Methylparatyrosine Induced Depressive Symptoms

ORIGINAL CONTRIBUTION

Regional Brain Metabolic Correlates of ␣-Methylparatyrosine–Induced Depressive Symptoms Implications for the Neural Circuitry of Depression

J. Douglas Bremner, MD Context We previously used positron emission tomography (PET) measurement of Meena Vythilingam, MD brain metabolism with 18fluorodeoxyglucose to show that patients receiving selective Chin K. Ng, PhD serotonin reuptake inhibitors (SSRIs) who have a tryptophan depletion–induced return of depressive symptoms have an acute decrease in metabolism in orbitofrontal Eric Vermetten, MD cortex, dorsolateral prefrontal cortex, and thalamus. Many patients with depression Ahsan Nazeer, MD in remission while taking norepinephrine reuptake inhibitors (NRIs) (but not SSRIs) experience a return of depressive symptoms with depletion of norepinephrine and do- Dan A. Oren, MD pamine using ␣-methylparatyrosine (AMPT). Robert M. Berman, MD Objective To assess brain metabolic correlates of AMPT administration in patients Dennis S. Charney, MD with depression in remission while receiving NRIs. Design, Setting, and Participants Randomized, controlled, double-blind trial in AJOR DEPRESSION IS AN IM- which 18 patients recruited in 1997-2000 from the general community who had de- portant public health pression in remission while taking NRIs had PET imaging in a psychiatric research unit following AMPT and placebo administration. problem that affects about 16% of people in the Interventions After initial medication with desipramine and follow-up until re- UnitedM States at some time in their lives.1 sponse, patients underwent active AMPT (five 1-g doses administered orally over 28 A major effort of the past few decades hours) and placebo (diphenhydramine hydrochloride, five 50- mg doses administered similarly) catecholamine depletion challenges in randomized order of assignment, af- has been elucidating biological mecha- ter which PET imaging was performed on day 3 of each condition. Both study condi- nisms that underlie symptoms of de- tions were performed 1 week apart. pression. Understanding the biology of Main Outcome Measures Regional brain metabolism rates in patients with and depression may help in the treatment of without AMPT-induced return of depressive symptoms. this disabling disorder. Two classic biological models for depression have been Results AMPT-induced return of depressive symptoms was experienced by 11 of the 18 patients and led to decreased brain metabolism in a number of cortical areas, the serotonin and norepinephrine hy- with the greatest magnitude of effects in orbitofrontal (P=.002) and dorsolateral potheses of depression. prefrontal (P=.03) cortex and thalamus (P=.006). Increased resting metabolism in Multiple lines of evidence support the prefrontal and limbic areas predicted vulnerability to return of depressive symp- serotonergic hypothesis of depres- toms. 2,3 sion. Serotonergic neurons have their Conclusions Different neurochemical systems that mediate depression may have cell bodies in the brain stem (dorsal effects on a common brain circuitry. Baseline metabolism in successfully treated de- raphe). These cell bodies give rise to pressed patients may predict vulnerability to future episodes of depression. long axons that project throughout the JAMA. 2003;289:3125-3134 www.jama.com brain. Treatment with selective serotonin reuptake inhibitors (SSRIs) is efficacious for treatment of depression. Author Affiliations: Departments of Psychiatry and (Drs Vythilingam and Charney); Department of Radi- Studies have found a decrease in the Behavioral Sciences and Radiology and the Emory Cen- ology, University of Louisville, Louisville, Ky (Dr Ng); and ter for Positron Emission Tomography, Emory Univer- Department of Psychiatry, Yale University School of serotonin metabolite 5-hydroxyindole- sity School of Medicine, Atlanta, Ga (Drs Bremner, Ver- Medicine, New Haven, Conn (Drs Oren and Berman). acetic acid in the cerebrospinal fluid4,5 metten, and Nazeer); Atlanta Veterans Affairs Medical Corresponding Author and Reprints: J. Douglas Brem- 6 Center, Decatur, Ga (Drs Bremner, Vermetten, and ner, MD, PET Center/Nuclear Medicine, Emory Uni- and in postmortem brain, decreased Nazeer); Mood and Anxiety Disorders Program, Na- versity Hospital, 1256 Briarcliff Rd, Atlanta, GA 30306 serotonin concentrations in platelets,7 tional Institute of Mental Health, Bethesda, Md (e-mail: [email protected]).

Downloaded From: https://jamanetwork.com/ on 09/27/2021 CORRELATES OF AMPT-INDUCED DEPRESSIVE SYMPTOMS

increased 5-hydroxytryptamine recep- which is the time frame of action of the phenotypical outcome of depression. It tor binding in frontal cortex in some antidepressive effects.25 These medica- is not clear if there are 2 separate eti- studies8-11 but not in others,12 and re- tions lead to a reduction in whole- ologies for depression (eg, serotoner- duction of prolactin response to the body norepinephrine turnover in pa- gic or noradrenergic dysfunction) or serotonin agents fenfluramine13 and clo- tients treated for depression.26 Other whether these diverse neurochemical mipramine14 and the serotonin precur- evidence for alterations in noradrener- systems act on common brain regions sor L-tryptophan15 in patients with de- gic function in depressed patients in- and/or have common postsynaptic ef- pression. Dietary depletion of cludes a well-replicated finding of fects that are involved in the media- tryptophan, an amino acid that is the blunted growth hormone response to tion of symptoms of depression.42 ␣ precursor of serotonin, is associated the 2 agonist clonidine (the growth Brain imaging studies have begun to with decreased levels of plasma tryp- hormone response to clonidine is be- map out a common circuit of brain re- tophan16 as well as a decrease in brain lieved to be related to postsynaptic gions that are thought to mediate symp- 17 ␣ 27-29 serotonin levels. Tryptophan deple- stimulation of the 2 receptor). Post- toms of depression. Deficits in hippo- tion also results in a transient return of mortem studies of the brains of de- campal structure and function have depressive symptoms in more than half pressed patients found decreased den- been hypothesized to underlie both the of patients who have been success- sity of the noradrenergic transporter 30 deficits in emotional dysregulation as fully treated with SSRIs,18-20 but deple- and upregulation of tyrosine hydroxy- well as cognitive dysfunction associ- tion does not worsen mood in healthy lase.31 Studies of norepinephrine me- ated with depression.43 Deficits in pre- persons16 or in patients with un- tabolites in urine and cerebrospinal frontal cortical function have also been treated depression.21 Patients with fluid in patients with major depres- hypothesized to underlie symptoms of treated depression who have stopped sion have been inconclusive.23 depression.44,45 Areas of prefrontal cor- SSRIs have a return of symptoms of de- ␣-Methylparatyrosine (AMPT) is a tex implicated in depression include pression with tryptophan depletion.22 competitive inhibitor of the rate- dorsolateral prefrontal cortex (in- These studies suggest that alterations limiting enzyme of catecholamine syn- volved in working memory) and me- in serotonergic function mediate, at thesis, tyrosine hydroxylase.32 AMPT re- dial prefrontal cortex (mediating least in part, symptoms of depression. sults in a decrease in urine,33,34 emotion), which consists of several re- Alterations in noradrenergic func- plasma,35-38 and cerebrospinal39 levels lated areas, including orbitofrontal cor- tion have also been hypothesized to un- of catecholamine metabolites; it has no tex, anterior cingulate (area 25, sub- derlie symptoms of depression.3,23 The effect on mood in healthy persons36,40 callosal gyrus; area 24, subgenual gyrus; majority of the noradrenergic neu- or in drug-free depressed patients,35 but and area 32, the Stroop area), and an- rons in the brain have their cell bodies more than half of patients with treated terior prefrontal cortex (area 9). Alter- in the pons (brain stem), with long depression who are taking NRIs have ations in the amygdala have also been axons that project throughout the brain, a transient return of depressive symp- hypothesized in depression. releasing transmitter in multiple cor- toms following AMPT administra- Imaging studies in depression tical and subcortical sites, including pre- tion.37,38 In one study, 71% of drug- showed alterations in hippocampus, frontal, parietal, and sensory cortex and free patients who remitted from amygdala, and prefrontal cortex.43,46,47 hippocampus. The noradrenergic hy- depression had a return of depressive Magnetic resonance imaging (MRI) pothesis of depression was originally symptoms.41 These studies are consis- studies in patients with depression based on the finding that the antihy- tent with alterations in noradrenergic showed smaller hippocampal vol- pertensive medication reserpine, which function underlying symptoms of de- umes and other alterations in hippo- acts by interfering with uptake and stor- pression. campal structure48-56 (although see Pan- age of norepinephrine and dopamine in Both noradrenergic and serotoner- tel et al57 and Ashtari et al58) and smaller intracellular storage vesicles, is associ- gic systems therefore appear to modu- subgenual (anterior cingulate) corti- ated with symptoms of depression in late symptoms of depression. Trypto- cal59 and orbitofrontal cortical60,61 vol- many patients.24 Treatment with selec- phan depletion–induced return of umes. Changes in both amygdala struc- tive norepinephrine reuptake inhibi- depressive symptoms has been ob- ture48,62 and function63,64 were found in tors (NRIs) is also efficacious for treat- served in patients who responded to depression. Multiple positron emis- ment of depression in the majority of SSRIs (but not patients treated with sion tomography (PET) and single- cases. Norepinephrine reuptake inhibi- NRIs),18 while patients with depres- photon emission computed tomogra- ␣ 65-68 tors decrease sensitivity of the 2 au- sion who were successfully treated with phy studies found decreased left and toreceptor and increase sensitivity of the NRIs (but not patients taking SSRIs) bilateral69-71 dorsolateral prefrontal cor- ␣ 1 receptor, leading to increased nor- were susceptible to AMPT-induced re- tex function at baseline in patients with adrenergic transmission. Norepineph- turn of depressive symptoms.38 These untreated major depression. Some PET rine reuptake inhibitors also lead to a studies suggest that different neuro- studies also found decreased metabo- reduction in ␤ receptors after 2 weeks, chemical systems may result in the same lism at baseline in medial prefrontal cor-

