Change Over Time in Brain Serotonin Transporter Binding in Major Depression: Effects of Therapy Measured With [¹²³I]-ADAM SPECT

Abstract

Several studies have reported low brain serotonin transporter (SERT) binding in individuals with major depression. We hypothesized that the SERT standardized uptake ratio (SUR) values using [¹²³I]-ADAM single photon emission computed tomography would increase in depressed subjects who responded to cognitive behavior therapy (CBT) compared to CBT nonresponders. [¹²³I]-ADAM scans were acquired before and after 12 weeks of CBT from 20 depressed subjects and on two occasions 12 weeks apart from 10 non-depressed, healthy volunteers. The primary outcome measure was change over time in SUR values in the midbrain, medial temporal lobe, and basal ganglia regions. Depressed subjects demonstrated low pre-treatment mean SUR values that significantly increased over time in the midbrain (p=0.011), right medial temporal lobe (p=0.008), and left medial temporal lobe (p=0.000) regions. Treatment responders showed a significant increase over time in SUR values in left medial temporal lobe (p=0.029) and right medial temporal lobe (p=0.007) regions. Partial and nonresponder subjects also showed a significant increase over time in SUR values in the left medial temporal region (p=0.040) (versus healthy volunteers), but to a lesser degree. The findings suggest that low pre-treatment SERT binding may increase over time in some depressed individuals who experience symptom improvement.

Introduction

Several studies have demonstrated low serotonin transporter (SERT) binding in specific brain regions of depressed individuals.^1-9 For example, Malison et al.¹ and Eggers et al.⁶ found low SERT binding in depressed subjects using [¹²³I]-β-CIT, a non-selective SERT and dopamine transporter radioligand using single photon emission computed tomography (SPECT). However, these observations have not been universally observed.^10-15 For example, Reivich et al.¹³ used the selective SERT radioligand, [¹¹C](+)McN5652 and positron emission tomography (PET) and found greater SERT binding in the temporal lobe region of depressed subjects (versus healthy volunteers). In contrast, other investigators using [¹¹C](+)McN5652 have reported low SERT binding in depressed subjects.^5,15 In addition, Ruhé et al.⁸ reported low midbrain SERT binding in depressed males (versus healthy volunteers) using [¹²³I] β-CIT SPECT; while Kalbitzer et al.¹⁶ found no difference in SERT binding in depressed (versus healthy) women, and an increase in midbrain SERT binding in the winter (versus other seasons).

We previously used the selective radioligand ¹²³I-labeled ((2-((dimethylamino)methyl) phenyl)thio)-5-iodophenylamine) ([¹²³I]-ADAM) to examine brain SERT binding in depressed individuals.³ In a preliminary study of 7 depressed subjects and 6 healthy volunteers, we reported low midbrain SERT binding in depressed (versus healthy) subjects (p=0.01).³ More recently, we replicated these findings in a new cohort of 20 depressed subjects versus 10 healthy volunteers¹⁷ - although other [¹²³I]-ADAM studies have not substantiated these findings.^18,19

The current study examines the change over time in SERT binding standardized uptake ratio (SUR) values before and after treatment with cognitive behavior therapy (CBT) in drug free depressed subjects. We compared these results with SUR values obtained from untreated, healthy volunteers studied with other [¹²³I]-ADAM SPECT on two separate occasions. We hypothesized that reduced SUR values would increase over time in treatment responders and would show little or no change over time in partial or nonresponders.

Materials and Methods

Informed Consent

Subjects provided written informed consent in accordance with the ethical standards of the Institutional Review Board (IRB) of the University of Pennsylvania. The study was conducted under IND #65,542 for [¹²³I]-ADAM using Good Clinical Practice guidelines with oversight by the local Office of Human Research and an independent Data & Safety Monitoring Board.

