Limitations of the use of the MP-RAGE to identify neural changes in the brain: Recent cigarette smoking alters gray matter indices in the striatum

T-1weighted MP-RAGE

Axial Sections . Sagittal Sections . Coronal Sections
How to use this website

Select one of the three views above to begin. The images will be displayed in a table on the left, with a slideshow panel on the right.

Clicking on an image on the left will set the slideshow to that image.

The slideshow can be used to manually or automatically flip through the images in order.

Supplementary Online Material HTML tutorial Figure 1
General Commentary

          The magnetization-prepared rapid gradient-echo (MP-RAGE) T1-weighted high resolution structural MRI is a mainstay tool used to identify morphometric biomarkers of disease conditions, progression and treatment effects despite a critical limitation: the relaxation signal on which inferences are based is nearly indistinguishable for gray matter versus blood flow (Lu et al., 2004; Wright et al., 2008). Thus, apparent reported morphometric findings might be at least partially related to transient changes in blood flow or other physiological signals.
          Consistent with this technical limitation, using a standard analysis technique, voxel based morphometry (VBM), we recently reported that a single dose of a medication had ‘apparent’ effects on T1-weighted MRIs (Franklin et al., 2013). Specifically, we observed medication-induced decreases in gray matter volume in the anterior cingulate and other regions that overlapped with changes in brain blood flow (perfusion). Similarly, others have shown effects of medication on T1-weighted scans that are likely transient. For example, acute levodopa administration altered gray matter indices on T1-weighted images in the midbrain (Salgado-Pineda et al., 2006). Further, in a well-controlled longitudinal VBM study of patients with attention deficit hyperactivity disorder (ADHD), Hoekzema and colleagues (2014) showed that stimulant drugs trigger transient volumetric changes in the ventral striatum, challenging previous VBM studies suggesting that these changes were an intrinsic feature of the disorder (Hoekzema et al., 2014). These examples highlight the importance of employing appropriate experimental conditions to ensure that morphometric findings are actually revealing long-term changes that are unrelated to transient states. Specifically, longitudinal investigations into the effects of a medication on neural structure using VBM may want to include a group that is administered an acute dose of the medication.
          The findings of altered gray matter induced by acute pharmacological manipulations highlighted above led us to hypothesize that cigarette smoking, which delivers the psychostimulant, nicotine, to the brain may transiently alter the MP-RAGE, based on the T1-weighted images. Thus, we examined T1-weighted structural MRIs acquired in a within-subjects design in 39 otherwise healthy nicotine-dependent individuals. Subjects meeting our standard MRI-specific inclusion/exclusion criteria (Franklin et al., 2013; Franklin et al., 2014) provided written informed consent to participate in the study, which adhered to the Declaration of Helsinki and was approved by the University of Pennsylvania Institutional Review Board. Screening procedures included a medical history, physical examination, and a psychological assessment. Subjects participated in two counterbalanced scanning sessions that occurred on separate days. For condition ‘smoke’ subjects smoked a cigarette to satiety approximately 45 minutes prior to obtaining structural MRI data. For condition ‘no smoke’, subjects smoked a cigarette and then remained abstinent for approximately five hours prior to data acquisition. To obtain quantitative cerebral blood flow (CBF) maps, five minute resting baseline pseudo-continuous arterial spin labeled perfusion fMRI scans were also acquired under both conditions. Nicotine’s terminal half-life is approximately two hours; thus, conservative estimates might place the onset of withdrawal symptomatology within the first two hours after last smoking (Benowitz et al., 1982). Thus, five hours of deprivation was chosen to ensure that the subjects would be experiencing withdrawal from nicotine. During the ‘no smoke’ condition, carbon monoxide (CO) levels were acquired after the baseline cigarette and prior to scanning. Abstinence was confirmed by monitoring the subjects and was biochemically confirmed by observing reductions in CO from baseline to time of scan acquisition. Structural data were analyzed using statistical mapping software (SPM8) and its available VBM tool with diffeomorphic anatomical registration using exponentiated lie algebra (DARTEL). Statistical parametric maps were generated to perform between-condition comparisons (paired t-test) using the gray matter tissue segmentation output by DARTEL. To obtain measures of brain blood flow mean CBF maps were calculated for each condition and between-condition comparisons (paired t-test) were performed. Image acquisition, data processing and analysis methods have been previously published (Franklin et al., 2013; Franklin et al., 2014).
          Once again, we report that transient state impacts gray matter indices. Compared to not smoking for several hours, recent smoking altered the T-1weighted MP-RAGE, bilaterally in the striatum. Differences were not observed in any other brain regions. Additionally, as in our previous study, the effects overlapped with changes in blood flow. These results unambiguously demonstrate that the use of VBM to identify morphometric biomarkers of disease conditions, progression and treatment effects must employ appropriate experimental conditions to ensure that morphometric findings are not related to acute states that may be introduced by drug or medication, but are indeed revealing long-term change. Importantly, the use of medications or substances (e.g., alcohol, cigarettes, concomitant medications) that affect the brain should be controlled (either through matching or statistical co-variation) across groups that are being compared. Third, these results show that the MP-RAGE should be acquired at the initiation of a scanning session, given the potential for it to be altered by subsequent activities (tasks, pharmacological manipulations) that occur over the course of scanning. These findings also underscore the crucial and immediate need to develop neuroimaging tools that can uniquely capture changes in neuronal structure dissociable from those related to blood flow or other physiological signals.

KEYWORDS:voxel based morphometry, ventral striatum, cigarette smoking, gray matter volume, longitudinal, T1-weighted, MP-RAGE.

AUTHORS: Teresa R. Franklin*, Reagan R. Wetherill, Kanchana Jagannathan, Nathan Hager, Charles P. O’Brien and Anna Rose Childress.

* Corresponding author: Teresa R. Franklin, Ph.D.
University of Pennsylvania, Department of Psychiatry
3900 Chestnut Street, Philadelphia, PA 19104, USA
Telephone:(215) 222-3200 ext. 119
Fax: (215) 386-6770


          The authors would like to thank the clinical and support staff, with special thanks to Melanie Maron (Study Coordinator), at the Center for the Studies of Addiction, Perelman School of Medicine at the University of Pennsylvania and the MRI technical staff at the Hospital of the University of Pennsylvania.

Funding Sources

          This work was supported by the National Institutes of Health, National Institute for Drug Abuse (CPO 5P60DA005186; TRF K01DA015426-02; TRF R21DA025882; TRF R01DA029845-01A1 and TRF R01DA03039401A1) and a TRF IIR from Pfizer Pharma. The funding organizations had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing Interests

          None of the authors report competing interests related to the study.