Downloaded From: https://jamanetwork.com/ on 09/27/2021 CORRELATES OF AMPT-INDUCED DEPRESSIVE SYMPTOMS

tex/anterior cingulate.47,59,70,72 Other PET fact that the orbitofrontal and dorso- This study was approved by the Yale studies in patients with depression and lateral prefrontal cortex receive impor- University Human Investigation Com- a range of comorbid conditions found tant innervation from the noradrener- mittee. All participants gave written in- decreased metabolism and/or blood gic system (and are involved in a circuit formed consent and were paid for their flow in the prefrontal cortex, includ- with the thalamus) and that decreases participation. Informed consent in- ing orbitofrontal cortex and anterior were observed in these areas with tryp- cluded a description of the possibility cingulate.73-76 Patients with depres- tophan depletion–induced depressive that AMPT would result in a return of sion showed a blunted metabolic re- relapse, we hypothesized that AMPT- depressive symptoms, as well as the pos- sponse to challenge with the seroton- induced relapse would be associated sible adverse effects of AMPT. This study ergic agent fenfluramine in left with a decrease in metabolism in orbi- included provision for supportive treat- prefrontal and temporal-parietal cor- tofrontal cortex, dorsolateral prefron- ment of depression in an inpatient psy- tex relative to controls.77 Successful re- tal cortex, and thalamus. chiatric research unit; however, all par- sponse to treatment was associated with ticipants had a return of normal mood changes in the prefrontal cortex.78,79 METHODS within 6 hours of the last AMPT dose, To assess neural correlates of sero- Participants consistent with prior studies of AMPT, tonergic contributions to depression, we Twenty-three men and women with a which have found no episodes among used PET to measure brain metabo- history of major depression in remis- individuals who had depressive symp- lism in conjunction with administra- sion who were taking NRIs partici- toms to last more than 8 hours after the tion of tryptophan depletion and pla- pated in the study. None had partici- last AMPT dose.35-38,41,81 cebo in patients with depression in pated in a previously published study Twenty-five participants were started remission while taking SSRIs. De- of depression. All were free of major on desipramine, 1 was switched to nor- pressed patients in remission with a medical illness on the basis of history triptyline because of nonresponse, 2 tryptophan depletion–induced return and physical examination, laboratory dropped out because of adverse ef- of depressive symptoms were com- testing, and electrocardiogram, were not fects or nonresponse, 23 started the PET pared with depressed patients with- actively abusing substances or alcohol imaging protocol, and 5 did not com- out a tryptophan depletion–induced re- (in the past 6 months), and were free plete the second test day (FIGURE 1). turn of depressive symptoms. Patients of all medications at initial recruit- Four of these noncompleting partici- with a tryptophan depletion–induced ment. Participants with symptoms of pants received AMPT on the first day return of depressive symptoms (but not depression were recruited by newspa- and 1 received placebo on the first day. nondepressed patients) had decreased per advertisement. Participants were One participant who received AMPT metabolism in dorsolateral prefrontal started on desipramine (if there was no was removed from the protocol be- cortex, orbitofrontal cortex, and thala- contraindication) and followed up by cause of adverse effects of AMPT; the mus on the tryptophan depletion day a psychiatrist in an outpatient research other 3 participants who received in comparison with the placebo day. Re- clinic in an academic setting until they AMPT on day 1 refused or did not fol- lapse-prone patients had increased rest- had a response to medication. After they low up with completion of day 2. These ing (ie, on the placebo day) limbic and had a response to medication as defined participants were clinically noted to prefrontal cortical metabolism.44 What by the Hamilton Depression Rating have had a return of depressive symp- was unanswered by this study was Scale (HDRS) criteria outlined herein, toms with AMPT. The participant who whether the findings were related to se- they were entered into the PET por- received placebo on day 1 did not fol- rotonergic alterations in these brain re- tion of the study. If they did not have a low up with completion of day 2. gions (eg, alterations in postsynaptic response, if possible, they were switched Eighteen participants completed both serotonin receptors or in serotonin neu- to another NRI (nortriptyline). Partici- test conditions (9 women and 9 men). rotransmitter release) or whether there pants with a serious medical or neuro- Of these, 11 had an AMPT-induced re- was a regional brain abnormality (not logical illness, organic mental disor- turn of depressive symptoms and 7 did linked to a specific neurochemical sys- der, comorbid psychotic disorder or not. Mean (SD) age was 43 (13) years. tem) that accounted for the findings. posttraumatic stress disorder based on No women were postmenopausal. All One way to address this question is the Structured Clinical Interview for the participants were right-handed. Par- to probe alternative neurochemical sys- Diagnostic and Statistical Manual of Men- ticipants had been treated with desi- tems that have been implicated in de- tal Disorders, Fourth Edition (SCID),80 pramine or nortriptyline in an open- pression, such as the noradrenergic sys- premenstrual dysphoric disorder, his- label fashion until they achieved tem. The purpose of the current study tory of alcohol or substance abuse or remission of depression with dosage ad- was to assess neural correlates of AMPT- dependence, retained metal, or his- justments made by a study psychia- induced return of depressive symp- tory of head trauma, loss of conscious- trist based on clinical response. Desi- toms in patients with depression in re- ness, cerebral infectious disease, or dys- pramine dosages ranged from 75 to 300 mission while taking NRIs. Based on the lexia were excluded. mg/d (mean dosage, 150 mg/d), with

Downloaded From: https://jamanetwork.com/ on 09/27/2021 CORRELATES OF AMPT-INDUCED DEPRESSIVE SYMPTOMS

AMPT after 3 days of administra- Figure 1. Patient Flow Through the Study tion.35,37,38 Medication was prepared by