Subjects

Outpatient subjects ≥18 years old with a DSM IV-TR Axis I diagnosis of major depressive disorder (single or recurrent episode) were enrolled. Subjects were self-referred from IRB-approved media advertisements. The diagnosis was verified using the Structured Clinical Interview for the DSM-IV-TR format.²⁰ Subjects were drug free ≥ 12 months, and had a pre-treatment 17-item Hamilton Depression Rating (HAM-D) ²¹ total score ≥16. Subjects underwent a physical examination and laboratory evaluation, and were in good health without meaningful medical conditions or laboratory abnormalities. Women of child-bearing potential had a negative pregnancy test. Subjects were excluded from the study if they met any of the following criteria: participant in prior [¹²³I]-ADAM study; primary Axis I diagnosis other than major depressive disorder; history of mania or psychosis; actively suicidal; substance abuse or dependence within the preceding 3 months; positive screen for illicit drugs; unstable medical condition; pregnant or nursing; or history of transient ischemic attack, cerebral infarction, hypertensive encephalopathy, intracranial hemorrhage, closed head trauma with loss of consciousness, encephalitis, neurotoxin exposure, normal pressure hydrocephalus, brain tumor, basal ganglia disease, polyneuropathy, or unable to provide informed consent.

Healthy volunteers were recruited from IRB-approved media advertisements. The purpose of including this control group was to demonstrate the natural variability in [¹²³I]-ADAM uptake over time and to compare this value with the change over time in SERT binding observed in depressed subjects. Healthy subjects were ≥18 years old, had no current DSM IV-TR Axis I disorder (verified using the SCID format), and had a baseline total HAM-D score ≤6. None of the healthy volunteers had any clinically meaningful medical conditions or laboratory abnormalities.

Imaging Procedures

Depressed subjects underwent [¹²³I]-ADAM scanning sessions on two separate occasions approximately 12 weeks apart: prior to initiating CBT and within 2 weeks of completing CBT. Healthy volunteers also underwent two separate [¹²³I]-ADAM scanning sessions approximately 12 weeks apart. [Note – Two depressed subjects prematurely discontinued CBT and had their post-CBT scanning session performed at week 9 of the study. To comport with this time frame, two healthy volunteers also underwent their second scanning session at week 9 of the study].

At each scanning session, subjects were administered 18 drops of concentrated Lugol’s solution in order to block ¹²³I uptake by the thyroid gland. [¹²³I]-ADAM 185 MBq (5 mCi) was injected through an intravenous catheter. Four hours after [¹²³I]-ADAM administration, SPECT images were acquired over 60 minutes. Prior kinetic modeling with [¹²³I]-ADAM indicated that the 4-hour delay in image acquisition allowed for the use of the reference region method for estimating SERT binding, without the need for arterial sampling.²²

Image Analysis

SPECT images were analyzed using previously validated methods.²³ Images were reconstructed using a low pass filter and attenuation correction. All scans were re-sliced in the same plane using oblique reformatting. Manual demarcation of the region of interest (ROI) was then performed on the baseline scans focusing on the basal ganglia, midbrain, and medial temporal lobe regions – as these were the regions with low SERT binding from our prior ¹²³I]-ADAM studies. Standardized templates containing ROIs were fit on each scan using previously reported techniques.³ These templates were originally developed using anatomically defined regions based on a magnetic resonance imaging atlas of ROIs. Within the x-y plane, the ROIs in the template are smaller than the actual structures they represent in order to minimize resolution induced problems with ill-defined edges. To reduce the effects of volume averaging in the axial direction, the small ROIs were not placed on the slices that contained the upper most and lower most portions of the structures they represented. This limits the small ROIs to the central aspect of the structures they represented. The ROIs are therefore placed on the pre-treatment scans within each brain structure in order to obtain the mean counts per voxel. ROIs were then directly placed onto the post-treatment scans since all scans were oriented and sliced in the same manner. All ROIs were placed by an expert in nuclear medicine image interpretation and analysis who was blinded to the diagnosis and treatment status.

The primary imaging outcome measure was the SUR value at 4 hours post [¹²³I]-ADAM injection, when the distribution of [¹²³I]-ADAM approached a near equilibrium state that reflected the ratio of k3/k4, which was related to [¹²³I]-ADAM binding potential. The SUR value was calculated as the ROI ÷ reference region where the reference region was the cerebellum which consists of non-specific binding, as described previously.²² [Note - Statistical Parametric Mapping (SPM) and other approaches to image analysis were considered. However, we chose the ROI technique because it has previously demonstrated highly accurate values in quantifying [¹²³I]-ADAM binding in humans.^3,17,22,23 This procedure allowed us to compare [¹²³I]-ADAM uptake in the target regions to uptake in regions of non-specific [¹²³I]-ADAM binding, and to calculate individual SUR values. Moreover, since the region of [¹²³I]-ADAM uptake is quite specific, we would have needed to use the ROI tool contained in the SPM program for our calculations. In addition, a voxel by voxel analysis would not have yielded additional information because [¹²³I]-ADAM binds to only a limited number of brain structures].

