Skip to main content
Search form
  • 00:00

    [MUSIC PLAYING][Introduction to Designing Clinical Trials in the Eraof Precision Medicine]

  • 00:14

    JAMES WASON: I'm James Wason, and Iam a professor of biostatistics at Newcastle University.[James Wason, Professor of Biostatistics,Newcastle University] And also, I'm a group leader at the MRCbiostatistics unit in Cambridge.And my research interests are to dowith how to make clinical trials more efficient.I'm a statistician by background,and I'm interested in using statistical methods

  • 00:35

    JAMES WASON [continued]: to make clinical trials more efficient and betterfor patients.So in this video, we're going to be talking about precisionmedicine and how we can use clinical trialsin the area of precision medicineto give us better information about which patients benefitfrom new treatments and how we can better match which patientsshould get which treatments.

  • 00:55

    JAMES WASON [continued]: [How did you become interested in precision medicine?]I did a PhD in statistical genetics,and that involved using genetic data to understand diseaseand looking at what genetic variables influencedifferent diseases.

  • 01:16

    JAMES WASON [continued]: After my PhD, I started working in clinical trials.And although I didn't immediatelymake the connection between genetics and clinical trials,I was always interested in how wecould use statistical approaches to makeclinical trials more efficient and better for patients.And over the last few years it's beenthe big move towards getting more information

  • 01:38

    JAMES WASON [continued]: about precision medicine.Precision medicine is basically lookingbeyond whether a treatment works, on average,for all patients and looking at which types of patientsor which groups of patients maybe especially benefitfrom a new treatment and which maybe don't receive any benefitor are even harmed by it.So I've been quite interested in how statistical methods can

  • 02:02

    JAMES WASON [continued]: be used to design and analyze clinical trials betterin order to give us more information about precisionmedicine and also how we can use informationabout which patients benefit from a treatmentduring a trial to ensure that patients who are recruitedto trials have a better chance of gettinga good treatment for them.

  • 02:23

    JAMES WASON [continued]: [What are some of the key characteristics of precisionmedicine?]In the past, it's all been about doesthe treatment work, on average?So in a clinical trial, often we usea randomized controlled trial wherewe recruit a wide selection of patientsand we randomize them between an experimental treatment

  • 02:45

    JAMES WASON [continued]: that we're interested in testing and some controltreatment, which might be the best current treatment,or it might be best current treatment plus a placebo,for example.And we basically do that randomization.And we don't necessarily use informationthat we're collecting during the trialto help give us more information at the end.

  • 03:07

    JAMES WASON [continued]: So one thing that we could do is to,if we have a treatment that we imagine will work differentlyfor different types of patients, then we can includethat information in the trial.And if we're interested in perhaps testingseveral new treatments that are available at the moment,we can help during the trial match which patients receive

  • 03:31

    JAMES WASON [continued]: which treatment in order to ensurethat more patients on the trial receivea good treatment for them but also that weget more information at the end of the trial about whichtreatments work best for which groups of patients.[What are some examples of trial designs in the era of precisionmedicine?]

  • 03:53

    JAMES WASON [continued]: The way in which these trials are different--so there's been a lot of work in howwe can do better clinical trials in this era of precisionmedicine.And I'll give a few examples of different types of trialsthat I've been interested in.So one example is what's called a baskettrial, which originated mostly in oncology, in cancer.And this was where there's a new drug that's targeted

  • 04:17

    JAMES WASON [continued]: at a particular tumor mutation.So there's a lot of mutations in different tumorsthat people have.And these mutations can be present in different typesof different locations.So you could have the same tumor mutation appearing in,for example, breast cancer and colorectal cancer.

  • 04:38

    JAMES WASON [continued]: And the idea of a basket trial isthat we can test a drug which has been developedto target that mutation and to work particularly wellfor patients who have that tumor mutation, that we canincorporate not just one type of cancer, like breast cancer,but we can also include other types of cancer

  • 04:58

    JAMES WASON [continued]: where that mutation appears.And the benefit of that is that we can bring togetherseveral trials into one.So normally we would do separate trials of those different typesof cancer.And that would be quite expensive to do because we'rerunning several trials.And it would mean a lot of inefficienciesin what we're doing.

  • 05:19

    JAMES WASON [continued]: So instead of doing that, we can do one trial called a baskettrial where we have the different types of cancer.Same drug and the same outcome on the trial,so we're determining whether it workswith the same information.So, for example, in cancer, we look at somethingcalled the response rate, which is

  • 05:39

    JAMES WASON [continued]: looking at how the tumor shrinks or not after treatment.So we get the efficiencies of running itall in one trial instead of multiple separate trials.But we also can use novel statistical methodsto allow us to potentially borrowinformation between the different types of cancer.So, for example, if one of the types of cancer

  • 06:00

    JAMES WASON [continued]: is much rarer than the offers, wemay not have much statistical powerto show that the new treatment is beneficial.But if we can borrow information,and a treatment is showing consistent effectsin other types of cancer where that mutation is present,and we can potentially allow borrowing of informationand give us a better answer about the rarer types of cancer

  • 06:25

    JAMES WASON [continued]: as well.An umbrella trial design also comes from oncology.[umbrella trial] And this is whenyou have multiple treatments thatare available for testing a specific type of cancersuch as lung cancer, for example.But you have different types of patients.Maybe they have different types of tumor mutations,for example.