25 Participants Started Desipramine Therapy an onsite research pharmacist who assigned participants to an AMPT- 2 Dropped Out placebo or placebo-AMPT test order us- 1 Had Adverse Effects 1 Had No Treatment Response ing a random number generator. 1 Changed to Nortriptyline Therapy Medication was dispensed in num- (No Response to Desipramine Therapy) bered bottles and administered by a psy- 22 Continued Desipramine With Positive chiatrist and research nurse in a re- Treatment Response 1 Continued Nortriptyline With Positive search study unit. All raters, participants, Treatment Response psychiatrists, nurses, imaging person- nel, and data analysts were blinded to as- 23 Randomized signment until the data analysis was com- 14 Randomly Assigned to Receive AMPT 9 Randomly Assigned to Receive Placebo pleted. Participants were recruited Challenge First Challenge First 14 Completed AMPT Challenge as Assigned 9 Completed Placebo Challenge as Assigned between 1997 and 2000. Each study condition involved outpa- 4 Withdrew After AMPT Challenge 1 Withdrew After Placebo Challenge tient visits on 4 days. Behavioral ratings and blood samples for monoamine me- 10 Completed Both AMPT and Placebo Challenges 8 Completed Both Placebo and AMPT Challenges tabolites were obtained daily (8:00 AM to 10 Included In Analysis 8 Included In Analysis 9:00 AM and 3:00 PM to 4:00 PM) during days 2 and 3 and once in the mornings Of 25 patients who started desipramine, 2 dropped out of the study because of adverse effects or nonre- of days 1 and 4. Medication capsules con- sponse, 1 did not respond to desipramine and was switched to nortriptyline, 23 started the positron emission tomography (PET) part of the study, and 18 completed both test days. AMPT indicates ␣-methylparatyrosine. taining AMPT (1 g) or diphenhydramine hydrochloride (50 mg) were given during day 2 (9:00 AM, noon, and 7:00 desipramine blood levels in the 150 to bia. Two patients (11%) had a lifetime PM) and day 3 (9:00 AM and noon). In- 300 ng/mL range. Nortriptyline dos- history of panic disorder without ago- vestigators were blinded to study con- age was 150 mg/d, with blood levels in raphobia; in one of these patients, the dition. Vital signs were obtained 3 times the 50 to 150 ng/mL range. Mean (SD) diagnosis was current. One patient had daily. Daily urinalysis was performed to number of depressive episodes was 17 lifetime (not current) cocaine depen- allow for early detection of potential uri- (28) (range, 1-99). Full remission of de- dence, 2 patients (11%) had lifetime nary crystal formation. To minimize the pression was achieved after a mean (SD) (not current) marijuana dependence, risk of urinary crystal formation, partici- of 7 (8) weeks (range, 2-36 weeks). Par- and 4 patients (22%) had lifetime (not pants drank2Lofwater per day. Posi- ticipants continued taking desipra- current) alcohol dependence. tron emission tomographic scanning was mine after remission until they com- performed at 11:00 AM on day 3. pleted PET scanning and then were AMPT and Placebo Administration Symptoms of depression were mea- referred for follow-up treatment. At the Participants underwent active AMPT sured using the HDRS.82 AMPT- time of scanning, participants had been (five 1-g doses administered orally over induced changes in mood were mea- treated for a mean (SD) of 13 (10) 28 hours) and placebo (diphenhydr- sured as the change in score on the weeks (range, 5-46 weeks). amine hydrochloride, five 50-mg doses HDRS between AMPT baseline (ie, Diagnosis of major depression was administered similarly) catecholamine AMPT day 1) and the time of the PET established using the SCID.80 All of the depletion challenges in a randomized scan (AMPT day 3), subtracting the depressed patients had current and life- double-blind design, with randomized HDRS change from day 1 to day 3 of time unipolar major depression. For 3 order of assignment.35,37,38,41 Both study the placebo session. We have used this patients (17%), this was the first epi- conditions were performed 1 week apart. method of analyzing response in prior sode of depression; the other patients Fourteen of 23 participants underwent studies of tryptophan depletion44 and had recurrent depression. Two of the the AMPT test first. Of the participants AMPT in depression.35,41 Return of de- 18 depressed patients (11%) fulfilled who completed both tests, 9 of 18 un- pressive symptoms was categorized as criteria for a lifetime history of dysthy- derwent the AMPT test first. Diphen- an increase with AMPT of 9 points and mia based on the SCID. One patient hydramine was used as an active pla- a 50% increase from baseline on the (6%) had a lifetime (not current) his- cebo to approximate the level of sedation HDRS defined in this fashion. tory of agoraphobia. One patient had induced by AMPT.35,37,38 The course of a lifetime (not current) history of so- AMPT administration and PET imag- PET and MRI Scanning Methods cial phobia. One patient had a lifetime ing was based on our prior studies show- Two PET scans were performed 1 week (not current) history of simple pho- ing the maximal behavioral effects of apart in conjunction with the AMPT

Downloaded From: https://jamanetwork.com/ on 09/27/2021 CORRELATES OF AMPT-INDUCED DEPRESSIVE SYMPTOMS

and placebo conditions. The PET scans ages were attenuation-corrected based symptoms,44 since a primary aim of the took place on day 3 of the study con- on the transmission scan and recon- current study was to replicate the brain dition at 11:00 AM, 3 hours after the last structed on a Sparc Workstation (Sun findings of that study using provoca- dose of AMPT or placebo. Participants Microsystems, Sunnyvale, Calif). A tion of a different neurochemical sys- were scanned with a Posicam 6.5 PET 20-cm cylindrical fluid-filled phan- tem. The only additional brain re- camera (Positron Corp, Houston, Tex). tom with a known amount of radioac- gions added were a region of The Posicam 6.5 is a 21-slice camera tivity was scanned on the same day to dorsolateral prefrontal cortex more su- with 5.125-mm interslice thickness. In- obtain calibration factors for each of the perior to the area measured in that prior herent resolution in plane is 5.8 mm 21 slices of the camera for conversion study (to correspond to the area where and 11.9 mm in the Z axis; system of radioactivity into units of millicu- alterations in neuronal morphology sensitivity is 165 kilocounts/s per ries per milliliter. were found in postmortem brain89) and µCi/cm3.83 An intravenous line was in- Magnetic resonance imaging scans the subgenual prefrontal cortex (for serted in the hand and warmed with a were obtained in all participants for similar reasons).90 Global brain me- heating pad for measurement of arte- coregistration with PET and determi- tabolism was calculated as the mean of rialized venous blood samples. This nation of regions of interest from MRI brain tissue activity in all slices. method has been shown to yield equiva- scans resliced to correspond to PET lent values of metabolism to arterial line slices. Magnetic resonance imaging Data Analysis placement.84 Participants were then scans in the same participants were ob- Specific regions of interest were se- placed in the scanner with the head held tained on a 1.5-T General Electric Signa lected based on the results of our prior in a head holder to minimize patient scanner (Milwaukee, Wis). Scan acqui- PET study of neural correlates of tryp- motion. The head was positioned with sition included axial slices, 3-mm slice tophan depletion–induced depres- the canthomeatal line parallel to the ex- thickness, T1-weighted images (rep- sion44 and on prior studies of neural cor- ternal laser light. Following position- etition time=25 ms; echo time=5 ms; relates of depression. Regions were ing within the camera gantry, a trans- number of excitations=2; matrix, separated into those that were and were mission scan of the head was obtained 256ϫ256; field of view, 24 cm). not hypothesized to change with AMPT- using an external gallium citrate Ga 67/ induced return of depressive symp- germanium Ge 68 rod source. These Image Processing toms. Regions hypothesized to change data were used to correct emission data and Analysis Methods included dorsolateral prefrontal cortex from attenuation due to overlying bone Positron emission tomography and MRI (middle frontal gyrus), orbitofrontal cor- and soft tissue. Participants then re- scans were transferred to a Sun Work- tex, and thalamus. Other regions from ceived an intravenous injection of 5 mCi station (Sun Microsystems, Mountain the prior study that showed reduction (185 MBq) of 18fluorodeoxyglucose in View, Calif) for analysis. A surface with depressive symptoms after correc- a single bolus followed by scanning of matching algorithm and the ANA- tion for multiple comparisons in- the brain from 30 to 50 minutes after LYZE (Mayo Clinic, Rochester, Minn) cluded parietal and temporal cortex, injection with eyes open in a dimly lit software package, version 5.0, was used postcentral gyrus, superior frontal gy- room. Twenty-three arterialized ve- for coregistration of images.86 Brain sur- rus, and anterior cingulate. nous blood samples were obtained at faces from PET and MRI were matched Data were analyzed for regional brain multiple points after injection for mea- using this program. The MRI was metabolic rates following administra- surement of radioactivity in plasma, resliced to correspond to the 21 PET tion of placebo or AMPT using repeated- used for construction of a plasma time- slices. Using this technique, we have measures analysis of variance with drug activity curve. Three blood samples shown a registration error of 2.86 mm.87 (AMPT vs placebo) as the repeated fac- were also obtained for measurement of Regions of interest were drawn on tor, depression status (those with vs plasma glucose concentrations. There resliced MRI scans using specific cri- without a return of depressive symp- were no differences between plasma teria based on anatomical landmarks toms) (FIGURE 2) and hemisphere (left glucose concentrations during the with a method that we have shown to vs right) as factors in the analysis, and AMPT (85 [SD, 13] mg/100 mL) and be highly reliable. These regions of in- baseline brain metabolism (during the placebo (85 [SD, 13] mg/100 mL) tests. terest use anatomical landmarks from placebo condition) as a covariate. Data Brain and tissue measurements were the MRI. We have previously pub- analysis examined the interaction be- used for calculation of cerebral glu- lished the details and the reliability of tween depression status and drug. Be- cose metabolic rate (in milligrams per these methods.88 Multiple brain re- cause of the highly variable measure- minute per 100 mL) using the for- gions were selected for analysis. These ment of glucose metabolic rate, glucose mula of Sokoloff et al.85 Glucose up- regions correspond to the same reuptake rates were used as the primary take rates were also calculated by di- gions measured in our prior study of outcome. A secondary analysis exam- viding brain activity by the product of neural correlates of tryptophan deple- ined the ratio of regional to whole- body weight and injected dose. Im- tion–induced return of depressive brain metabolism. Bonferroni correc-

