Abstract

1. INTRODUCTION

Striatal dopamine may play an important role in the pathophysiology of depression by modulating emotional and motor symptoms (Rogers et al., 1998; Byrum et al., 1999; Newberg et al., 2007). Cortical dopamine pathways have been identified that project to the striatum, globus pallidus, substantia nigra, and back to the cortex (Rogers et al., 1998). Some studies have suggested that dopamine transporter (DAT) activity may reflect the general state of dopamine function in the brain, and that alterations in DAT activity may reflect abnormalities in dopamine function (Jaber et al., 1997). Several anatomic imaging studies have reported reduced basal ganglia volume in depression (Husain et al., 1991; Parashos et al., 1998), and this observation has been confirmed in at least one postmortem study (Baumann et al., 1999). In addition, several magnetic resonance imaging studies have found greater signal intensities in the striatum of depressed subjects (Greenwald et al., 1996; Lenze et al., 1999).

A large number of single photon emission computed tomography (SPECT) and positron emission tomography (PET) studies were reviewed by Soares and Mann (1997) and generally found reduced regional cerebral blood flow and reduced glucose metabolism, respectively, in the basal ganglia in depressed subjects - although these observations have not been universally confirmed. The authors of this review suggested that the findings in the basal ganglia, in conjunction with generally decreased cerebral blood flow and metabolism in the frontal regions, limbic structures and the thalamus, indicate an overall network of structures involved in the pathophysiology of mood disorders. Further, given the important role of the striatum, it is likely that the dopaminergic system is an important mediator of mood disorders.

The potential role of the dopamine system in depression has been studied using a variety of imaging studies. For example, abnormalities in the striatal dopamine (D₂/D₃) receptor binding in depression has been reported as being either greater (D’haenen & Bossuyt, 1994; Shah te al., 1997) or no different than that seen in non-depressed subjects (Klimke et al., 1999). Several SPECT studies have examined striatal DAT density in depression. Laasonen-Balk et al. (1999) found approximately 20% greater striatal DAT density in 15 depressed subjects (versus 18 non-depressed subjects) using [¹²³I]β-CIT (a radioligand with mixed serotonin transporter and DAT binding affinity). In contrast, Malison et al. (1998) reported no difference in striatal [¹²³I]β-CIT binding in 15 depressed versus 15 non-depressed subjects. Similarly, a postmortem study found no difference in striatal DAT density of depressed suicide victims versus non-depressed subjects, using the non-selective radioligand, [³H]GBR12935 (Bowden et al., 1997).

We have utilized the selective DAT radioligand, [^99mTc]TRODAT-1, to quantify striatal DAT density in depressed versus non-depressed subjects (Brunswick et al., 2003; Weintraub et al., 2005; Amsterdam & Newberg 2007; Newberg et al., 2007). [^99mTc]TRODAT-1 is highly selective for DAT sites, utilizes a less expensive isotope (i.e., ^99mTc), and has excellent imaging characteristics. In a study of DAT density of emotional affect in 73 healthy, non-depressed volunteer subjects, we observed greater striatal DAT density in individuals with higher self-report scores of depressive affect on the Profile of Mood Scale (p=0.02) (Newberg et al., 2007). In another study of 15 depressed versus 46 non-depressed subjects, we observed significantly greater age and gender adjusted striatal [^99mTc]TRODAT-1 uptake (i.e., DAT density) in depressed subjects in the right anterior putamen (p<0.001), right posterior putamen (p<0.001), left posterior putamen (p<0.05), and left caudate (p<0.05) (Brunswick et al., 2003). These finding were subsequently replicated by Yang et al. (2008) using [^99mTc]TRODAT-1 in 10 depressed (versus 7 non-depressed) subjects.

In the current study, we examined a new cohort of depressed subjects in an effort to validate our earlier observation of greater striatal DAT density in depressed (versus non-depressed) subjects using [^99mTc]TRODAT-1, and to test the hypothesis that greater striatal DAT density may represent a putative biomarker to distinguish depressed from non-depressed subjects.

