N of 1 Clinical Trial Design and its Use in Rare Disease Studies

In the beginning (February) of this year, I attended a workshop titled “Clinical Trial Design for Alpha-1 Deficiency: A Model for Rare Diseases”. During the meeting, the N of 1 design was mentioned as one of the study methods to address the challenges in clinical trials in rare disease areas.

This was echoed in FDA’s “Public Workshop – Complex Issues in Developing Drug and Biological Products for Rare Diseases”. Session 2: “Complex Issues for Trial Design: Study Design, Conduct and Analysis” had some extensive discussions about the N of 1 trial design and its potential use in rare disease clinical trials.

In a presentation by Dr. Temple in FDA titled “The Regulatory Pathway for Rare Diseases Lessons Learned from Examples of Clinical Study Designs for Small Populations”, N of 1 study design was mentioned along with other methods such as randomized withdrawal, enrichment, crossover designs.  

According to Wikipedia, “an N of 1 trial is a clinical trial in which a single patient is the entire trial, a single case study. A trial in which random allocation can be used to determine the order in which an experimental and a control intervention are given to a patient is an N of 1 randomized controlled trial. The order of experimental and control interventions can also be fixed by the researcher. “

While N of 1 is not commonly used in clinical trials, the concept of the N of 1 method with focusing on the single patient is actually pretty common in the clinical trial setting. There is some similarities between Aggregated N of 1 and the typically crossover design, especially the high order cross over design. For safety assessment in clinical trials, challenge – dechallenge – rechallenge (or CDR) is often used to assess if an event is indeed caused by the drug. CDR can be considered as an simple N of 1 design.

“Challenge–dechallenge–rechallenge (CDR) is a medical testing protocol in which a medicine or drug is administered, withdrawn, then re-administered, while being monitored for adverse events at each stage. The protocol is used when statistical testing is inappropriate due to an idiosyncratic reaction by a specific individual, or a lack of sufficient test subjects and unit of analysis is the individual. During the withdraw phase, the medication is allowed to wash out of the system in order to determine what effect the medication is having on an individual.
 CDR is one means of establishing the validity and benefits of medication in treating specific conditions as well as any adverse drug reactions. The Food and Drug Administration of the United States lists positive dechallenge reactions (an adverse event which disappears on withdrawal of the medication) as well as negative (an adverse event which continues after withdrawal), as well as positive rechallenge (symptoms re-occurring on re-administration) and negative rechallenge (failure of a symptom to re-occur after re-administration). It is one of the standard means of assessing adverse drug reactions in France.”

While N of 1 is the experiment on a single patient, using aggregated single patient  (N-of-1) trials will involve multiple patients – quantitative analyses become more feasible. See examples below for using aggregated N of 1 trials.  

• Using aggregated single patient (N-of-1) trials to determine the effectiveness of psychostimulants to reduce fatigue in advanced cancer patients: a rationale and protocol
• Aggregated n-of-1 trials of central nervous system stimulants versus placebo for paediatric traumatic brain injury – a pilot study

N of 1 clinical trials could involve some complicated statistical analyses. See the discussions below:

• the presentation of “N-of-1 Randomized Clinical Trials” on Henry Steward Talks  
• A Comparison of Four Methods for the Analysis of N-of-1 Trials
• Aggregating single patient (n-of-1) trials in populations where recruitment and retention was difficult: The case of palliative care
• N-of-1 Trials by Christopher H Schmid

N of 1 clinical trial design is rarely discussed in statistical conferences, perhaps because of the perception that not too much statistics is involved in the analysis of N of 1 study data. However, we do see that N of 1 study can be a very effective method in demonstrating the efficacy if the characteristics of the indication/drug fit.

• Feasibility Study of N-of-1 Trials With Blood Pressure Self-Monitoring in Hypertension
• N of 1 Tests in Primary Care: A New and Better Way to Prospectively Monitor Chronic Care Drugs for Safety
• Don't Overlook "n-of-1" Trials, Urges UK Expert
• Clinical Trials for Cancer, One Patient at a Time

One of the key questions is that the N of 1 study design is only applicable in certain situations – it  depends on the disease characteristics, treatment (short washout period), endpoint (quick measurements). We can see some of the discussions about the situations where the N of 1 study design may be used from the transcripts of the FDA Public Workshop on Complex Issues in Rare Disease Drug Development Complex Issues for Trial Design: Study Design, Conduct  and Analysis:

“Ellis Unger: We have no ... No comments right now, so let me put a question to the group. I  presented us a slide on the N of 1 study, which we almost never see. Just to remind you, the N of 1 study is a scenario where a patient doesn't contribute and end, but of course the treatment contributes an end, and of course the treatment can be capped in a certain number above ... weeks.
 Unless someone has the amount of interest, in which case, you give up on that course, that aborts that course to treatment and then they re-randomize. Are there therapies, disease states people around the table can think off that would be ... where this design could be applicable, because we don't see these studies. Dr. Walton?
 Marc Walton: I'll just mention that by firing away in all the clinical trials I've reviewed, the most powerful piece of evidence about the effectiveness of a drug came from a N of 1 type of study where it was a study with Pulmozyme cystic fibrosis where patients were treated Pulmozyme that are pulmonary function tested, then the Pulmozyme was discontinued and then tested again, and then several cycles, and I think it was maybe five cycles and you saw such remarkably reproducible effects that it was utterly convincing that the drug was effective for that.
The utility comes about though when you have, as you have said, a disorder that has enough stability and drugs that have a short enough washout period, that you are able to have that repeatedly look as if it was a new exposure to the patient. In disorders where we have that and treatments that are expected to have that sort of reversible effect, this N of 1 becomes a truly powerful piece of information, as well worth considering when those circumstances present themselves.
 Ellis Unger: Typically, a company will come in and say, "You randomize to our treatment or placebo. We're going to count the number of exacerbations or pain episodes or whatever over the course of the study." This is basically saying once you have one of these events, we're going to re-randomize you. Just again, so anybody around the room ... Okay, Dr. Summar.
 Marshall Summar: Yeah, it seems like from the intermediary metabolism, the effects where you have frequent attacks of hyperammonemia, acidosis, things like that, that actually might be a fairly ideal group washout for most of the treatment is pretty fast. That seems like a group where that might actually play out pretty well. I have to think about that but it seems to make some sense.
 Ellis Unger: Dr. Kakkis?
Nicole Hamblett: Thanks. I think the N of 1s, studies are incredibly intriguing and I think one thing I need to wrap my head around is the consigning by commons in medications, for instance, if you're measuring exacerbations during on and off periods in their treatment for that event could alter what's going to happen during the next events. I think that's a little bit difficult to the chronic study, but I guess I also wonder what are the parameters for being able to use an N of 1 study or N of 1 studies for your pivotal trial, as well as difficult enough to conduct confirmatory study. How would we define that for these types of newer or more customized study designs?
 Ellis Unger: Well, I think the N of 1 study, again, you have to have a treatment where there's an offset that's reasonably rapid and you're not expecting the effect on the disease to be ... the effect is not lasting. It's not like Dr. Hyde was mentioning in a gene therapy, as that would be the extreme opposite where you couldn't do this, but if you have something where there's an offset in a reasonable amount of time and patients are subjected to repeated events, I think that's the key.
 If it's progression and it happens slowly with time, you're not going to be able to do an N of 1 study, but if you have some episodic issue and you have a drug with a reasonable offset, I think it will lends itself to this and we're talking about a dozen patients to do a study, the whole deal, and that could be your phase 3 study. I mean that the example I showed was just about a dozen patients. You don't need a lot of patients. “

It is clearly that the N of 1 study design is not appropriate for a study with the efficacy endpoint measured for very long period of time. N of 1 study design may be applicable for short–term endpoints (bio-markers, metabolites, …). However, over the last 10-20 years, the direction of regulatory agencies is moving toward to the long-term endpoints. For an enzyme replacement therapy, a drug showing the increase in enzyme level would be considered sufficient for approval 20 years ago. Nowadays, an endpoint measuring the long term clinical benefit may be required. Similarly, for a thrombolytic agent, it is not sufficient to show the thrombolysis in short term, the long-term benefit of the thrombolytic agent will be required. This trend of requiring the long-term measures in efficacy endpoints make the N of 1 study design unlikely to be used in the licensure studies.

Full Analysis Set and Intention-to-Treat Population in Non-randomized Clinical Trials?

Intention to treatment principle has now been an routine term in the statistical analysis for randomized, controled clinical trials. If a publication is for a randomized, controled clinical trial, it is almost universal that the intention to treatement principle will be mentioned even though the actual analysis may not exactly follow the intention to treat principle in some studies.

Strictly speaking, the intention to treatment principle indicates that the intention to treatment population includes all randomized patients in the groups to which they were randomly assigned, regardless of their adherence with the entry criteria, regardless of the treatment they actually received, and regardless of subsequent withdrawal from treatment or deviation from the protocol. See one of my early articles and the presentation on ITT versus mITT.