Downloaded From: https://jamanetwork.com/ on 09/27/2021 CORRELATES OF AMPT-INDUCED DEPRESSIVE SYMPTOMS

eral prefrontal cortex, orbitofrontal cor- Figure 2. Effects of AMPT and Placebo (Diphenhydramine) on Symptoms of Depression as Measured With the HDRS tex, and thalamus). There were mod- est negative correlations between 35 Challenge increased symptoms of depression and AMPT Challenge ∗ With Relapse decreased brain metabolism with AMPT 30 Without Relapse ∗ for middle frontal gyrus (r =−0.46; Placebo Challenge 25 With Relapse df=17; P=.05), orbitofrontal cortex Without Relapse (r=−0.40; df=17; P=.10), thalamus 20 ∗ (r=−0.44; df=17; P=.07), temporal cortex (r=−0.47; df=17; P=.045), and hip- 15

HDRS Score pocampus (r=−0.53; df=17; P=.03). 10 There were positive correlations between change in brain metabolism with 5 AMPT in middle frontal gyrus, orbito-

0 frontal cortex, and thalamus (range of 9 AM 11 AM 4 PM 9 AM 11 AM 4 PM 9 AM 11 AM 4 PM 9 AM 11 AM 4 PM r values, 0.95-0.97). Day 1 Day 2 Day 3 Day 4 Relationship Between Clinical ␣ Patients are divided into those with (n=11) and without (n=7) an -methylparatyrosine (AMPT)–induced re- Factors and Biological Outcomes turn of depressive symptoms. AMPT (but not placebo) resulted in an increase in depressive symptoms at all points on day 3 of the study (the day of the positron emission tomography scan) in patients defined as having There were no significant relation- a return of depressive symptoms (PϽ.001 as indicated by asterisks). HDRS indicates Hamilton Depression Rat- ing Scale. Error bars indicate SDs. ships between brain metabolism and age or sex. Number of lifetime depressive episodes was significantly correlated tions were performed to correct for toms tended to show the opposite re- with decreased orbitofrontal metabo- multiple comparisons for nonhypoth- sponse (eg, in orbitofrontal cortex; Figure lism with AMPT (r =−0.71; df=11; esized regions (P=.05/10=.005). This 3B). The value for the interaction be- PϽ.05); this effect was primarily re- study was not analyzed as an intention- tween depression status and drug (AMPT lated to a strong correlation in pa- to-treat analysis. vs placebo) for these regions is pre- tients who had a return of depressive Brain behavioral correlations were sented in the Table. There were no sig- symptoms with AMPT (r=−0.74) and performed by comparing the relation- nificant interactions between relapse sta- was not significant in patients who did ship between the placebo-corrected tus and drug for the ratio of regional to not experience a return of depressive change in HDRS with AMPT (as de- global brain metabolic rates. symptoms (r=−0.28). There were no scribed herein) and the changes in re- Patients who had an AMPT-induced differences in brain metabolic regional brain metabolism with AMPT return of depressive symptoms had in- sponses to AMPT by order of assign- and placebo. creased resting metabolic rates in sev- ment to AMPT or placebo. Data were analyzed using SAS statis- eral prefrontal and limbic brain re- tical software, version 8 (SAS Institute gions. These included orbitofrontal COMMENT Inc, Cary, NC). cortex, middle frontal gyrus, hippocam- AMPT-induced return of depressive pus, parahippocampus, and amygdala. symptoms was associated with de- RESULTS There were also elevations in temporal creased metabolism in multiple corti- Effects of AMPT on Regional and parietal cortex not observed in our cal regions, with the greatest effects in Cerebral Metabolism prior study (PϽ.005).44 orbitofrontal cortex, dorsolateral pre- Metabolic (glucose uptake) rate data are frontal cortex, and thalamus. These presented in the TABLE. Patients with Relationship Between Depression frontal regions receive important nor- AMPT-induced return of depressive and Brain Metabolism adrenergic innervation and are in- symptoms showed decreased brain me- The relationship between symptoms of volved in a functional circuit with the tabolism after controlling for baseline depression during PET scanning and thalamus. These findings are consis- metabolism (measured during the pla- brain metabolism was examined by cal- tent with our prior study of trypto- cebo condition) in a number of cortical culating the baseline-corrected change phan depletion–induced return of de- regions with AMPT in comparison with in HDRS score for the AMPT vs pla- pressive symptoms, which found placebo, with the greatest magnitude of cebo tests and correlating that with the decreases in these areas with return of effect in dorsolateral prefrontal cortex, change in regional brain metabolism for depressive symptoms. In the current orbitofrontal cortex (FIGURE 3A), and AMPT and placebo tests. Correlations study, we also found greater decreases thalamus. Patients without AMPT- were examined for brain regions with in temporal and parietal cortex than we induced return of depressive symp- the most significant change (dorsolat- found in that prior study. Greater de-

Downloaded From: https://jamanetwork.com/ on 09/27/2021 CORRELATES OF AMPT-INDUCED DEPRESSIVE SYMPTOMS

Table. Metabolic Rates in Patients With and Without an AMPT-Induced Return of Depressive Symptoms* Depletion- Depression With Relapse (n = 11) Without Relapse (n = 7) Interaction