2. METHODS

3. RESULTS

3.2 DVR Values

Table 1 displays the intra-subject correlation coefficient of DVR values for striatal ROIs within depressed and non-depressed subjects. DVR values were highly correlated within all striatal ROIs in depressed (p<0.005) and non-depressed (p<0.005) subject groups.

Table 1

Intra-subject correlation of striatal DVR values in 123 depressed and non-depressed subjects ^*

Depressed subjects had significantly greater mean (SD) DVR values (versus non-depressed subjects) for the right putamen (p<0.003) and left putamen (p<0.0005) (Table 2).

Table 2

Mean (SD) DVR values in depressed versus non-depressed subjects ^*

There were no significant differences in mean DVR values for unipolar versus bipolar II subjects, or for depressed subjects with or without prior psychotropic medication exposure (Table 3).

Table 3

Mean (SD) DVR values in depressed subjects by diagnosis and prior psychotropic drug exposure

Figure 1 displays the ROC curve for the total putamen plus caudate region with an AUC value of 0.79. The ROC curve plots the sensitivity versus 1 minus the specificity for tests based on all possible predicted probability values from the logistic regression models (where each probability value is used as a cutoff point for distinguishing depressed from non-depressed subjects). As sensitivity increases in value, specificity decreases in value.

Figure 1

Receiver operator characteristic (ROC) curve of the logistic regression models for the Total Putamen and Caudate DVR values.

Table 4 displays the estimated probabilities with associated sensitivity and specificity values for the first ten points along the ROC curve, when the values are ordered according to the distance between each (sensitivity, specificity) pair and the point (1,1).

Table 4

Sensitivity and specificity values for putamen DVR measures along the ROC curve ^†

The Hosmer-Lemeshow test indicates that the fit of the regression models are good for the left putamen, right putamen, total putamen, and total putamen plus caudate regions (Table 5). The predictive ability of each ROI to distinguish depressed from non-depressed subjects is high because the AUC values are close to 1.0.

Table 5

Predictive ability of DVR value to distinguish depressed from non-depressed subjects

4. DISCUSSION

The dopamine system may be involved in the pathophysiology of depression (Rogers et al., 1998; Klimke et al., 1999; Amsterdam & Newberg 2007; Newberg et al., 2007). Several studies have suggested that DAT density may be greater in depression (Jaber et al., 1997; Laasonen-Balk et al., 1999; Brunswick et al., 2003; Yang et al., 2008). For example, Laasonen-Balk et al. (1999) measured striatal DAT density in 15 depressed and 18 non-depressed subjects using [¹²³I]β-CIT and found greater DAT density in depressed subjects. These investigators speculated that this alteration may represent a key mechanism by which dopamine is modulated in depression. Similarly, we previously reported a greater age and gender adjusted DAT density using [^99mTc]TRODAT-1 in 15 depressed (versus 46 non-depressed) subjects (Brunswick et al., 2003). In that study, we found that [(99m)Tc]TRODAT-1 binding was significantly higher in the right anterior putamen (23%), right posterior putamen (36%), left posterior putamen (18%), and left caudate nucleus (12%) of the patients with depression than in control subjects. This finding was subsequently replicated by Yang et al. (2008) in 10 depressed (versus 7 non-depressed) subjects. In contrast, Malison et al. (1998) reported no difference in DAT density using [¹²³I]β-CIT in 15 depressed (versus 15 non-depressed) subjects, while Weintraub et al., (2005) found reduced DAT density in subjects with co-morbid depression plus Parkinson’s disease which also correlated with affect scores.

The current study validates our prior observation of greater striatal DAT density in depression (Brunswick et al., 2003). Moreover, the current observation also suggest that greater DAT density may represent a putative biomarker of altered dopamine function in depression with the ability to distinguish depressed from non-depressed subjects. This has implications for the clinical evaluation of patients with depression as the DAT binding may ultimately be useful to evaluate depression in patients or to help determine prognosis and response to various interventions. However, future studies will have to be performed in order to better evaluate the potential clinical utility of DAT imaging in patients with depression.

