SEARCH

SEARCH BY CITATION

Keywords:

  • Brain imaging;
  • fMRI;
  • prediction;
  • relapse;
  • risk;
  • treatment

Neuroscientists have collected valuable information on neurofunctional and neurostructural characteristics accompanying addiction. This editorial argues for three prerequisites to making neuroimaging practically relevant for addiction medicine: the integration of imaging findings, the identification of concrete neuroimaging risk factors for the individual and the ability to inform the development of treatment regimens.

Substance-using individuals are a heterogeneous group, which is due—among other aspects—to drug-related factors such as drug type or duration of use, as well as drug-unrelated factors such as demography or neuropsychiatric comorbidities. Several models have conceptualized addiction, but the identification of risk factors for the development of substance use disorders (SUD) or the likelihood to relapse for an individual substance user has poor predictive value [1]. Extending beyond socio-demographic, cognitive and psychological measures, neuroimaging methods such as (functional) magnetic resonance imaging [(f)MRI], electroencephalography (EEG) and positron emission tomography (PET) have characterized the functional, structural and molecular basis of SUDs, vastly exceeding knowledge acquired through behavioural or experience-based measures such as reaction times and subjective craving. For example, cognitive neuroscience has established a strong link between frontal and subcortical brain regions and prominent addiction-related symptoms, e.g. the compulsive intake and the intense impulse to use drugs at the expense of more advantageous behaviours [2, 3]. Neuroscientists now have to demonstrate how our findings can extend and support existing models and therapies successfully. Specifically, we have to take three critical steps to implement neuroimaging as a new basis for diagnostics and treatment of SUDs: first, we need to merge diverse imaging findings into one comprehensive brain imaging perspective of addiction. Next, we need to identify prediction algorithms for individual substance users. The ultimate goal has to be the development of treatment regimens based on neuroimaging results.

Most urgently, the integration of independent neuroimaging findings into one comprehensive brain model of addiction needs to be accomplished. Neuroscientists need to merge results from functional studies probing the same, similar or distinct mental processes by means of appropriate meta-analysis approaches [4] and relate these to structural and molecular characteristics of addiction. Importantly, researchers need to accept that advantages of their passionately preferred neuroimaging technique come at the expense of limitations. fMRI, for example, maps mental processes with high spatial resolution, but is a relative and indirect measure of brain activation, unable to resolve adequately the temporal dynamics or the associated molecular basis of these processes. Only the combination of complementary neuroimaging methods (including behavioural measures and participants' self-reports) is likely to advance addiction medicine by shaping a neurophysiological picture of SUDs. Besides recent methodological and statistical developments allowing for analyses of structural, temporal and causal relationships between brain regions [5, 6], modern neuroscience provides multi-modal approaches to overcome technological shortcomings. Simultaneous EEG–fMRI, for instance, combines high temporal and spatial resolution of exactly the same mental process [7], and hybrid MR–PET imaging allows for functional/structural and molecular characterizations [8]. While multi-modal imaging techniques appear to be an ideal solution, their limited use will be related not only to costs or lack of expertise and availability, but will also reflect the existing researchers' hesitation, unwillingness or inability to consolidate findings from different imaging modalities. If neuroscientists in addiction research remain unable (or unwilling?) to link findings from different techniques and studies, how will a true integration with traditional, well-established non-imaging models ever be achieved?

Along with multi-modal imaging approaches, neuroscience needs to further diagnostics and outcome prediction to reach clinical relevance. One crucial driving force proving the relevance of neuroimaging in addiction medicine will be the delineation of characteristic stages of SUDs, e.g. occasional versus habitual versus compulsive use or intoxication versus abstinence versus relapse. Although neurobiological predictors of relapse as a central hallmark of addiction have attracted surprisingly little attention, recent studies identifying neural differences between subsequently relapsing and non-relapsing users are promising [9, 10]. Moreover, neuroimaging has successfully characterized subjects at high risk for SUDs [11]. Such developments in fMRI study designs and analyses verify that neuroimaging has become a useful tool to identify people at high risk for SUDs or susceptibility to relapse, at stages where patients and therapists may have few other means for predictions. Predictive values of biomarkers now need to be determined on individual levels, which will require skills beyond those of a typical neuroscientist in relation to ethical issues (what does it mean for the individual and society if neuroscience identifies high-risk subjects?) and statistical/methodological expertise (do we really believe enough in our methods and data to predict individual outcomes?). As long as specificity and sensitivity have not been proved, brain imagers themselves, let alone clinicians, will remain sceptical of neuroimaging-based predictions. One important barrier potentially preventing significant progress in this domain is the need for longitudinal studies; ideally, huge samples of participants would be included prior to substance initiation and followed for decades. Such studies, however, are methodologically challenging, expensive and not promising in terms of short-term publication of results—the latter showing how the nature of science and its funding can impact directly upon the utility of its own findings. Alternative, collaborative approaches may be more promising on a shorter time-scale, wherein existing tools developed for multi-centre fMRI studies already enable researchers to pool data into large databases and analyses [4, 12].