Placebo, AMPT, % Placebo, AMPT, % F P Brain Regions Mean (SD) Mean (SD) Change Mean (SD) Mean (SD) Change Value Value Brain Regions Hypothesized to Change With AMPT Orbitofrontal cortex 10.78 (6.04) 8.27 (3.30) −23 6.28 (3.64) 9.53 (4.71) 52 10.27 .002 Middle frontal gyrus 12.65 (6.87) 9.87 (3.82) −22 7.68 (4.42) 11.32 (4.79) 47 4.68 .04 Superior dorsolateral prefrontal cortex 13.4 (6.65) 9.36 (4.17) −30 8.04 (4.93) 11.00 (5.91) 37 5.32 .03 Thalamus 9.48 (3.77) 7.41 (2.58) −22 5.59 (3.58) 8.15 (3.64) 46 8.76 .006 Brain Regions Not Hypothesized to Change or With Less Evidence for Change With AMPT Temporal cortex 12.13 (5.94) 9.01 (3.44) −26 7.17 (4.37) 10.10 (4.44) 41 11.76 .001 Superior frontal gyrus 12.39 (6.56) 9.11 (3.52) −26 7.04 (4.11) 10.87 (5.34) 54 6.08 .02 Inferior frontal gyrus 12.59 (6.47) 9.52 (3.51) −24 7.66 (4.73) 11.15 (5.62) 46 4.07 .05 Parietal cortex 12.53 (4.66) 9.16 (3.58) −27 7.57 (4.46) 10.65 (6.26) 41 5.55 .03 Caudate 10.26 (6.07) 8.63 (3.85) −16 6.81 (5.02) 10.60 (5.56) 56 1.35 .25 Putamen 11.66 (5.03) 9.95 (3.63) −15 7.07 (5.10) 11.38 (5.07) 61 4.53 .04 Anterior cingulate 12.04 (6.00) 9.10 (3.68) −24 6.76 (4.17) 9.84 (4.91) 46 6.50 .02 Postcentral gyrus 12.03 (5.16) 9.13 (3.83) −24 7.59 (4.73) 10.42 (5.77) 37 5.21 .03 Hippocampus 7.39 (3.12) 5.73 (2.25) −22 4.55 (3.06) 6.73 (2.69) 48 7.63 .009 Subcallosal region 9.47 (5.27) 7.86 (3.26) −17 6.16 (4.03) 9.18 (4.08) 49 1.49 .24 Parahippocampus 7.23 (2.72) 5.83 (2.13) −19 4.34 (2.41) 6.57 (3.11) 51 7.13 .01 Amygdala 7.34 (3.42) 5.83 (2.30) −21 4.64 (3.06) 6.61 (3.05) 42 3.42 .07 Midbrain 7.08 (2.50) 6.01 (1.79) −15 4.42 (2.48) 6.58 (3.07) 49 5.32 .03 Global 7.64 (3.25) 6.00 (2.85) −21 4.42 (2.54) 6.46 (2.85) 46 2.27 .15 Abbreviation: AMPT, ␣-methylparatyrosine. *Metabolic rates shown in this table are glucose uptake rates, calculated as regional activity in microcuries per cubic centimeter/[injected dose in microcuries × body weight in pounds].

creases in metabolism in these regions Figure 3. Orbitofrontal Glucose Uptake Rates in Patients With (n=11) and Without (n=7) could not be explained by greater mag- AMPT-Induced Return of Depressive Symptoms nitude of catecholamine depletion with AMPT in the relapsing patients. De- A With Relapse B Without Relapse creased metabolism in these regions was 30 30 not significant when the ratio of re- 25 25 gional to whole-brain metabolism was examined, demonstrating the strong 20 20 global changes with AMPT-induced return of depressive symptoms. In- 15 15 creased resting metabolism during the 10 10 placebo test was found in patients vulnerable to AMPT-induced return of de- 5 5 Orbitofrontal Glucose Uptake Rate Orbitofrontal Glucose Uptake Rate Orbitofrontal pressive symptoms in prefrontal and 0 0 limbic regions (orbitofrontal cortex, Placebo AMPT Placebo AMPT

middle frontal gyrus, hippocampus, Glucose uptake rates were calculated as regional activity in microcuries per cubic centimeter/[injected dose in parahippocampus, and amygdala). This microcuries ϫ body weight in pounds]. Patients with a return of depressive symptoms showed decreased or- was also reported in our prior trypto- bitofrontal metabolism with ␣-methylparatyrosine (AMPT). Patients without an AMPT-induced return of depressive symptoms showed a pattern of increase in orbitofrontal cortical metabolism with AMPT. phan depletion study. In the current study, number of lifetime episodes of depression was correlated with de- induced return of depressive symp- transmitter systems have their cell bod- creased AMPT-induced metabolism in toms, the results suggest that norepi- ies in the brainstem, with long neu- orbitofrontal cortex. nephrine and serotonin modulate rons that project throughout the brain, Interpreted in conjunction with our common brain regions to mediate globally modulating neuronal func- prior PET study and prior studies of symptoms of depression. Both of the tion. The research suggests that dis- AMPT and tryptophan depletion– serotonin and norepinephrine neuro- ruption of either neurochemical sys-

Downloaded From: https://jamanetwork.com/ on 09/27/2021 CORRELATES OF AMPT-INDUCED DEPRESSIVE SYMPTOMS

tem can lead to depression, probably by portant to more carefully describe the effects of AMPT on mood in healthy, affecting the common brain regions that brain-related variables that underlie this never-depressed individuals. We did not they innervate. As reviewed in the in- vulnerability to depressive recurrence. control for stage of menstrual cycle. It troduction to this article, disruption of Brain imaging may also be developed to would have been difficult, however, to these systems causes mood changes in predict who will be vulnerable to de- perform all scans at the same stage of the patients with a history of depression pressive recurrence if antidepressant menstrual cycle because the PET study who experience remission while tak- medications are discontinued. days were 1 week apart and there was a ing antidepressants, or in patients with Prior studies in depression have random assignment to AMPT or pla- a history of depression who are in re- found structural alterations in the brain cebo on the first scan day. We did not mission and recently stopped medica- regions implicated in this study. For ex- assess smoking status, which could affect tion.22 However, patients with current ample, studies have found reductions plasma glucose levels, among other depression and healthy persons within volume of orbitofrontal cortex in pa- things. out a history of depression are unaf- tients with depression,60,61 while other The mechanisms by which AMPT af- fected by disruption of these neuro- studies found hippocampal volume re- fects regional brain metabolism are not transmitter systems. This is primarily ductions.48,56 However since this study entirely clear. Administration of AMPT related to a ceiling effect, whereby pa- has a within-subject design, changes in with depletion of norepinephrine and tients with depression have a limit in metabolism between placebo and dopamine resulted in decreased metabo- the degree to which they can develop AMPT tests are not likely to be af- lism (from an elevated baseline) in pa- additional symptoms of depression. fected by structural alterations. tients who developed depressive symp- We found that increased resting base- There are number of limitations of this toms, and increased metabolism (from line metabolism in prefrontal and lim- study that are worth mentioning. The re- a relatively lower baseline) in patients bic regions was associated with vulner- sults relied on the use of glucose up- who did not develop depressive symp- ability to return of depressive symptoms. take rates, since the plasma data were too toms. The findings suggest a dose- This also was reported in our prior study unreliable for use in quantitation of glu- response effect or a variation in recep- of tryptophan depletion. Prior studies cose metabolic rates in the brain. Sev- tor responsiveness that differs depending have shown that vulnerability to tryp- eral patients dropped out of the study af- on vulnerability to return of depressive tophan depletion–induced return of de- ter the first test day. This was largely symptoms in individual participants. At pressive symptoms predicts long-term related to the fact that patients had a re- the level of the postsynaptic receptor, vulnerability to relapse when medica- turn of depressive symptoms with AMPT norepinephrine has a stimulatory effect 91 ␣ tions are withdrawn. It appears that al- on that first day and did not wish to con- with binding to 1 receptors, an inhibi- ␣ terations in brain function continue to tinue. Sample size is limited and the cur- tory effect with 2 receptors, and excit- persist, even after patients have been suc- rent study needs to be replicated with a atory and inhibitory effects with ␤ re- cessfully treated for depression. Con- larger sample. AMPT has sedating ef- ceptors. Animal studies have shown that sistent with this idea is the finding that fects that may confound the depresso- stimulation of the norepinephrine sys- some patients with a history of depres- genic effects and may interfere with the tem (with medications such as the ␣-ad- sion are vulnerable to tryptophan deple- blinded administration of the medica- renergic receptor blocker phentol- tion–induced return of depressive symp- tion. We attempted to control for this by amine and phenylethylamine) resulted toms, even when they are not taking using the sedating medication diphen- in a dose-dependent effect, with in- medications, although healthy partici- hydramine as the placebo, which, in our creased cerebral blood flow at lower pants without a history of depression are prior studies, resulted in an equally se- doses and decreased blood flow at high 92,93 ␣ not vulnerable to the development of de- dating effect. Diphenhydramine may also doses. Administration of the 2 an- pressive symptoms. The number of epi- have effects on brain metabolism that are tagonist yohimbine, which stimulates re- sodes of depression has been corre- currently unknown. However, it is an- lease of norepinephrine in the brain, related with hippocampal atrophy,48,56 and ticipated that sedating medications would sulted in decreased brain metabolism in it is known that the risk of depressive result in decreased global brain metabo- frontal, parietal, sensory, and olfactory recurrence increases with the number of lism. It is possible that the changes in cortex.94,95 Electrical stimulation of the episodes of depression experienced dur- brain function are not related to symp- locus coeruleus (brainstem site of most ing an individual’s lifetime. This sug- toms of depression. This study is cross- of the noradrenergic cell bodies) re- gests that changes in the brain underlie sectional, and other factors may contrib- sulted in a decrease in metabolism or the development of vulnerability to de- ute to the symptoms of depression blood flow in several cortical and sub- pressive recurrence. Consistent with that observed in this study. The current study cortical regions,96-99 while lesions of the idea is the finding in the current study did not include a never-depressed group locus resulted in increased metabo- that number of episodes of depression of healthy persons for comparison. How- lism.100-102 Therefore, the brain re- correlate with decreased orbitofrontal ever, prior studies described in the in- sponse to norepinephrine depletion with metabolism with AMPT. It will be im- troduction to this article have not found AMPT in patients without a return of de-