  • 06:45

    JAMES WASON [continued]: And the aim of the trial is to, again,do what would normally be several trials in one trialso we get the efficiencies of running it all as one trial.But we get some extra benefits of patients on the trialwill be more likely to be matched to a treatment that'sgoing to work well for them because they

  • 07:06

    JAMES WASON [continued]: have different tumor mutations.And there's different treatments that are targetingthose different mutations.So hopefully more patients on the trialwill benefit from the trial.And we'll also get more high-quality informationat the end of a trial about which treatmentswork best for which types of the subgroups, of which typeof patient, which types of treatments work best for which

  • 07:30

    JAMES WASON [continued]: subgroups of patients.And I've particularly been interested--I've talked a lot about oncology.Oncology is really where we have seen the most innovationin trial design with these basket trials and umbrellatrials and other types of novel trial design as well.But I'm particularly interested in how

  • 07:51

    JAMES WASON [continued]: we can use ideas from oncology and use themin other quite different disease areas, maybe rheumatology,for example.This is an area where I've been particularly interestedin the future and making a difference in howtrials are done in this area.So an example of the basket trial in rheumatologymight be where you have a new biologic treatment that youexpect could work in different, similar rheumatology conditions

  • 08:15

    JAMES WASON [continued]: where you have similar symptoms.And you can run what would normallybe several different trials as one basket trial,allowing more efficiency and alsoallowing that borrowing of informationwith the innovative statistical methodsthat we can apply at the end of the trial.[Are there disadvantages to this approach?]

  • 08:41

    JAMES WASON [continued]: There are some limitations.So to get information about precision medicine,we need to have good subgroups of patients.[need for subgroups] We need to knowthat different types of patients will respond differentlyto the different treatments that might be availableor to one treatment if we're testing one treatment.And to do that, we need good information

  • 09:03

    JAMES WASON [continued]: to what we call categorize patients into different groups.So in oncology, for example, we have these tumor mutations,which generally are very good at distinguishing who benefitsfrom a new oncology treatment.But in other areas, there are not somany good ways of dividing patients

  • 09:24

    JAMES WASON [continued]: into groups that might benefit differentlyfrom the different types of treatments available.So that's the limitation.A disadvantage is that these trials are oftenvery complex to run. [complexity]And they require a lot more effort to set up.So if we're getting funding for a trial, a basket trial,

  • 09:45

    JAMES WASON [continued]: for example, it might look more expensivethan the normal randomized controlled trial just testingone treatment in one type of patient.But it should be a lot cheaper than runninga lot of separate ones, which we're combining into one.But still, there's a lot of effort to set up these trials,and that's certainly a limitation.

  • 10:07

    JAMES WASON [continued]: [What are the main challenges when designing these trials?]When designing the trials, I guessit just takes a lot longer normallyto design, from a statistical point of view,these type of trials.Often, they're using what we call an adaptive design where

  • 10:28

    JAMES WASON [continued]: an adaptive design uses informationthat we collect during the trial on how well patientsare responding to the different treatments.And it can make changes to the trial, whichis very beneficial to the efficiencyand good for the patients on the trial.For example, if a treatment is not showing much promise,then it could be dropped from the trial and more effort,

  • 10:50

    JAMES WASON [continued]: more resources put into the other treatmentsthat are being tested.But the disadvantage of that is that it takes a lot longerto plan because it's more complex.There's no single sample size formula available.We need to do, generally, statistical simulations,which can take a lot of time and a lot of communication

  • 11:12

    JAMES WASON [continued]: with different people involved in the trial.Sometimes getting these trials fundedcan be more difficult because they generally are moreexpensive than a normal trial.Although, as I've said, they're efficient.They combine a lot of separate trials into one.But they might look expensive to a funder.So it's about sharing the funder that this presents

  • 11:33

    JAMES WASON [continued]: a lot of value for money compared to just fundinga lot of separate trials.So for these trials to be successful,we really need lots of expertise in different areassuch as statistics, the clinical area, IT support, datamanagement, the trial managers, every type of expertthat we need to run the trial.

  • 11:55

    JAMES WASON [continued]: We need good communication between them.Everyone needs to be on board with the design.It's more complex for the regular design.Of course the patient needs to be involved.And often patients find these trial designsmore appealing than regular trialsbecause they, hopefully, have a greater chanceof benefiting from the trial.They are more likely to get a treatment that's

  • 12:17

    JAMES WASON [continued]: working better for them.So they are popular with patients.But patients do have to be on board.And it means that these complex ideasneed to be explained in intuitive ways.[What are the ethical considerations for these typesof trials?]

