[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Views 903
Citations 0
Viewpoint
March 25, 2020

A Framework for Advancing Precision Medicine in Clinical Trials for Mental Disorders

Author Affiliations
  • 1Washington University School of Medicine in St Louis, St Louis, Missouri
  • 2Washington University, St Louis, Missouri
JAMA Psychiatry. Published online March 25, 2020. doi:10.1001/jamapsychiatry.2020.0114

Why Aren’t We Getting to Precision Mental Health?

Getting the right treatment to the right patient is a frustrating goal in mental health: treatment is a trial-and-error endeavor, often yielding disappointing outcomes. Why? Traditional randomized clinical trials (RCTs) do not adequately capture the dynamic complexity of the brain and behavior during treatment. Our increasing understanding of the pathophysiology of mental disorders will not translate to legitimate movement toward precision medicine unless we overcome 3 limitations of RCTs methods: (1) measuring only static baseline characteristics is unlikely to identify treatment responders, (2) a treatment that does not adapt to the individual patient is unlikely to be optimal, and (3) infrequently administered measures do not have the precision to predict or measure outcomes.

Precision Clinical Trials

In this Viewpoint, we present a framework for precision clinical trials (PCTs), which incorporate design features to account for dynamic complexity in treatment, addressing the core question of precision medicine: “How can treatment get the most benefit for this patient?” Those features are (1) treatment-targeted enrichment, (2) adaptive treatment, and (3) precision measurement. The PCT framework requires new approaches and new methodological research, so this Viewpoint is a call to action for trialists, funders, and regulators, such as the US Food and Drug Administration. We explain the PCT features and provide an example in the Figure.

Figure.  Precision Clinical Trial Methods
Precision Clinical Trial Methods

Using the example of transcranial magnetic stimulation (TMS) for depression, this figure depicts the 3 features of precision clinical trials. EEG indicates electroencephalogram; EMA, electromagnetic acoustic.

Feature 1: Treatment-Targeted Enrichment

Treatment response in mental disorders involves a complex, dynamic interplay between the patient and the treatment. Clinical practice and research have taught us that a short-term response is strongly associated with long-term response.1 Accordingly, treatment-targeted enrichment involves measuring an individual’s brain and/or behavioral change in response to a brief exposure to the intervention under study to determine their associated outcome with a full course of treatment. In other words, the method first exposes all participants briefly to an intervention’s key active ingredient and then measures the change in symptoms and/or brain and behavior changes based on the mechanistic understanding of the treatment’s target. This information is then incorporated into randomization to a full course of the therapy or placebo as a stratification variable or inclusion criterion.

Many treatments have mechanisms of action that can be tested using noninvasive neurobiological measures, including transcranial magnetic stimulation (TMS) paired with electroencephalography (EEG)2 and functional magnetic resonance imaging,3 and cognitive and behavioral phenotyping with mobile technology.4 For example, a single session of repetitive TMS may produce immediate changes detectable via neuroimaging or neurophysiological or cognitive testing that differentiate patients who are most likely to benefit from a full course of repetitive TMS (Figure). Using such mechanistic assessment tools to determine the associated treatment response is compatible with the National Institute of Mental Health’s experimental therapeutics agenda, which requires target engagement tests within clinical trials.5 However, short-term changes may differ from the long-term intervention mechanism that produces a response yet still be associated with that response; examples include participant engagement, acceptability, and tolerability.

Feature 2: Adaptive Treatment

Optimally effective treatments adapt to the individual patient over time, which is why clinicians adjust dosage and other parameters of treatment to optimize each patient’s outcome. Precision clinical trials embrace this concept, systematically adapting intervention parameters over time for individual participants. This approach departs from a traditional RCT design in which all participants with the active condition typically receive the same intervention parameters.

Treatments in PCTs need to adapt to each patient based on their response, engagement, and preferences. Stepped-care interventions and just-in-time adaptive interventions are examples of iteratively assessing responses and shifting treatment parameters. Smartphone, sensor, and brain measurements (such as TMS-EEG) can provide real-time feedback informing the need for adaptation in the case of nonresponse or lack of engagement. Options range from intuitive stepwise changes or intensifications in treatment to systematic auto-adaptation using Sequential Multiple Assignment Randomized Trial methods such as introducing a second randomization in treatment nonresponders, testing a predefined augmentation strategy. New inference and analysis methods will be needed accompany this shift. We expect that a combination of machine learning and bayesian inference approaches will help solve these challenges as they have the capacity to handle many parameters and complicated data structures.6

Feature 3: Precision Measurement of Predictors and Outcomes

In traditional RCTs, outcome assessments are obtained infrequently and often only baseline and end point values are analyzed for outcome analysis. Such assessments rely heavily on retrospective reporting and averaging (ie, “In the past 7 days…”), which are prone to error introduced by cognitive biases. But even if such measures were not distorted by cognitive biases, they still do not possess the temporal resolution necessary to assess individual-level changes in neurobehavioral symptoms that occur rapidly and are context dependent (eg, symptom presentation differs at home vs in the community or workplace).7 A solution is smartphone-based ecological momentary assessment (EMA),8 which provides frequent, real-time, lived-in assessments. The EMA makes it possible to reliably assess changes over time, potentially reducing the placebo response to the extent that it is caused by measurement error. The EMA can also monitor the course of treatment as it unfolds, facilitating treatment-targeted enrichment and treatment adaptation. With enough measurement points, statistical tests can detect or predict important changes in symptom presentation or treatment responses in a single individual without reference to a larger group.9

Other emerging techniques for association or outcome assessment include precision brain mapping,10 TMS-EEG, and digital phenotyping. These techniques are in their infancy—we cannot predict treatment responses or adapt treatments based on digital phenotyping, for example— and they need development and validation. But, like EMA, these approaches can provide an accurate dynamic measurement of individual-level brain and behavior changes.