Downloaded From: https://jamanetwork.com/ on 09/27/2021 CORRELATES OF AMPT-INDUCED DEPRESSIVE SYMPTOMS

pressive symptoms can be considered to HVA concentrations in elderly depressed patients who serpine: a reevaluation. Semin Psychiatry. 1971;3: attempted suicide. Am J Psychiatry. 1990;147:1225- 435-448. be the “normal” response. This re- 1227. 25. Charney DS, Menkes DB, Heninger GR. Recep- sponse was similar to what was found 6. Mann JJ, Arango V, Marzuk PM. Evidence for the tor sensitivity and the mechanism of action of anti- 5-HT hypothesis of suicide: a review of post-mortem depressants. Arch Gen Psychiatry. 1981;38:1160- in animal studies; ie, increased metabo- studies. Br J Psychiatry Suppl. December 1989:7-14. 1180. lism with removal of norepinephrine. 7. Mann JJ, McBride A, Anderson GM, Mieczkowski 26. Golden RN, Markey SP, Risby ED, Rudorfer MV, The question remains: Why do pa- TA. Platelet and whole blood serotonin content in de- Cowdry RW, Potter WZ. Antidepressants reduce pressed inpatients: correlations with acute and life- whole-body norepinephrine turnover while enhanc- tients who have a return of their depres- time psychopathology. Biol Psychiatry. 1992;32:243- ing 6-hydroxymelatonin output. Arch Gen Psychia- sive symptoms have decreased func- 257. try. 1988;45:150-154. 8. Mann JJ, Stanley M, McBride A, McEwen BS. In- 27. Charney DS, Heninger GR, Sternberg F, Hafsted tion? It appears that the decreased creased serotonin-2 and beta-adrenergic receptor bind- KM, Giddings S, Landis DH. Adrenergic receptor sen- function in the prefrontal and orbito- ing in the frontal cortices of suicide victims. Arch Gen sitivity in depression. Arch Gen Psychiatry. 1982;39: frontal cortex represents a neural cor- Psychiatry. 1986;43:954-959. 290-294. 9. Stanley M, Mann JJ. Increased serotonin-2 bind- 28. Matussek N, Ackenheil M, Hippius H, et al. Effect relate of the symptoms of depression. ing sites in frontal cortex of suicide victims. Lancet. of clonidine on growth hormone release in psychiat- The fact that depressive symptoms can 1983;1:214-216. ric patients and controls. Psychiatry Res. 1980;2:25- 10. Arango V, Ernsberger P, Marzuk PM, et al. Au- 36. recur with serotonergic or noradrener- toradiographic demonstration of increased serotonin 29. Siever LJ, Uhde TW, Silberman EK, et al. Evalu- gic depletion suggests that this may not 5-HT-2 and alpha-adrenergic receptor binding sites ation of alpha-adrenergic responsiveness to cloni- in the brain of suicide victims. Arch Gen Psychiatry. dine challenge and noradrenergic metabolism in the be secondary to a change in the post- 1990;47:1038-1047. affective disorders and their treatment. Psychophar- synaptic receptor in vulnerable pa- 11. Ferrior IN, McKeith IG, Cross AJ, Perry EK, Candy macol Bull. 1982;18:118-119. tients but rather a common alteration in JM, Pemy RH. Postmortem neurochemical studies in 30. Klimek V, Stockmeier C, Overholser JC, et al. Re- depression. Ann N Y Acad Sci. 1986;487:128-142. duced levels of norepinephrine transporters in the lo- an intracellular process in specific brain 12. Lowther S, De Paermentier F, Crompton MR, Ka- cus coeruleus in major depression. J Neurosci. 1997; areas, including orbitofrontal cortex. tona CL, Horton RW. Brain 5-HT-2 receptors in sui- 17:8451-8458. cide victims: violence of death, depression and ef- 31. Ordway G, Smith I, Haycock J. Elevated tyrosine Author Contributions: As principle investigator, Dr fects of antidepressant medication. Brain Res. 1994; hydroxylase in the locus coeruleus of suicide victims. Bremner had full access to all of the data in the study 642:281-289. J Neurochem. 1994;62:680-685. and takes responsibility for the integrity of the data 13. Muhlbauer HD, Muller-Oerlinghausen B. Fenflu- 32. Nagatsu T, Levitt M, Udenfriend S. Tyrosine hy- and the accuracy of the data analysis. ramine stimulation of serum cortisol in patients with droxylase: the initial step in norepinephrine biosyn- Study concept and design: Bremner, Charney. major affective disorders and healthy controls: fur- thesis. J Biol Chem. 1964;239:2910-2917. Acquisition of data: Bremner, Vythilingam, Ng, ther evidence for a central serotonergic action of lithium 33. Engelman K, Horwitz D, Jequier E, Sjoerdsma A. Vermetten, Nazeer, Oren, Berman. in man. J Neural Transm. 1985;61:81-94. Biochemical and pharmacological effects of alpha- Analysis and interpretation of data: Bremner, Ng. 14. Golden RN, Hsiao J, Lane E, et al. Abnormal neu- methyltyrosine in man. J Clin Invest. 1968;47:577- Drafting of the manuscript: Bremner, Berman, roendocrine responsivity to acute i.v. clomipramine 594. Charney. challenge in depressed patients. Psychiatry Res. 1990; 34. Mendels J, Frazer A. Brain biogenic amine deple- Critical revision of the manuscript for important in- 3:39-47. tion and mood. Arch Gen Psychiatry. 1974;30:447- tellectual content: Vythilingam, Ng, Vermetten, 15. Heninger GR, Charney DS, Sternberg DE. Seroton- 451. Nazeer, Oren, Charney. ergic function in depression: prolactin response to in- 35. Miller HL, Delgado PL, Salomon RM, Heninger Statistical expertise: Bremner. travenous tryptophan in depressed patients and healthy GR, Charney DS. Effects of alpha-methyl-para- Obtained funding: Bremner, Charney. subjects. Arch Gen Psychiatry. 1984;41:398-402. tyrosine (AMPT) in drug-free depressed patients. Neu- Administrative, technical, or material support: 16. Delgado PL, Charney DS, Price LH, Landis H, ropsychopharmacology. 1996;14:151-157. Bremner, Ng, Nazeer, Oren, Berman, Charney. Heninger GR. Neuroendocrine and behavioral ef- 36. Salomon RM, Miller HL, Krystal JH, Heninger GR, Study supervision: Bremner, Vythilingam, Oren, fects of dietary tryptophan restriction in healthy sub- Charney DS. Lack of behavioral effects of mono- Charney. jects. Life Sci. 1989;45:2323-2332. amine depletion in healthy subjects. Biol Psychiatry. Funding/Support: This study was supported by a VA 17. Young SN, Ervin FR, Pihl RO, Finn P. Biochemi- 1997;41:58-64. Merit Review Grant to Dr Charney, the National Al- cal aspects of tryptophan depletion in primates. Psy- 37. Delgado PL, Miller HL, Salomon RM, et al. Mono- liance for Research in Schizophrenia and Affective Dis- chopharmacology. 1989;98:508-511. amines and the mechanism of antidepressant action: orders (NARSAD) Young Investigator Award and Na- 18. Delgado PL, Miller HL, Salomon RM, et al. effects of catecholamine depletion on mood of pa- tional Institute of Mental Health grant 1R01MH56120 Tryptophan-depletion challenge in depressed tients treated with antidepressants. Psychopharma- to Dr Bremner, and a Veterans Administration Career patients treated with desipramine or fluoxetine: col Bull. 1993;29:389-396. Development Award grant to Dr Bremner. implications for the role of serotonin in the mecha- 38. Miller HL, Delgado PL, Salomon RM, et al. Clini- Acknowledgment: We thank Helen Sayward, MS, for nism of antidepressant action. Biol Psychiatry. 1999; cal and biochemical effects of catecholamine deple- image processing and data analysis; Robert Stur- 46:212-220. tion on antidepressant-induced remission of depres- wold, PharmD, for performing randomization and 19. Delgado PL, Charney DS, Price LH, Aghajanian sion. Arch Gen Psychiatry. 1996;53:117-128. medication preparation; and Cathy Colonese, RN, and GK, Landis H, Heninger GR. Serotonin function and 39. Brodie H, Murphy DL, Goodwin F, Bunney W. Cat- Jacque Piscitelli, MS, for assistance in data collection. the mechanism of antidepressant action: reversal of echolamines and mania: the effect of alpha-methyl- antidepressant-induced remission by rapid depletion para-tyrosine on manic behavior and catecholamine of plasma tryptophan. Arch Gen Psychiatry. 1990;47: metabolism. Clin Pharmacol Ther. 1971;12:218- REFERENCES 411-418. 224. 1. Kessler RC, Berglund P, Demler O, et al. The epi- 20. Spillmann MK, Van der Does AJ, Rankin MA, et 40. McCann UD, Thorne D, Hall M, et al. The ef- demiology of major depressive disorder: results from al. Tryptophan depletion in SSRI-recovered de- fects of l-dihydroxyphenylalanine on alertness and the National Comorbidity Survey Replication (NCS-R). pressed outpatients. Psychopharmacology. 2001;155: mood in alpha-methyl-para-tyrosine-treated healthy JAMA. 2003;289:3095-3105. 123-127. humans: further evidence for the role of catechol- 2. Risch SC, Nemeroff CB. Neurochemical alterations 21. Delgado PL, Price LH, Miller AH, et al. Serotonin amines in arousal and anxiety. Neuropsychopharma- of serotonergic neuronal systems in depression. J Clin and the neurobiology of depression: effects of tryp- cology. 1995;13:41-52. Psychiatry. 1992;53:3-7. tophan depletion in drug-free depressed patients. Arch 41. Berman RM, Narasimhan M, Miller HL, et al. Tran- 3. Delgado PL. Depression: the case for a mono- Gen Psychiatry. 1994;51:865-874. sient depressive relapse induced by catecholamine amine deficiency. J Clin Psychiatry. 2000;61(suppl 6): 22. Moreno FA, Gelenberg AJ, Heninger GR, et al. depletion: potential phenotypic vulnerability marker? S7-S11. Tryptophan depletion and depressive vulnerability. Biol Arch Gen Psychiatry. 1999;56:395-403. 4. Asberg M, Thoren P, Traskman L, Bertilsson L, Ring- Psychiatry. 1999;46:498-505. 42. Heninger GR, Delgado PL, Charney DS. The re- berger V. Serotonin depression: a biological sub- 23. Delgado PL, Moreno FA. Role of norepinephrine vised monoamine theory of depression: a modula- group within the affective disorders? Science. 1976; in depression. J Clin Psychiatry. 2000;61(suppl 1):S5- tory role for monoamines, based on new findings from 191:478-480. S12. monoamine depletion experiments in humans. Phar- 5. Jones JS, Stanley B, Mann JJ, et al. CSF 5-HIAA and 24. Goodwin F, Bunney W. Depressions following re- macopsychiatry. 1996;29:2-11.

