Brain SPECT Imaging in Neuropsychiatry

Brain Trauma, Violence, and Substance Abuse

Daniel G. Amen, MD
Director, Amen Clinics, Inc.

Psychiatrists are the only medical specialists who rarely look at the organ they treat.  In my opinion, the lack of brain imaging has kept psychiatry behind medicine’s other specialties, decreasing our effectiveness with patients, and helping to maintain stigma and non-compliance with needed treatment.  The odds are if a patient is having serious problems with his feelings (depression or anxiety), thoughts (schizophrenia or bipolar disorder), or behavior (violence, pedophilia, or substance abuse) the psychiatrist will never order a brain scan.  He will prescribe medications, psychotherapy, ECT, or a host of other treatments that will change brain function without ever looking at how the patient’s specific brain works.  He will not know which areas of his brain work well, which areas of his brain work too hard, or which areas of his brain do not work hard enough. 

Can you imagine the outcry if other medical specialties acted without looking?  If orthopedic doctors set broken bones without x-rays?  If cardiologists diagnosed coronary artery blockages without doing angiograms or fast CT scans?  If urologists diagnosed prostate cancer and prescribed treatment with biopsies.  Or, if internists diagnosed pneumonia without ordering chest x-rays or doing sputum cultures?  For the most part, psychiatrists diagnose and treat patients based on symptoms clusters, not underling brain dysfunction.  Think about taking your car to the mechanic because it is smoking, using too much gasoline, and stalling in the middle of intersections.  The mechanic listens to the symptoms, looks at the outside of the car and decides to change the fuel pump, without ever popping the car’s hood and turning it on.  How would you feel?  Probably, like going somewhere else.  Yet, the state-of-the-art in psychiatry is not to look at the organ it treats. 

Researchers tell us it’s not time yet to use imaging tools, even though high resolution brain SPECT (single photon emission computed tomography) and PET (positron emission tomography) scans have been available for more than 15 years.  They say imaging tools need more research to determine specificity and sensitivity, even though there are hundreds of articles on functional brain imaging for a host of neuropsychiatric disorders (see www.brainplace.com for a reference list).  And, by the way, they ask how would imaging tools help the practicing clinician?  This article will discuss the practical use of brain SPECT imaging for three common psychiatric areas: brain trauma, substance abuse, and violence.  The opinions in this article are based on my experience with over 10,000 SPECT studies over the past 11 years, both as a treating psychiatrist and a physician who is also licensed in California to perform these studies for others. 

Why SPECT?

Brain SPECT imaging is a nuclear medicine study, which uses minute doses of isotopes bound to neurospecific pharmaceuticals to study regional cerebral blood flow (rCBF) and thus indirectly brain metabolic activity (1).  The course of rCBF measurement has evolved greatly over the past two decades.  Imaging detectors have been developed, which provide three-dimensional representations of rCBF similar to the images created by positron emission tomography (PET) technology.  Like PET, more advanced SPECT cameras have multiple heads, which increase the resolution of images and decrease the time necessary to complete scans.  Although the resolution of SPECT is a limitation cited by some researchers (2), the resolution of contemporary SPECT multi-head cameras has entered the same realm as PET at considerably less cost (3).  SPECT imaging cameras are also located in many hospitals in the United States, in contrast to PET cameras, which tend to be located mainly in university research centers.  The radiation exposure from SPECT scans has been cited as a reason to avoid its use, especially with children.  However, according to the American Academy of Neurology, SPECT is a safe procedure (4).  The average radiation exposure for one SPECT scan is 0.7 rem.  This amount of radiation is similar to a routine radionuclide bone scan or brain CT.  The radiation exposure concern has limited community normative samples using SPECT.  To address this issue, the Society of Nuclear Medicine has undertaken storing a large normative database at Yale University for clinician and researcher use.   Also, Chiron et al. (5) published a normative database for children, demonstrating that the SPECT study of a two year old child shows the same relative rCBF pattern as the adult brain.  Ultimately, symmetry is the hallmark of a normal study (6). In contrast to PET, QEEG, and fMRI, brain SPECT imaging is more frequently covered by third party carriers (i.e., Medicare in California) when ordered properly for medical indications (i.e., to rule out frontal or temporal lobe disorders, dementia, and brain trauma).  SPECT is typically less expensive than PET or fMRI studies (approximately half the cost in California).   

The Society of Nuclear Medicine currently lists the common indications for brain SPECT imaging as the evaluation of suspected brain trauma, evaluation of suspected dementia, localization of epileptic foci, and the detection and evaluation of cerebral vascular disease (7).  Brain SPECT imaging, as with any kind of medical test, should only be ordered when it adds clinically relevant information that helps medical practitioners provide improved care.    These concerns are highlighted in different phases of patient care, including diagnostic, prognostic, and treatment considerations.  

