Clinical Considerations for First-in-Human (FIH) and Early Phase Clinical Trials
You are coming close to having completed your nonclinical studies and you are beginning to think about your next steps. But exactly what are those? Chances are you are moving into unknown territory with your asset, and you may even have a crucial milestone to consider. How do you navigate your way through the regulatory maze and are there lessons you can learn from those who went before you? In this paper we will be sharing some tips and recommendations that might help you circumvent a hurdle or save you from making a costly and time-consuming error.
A good starting point is the ‘Guideline on Strategies to Identify and Mitigate Risks for First-in-Human and Early Clinical Trials with Investigational Medicinal Products’. With this guidance document, one of the key things to remember is that it is not a recipe book and does not have to be followed to the letter, it is possible to have flexibility. Read it but do be aware that there may be elements of your planned trial to which it may not apply, and that’s absolutely fine.
There are also lots of places where you can get guidance if you are unsure of your plan, such as the MHRA Helpline or website in the UK. The MHRA Clinical Trials Helpline has both a phone (+44 (0) 20 3080 6456) and email address, which connects you not only with Assessors who are very experienced in the conduct of all types of clinical trials, it also gives you access to Good Clinical Practice (GCP) inspectors. There are also the more formal advice routes throughout Europe and US including regulatory advice meetings, more formal scientific advice meetings and the innovation office at the MHRA. The EMA has an innovation taskforce where you can arrange meetings as well. So, if something that you’re doing isn’t covered by the guideline, or you haven’t been able to find anything else published, it is important that you do not try and work it out yourself. The regulators, and other people are all ready and willing to help. If you are in Europe there is also help for novel trials via the Recommendation Paper, Initiation and Conduct of Complex Clinical Trials, but more on novel trial design later.
The FDA has recently published guidance on expansion cohorts in FIH oncology trials (March 2022) which covers a lot of what is already done in the UK and EU, but which is not necessarily covered in the EU guidance. Of course, in the US, evaluation in a different indication would often mean a new IND, an aspect which is included in this guidance. The FDA has an additional guideline on the starting dose in Healthy Volunteers, which also includes some good principles to consider. It is worth noting, however, that the FDA generally like to follow their own guidance in preference to the EU guideline and there is more focus on the NOAEL for the starting dose.
All of these documents can be used together, and it is definitely a good idea to consult all of them while you design your trial, regardless of whether you intend to conduct your studies in the EU, UK, or the US. You will not go far wrong in following all of them.
Common Issues Documents
Make sure that you review the MHRA ‘Common Issues’ documents on the MHRA website! They are not specific for First in Human (FIH) trials, but it is well worth having a look at them before you start completing your application as they cover multiple areas from validation, nonclinical, clinical pharmaceutical and some other resources as well. The information is there, so the same mistakes are easily avoided.
Don’t make assumptions and be consistent!
It is important that you don’t make any assumptions in the documentation that you submit for your clinical trial. The assessors have a specific legislative timeline for their review, and so it is important that you do not assume anything, include all information from the basic level upwards. Ask yourself questions e.g., if you were to get a question back from the regulator asking for a more robust dose justification, what would you include? Is all of that information available and already included in the documentation? If it is not, then include it from the start.
Be aware that the rules of engagement are very different for a clinical trial review compared to a license application further down the development pathway. There are different assessment teams for these reviews in the UK and most European countries. The dedicated teams of clinical trials reviewers use the Clinical Trial Authorisation (CTA) Application documents as a key part of the review process. As you might imagine, this is about safety, risk versus benefit and ensuring that the design protects the scientific integrity of the data that you’re trying to generate.
There are certain things that the MHRA CTU assessors will not do: for example, they will not optimise protocols so if they see a design that will deliver the data that you want but not in the most streamlined way, they will not comment on that or ask any questions. They will not advise on formulation development; if you decided to go with an intravenous route, but actually an oral route could be added in for example, they will not advise on that at the time of your application. And another key aspect is that they do not review the documents in the context of the complete clinical development plan. Each individual trial is reviewed in its own right. So if, for example, you have chosen a particular set of end points for a FIH trial, or one of your early phase trials, but those are not necessarily the right kind of endpoints that will get you through to a Phase II or into Phase III they will not automatically comment on that, unless there is something that is inherently unsafe or that will affect the integrity of the data.