Lastly, neuroscientists need to develop concrete treatment strategies based on their findings. Respective approaches are already in development for diverse neuropsychiatric disorders, and several scenarios can be envisioned. First, linkage of neuroimaging and pharmacological studies will prove useful for predicting response to medication. Secondly, knowledge of the biological differences between responders and non-responders to available treatments might facilitate identification of the best-suited therapy for that particular individual [13, 14]. Thirdly, understanding which brain regions show alterations in functioning should spur the development of specific medications, cognitive–behavioural or neuroimaging-based trainings (‘neurofeedback’ [15, 16]) that target optimal activation levels in these regions. For example, integration of fMRI and incentive-based interventions could alter striatal reward-related processing so that cognitive control processes reduce the propensity of drug-taking behaviour.

Clearly, neuroimaging is playing an increasingly important role in addiction medicine. It has huge potential to change fundamentally the way in which we treat addictions once remaining conceptual and technological challenges have been addressed. To become clinically relevant, neuroscientists now need to push the limits of imaging technologies to prove their abilities to define clinically relevant information on a single-subject basis. The ultimate goal has to be the integration of imaging findings into the development of concrete treatment regimens. Only interdisciplinary research teams and projects will accomplish these steps.

Declaration of interests

None.

Acknowledgements

  1. Top of page
  2. Acknowledgements
  3. References

I am particularly grateful to all subjects, substance users and non-users, for participating in research projects providing valuable information about the neurophysiological basis of addiction. Additionally, I would like to thank all neuroscience colleagues for countless discussions on the relevance and limitations of fMRI.

References

  1. Top of page
  2. Acknowledgements
  3. References
  • 1
    Reske M., Paulus M. P. A neuroscientific approach to addition: ethical concerns. In: Illes J. , Sahakian B. J. , editors. Oxford Handbook of Neuroethics. Oxford: Oxford University Press; 2011, p. 177202.
  • 2
    Goldstein R. Z., Craig A. D., Bechara A., Garavan H., Childress A. R., Paulus M. P. et al. The neurocircuitry of impaired insight in drug addiction. Trends Cogn Sci 2009; 13: 372380.
  • 3
    Goldstein R. Z., Volkow N. D. Dysfunction of the prefrontal cortex in addiction: neuroimaging findings and clinical implications. Nat Rev Neurosci 2011; 12: 652669.
  • 4
    Eickhoff S. B., Bzdok D., Laird A. R., Kurth F., Fox P. T. Activation likelihood estimation meta-analysis revisited. Neuroimage 2012; 59: 23492361.
  • 5
    Friston K. J., Harrison L., Penny W. Dynamic causal modelling. Neuroimage 2003; 19: 12731302.
  • 6
    Zalesky A., Solowij N., Yucel M., Lubman D. I., Takagi M., Harding I. H. et al. Effect of long-term cannabis use on axonal fibre connectivity. Brain 2012; 135: 22452255.
  • 7
    Warbrick T., Mobascher A., Brinkmeyer J., Musso F., Stoecker T., Shah N. J. et al. Nicotine effects on brain function during a visual oddball task: a comparison between conventional and EEG-informed fMRI analysis. J Cogn Neurosci 2012; 24: 16821694.
  • 8
    Shah N. J., Oros-Peusquens A. M., Arrubla J., Zhang K., Warbrick T., Mauler J. et al. Advances in multimodal neuroimaging: hybrid MR-PET and MR-PET-EEG at 3T and 9.4T. J Magn Reson 2012; 229: 101115.
  • 9
    Kosten T. R., Scanley B. E., Tucker K. A., Oliveto A., Prince C., Sinha R. et al. Cue-induced brain activity changes and relapse in cocaine-dependent patients. Neuropsychopharmacology 2006; 31: 644650.
  • 10
    Paulus M. P., Tapert S. F., Schuckit M. A. Neural activation patterns of methamphetamine-dependent subjects during decision making predict relapse. Arch Gen Psychiatry 2005; 62: 761768.
  • 11
    Stewart J. L., Flagan T. M., May A. C., Reske M., Simmons A. N., Paulus M. P. Young adults at risk for stimulant dependence show reward dysfunction during reinforcement-based decision making. Biol Psychiatry 2013; 73: 235241.
  • 12
    Stöcker T., Schneider F., Klein M., Habel U., Kellermann T., Zilles K. et al. Automated quality assurance routines for fMRI data applied to a multicenter study. Hum Brain Mapp 2005; 25: 237246.
  • 13
    Brewer J. A., Worhunsky P. D., Carroll K. M., Rounsaville B. J., Potenza M. N. Pretreatment brain activation during Stroop task is associated with outcomes in cocaine-dependent patients. Biol Psychiatry 2008; 64: 9981004.
  • 14
    Moeller F. G., Steinberg J. L., Schmitz J. M., Ma L., Liu S., Kjome K. L. et al. Working memory fMRI activation in cocaine-dependent subjects: association with treatment response. Psychiatry Res 2010; 181: 174182.
  • 15
    deCharms R. C., Maeda F., Glover G. H., Ludlow D., Pauly J. M., Soneji D. et al. Control over brain activation and pain learned by using real-time functional MRI. Proc Natl Acad Sci USA 2005; 102: 1862618631.
  • 16
    Sitaram R., Caria A., Veit R., Gaber T., Rota G., Kuebler A. et al. 2007; FMRI brain-computer interface: a tool for neuroscientific research and treatment. Comput Intell Neurosci 2007: 25487. doi: 10.1155/2007/25487.