Brain Trauma

Brain trauma frequently impacts behavior, cognition, and emotion.  Current literature indicates that SPECT is useful in the clinical evaluation of brain trauma.  Documentation of the extent and location of brain injuries is essential for several clinical reasons, including accurate diagnosis, prognosis, treatment planning, and forensic evaluation.  SPECT is a sensitive tool for evaluating perfusion abnormalities not only in cases evaluating blunt brain trauma, but also in cases of post-concussive syndrome (8) and whiplash (9).  Even minor brain injuries that leave brain tissues grossly intact can cause perfusion abnormalities, which can be seen on SPECT for extended periods of time, even when there is no loss of consciousness (10).  For example, patients with headaches, memory loss, concentration difficulties, dizziness, perceptual sensitivities, and emotional lability may be inaccurately labeled as malingering when EEG, CT, and MRI scans are normal after a brain injury.

Researchers have demonstrated that functional imaging may be more sensitive than MRI or CT in detecting the effects of mild to moderate brain injury (11-12).  Abdel-Dayem et. al. (11) acquired SPECT images of 228 patients with no other remarkable medical history who suffered mild to moderate TBI.  All patients experienced significantly disruptive symptoms subsequent to their injury including headache, memory problems, dizziness, and/or sleep disorders.  Focal areas of hypoperfusion were seen in 77% (176/228) of patients.  However, of the 41 patients with normal CT scans who experienced mild traumatic brain injuries (i.e., no loss of consciousness), 28 had abnormal SPECT results. 

There is an association between SPECT imaging and clinical outcome.  Jacobs et. al. (13) used SPECT to prospectively evaluate 67 patients who experienced mild or moderate TBI.  Each patient received a SPECT scan within four weeks of the initial injury and 3 months after the first scan.  Clinical evaluations were conducted at both times in order to determine the presence and severity of the patients’ clinical symptoms.  Of the 33 patients who showed negative results on the first scan (i.e., no significant abnormalities or lesions), 32 patients’ clinical symptoms had resolved within 3 months.  These authors suggested that negative initial SPECT studies are a reliable predictor of a favorable clinical outcome.  Twenty of the 34 patients with initial SPECT scan abnormalities had persistent clinical symptoms.

Typically, SPECT findings in brain trauma include focal areas of decreased perfusion, often in a coup-contra-coup pattern (such as decreased perfusion in the left anterior prefrontal cortex and right occipital lobe or the anterior and posterior aspects of a temporal lobe) and, in some cases, marked hyperperfusion over the site of the injury.  How can SPECT help in clinical patient management?  Abnormal SPECT studies may help with treatment compliance, patients more readily accept help for something they can see.  SPECT can also uncover forgotten brain trauma in clinically confusing or complex cases.  It is common for patients to forget even significant brain trauma.  This may be due to amnesia surrounding the trauma, psychological repression, or the fact that it occurred at a relatively young age.  SPECT can also help target medical treatment, guiding physicians to prescribe anticonvulsants if there is focal areas of hypo or hyperperfusion in the temporal lobes, psychostimulants if there is diffuse hypoperfusion in the prefrontal cortex, antidepressants if there is hyperperfusion in the limbic system, or anti-obsessive antidepressants if there is diffuse hyperperfusion in the prefrontal cortex.  

Here are two examples of how SPECT can be useful in the clinical evaluation of brain trauma.
T, exhibited severe conduct problems.  By the age of 15, he had already been arrested for shoplifting, he frequently cut school, and he was defiant and hostile toward his parents.  He did poor socially and never fit in with his peers.  He had already been in two residential treatment programs and was on his way to a third program when his parents brought him to our clinic for evaluation.  From an early age, T was hyperactive, impulsive, moody, and had learning problems in school.  When he was 18 months old he fell down a flight of stairs with a brief loss of consciousness.  He was taken to the emergency room where a CT scan was read as normal.  His mother said he was never quite the same.  He was tried on numerous medications without success.  His brain SPECT study showed severe damage to his left prefrontal cortex, temporal lobe, and occipital lobe.  T was placed on a combination of an anticonvulsant medication, to stabilize his temporal lobe, and a psychostimulant to enhance prefrontal cortex perfusion.  The combination enhanced mood stability and impulse control.  He was able to live at home with less turmoil.  T and his parents had a better understanding of his problems, allowing emotional healing to begin, and the school placed him in a program for brain injured children rather than just labeling him as behaviorally disturbed.

C, a 32-year-old dentist, was rear ended by a semi-truck on the freeway.  Her head quickly jerked back and forth with strong force but never hit anything.  She felt dazed for several days after the accident.  She complained of a headache for 3 months, reported memory problems, struggled with procedures at work, and was more irritable at home and with her employees.  Her insurance company denied disability benefits because the evaluating physician wrote that she never hit her head and therefore could not have sustained a brain injury.  Her treating psychologist, whom she saw for new onset family problems, referred her for evaluation.  Her SPECT study showed marked decreased perfusion in the anterior and posterior aspects of her left temporal lobe, a coup-contra-coup injury.  This information was useful in obtaining disability benefits, in helping her family understand she had sustained a brain injury, and in directing treatment, neurofeedback over the site of her left temporal lobe.  Within 6 months she was back to work with improved functioning overall.