  • 12:37

    JAMES WASON [continued]: Ethics is definitely a really important partof clinical trials.And to do a clinical trial and to randomize patients,we really need to ensure that we'retesting things in an ethical way and that we don'tknow for sure which treatment approach is bestbecause if we did, we would have to do that

  • 12:58

    JAMES WASON [continued]: rather than randomize patients.s in these types of trials, theremight be additional ethical issuesto do with using information to change thingsas we're going along such as allocatingmore patients to a certain arm that's doing better.Patients often ask if you're doing that,

  • 13:18

    JAMES WASON [continued]: why don't you just give all patients that treatment?And it's because we're still not sure.But it's definitely better.But we think it's more likely to be better.So in a way, these types of trial designs can, in a sense,be more ethical for the patients on the trialbecause the patients on the trialare more likely to benefit from the trial.But that's a controversial thing that people are arguing about,

  • 13:42

    JAMES WASON [continued]: whether it's really an ethical--well, I think no one thinks it's unethical to usethese approaches, but whether it providesmore ethical management of patients is more controversial.[What is an example of a study that has used this approach?]I'll tell you about a trial that we're

  • 14:04

    JAMES WASON [continued]: about to start in Newcastle.And this is outside of oncology.So this is an example of the basket trial, which hasbeen done outside of oncology.So far about 95% of basket trials are in oncology.So it's quite rare to see them outside.So we're testing a new drug.

  • 14:25

    JAMES WASON [continued]: I can't really say too much about the detailsbecause it's funded by a company,and there's certain agreements in place.But this drug is for patients with primary biliarycholangitis, which are called PBC.And this is an autoimmune liver disease.And it's associated with not only problems with the liver

  • 14:46

    JAMES WASON [continued]: but also cognitive dysfunction.So patients who have this conditioncan experience decline in their cognitive abilities over time.So this drug is aiming to stop the declinein cognitive abilities over time.PBC has different stages.

  • 15:07

    JAMES WASON [continued]: So there's early PBC, patients whohave recently been diagnosed with it, and later PBC.And it's thought that the treatmentmight have different effects in these two groups.So that's one thing that we're dividingearly- and late-stage PBC.And since we're not sure about the effect of this treatment,

  • 15:30

    JAMES WASON [continued]: whether it might be similar between those two groupsor different, we're going to treat this as a basket trialwhere effectively we're doing two trials in parallelin the early and late PBC.And each one will be--each one, in itself, will be a randomized controlled trialwith this new drug and placebo.

  • 15:51

    JAMES WASON [continued]: And it will be analyzed as if it was its own trial.But it will also use an innovative statistical methodthat will allow the borrowing of information.So if the treatment difference between the new drugand placebo is similar between the two groups,then this method will allow what wecall borrowing of information.And that will improve the power to show

  • 16:12

    JAMES WASON [continued]: that the new drug is effective.Now, the company that was making this drugwas interested in whether it would alsowork in Parkinson's disease.Parkinson's disease is different to PBC,but it has a similar symptom in that patientswith Parkinson's disease experience cognitive declines.

  • 16:33

    JAMES WASON [continued]: And so instead of doing this trial in PBC and doinga separate trial in Parkinson's disease,we convinced the company-- and they were very supportiveof this--of including an extra module, as we call it,of the trial, so effectively a third partof the trial, which would be Parkinson's disease patients.

  • 16:54

    JAMES WASON [continued]: And it would have the same drug placebo as well,have the same endpoint, which is a measure of the differencebetween the new drug and placebo in the cognitive score.And now we can potentially borrow informationbetween all three of those.So the early-stage, the late-stage PBC,and the Parkinson's disease.

  • 17:15

    JAMES WASON [continued]: And it's much cheaper to do this fundto run the separate trials.[What resources do you recommend for others considering thisapproach?]There's a really nice website called,B-I-G-T-E-D dot org.And this is an online repository of different trial designs

  • 17:39

    JAMES WASON [continued]: that are available for using patient subgroups in trials,which is effectively precision medicine trials.So this was a lot of work to put together from researchersin the University of Liverpool.And they did a really good job making it easy to navigateand having nice figures for each type of trial design available.

  • 18:01

    JAMES WASON [continued]: So you can click on the type of trialyou want, and it shows a nice picture, gives youmore information, gives you a reference to a paper thattalks more about that particular designif you want more details.And it also gives you some help with things like sample sizecalculation.So this is a really good resource.I am co-leading the Stratified Medicine Working

  • 18:23

    JAMES WASON [continued]: Group, which is part of the trial's Methodology ResearchPartnership.And we aim, in the future, to produce more guidancein this area and to help people with selectingthe best approach when precision medicine is the goal.[MUSIC PLAYING]


James Wason, Professor of Biostatistics at Newcastle University, discusses designing clinical trials in the era of precision medicine, including challenges, ethical considerations, and resources.

Looks like you do not have access to this content.

Introduction to Designing Clinical Trials in the Era of Precision Medicine

James Wason, Professor of Biostatistics at Newcastle University, discusses designing clinical trials in the era of precision medicine, including challenges, ethical considerations, and resources.

Copy and paste the following HTML into your website