CBT Procedures

Treatment was conducted by a senior therapist at the Center for Cognitive Therapy at the University of Pennsylvania. CBT was administered in a structured fashion according to Beck et al.²⁴, and consisted of a series of active directive sessions targeted at promoting behavioral activation and counter-acting maladaptive cognitive biases. As CBT progressed, the emphasis shifted to identification and evaluation of underlying beliefs and schemas. There was an emphasis on cognitive-behavioral skills-training (including homework assignments) designed to allow individuals to gain more independent functioning as their own ‘therapists’ by the end of treatment. These strategies were supplemented by techniques developed to prevent relapse following response to CBT.²⁵ Within this framework, CBT was provided in a flexible fashion as determined by the needs of the individual. Sessions were scheduled twice weekly during the initial 4 weeks (when possible) and weekly thereafter through week 12. Sessions typically lasted 50 minutes.

Outcome Measures

The primary outcome measure was the change over time in mean SUR values of the midbrain, medial temporal lobe, and basal ganglia regions in CBT responders versus healthy volunteers compared to CBT partial and nonresponders versus healthy volunteers. Response was defined as ≥ 50% reduction in baseline total HAM-D score. Partial and nonresponse was defined as < 50% reduction in baseline total HAM-D score.

Statistical Procedures

Analyses were implemented with the realization that the limited sample size may only allow for the detection of large differences between groups. All analyses were conducted using Stata 11.0 (College Station, TX) with two-sided tests of hypotheses and a p-value <0.05 as the criterion for statistical significance. Analyses included means, medians, ranges, and standard deviation (SD) of continuous covariates (e.g. age) and SUR values. The ‘sktest’ procedure in Stata was used to assess the normality of SUR values for each ROI. The intra-subject association of SUR values was estimated using the Spearman rank correlation coefficient test for each ROI.

T-tests and non-parametric Wilcoxon rank sum tests were used to compare the change over time in mean SUR values for depressed subjects and for depressed subjects versus healthy volunteers. ANOVA was used to compare the change over time in mean SUR values before and after CBT in responders versus partial and nonresponders (relative to the change over time in mean SUR values in healthy volunteers). Where significant differences occurred in the change over time in SUR values between depressed and healthy subjects, post hoc group comparisons were examined using the Scheffe multiple comparison test.

Sample Size Justification

Sample size calculations for the primary study outcome was conducted using Nquery Advisor sample size software and Diggle et al.²⁶ to obtain estimates that could be used to power a larger follow up study. Preliminary analyses yielded mean (SD) SUR values of 1.81 (0.07) and 1.95 (0.13) in depressed and healthy subjects, respectively. Assuming a common SD of 0.10 would yield an effect size difference of |1.81 − 1.95|/0.10 = 1.40. A sample size of 20 depressed subjects versus 10 healthy volunteers would have 80% power to detect an effect size of 1.124 using a two group t-test with a 0.05 two-sided significance level.

Results

Enrollment

We examined 20 depressed subjects and 10 healthy volunteers. There were no screen failures. Demographic and clinical characteristics of the subject cohorts are displayed in Tables 1 and 2. Ten depressed subjects (2 women) were drug naïve and 10 (5 women) previously received 3.0 (2.0) (range 1 - 10) antidepressant treatments over the course of their illness.

Table 1.

Depressed subjects’ demographic and clinical characteristics.

Subject	Age	Gender	Age 1st Episode	Episode #*	Episode Duration**	Pre-CBT HAM-D	Post-CBT HAM-D
1	29	M	20	4	3	20	8
2	39	M	27	5	7	16	4
3	28	M	26	0	3	23	24
4	35	F	14	10	2	21	18
5	62	M	19	6	40	16	8
6	40	M	32	0	96	25	24
7	44	M	40	0	42	16	9
8	28	F	16	4	3	23	6
9	41	M	32	5	6	21	9
10	53	M	15	10	8	19	11
11	26	M	17	1	18	21	21
12	61	M	17	5	7	17	14
13	26	M	15	0	120	23	4
14	56	M	16	4	6	22	19
15	31	M	13	0	12	20	6
16	58	M	20	0	3	20	0
17	25	F	25	0	34	19	14
18	33	F	17	1	8	22	7
19	26	M	20	4	30	20	22
20	43	F	20	7	6	21	4
Mean	41.0		21.1	3.3	22.7	20.3	11.6
SD***	12.8		7.1	3.3	32.1	2.5	7.4
Range	25–63		14-40	0-10	2-120	16-25	4-24

^*Episode # = Number of prior major depressive disorder episodes

^**Episode duration = Duration in months of the current major depressive disorder episode

^***SD = Standard deviation

Table 2.