It is also worth bearing in mind that whilst the assessors are scientific, they are not experts in every field. They are certainly not as expert as you are on the trial that you are wanting to conduct, particularly on areas such as the mechanism of action. Once again, it comes back to not making any assumptions, make sure all of the information is there in your documentation, with lots of justification. Use diagrams if they would help to clarify any elements, particularly the trial design.
It is also important to ensure consistency. Grounds for non-acceptance simply because the information is not consistent, either between documents, or even worse, within the same document, is inexcusable and wastes valuable time for everyone.
Learn from your peers
Sharing information and learning from each other is critical. It is not uncommon to see an application from a company for two similar trials where it is evident there has been no internal communication. There might have been previous grounds for non-acceptance, and then exactly the same mistakes are made again, wasting time, money, and effort. So do take the time to learn and share when you have had feedback, particularly regulatory feedback.
Risk vs Benefit
As has been mentioned before, the assessors are constantly asking the question ’Do the data supplied support the use of this product administered in this way, at the proposed dose, for the proposed duration, to this type of participant?’ And by type of participants, they mean healthy volunteers or patients, or whichever type of participant is in your inclusion and exclusion criteria. We know that there are risks associated with all trials, but it is the degree of acceptable risk that is important in determining the risk verses benefit outcome. And that depends on a whole number of factors.
The population is just one of those factors. It should be easy to see how a healthy volunteer trial will have a very different risk-benefit profile to for example, a Phase III cancer study or even another FIH trial in patients. It is always a good idea to try to be aware of what data might be missing and those known unknowns, and unknown unknowns. What don’t you know? It is all right not to know absolutely everything, but it is about awareness and how could you mitigate for those areas that are missing Information. Ensure you are constantly re-evaluating the risk-benefit and making sure that nothing has changed, including whilst the trial is ongoing.
There is a requirement to do a risk-benefit analysis for every CTA application. And that should be there front and centre in the documentation. Far too often it’s missing or hidden when it should be the right at the forefront of your protocol.
Mechanism of Action
Some other factors about the risk-benefit includes the mechanism of the drug that you are administering. Understanding the mechanism will underpin much of how you design the protocol. Some events will almost always be drug related if you see them, such as anaphylaxis or Stevens-Johnson Syndrome. But you should constantly ask yourself questions on the multiple possible signaling pathways involved. How do they interact with each other? Immune-related aspects play a critical role, particularly if there are any kind of amplification pathways as part of the mechanism. Is there a cascade mechanism involved? What kind of cytokines are involved? Are they going up or down?
Trying to think through all of these areas ahead of your trial application is very important. And if you do come up with a possible risk, then it is all about what can be monitored. Have another look at the nonclinical data and go back to the information that you had in animal models. Are all the signals there? Are they reversible or not? Do you need to look at it from a slightly different angle? Perhaps by looking at biomarkers or surrogate endpoints, for example, so that whilst they are not directly evaluating a clinical parameter, they can still be providing a monitoring system and protect the safety of the participants in the trial. For FIH trials you very much rely on nonclinical data. And the mechanism of action can tell you a lot about what you want to be doing in terms of dosing, safety, and mitigation in your clinical trial protocol.
Once again it comes back to asking yourself why: why have we seen raised liver enzymes, why is the PK profile not quite as we predicted? Why do we need sentinel dosing (where you are dosing a number of subjects ahead of the remainder of a cohort) why do we need it? Do we really need it? Sentinel dosing in particular is generally used to identify acute toxicity and that includes thinking about the timing. How long to wait before dosing other subjects and what are the criteria that will stop you from dosing the other subjects in your trial?
In fact, the stopping rules are a crucial aspect of your study design, as was tragically demonstrated in the BIA 10-2474 trial in France in 2016. The death occurred in the multiple ascending dose part of the trial and the stopping criteria do not appear to have been clear enough to stop the dosing when a safety issue was observed. You need clear stopping rules not just for the whole trial, or cohorts, but for multiple areas. There should be stopping rules for an individual subject, after either one or more sentinel subjects, for a cohort, for dose escalation, for previous dose levels and cohorts (if a safety signal is seen), and of course, for the whole study. They may indeed all be the same, you don’t have to have different stopping rules for each of these elements, but if that is the case, it is really important that it is clear in the protocol.
