated and control groups. A series of N=1 trials can address this. Because each patient serves as their own control, it becomes possible to make statements about treatment effectiveness despite individual variability, particularly when multiple crossover blocks are used. Depending on the expected effect size and other variables, we can estimate how many patients and how many crossover blocks are required to achieve a sufficient level of certainty.”

Unlike an individual N=1 experiment, a study consisting of a series of N=1 trials must be reviewed by a medical ethics committee.

N=1 series can play an important role in validating new treatment approaches for rare diseases. Positive outcomes may serve as a stepping stone towards further drug development in a commercial setting, or towards the development or repurposing of treatments within academia. In ’t Hout notes: “We are increasingly seeing N=1 series being used as supportive evidence. To my knowledge, there are no examples yet of market authorisation based solely on a series of N=1 studies, but the regulatory landscape is evolving, particularly in the field of ultra-rare diseases, and there is growing willingness to consider evidence from innovative study designs.”

Conditions

Not every condition or treatment is suitable for this approach. In ’t Hout explains: “An N=1 trial or series is most appropriate for treatments with a relatively rapid onset of effect. If it takes months before any effect becomes apparent, the study quickly becomes impractically long. You also want the effect to diminish within a reasonable time after discontinuation of treatment. Irreversible treatments, such as personalised gene therapies (N=1 therapies), cannot be studied in an N=1 crossover trial. They may sometimes be applied and evaluated in an N=1 treatment setting, but that is a different context.” A third requirement is that the disease should be relatively stable. While some degree of fluctuation is inherent to biological processes, rapidly progressive diseases are generally not well suited for N=1 trials.

The interpretation of results from an N=1 series also depends on the magnitude of the observed effect. In ’t Hout: “Sometimes the added value of a treatment is very clear and consistent across nearly all patients. In other cases, there is more variability, and results must be interpreted within the so-called totality of evidence. This means considering not only clinical outcomes, but also mechanistic evidence, such as whether the treatment reaches its intended site of action. Especially in rare diseases, it is often necessary to take this broader perspective in order to draw sufficiently robust and generalisable conclusions about causality and effectiveness.”

Interested in N=1 trial design and keen to discuss further? Or are you applying an N=1 series in a regulatory context yourself? Feel free to contact Joanna and her colleagues at info@rare-nl.nl.