Healthy volunteers’ demographic and clinical characteristics.

Subject	Age	Gender	HAM-D Score #1	HAM-D Score #2
1	34	F	2	2
2	53	F	0	0
3	56	F	0	0
4	40	M	0	0
5	48	M	1	1
6	26	M	0	0
7	36	M	0	0
8	62	M	0	0
9	47	M	0	0
10	46	M	0	0
Mean	44.8	-	0.3	0.3
SD	10.93	-	0.7	0.7
Range	20-62	-	0-1	0-2

Baseline SUR Values

Depressed subjects demonstrated significantly lower SUR values for the midbrain (p<0.005), right medial temporal lobe (p<0.0005), left medial temporal lobe (p<0.004), right basal ganglia (p<0.03), and left basal ganglia (p=0.016) regions (versus healthy volunteers).¹⁷ There was no effect of age, gender, illness duration, prior antidepressant drug exposure, or symptom severity on mean baseline SUR values.¹⁷

Change in Mean SUR Values

Overall, depressed subjects demonstrated a significant increase in mean SUR values for the midbrain (p=0.011), right medial temporal lobe (p=0.008), and left medial temporal lobe (p=0.0001) regions versus healthy volunteers (Table 3). Healthy volunteers demonstrated a modest (albeit non-significant) decrease over time in mean SUR values after repeated [¹²³I]-ADAM testing (Figure 1 and Table 3).

Table 3.

Mean change over time in SUR values in all depressed subject vs. healthy volunteers.

	Healthy Volunteers (n=10)		Depressed Subjects (n=20)		T test	Wilcoxon test
	SUR Change	95% CI	SUR Change	95% CI	Pr (T > t)	Pro >|z|
RBG †	-0.035 (0.159)	-0.148; 0.079	0.069 (0.25)	-0.049; 0.187	0.123	0.455
LBG †	0.021 (0.156)	-0.091; 0.133	0.098 (0.20)	0.004; 0.192	0.150	0.356
Midbrain	-.227 (0.227)	-0.390; -0.065	0.029(0.29)	-0.107; 0.166	0.011	0.008
RMTL †	-0.085 (0.464)	-0.417; 0.248	0.33(0.40)	0.147; 0.522	0.008	0.003
LMTL †	-0.127 (0.176)	-0.253; -0.001	0.17 (0.24)	0.059; 0.285	0.000	0.002

^†RBG = Right Basal Ganglia; LBG = LEFT Basal Ganglia; RMTL = Right Medial Temporal Lobe; LMTL = Left Medial Temporal Lobe

Figure 1.

SUR values for the right medial temporal lobe for CBT responders, CBT nonresponders, and healthy volunteers.

Change in Mean SUR Values Relative to Treatment Response

ANOVA demonstrated significant group differences over time in mean SUR values for the left medial temporal lobe (p=0.029) and right medial temporal lobe (p=0.007) regions, and a small (albeit non-significant) difference for the midbrain region (p=0.076) (Table 4). Scheffe post hoc tests showed a significant increase over time in mean SUR values for treatment responders (versus healthy volunteers) for the right medial temporal lobe (p=0.029) and left medial temporal lobe (p=0.012) regions. Scheffe test also showed a significant increase over time in mean SUR values for partial and nonresponders (versus healthy volunteers) for the left medial temporal lobe region (p=0.040) (Table 3). [Note - Some depressed subjects classified as partial and nonresponders demonstrated a clinically meaningful reduction in total HAM-D scores, but failed to achieve the criterion for ‘response’].

Table 4.

Mean (SD) change in SUR values in responders and nonresponders vs. healthy subjects.