While the mechanism of increased DAT density in depression is not known, it is possible that the greater DAT density may compensate for a diminished extra-neuronal dopamine concentration. It is also possible that the greater DAT density acts as a compensatory clearing mechanism for excessive central dopamine concentrations, or compensates for reduced synaptic D₂/D₃ receptor function. However, several studies have examined the relationship of D₂/D₃ receptor function on DAT density in depression and have reported no difference in D₂/D₃ receptor activity in depression (Klimke et al., 1999; Kimmel et al., 2001; Yang et al., 2008). Altered DAT density may also represent a neuronal reflection of psychomotor or affective changes that occur in depression (Byrum et al., 1999; Caligiuri & Ellwanger, 2000; Meyer et al., 2001, 2006; Newberg et al., 2007). In this framework, the putamen is part of a complex system in which cortical pathways project to the striatum (along with neuronal inputs from the thalamus and amygdala) and then project back to the cortex. Several of these pathways involve circuits that modulate mood and behavior, and alterations in striatal dopamine could impair the cortico-striatal circuits disrupting limbic input back to the cortex in depression (Byrum et al., 1999; Drevets, 2000). Thus, the abnormalities observed in our study of DAT binding in depression may represent striatal dopamine mediated dysregulation of the cortical and limbic structures.

Several caveats should be considered in the interpretation of the current findings. For example, DAT density may be influenced by gender (Lammers et al., 1999; Koch et al., 2007) and age (Malison et al., 1998; Laasonen-Balk et al., 1999; Mozley et al., 2000, 2001). In the current study, women were included without regard to menopausal status or menstrual cycle phase. This may have influenced DVR values (Lammers et al., 1999). Similarly, the broad age range of our subjects could have influenced DVR values - although analyses were adjusted for age and gender.

It is possible that diagnostic heterogeneity may have influenced DVR values. We included subjects with unipolar or bipolar II major depression – although subjects with prior mania (i.e., bipolar I disorder) were specifically excluded. Prior examination of [^99mTc]TRODAT-1 binding in 10 unipolar versus 5 bipolar II depressed subjects found no significant difference in putamen DVR values between diagnostic groups (Amsterdam & Newberg, 2007) suggesting that the increased DAT binding in both groups compared to controls is more likely a biomarker of depression. This appears to be confirmed in the current study which found no difference in DVR values between unipolar and bipolar II subjects. Thus, there appears to be a specific relationship between depression and higher DAT binding suggesting that there is altered dopamine function in depression.

It is possible that prior antidepressant exposure may have influenced DVR values, despite the fact that subjects in the current study were drug free for ≥6 months. While we found no influence of prior psychotropic medication exposure on DVR values, Kugaya et al. (2003) found that citalopram increased striatal DAT density by 15% after 8 days of administration and they reported increased DAT density during paroxetine administration. Similarly, Shang et al. (2007) reported increased [¹²³I]β-CIT binding to DAT sites with venlafaxine, while Pirker et al. (1995) found no change in DAT binding during antidepressant administration.

It is possible that differences in illness length, episode duration, symptom severity, or psychosocial stress may have influenced DVR values. Differences in stamina, daily activity level, sleep, and circadian rhythms may influence DVR values. Finally, the extent of tobacco and social alcohol consumption was not specifically controlled in the current study. It is possible that smoking and social alcohol use may have influenced DVR values (Middleton et al., 2004; Cosgrove et al., 2009).

5. CONCLUSION

We used [^99mTc]TRODAT-1 and SPECT imaging to measure striatal DAT density in depressed versus non-depressed subjects. The finding of greater striatal DAT density in depressed subjects validates our prior observation, and suggests that depression may be associated with an increase in DAT density. Moreover, greater striatal DAT density may represent a putative biomarker of depression with the predictive ability to distinguish depressed from non-depressed individuals. Future studies will be needed to confirm these observations.

Acknowledgments

Footnotes

References