According to ICH E9 “STATISTICAL PRINCIPLES FOR CLINICAL TRIALS”, Full Analysis Set (FAS) is identical to the Intentio-to-Treat (ITT) population. It states:

“The intention-to-treat (see Glossary) principle implies that the primary analysis should include all randomised subjects. Compliance with this principle would necessitate complete follow-up of all randomised subjects for study outcomes. In practice this ideal may be difficult to achieve, for reasons to be described. In this document the term 'full analysis set' is used to describe the analysis set which is as complete as possible and as close as possible to the intention-to-treat ideal of including all randomised subjects.”

Here both FAS and ITT population are tied to the randomization. However, in the real world, there are also a lot of non-randomized trials, for example, a clinical study without concurrent control, an early phase dose escalation study without a concurrent control, a long-term safety follow up study where all subjects receive the experimental medication. In these situations, since there is no randomization, it is inappropriate to define an ITT population even though the general principle should remain the same, ie, to preserve as many subjects as possible to avoid bias. The issue is that without randomization, what will be trigger point for defining the ITT population? It looks like that the trigger point could be the time of the administration of the first study medication. Instead of allocating subjects in ITT once randomized’, the subject is in ITT ‘once dosed’. This seems to be the case in the following example, according to CSL’s RIASTAP summary basis of approval, the ITT population was defined for a study without concurrent control and without randomization. It implied that the ITT population includes all subjects who received the study medication, which is essentially the same as the Safety population.

For non-randomized studies, it may be better to use Full Analysis Set instead of ITT population. It seems to be logical to define the full analysis set to include any subjects who receive any amount of the study medication. If this definition is used with the trigger point being the first dose of the study medication, most likely, the full anlaysis set and the safety population will be identical. It is not uncommon that we define two populations that is identidical, but use it for different analyses. For safety analyses, the safety population is used; for efficacy analyses, full analysis set is used.

Another term we can use in non-randomized studies is Evaluable Population which is usually defined as any subjects who receive any amount of the study medication and have at least one post-baseline efficacy measurement. Evaluable population in non-randomized clinical trials is similar to the modified ITT population in randomized clinical trials where some randomized subjects are excluded from the analysis with justifiable rationales.

While the ICH E9 did not use the term ‘modified Intention-to-Treat’, the following paragraphs are intented to provide the guidelines or examples when the subjects can be excluded from the full analysis data set or Intention to treatment population:

“There are a limited number of circumstances that might lead to excluding randomised subjects from the full analysis set including the failure to satisfy major entry criteria (eligibility violations), the failure to take at least one dose of trial medication and the lack of any data post randomisation. Such exclusions should always be justified.Subjects who fail to satisfy an entry criterion may be excluded from the analysis without the possibility of introducing bias only under the following circumstances:(i) the entry criterion was measured prior to randomisation;
(ii) the detection of the relevant eligibility violations can be made completely objectively;
(iii) all subjects receive equal scrutiny for eligibility violations; (This may be difficult to ensure in an open-label study, or even in a double-blind study if the data are unblinded prior to this scrutiny, emphasising the importance of the blind review.)
(iv) all detected violations of the particular entry criterion are excluded.
In some situations, it may be reasonable to eliminate from the set of all randomised subjects any subject who took no trial medication. The intention-to-treat principle would be preserved despite the exclusion of these patients provided, for example, that the decision of whether or not to begin treatment could not be influenced byknowledge of the assigned treatment. In other situations it may be necessary to eliminate from the set of all randomised subjects any subject without data post randomisation. No analysis is complete unless the potential biases arising from these specific exclusions, or any others, are addressed.
In some situations, it may be reasonable to eliminate from the set of all randomised subjects any subject who took no trial medication. The intention-to-treat principle would be preserved despite the exclusion of these patients provided, for example, that the decision of whether or not to begin treatment could not be influenced byknowledge of the assigned treatment. In other situations it may be necessary to eliminate from the set of all randomised subjects any subject without data post randomisation. No analysis is complete unless the potential biases arising from these specific exclusions, or any others, are addressed.
Because of the unpredictability of some problems, it may sometimes be preferable to defer detailed consideration of the manner of dealing with irregularities until the blind review of the data at the end of the trial, and, if so, this should be stated in the protocol.”

In summary, while the general principle is the same, the different terms may be preferred to be used depending on a study being a randomized or non-randomized.

Randomized studies	Non-randomized studies
ITT	Full Analysis Set
Safety	Safety
mITT	Evaluable

Interesting talks about the Intention to treatment principle:

• Analysis Issues, ITT, Post-Hoc, and Subgroups by Dr Somon Day
• Intention-To-Treat is a fraud by Dr Gerard E. Dallal