	Control (n=10)	Nonresponder (n=10)	Responder (n=10)	F	p-value
RBG †	-0.035 (0.158)	-0.001 (0.226)	0.139 (0.267)	1.73	0.196
LBG †	0.021 (0.156)	0.081 (0.188)	0.114 (0.222)	1.61	0.196
Midbrain	-0.227 (0.227)	0.026 (0.361)	0.033 (0.223)	2.85	0.076
RMTL †	-0.085 (0.464)	0.222 (0.369)	0.448 (0.418)	4.07	0.029*
LMTL †	0.127 (0.176)	0.144 (0.233)	0.200 (0.259)	6.04	0.007**,#

^†RBG = Right Basal Ganglia; LBG = Left Basal Ganglia; RMTL = Right Medial Temporal Lobe; LMTL = Left Medial Temporal Lobe

^*Scheffe multiple comparison test showed a significant change over time in mean SUR values in CBT responders versus healthy volunteers for the right medial temporal lobe region (p=0.029).

^**Scheffe multiple comparison test showed a significant change over time in mean SUR values in CBT responders versus healthy volunteers for the left medial temporal lobe region (p=0.012).

^#Scheffe multiple comparison test showed a significant change over time in mean SUR values in CBT nonresponders versus healthy volunteers for the left medial temporal lobe region (p=0.040).

Adverse Events

There were no serious adverse events. Five subjects did experience tingling at the site of [¹²³I]-ADAM injection. Thirteen subjects reported an unusual taste and 15 subjects reported unusual smell after ¹²³I-ADAM injection, which dissipated within one minute.

Discussion

We previously reported significantly lower baseline mean SERT binding levels in the midbrain, medial temporal lobe, and basal ganglia regions of drug free depressed subjects (versus healthy volunteers) using [¹²³I]-ADAM.^3,17 These findings support other studies showing low SERT binding in depressed subjects using other SERT radioligands (e.g., [¹²³I]-β-CIT)^1,2,4-9 - although this has not been a universal observation.^11-14 In the current study, we examined the effects of treatment outcome on SERT binding in depressed subjects. We selected CBT as a non-pharmacological treatment intervention that would not directly affect SERT binding.

Several studies have examined the relationship of SERT binding to treatment response with selective serotonin reuptake inhibitor (SSRI) antidepressants. In general, these studies have reported a significant reduction in baseline SERT binding during SSRI treatment.^18,27 For example, Kugaya et al.¹¹ used [¹²³I] β-CIT to examine the relationship between baseline SERT binding and response to fluoxetine therapy in 23 depressed subjects and found a significant relationship after 4 weeks of treatment (β=9.30; p=0.028), but not after 6 weeks of treatment (β=2.22; p=0.42). Similar results were reported in 10 depressed subjects treated with paroxetine.¹¹ Hsieh et al.²⁸ used [¹²³I]-ADAM to examine SERT binding in 13 drug free subjects whose depression had responded to SSRI treatment and found no significant difference in mean SUR values between euthymic depressed subjects (versus 26 healthy volunteers). These investigators suggested that their observations provided indirect evidence that low baseline SERT binding increased (i.e., ‘normalized’) after response to treatment.

Other studies have examined the effects of psychotherapy on SERT binding in depression. For example, Viinamýki et al.²⁹ used [¹²³I] β-CIT to examine the effect of dynamic psychotherapy on SERT binding in 2 depressed subjects. They reported low baseline SERT binding in the prefrontal cortex and thalamus that ‘normalized’ after one year of psychotherapy (relative to non-psychotherapy subjects). Martin et al.³⁰ used [^99mTc]-exametazime hexamethylpropyleneamine oxime SPECT to examine regional cerebral blood flow (rCBF) in 28 depressed subjects. They found an increase in rCBF in the basal ganglia region following response to either interpersonal psychotherapy or venlafaxine therapy. Increases in rCBF were observed in the temporal lobe region with venlafaxine and in the posterior cingulate gyrus with interpersonal psychotherapy. However, neither of these studies employed controlled methodology, and the response to pharmacotherapy was superior to interpersonal psychotherapy in both studies.

Finally, several studies used [¹⁸F]-fluoro-deoxyglucose (FDG) PET to examine the effect of psychotherapy or pharmacotherapy on brain metabolism in depression. Brody et al.³¹ studied the relationship between baseline glucose metabolism and response to treatment with either interpersonal psychotherapy or paroxetine in 24 depressed subjects. They reported that low baseline prefrontal glucose metabolism predicted response to both treatment modalities. Goldapple et al.³² examined changes over time in glucose metabolism in 14 responders to CBT versus 13 responders to paroxetine. Paroxetine responders demonstrated an increase in glucose metabolism in the dorsolateral-prefrontal cortex and a reduction in glucose metabolism in the hippocampus, while CBT responders showed a reduction in glucose metabolism in the dorsolateral-prefrontal cortex and an increase in glucose metabolism in the hippocampus and dorsal cingulate regions. These investigators speculated that CBT produced a ‘top down’ effect while antidepressants produced a ‘bottom up’ effect, with initial change occurring in midbrain neurotransmitter activity.³²