Novel trial design
Regarding novel trial designs it is important to consider what is needed to balance the risk/ benefits with participant safety whilst also obtaining the data you need from the trial in as efficient a manner as possible.
A good example of this is the COV001 trial, which was the FIH trial for the University of Oxford ChAdOx nCoV-19 vaccine (later known as the Astra Zeneca Vaxzevria vaccine) for COVID 19, the protocol for which is in the public domain. They proposed the objectives and the end points and then determined the schedule that they needed to work to, considering the time pressure that they were under to deliver a vaccine. Then they discussed it with the MHRA and worked on how to balance all of those aspects. The actual trial design included multiple overlapping groups. Some parts had sentinel dosing, but not all of them and then there were a series of amendments. They added cohorts later on, based on some of the safety issues that they thought they might potentially have, for example, giving pre-dose paracetamol, both to overcome some of the side effects and ensure that there wouldn’t be any interaction. The protocol itself was very clearly written. There were very clear safety reviews in terms of the timing, the data that would be reviewed, very clear dosing decisions, and justifications for why they had designed the trial in the way that they did.
There is also another paper that is helpful to review when considering novel trial designs. Such trials come under many different names and this particular paper refers to complex innovative designs. ‘Effective Delivery of Complex Innovative Design (CID) Cancer Trials—A Consensus Statement is aimed at oncology trials which are not necessarily FIH but nevertheless the recommendations and consensus statements could be very useful for any first in human or early-stage trial.
Figure 1 (Clin Transl Sci (2019) 12, 6–19; doi: 10.1111/cts.12582) highlights what can be done in a FIH trial. Traditionally, there would be a single ascending dose followed by multiple ascending dose. You might have had some expansion cohorts, but what else can you learn from your FIH trial?
Figure 1: Example study schema for a first‐in‐human trial with multiple objectives
For example, can you add other populations or different age groups? If you’ve included healthy volunteers, can you go into patients? There are all sorts of different populations where you can look into drug interactions so is it possible to include them in the first in human trial? You might also need to consider challenge agents, particularly in infectious disease. Food effect is also a common aspect of FIH trials, and it’s always worth trying to think about whether you can evaluate multiple different formulations.
Increasingly a lot of trial designs are not simply FIH, or even Phase I, they are including some early Phase II elements as well, such as some early efficacy by including cohorts of a patient population.
Of course, the FIH population is a critical question. Do you include healthy volunteers or patients? There are lots of different questions that you can ask yourself to try and reach the right decision. Again, you should go back to the data that you have seen in animals, the severity of any toxicity, and any other safety events and the overall safety profile; are those safety signals or adverse events that you saw reversible or not? Can they be monitored and/or can you treat them if you need to? What about off-target effects? Truly understanding the safety profile is as important as simply having a safety profile as a list of adverse events. And also critical is whether the target that you saw in your animal model, in terms of the mechanism of action, is even expressed in healthy volunteers. Quite often, it’s not.
For healthy volunteer populations, the eligibility criteria would usually be pretty strict, but you could also consider including both healthy and patient populations in a trial and it is even possible to switch mid dose escalation or in between cohorts. A good example would be a protocol for a drug that affects coagulation. The protocol could combine a single ascending dose phase which includes healthy volunteers for the first few cohorts but as soon as a certain coagulation signal is reached, or at a particular predefined dose level, the study switches to patients in order to protect the volunteers from significant thrombotic effects. This also demonstrates another critical factor; you need to consider the toxicity as the trial progresses. Do you need to change anything as the trial moves forward?
For the starting dose, it is all about the point when your drug becomes pharmacologically active. There is an expectation that in healthy volunteers you should start well below the pharmacologically active dose, and you need to have a justified safety margin. In patients it can be a little bit different, it depends on the patient population, but generally there is an expectation that you would start either at the lower limits of the pharmacologically active dose (especially in oncology trials), or at least that you escalate quickly up to that level. Unless, of course, there are very clear nonclinical safety concerns which means the more cautious approach in any scenario is warranted.