Downloaded From: https://jamanetwork.com/ on 09/27/2021 CORRELATES OF AMPT-INDUCED DEPRESSIVE SYMPTOMS

43. Bremner JD. Structural changes in the brain in de- jects resolves with antidepressant treatment: an fMRI 84. Brownell GL, Kearfott KJ, Kairento AL, et al. Quan- pression and relationship to symptom recurrence. CNS study. Biol Psychiatry. 2001;50:651-658. titation of regional cerebral glucose metabolism. J Com- Spectrums. 2002;7:129-139. 65. Ebert D, Feistel H, Barocka A. Effects of sleep dep- put Assist Tomogr. 1983;7:919-924. 44. Bremner JD, Innis RB, Salomon RM, et al. PET mea- rivation on the limbic system and the frontal lobes in 85. Sokoloff L, Reivich M, Kennedy C, et al. The surement of cerebral metabolic correlates of depres- affective disorders: a study with Tc-99m-HMPAO [14]deoxyglucose method for the measurement of lo- sive relapse. Arch Gen Psychiatry. 1997;54:364-374. SPECT. Psychiatry Res. 1991;40:247-251. cal cerebral glucose utilization: theory, procedure, and 45. George MS, Ketter TA, Post RM. Prefrontal cor- 66. Bench CJ, Friston KJ, Brown RG, Scott LC, Frack- normal values in the conscious and anesthetized al- tex dysfunction in clinical depression. Depression. 1994; owiak RS, Dolan RJ. The anatomy of melancholia: fo- bino rat. J Neurochem. 1977;28:897-916. 2:59-72. cal abnormalities of cerebral blood flow in major de- 86. Robb RA, Hanson DP, Karwoski RA, Larson AG, 46. Steffens DC, Krishnan KR. Structural neuroimag- pression. Psychol Med. 1992;22:607-615. Workman EL, Stacy MC. Analyze: a comprehensive, ing and mood disorders: recent findings, implications 67. Baxter LR, Schwartz JM, Phelps ME, et al. Reduc- operator-interactive software package for multidi- for classification, and future directions. Biol Psychia- tion of prefrontal cortex glucose metabolism com- mensional medical image display and analysis. Com- try. 1998;43:705-712. mon to three types of depression. Arch Gen Psychia- put Med Imaging Graph. 1989;13:433-454. 47. Mayberg HS, Brannan SK, Mahurin RK, et al. Cin- try. 1989;46:243-249. 87. Zubal IG, Zhang L, Tagare B, Duncan JS. 3-D reg- gulate function in depression: a potential predictor of 68. Martinot JL, Hardy P, Feline A, et al. Left pre- istration of SPECT and MRI brain images. J Nucl Med. treatment response. Neuroreport. 1997;8:1057- frontal glucose hypometabolism in the depressed state: 1993;34:187. 1061. a confirmation. Am J Psychiatry. 1990;147:1313- 88. Bremner JD, Bronen RA, de Erasquin G, et al. De- 48. Bremner JD, Narayan M, Anderson ER, Staib LH, 1317. velopment and reliability of a method for using mag- Miller H, Charney DS. Hippocampal volume reduc- 69. Mayberg HS, Lewis PJ, Regenold W, Wagner HN. netic resonance imaging for the definition of regions tion in major depression. Am J Psychiatry. 2000;157: Paralimbic hypoperfusion in unipolar depression. J Nucl of interest for positron emission tomography. Clin Posi- 115-117. Med. 1994;35:929-934. tron Imaging. 1998;1:145-159. 49. Sha PJ, Ebmeier KP, Glabus MF, Goodwin GM. 70. Mayberg HS. Frontal lobe dysfunction in second- 89. Rajkowska NA, Miguel-Hidalgo JJ, Wei J, et al. Cortical grey matter reductions associated with treat- ary depression. J Neuropsychiatry Clin Neurosci. 1994; Morphometric evidence for neuronal and glial pre- ment-resistant chronic unipolar depression. Br J Psy- 6:428-442. frontal cell pathology in major depression. Biol Psy- chiatry. 1998;172:527-532. 71. Biver F, Goldman S, Delvenne V, et al. Frontal and chiatry. 1999;45:1085-1098. 50. Krishnan KR, Doraiswamy PM, Figiel GS, et al. Hip- parietal metabolic disturbances in unipolar depres- 90. Ongur D, Drevets WC, Price JL. Glial reduction pocampal abnormalities in depression. J Neuropsy- sion. Biol Psychiatry. 1994;36:381-388. in the subgenual prefrontal cortex in mood disorders. chiatry Clin Neurosci. 1991;3:387-391. 72. Mayberg HS, Liotti M, Brannan SK, et al. Recip- Proc Natl Acad Sci U S A. 1998;95:13290-13295. 51. Sheline YI, Sanhavi M, Mintun MA, Gado MU. rocal limbic-cortical function and negative mood: con- 91. Moreno FA, Heninger GR, McGahuey CA, Del- Depression duration but not age predicts hippocam- verging PET findings in depression and normal sad- gado PL. Tryptophan depletion and risk of depres- pal volume loss in medically healthy women with re- ness. Am J Psychiatry. 1999;156:675-682. sion relapse: a prospective study of tryptophan deple- current major depression. J Neurosci. 1999;19:5034- 73. Mayberg HS, Starkstein SE, Peyser CE, Brandt J, tion as a potential predictor of depressive episodes. 5043. Dannals RF, Folstein SE. Paralimbic frontal lobe hy- Biol Psychiatry. 2000;48:327-329. 52. Vakili K, Pillay SS, Lafer B, et al. Hippocampal vol- pometabolism in depression associated with Hunting- 92. Raichle ME, Hartman BK, Eichling JO, Sharpe LG. ume in primary unipolar major depression: a mag- ton’s 99mTc-exametazine with single photon emis- Central noradrenergic regulation of cerebral blood flow netic resonance imaging study. Biol Psychiatry. 2000; sion tomography in major depression before and after and vascular permeability. Proc Natl Acad SciUSA. 47:1087-1090. treatment. J Affect Disord. 1993;29:245-255. 1975;72:3726-3730. 53. Steffens DC, Byrum CE, McQuoid DR, et al. Hip- 74. Ring RA, Bench CJ, Trimble MR, Brooks DJ, Frack- 93. McCulloch J, Harper AM. Factors influencing the pocampal volume in geriatric depression. Biol Psy- owiak RS, Dolan RJ. Depression in Parkinson’s dis- response of the cerebral circulation to phenylethyl- chiatry. 