We speculate that the current [¹²³I]-ADAM observations may support the notion of Goldapple et al.³² of a ‘top-down’ effect of CBT on SERT binding. The mechanism by which this physiological process occurs is not well understood. However, the SERT site is known to remove excess serotonin from the synaptic cleft,¹⁷ and it.is possible that the low SERT binding seen during depression may reflect lower brain serotonin levels via a compensatory SERT down-regulation. Thus, low SERT binding in depression may reflect an overall low serotonin function in depression that “normalizes’ during response to treatment. A comparative study of CBT and SSRI therapy with repeated [¹²³I]-ADAM scan sessions over time would be necessary to confirm a ‘top down’ (versus ‘bottom up’) effect of CBT.

Several caveats should be considered when interpreting the current observations. For example, it is possible that the increase in low SERT binding in depressed subjects may have occurred independently of CBT. As no comparative treatment intervention (e.g., SSRI) was included in the current study, it is possible that the observed increase in SERT binding occurred as a result of symptom reduction per se or from non-specific aspects of treatment (i.e., placebo effect).

The current study did not include comparison groups of depressed subjects who did not receive CBT and healthy volunteers who did receive CBT. While the inclusion of these groups would have been of heuristic interest, we believe that their inclusion would have raised substantial ethical and procedural difficulties. For example, the inclusion of an untreated depressed subject group would have been unethical. Moreover, the inclusion of healthy volunteers ‘treated’ with CBT would have raised ethical concerns such as exposing non-depressed, healthy subjects to a therapy specifically designed to treat depressive symptoms with no expected benefit or measurable outcome.

We also note that an increase in SERT binding after treatment occurred in both CBT responder and partial and nonresponder groups. This may have resulted from some subjects in the partial and nonresponder group having a clinically meaningful reduction in depressive symptoms and an increase in SERT binding – despite the fact that these subjects failed to achieve a priori criteria for ‘response’. Similarly, it is possible that a larger difference in SERT binding between groups would have been more evident had a larger sample size been studied, or a different a priori criteria for ‘response’ had been employed.

There was a modest, albeit non-significant, reduction over time in mean SUR values for healthy volunteers, and it is possible that this reduction in SERT binding contributed to the significance of the increase in SUR values seen in depressed subjects.

It is likely that the change over time in SERT binding occurred gradually and at varying rates among depressed subjects. Thus, the change over time in SERT binding in some depressed subjects may have lagged behind symptom improvement. As some depressed subjects were more severely ill than others, and some subjects did not receive a full course of CBT, it is also possible that some of these subjects may have had a more gradual increase in SUR values than other depressed subjects. It is also possible that the group differences would have been greater had a longer treatment course of CBT been applied.

Other factors may have contributed to variability in the SERT binding results. For example, there were differences in the time that elapsed between [¹²³I]-ADAM imaging sessions in 2 subjects. In addition, we did not control for possible seasonal effects of SERT binding.⁸ Although the current study found no influence of age on mean SUR values¹⁷, other studies have reported an effect of age on SERT binding.^1,3

It is possible that comorbid anxiety and/or prior exposure to antidepressant medication could have influenced SERT binding^18,33, although an exclusion criteria was that patients could not have antidepressant medication for at least 12 months. In addition, differences in illness length, episode duration, symptom severity, environmental stress, sleep, carbohydrate intake, circadian rhythms, smoking, and alcohol use may also have influenced the current SERT binding results.^2,8,16,34-38

Finally, it is possible that the current observations occurred by chance alone and were not related to treatment response or the degree of symptom improvement. Future placebo-controlled studies will be needed to determine whether or not treatment response is related to change over time in SERT binding in depression.

Conclusion

The findings of the current study suggest that low pre-treatment SERT binding in depression may increase over time with treatment response. However, a limited subject sample size and large treatment response range hampered our ability to determine whether or not the increase over time in SERT binding in depressed subjects was due to treatment response per se. Future studies with larger sample sizes will be needed to better determine the relationship between SERT binding and treatment in patients with depression.