Another crucial consideration is about how high the dose should go and this is an area that is very often missed or very poorly outlined in protocols, particularly in healthy volunteers. Around 2 to 3 times above the predicted clinical efficacy range is about as far as you should escalate based on the nonclinical data. To consider dosing much above that really needs a robust justification. Going high in the dosing just because you can, is not an option, there needs to be a reason for it. Something like consideration for multiple potential future clinical programs could be an acceptable way forward, but it comes down to safety. Remember it is always possible to have an amendment in the FIH trial to go higher if you need to later or to enable higher doses in patients in subsequent trials. You still need to be aware of where those top doses sit in terms of the safety margin. And it can also come back to those novel designs, try thinking about it differently. As long as it is safe, is clearly outlined in the protocol and is within well-defined limits, then you should be fine.
What about those intermediate doses? It’s all very well having a starting dose and a top dose, but how are you going to get from one to the other? Defining the dose increments between those doses or having absolute doses stated in between is absolutely fine. But don’t miss that critical stage out, be as transparent as possible for justifying all aspects of the dosing, not just the actual dose levels, but also the regimen that you’re using and the route of administration.
Remember that the dose escalation decisions also need to be very clearly outlined in the protocol as well. In particular, the amount of data that is going to be reviewed at each of the dose escalation decision points needs to be very clear. That is in terms of the number of subjects, how many are active or how many are on placebo, how many out of a cohort, and whether you’re wanting to preserve any blinding etc. Another area that is often missed out, or not done very well, is the amount of data for each participant that will be reviewed in terms of how many doses, how many days, how long after dosing, with consideration for the safety, pharmacodynamic and PK data.
The FIH guideline encourages a cumulative review which again is something that is often missed out. For example, if you are dose escalating between cohorts 3 and 4 it is about not just looking at the data that is emerging from cohort 3, it is about looking back at what you have seen for cohorts 1 and 2 and looking for patterns or trends. Sometimes you can get far more information from those cumulative reviews than you can from spot reviews when it comes to a small number of participants.
Another area that a lot of sponsors often don’t like to include is reviewing the pharmacokinetic data to allow for dose escalation. There might be various reasons for that. It could be that there are complex assays, and that they take a very long time to get the data, however that on its own, would not necessarily be acceptable to a regulator since the PK data can be very important. That data can feed back into modeling; you can update your models, get information quickly, and it provides valuable information about how your drug is behaving. It can be also useful for fully evaluating the safety profile. Putting all that together with the PK profile provides you with a much more rounded picture of your drug. The assays that you’re using, and their lower limits of quantification is something that a regulator would be happy to see as a justification. However, it is also about making sure that you are setting all of your doses correctly to make sure that you can take the PK data into account in those dose escalation decisions.
We have become familiar with FIH trials having cohorts of 6 on active, plus 2 on placebo, but more often people are questioning that design and are trying to come up with something different. It comes back to that risk-benefit evaluation and a justification. You don’t have to stick with 6 plus 2, but where did that come from? The rationale is shown in Figure 2 (Clin Transl Sci (2019) 12, 6–19; doi: 10.1111/cts.12582)
Figure 2: Detectable adverse event rate as a function of active cohort size and power
For an event that would occur in about 25% of subjects receiving a new therapy, one has an 80% probability of observing the event at least one time in a cohort of six subjects. These cohorts sizes are not going to pick up the very rare events but in general, a cohort size of six will pick up some of the slightly more common adverse events that you would want to be catching at this stage of development.
Another area to consider is blinding. Again, we are used to a 6 plus 2 design with some kind of blinding, but there are lots more publications and questions about whether you do actually need placebo subjects in FIH trials, particularly given that they are not providing critical data. And even though you are picking up some adverse events, you’re not picking up some of the rarer ones. The inclusion of placebo clearly increases the complexity of a study, it requires additional procedures, documentation, and there are the additional manufacturing considerations. You could argue that having pre-dose or time matched baseline measurements could be used in place of placebo subjects. We may be some way off scrapping the use of placebo cohorts completely but just be aware that you can consider alternative options, provided it is safe and has been well thought through and justified.