2000;48:301-309. ease: a positron emission study. Br J Psychiatry. 1994; amine. Neurology. 1979;29:201-207. 54. Mervaala E, Fohr J, Konogen M, et al. Quantita- 165:333-339. 94. Savaki HE, Kadekaro M, McCulloch J, Sokoloff L. tive MRI of the hippocampus and amygdala in se- 75. Mayberg HS, Starkstein SE, Sadzot B, et al. Se- The central noradrenergic system in the rat: meta- vere depression. Psychol Med. 2000;30:117-125. lective hypometabolism in the inferior frontal lobe in bolic mapping with alpha-adrenergic blocking agents. 55. Sheline YI. 3D MRI studies of neuroanatomical depressed patients with Parkinson’s disease. Ann Neu- Brain Res. 1982;234:65-79. changes in unipolar major depression: the role of stress rol. 1990;28:57-64. 95. Inoue M, McHugh M, Pappius HM. The effect and medical comorbidity. Biol Psychiatry. 2000;48: 76. George MS, Ketter TA, Parekh PI, et al. Blunted of alpha-adrenergic receptor blockers prazosin and yo- 791-800. left cingulate activation in mood disorder subjects dur- himbine on cerebral metabolism and biogenic amine 56. Sheline YI, Wang P, Gado M, Csernansky J, Van- ing a response interference task (the Stroop). J Neu- content of traumatized brain. J Cereb Blood Flow nier M. Hippocampal atrophy in major depression. Proc ropsychiatry Clin Neurosci. 1997;9:55-63. Metab. 1991;11:242-252. Natl Acad Sci U S A. 1996;93:3908-3913. 77. Mann JJ, Malone KM, Diehl DJ, Perel J, Cooper 96. Abraham WC, Delanoy RL, Dunn AJ, Zornetzer 57. Pantel J, Schroder J, Essig M, et al. Quantitative TB, Mintun MA. Demonstration of in vivo of reduced SF. Locus coeruleus stimulation decreases deoxyglu- magnetic resonance imaging in geriatric depression and serotonin responsivity in the brain of untreated de- cose uptake in ipsilateral mouse cerebral cortex. Brain primary degenerative dementia. J Affect Disord. 1997; pressed patients. Am J Psychiatry. 1996;153:174- Res. 1979;172:387-392. 42:69-83. 182. 97. De la Torre JC, Surgeon JW, Walker RH. Effects 58. Ashtari M, Greenwald BS, Kramer-Ginsberg E, et 78. Brody AL, Saxena S, Stoessel P, et al. Regional brain of locus coeruleus stimulation on cerebral blood flow al. Hippocampal/amygdala volumes in geriatric de- metabolic changes in patients with major depression in selected brain regions. Acta Neurol Scand. 1977; pression. Psychol Med. 1999;29:629-638. treated with either paroxetine or interpersonal therapy: 64(suppl):104-105. 59. Drevets WC, Price JL, Simpson JR, et al. Sub- preliminary findings. Arch Gen Psychiatry. 2001;58: 98. Katayama Y, Ueno Y, Tsukiyama T, Tsubokawa genual prefrontal cortex abnormalities in mood dis- 631-640. T. Long lasting suppression of firing of cortical neu- orders. Nature. 1997;386:824-827. 79. Mayberg HS, Brannan SK, Tekell JL, et al. Re- rons and decrease in cortical blood flow following train 60. Lai TJ, Payne ME, Byrum CE, Steffens D, Krishnan gional metabolic effects of fluoxetine in major depres- pulse stimulation of the locus coeruleus in the cat. Brain KR. Reduction of orbital frontal cortex volume in geri- sion: serial changes and relationship to clinical re- Res. 1981;216:173-179. atric depression. Biol Psychiatry. 2000;48:971-975. sponse. Biol Psychiatry. 2000;48:830-843. 99. Goadsby PJ, Duckworth JW. Low frequency stimu- 61. Bremner JD, Vythilingam M, Vermetten E, et al. 80. First MB, Spitzer RL, Williams JB, Gibbon M. Struc- lation of the locus coeruleus reduces regional blood Reduced volume of orbitofrontal cortex in major de- tured Clinical Interview for DSMIV-Patient Edition flow in the spinalized cat. Brain Res. 1989;476:71- pression. Biol Psychiatry. 2002;51:273-279. (SCID-P). Washington, DC: American Psychiatric Press; 77. 62. Tebartz van Elst L, Woermann FG, Lemieux L, 1995. 100. Schwartz WJ. 6-Hydroxydopamine lesions of rat Trimble MR. Amygdala enlargement in dysthymia: a 81. McCann UD, Thorne D, Hall M, et al. The ef- locus coeruleus alter brain glucose consumption, as volumetric study of patients with temporal lobe epi- fects of l-dihydroxyphenylalanine on alertness and measured by the 2-deoxy-D-[C-14]glucose tech- lepsy. Biol Psychiatry. 1999;46:1614-1623. mood in alpha-methyl-para-tyrosine-treated healthy nique. Neurosci Lett. 1978;7:141-150. 63. Drevets WC, Price JL, Bardgett ME, Reich T, Todd humans: further evidence for the role of catechol- 101. Bates D, Weinshilboum RM, Campbell RJ, Sundt RD, Raichle ME. Glucose metabolism in the amyg- amines in arousal and anxiety. Neuropsychopharma- TM. The effect of lesions in the locus coeruleus on the dala in depression: relationship to diagnostic subtype cology. 1995;13:41-52. physiological responses of the cerebral blood vessels and plasma cortisol levels. Pharmacol Biochem Be- 82. Hamilton M. A rating scale for depression. J Neu- in cats. Brain Res. 1977;136:431-443. hav. 2002;71:431-447. rol Neurosurg Psychiatry. 1960;12:56-62. 102. Savaki HE, Graham DI, Grome JJ, McCulloch J. 64. Sheline YI, Barch DM, Donnelly JM, Ollinger JM, 83. Mulaney NA, Gould LK, Hartz RK, et al. Design Functional consequences of unilateral lesion of the lo- Snyder AZ, Mintun MA. Increased amygdala re- and performance of Posicam 6.5 BGO Positron cam- cus coeruleus: a quantitative [14C]2-deoxyglucose in- sponse to masked emotional faces in depressed sub- era. J Nucl Med. 1990;31:610-616. vestigation. Brain Res. 1984;292:239-249.

Downloaded From: https://jamanetwork.com/ on 09/27/2021