Conclusions

Precision clinical trials provide a framework for developing and testing personalizable treatments, providing an evidence base for precision medicine in psychiatry. Although some methods are newly formed, embracing these techniques now will allow for a rapid refinement for optimal application. These methods will also accelerate the adoption of precision medicine in practice, as today’s precision research measurements become tomorrow’s real-world assessments. Ultimately, the most salient argument for PCTs is a patient-centered one: patients want and deserve better than the current trial-and-error approach to treatment that has emerged from traditional RCTs.

Back to top
Article Information

Corresponding Author: Eric J. Lenze, MD, Department of Psychiatry, Washington University School of Medicine in St Louis, 660 S. Euclid, PO Box 8134, St Louis, MO 63110 (lenzee@wustl.edu).

Published Online: March 25, 2020. doi:10.1001/jamapsychiatry.2020.0114

Conflict of Interest Disclosures: Dr Lenze reported personal fees from Janssen and Jazz Pharmaceuticals as well as grants from McKnight Brain Research Foundation, the National Institutes of Health (NIH), Patient-Centered Outcomes Research Institute, Takeda/Lundbeck, and Aptinyx outside the submitted work. Dr Rodebaugh reported grants from the NIH (National Institute on Deafness and Other Communication) and Brain Behavior Research Foundation outside the submitted work. Dr Nicol reported grants and personal fees from Alkermes and personal fees from Sunovion and Supernus outside the submitted work.

Funding/Support: The research reported in this article was supported by the Washington University Institute of Clinical and Translational Sciences grant UL1TR002345 from the National Center for Advancing Translational Sciences of the NIH. Additional funding is from the Taylor Family Institute for Innovative Psychiatric Research and Center for Brain Research in Mood Disorders (Washington University, St Louis).

Role of the Funder/Sponsor: The funding sources had no role in the preparation, review, or approval of the manuscript.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official view of the NIH.

References
1.
Reynolds  CF  III, Dew  MA, Pollock  BG,  et al.  Maintenance treatment of major depression in old age.  N Engl J Med. 2006;354(11):1130-1138. doi:10.1056/NEJMoa052619PubMedGoogle ScholarCrossref
2.
Tremblay  S, Rogasch  NC, Premoli  I,  et al.  Clinical utility and prospective of TMS-EEG.  Clin Neurophysiol. 2019;130(5):802-844. doi:10.1016/j.clinph.2019.01.001PubMedGoogle ScholarCrossref
3.
Dichter  GS, Gibbs  D, Smoski  MJ.  A systematic review of relations between resting-state functional-MRI and treatment response in major depressive disorder.  J Affect Disord. 2015;172:8-17. doi:10.1016/j.jad.2014.09.028PubMedGoogle ScholarCrossref
4.
Jacobson  NC, Weingarden  H, Wilhelm  S.  Using digital phenotyping to accurately detect depression severity.  J Nerv Ment Dis. 2019;207(10):893-896. doi:10.1097/NMD.0000000000001042PubMedGoogle ScholarCrossref
5.
Graham  AK, Lattie  EG, Mohr  DC.  Experimental therapeutics for digital mental health.  JAMA Psychiatry. 2019. doi:10.1001/jamapsychiatry.2019.2075PubMedGoogle Scholar
6.
Barbour  DL, DiLorenzo  JC, Sukesan  KA,  et al.  Conjoint psychometric field estimation for bilateral audiometry.  Behav Res Methods. 2019;51(3):1271-1285. doi:10.3758/s13428-018-1062-3PubMedGoogle ScholarCrossref
7.
Mofsen  AM, Rodebaugh  TL, Nicol  GE, Depp  CA, Miller  JP, Lenze  EJ.  When all else fails, listen to the patient: a viewpoint on the use of ecological momentary assessment in clinical trials.  JMIR Ment Health. 2019;6(5):e11845. doi:10.2196/11845PubMedGoogle Scholar
8.
Shiffman  S, Stone  AA, Hufford  MR.  Ecological momentary assessment.  Annu Rev Clin Psychol. 2008;4:1-32. doi:10.1146/annurev.clinpsy.3.022806.091415PubMedGoogle ScholarCrossref
9.
Fisher  AJ.  Toward a dynamic model of psychological assessment: implications for personalized care.  J Consult Clin Psychol. 2015;83(4):825-836. doi:10.1037/ccp0000026PubMedGoogle ScholarCrossref
10.
Wolfers  T, Doan  NT, Kaufmann  T,  et al.  Mapping the heterogeneous phenotype of schizophrenia and bipolar disorder using normative models.  JAMA Psychiatry. 2018;75(11):1146-1155. doi:10.1001/jamapsychiatry.2018.2467PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    ×