Related to adverse events, another common ground for non-acceptance (GNA) surrounds the use of the CTCAE particularly in healthy volunteer trials. It is very easy to have a false sense of security when you are using the grading scale, but why do regulators often pushback on the use of CTCEA in healthy volunteers? Liver function is a good example of why this might happen. For healthy volunteers, you would expect them to have normal liver function to be able to enter into the trial but then if you’re using CTCAE to grade adverse events and determine stopping of dosing, Grade 2 liver dysfunction is actually 3 to 5 times the upper limit of normal. Grade 3 is up to 20 times the upper limit of normal for the liver enzymes which is clearly not acceptable. Evidently, this is a very different scenario to using that kind of scale in an oncology trial. Therefore, if you are going to use it, make sure that you justify it and be aware that there are alternatives. For example, there is an FDA grading scale for vaccine trials, and whilst it is stated as being useful for vaccine trials, it covers lots of laboratory and other adverse events. So it can be very helpful and could certainly be used for the healthy volunteer or FIH trials, it doesn’t have to be restricted to vaccines.
The application process
In the UK, the MHRA and the Research Ethics Committee (REC) review is now a combined process, all clinical trials now go through a single portal, with all documentation for both the MHRA and REC submitted at the same time. There is also a parallel timeline; the previous 30 days and then 60 days for the review time points but now in one streamlined process.
The Commission on Human Medicines (CHM) is an expert advisory group for higher risk first in human studies which may require additional external independent advice. Typically, this not something that the sponsor needs to be too concerned about as the assessors will perform those reviews with the experts themselves, but there is some additional documentation required in the clinical trial authorisation application. If you are not sure whether your trial would qualify as a higher risk trial, then look at the MHRA website where you will find all of the information as to what counts as a high-risk trial.
Another topic that now often comes up when you’re talking about clinical trials is about how the UK now links to the European Union. In the EU, clinical trials have transitioned to the Clinical Trials Regulation that came into force in 2014 and this became applicable from 1st January 2022. Do be aware that from the end of January 2023, there is no option but to use the new Clinical Trials Regulation. Any new applications will have to be done under the regulation and you will require access to the portal. There is extensive training and support online. The EMA has lots of Q&A documents and video training, including practice sessions using the system.
Multiple trials have gone through that process already, but as the UK is no longer in the EU they will not be implementing the Clinical Trials Regulation. Although there are aims to align wherever possible, since it is not in anyone’s interests for the UK and the EU to divert from each other in terms of how they approach clinical trials, in the UK they are still using the statutory instrument under the medicines legislation.
It is also important to note that for Marketing Authorisations Northern Ireland is currently part of the EU but for the purposes of clinical trials it is not, and it remains under the remit of the MHRA, so if you put in your Clinical Trial Authorisation application it will automatically be valid in England, Wales, Scotland and Northern Ireland.
What is the future of clinical trials in the UK?
There was a public consultation in 2022, providing the opportunity to improve and strengthen the UK clinical trials legislation, to help make the UK the best place to research and develop safe and innovative medicines. There are five key domains to the proposals.
- Making sure that there are additional transparency provisions and patients are provided with information when they should be, and about involving those with lived experience throughout a trial in supporting diversity in clinical trials. And that absolutely applies to FIH trials as well.
- There’s a section about being more streamlined with competitive timelines and proportionate safety reporting.
- The whole proposal is about being more risk proportionate, and about expanding and continuing the previous notification scheme for those lower risk trials.
- Overall, it is about being agile in the UK, being flexible and pragmatic wherever we can be. Trying to use guidance wherever possible, rather than hard and fast legislation, where there cannot be any flexibility.
- There is also a focus on international partnerships and maintaining international standards, making sure that aspects, such as definitions, are harmonised not just between the UK and the EU, but on a global scale.
There is lots to think about when planning a FIH or early phase trial, but hopefully we have been able to highlight some of the key areas to think about to streamline the process for you. Consideration for safety and the risk-benefit is critical, and ensuring the areas outlined here have been covered in the protocol can only make for a more positive regulatory review and a better outcome for the development programme.
 Note: Following the Windsor Protocol agreement in March 2023 this is due to change: https://commission.europa.eu/system/files/2023-02/COM_2023_122_1_EN_ACT_part1_v2.pdf
Adapted from a webinar presentation by Dr Kirsty Wydenbach Head of Regulatory Strategy, Weatherden