Violent or Aggressive Behavior

In my experience, SPECT is a valuable tool in evaluating patients with violent behavior.  SPECT helps to evaluate the functional integrity of those systems frequently implicated in violence, such as the prefrontal cortex (14-16), temporal lobes (15, 17-19), and anterior cingulate gyrus (15).   Different kinds of deficits may require different interventions and treatment.  Therefore, identifying an individual’s type of brain dysfunction may have a significant impact on clinical management (20).

According to Raine et. al. (21), emotional and unplanned acts of violence may be the result of a person’s inability to utilize their prefrontal cortex and regulate the aggressive impulses that can be generated by subcortical structures of the brain.  Raine et. al. (16) also found structural deficits in the prefrontal cortex associated with antisocial personality disorders (APD).  The authors studied 21 community volunteers with APD, half of whom committed aggressive attacks on strangers.  They reported that APD subjects had an 11% reduction in prefrontal gray matter volume (as measured by MRI) when compared to two different control groups (i.e., 34 healthy subjects and 26 matched subjects for other psychiatric disorders).  Although structural deficits in the frontal lobes have been associated with violent behaviors, discrepancies between structural and functional imaging indicate that a person can have a structural evaluation that is within normal limits and still have significant abnormalities of the function of those tissues that appear intact.  Therefore, MRI, CT, and other structural techniques may not provide clinical information with the same sensitivity as SPECT and PET.

Studies have reported blood flow and metabolic abnormalities in the temporal lobes; usually in the left or dominant hemisphere (15, 17-19) of violent individuals.  Amen et al. (15) found temporal lobe abnormalities, usually left sided, in 72.5% of 40 adolescents and adults who exhibited violent or aggressive behavior.  Soderstrom et. al. (19) retrospectively compared violent subjects’ SPECT and MRI scans from pretrial forensic psychiatric investigations with control subjects.  Even after controlling for major mental disorders, substance abuse, and current medications in both groups, the violent group in this study displayed a significantly different functional brain pattern.  Sixteen of the 21 impulsively violent subjects showed some hypoperfusion in the temporal and/or frontal lobes.  However, MRI failed to show any corresponding structural damage or abnormalities in this sample of violent offenders.  A similar study of recidivistic violent offenders by Volkow and Tancredi (22) revealed that the largest functional deficits in the temporal lobe were found in their subjects whose structural CT scans were normal.

Violent individuals have also shown different SPECT patterns in the medial aspects of the frontal lobes.  Amen et. al. (15) found not only decreased perfusion if the prefrontal cortex and left temporal lobe in violent psychiatric patients, but also increased perfusion of the anterior cingulate gyrus, which has been associated with problems of cognitive inflexibility and repetitive negative thoughts.  Tiihonen et. al. (23) similarly relate dysfunction of the medial frontal lobes with violent behavior.  These authors evaluated serotonin (5-HT) and dopamine (DA) transporter specific binding in 52 subjects (21 impulsive violent offenders, 21 age- and sex-matched healthy control subjects, and ten non-violent alcoholic control subjects) by using SPECT.  They reported that violent offenders in their study had significant 5-HT specific binding reductions (i.e., including the anterior cingulate gyrus, a structure dense with 5-HT receptors) compared to healthy control subjects and non-violent alcoholics.

Considering the different types of brain dysfunction associated with violent behavior, there is existing literature that indicates that specific medical treatments may be effective for treating different patterns of brain dysfunction (i.e., temporal lobe and medial frontal lobe dysfunction).  Several authors suggest that SSRI medications are effective in treating disorders that include perfusion abnormalities in the medial aspects of the frontal lobes (i.e., obsessive compulsive disorder) (24-25).  Significant reductions of rCBF in the medial frontal lobes (i.e., anterior cingulate gyrus) of violent patients may help reduce these patients’ tendency toward cognitive inflexibility and repetitive negative thoughts.  Anticonvulsant medications have also been used effectively to treat violence and aggression (26, 27).  Marshall et. al. (27) evaluated 36 aggressive adolescents (i.e., impulsive rage attacks) with neuropsychiatric disorders over a 12 month period.  All subjects exhibited subcortically generated electrophysiological limbic abnormalities on long latency evoked potentials in response to auditory and visual stimuli; commonly an indication of metabolic abnormalities in the temporal lobes. Patients were treated with therapeutic doses of an anticonvulsant medication as part of a residential treatment program.  Upon discharge, two groups of patients were naturally formed according to their adherence to treatment (i.e., ON or OFF their medical regimen with their aftercare physician).  These authors reported that 82% of the patients who remained ON the medical regimen were both stable at discharge and continued to show good outcome one year post discharge.  Only 14% of the patients who went OFF of the medical regimen were stable at follow up.  It seems likely that reductions in prefrontal blood flow may correspond to executive dysfunction and poor impulse control in violent and aggressive patients.   When hypoperfusion in the prefrontal cortex is coupled with dysfunction in the temporal lobes and/or anterior cingulate gyrus, impulsive, violent, and aggressive behaviors may be more frequent.   

Here are two examples of how SPECT can be helpful in the clinical evaluation of aggressive behavior.
A, a nine-year-old boy, attacked a little girl on the baseball field for no obvious reason.  He also began to experience spontaneous suicidal and homicidal thoughts, drawing pictures of himself hanging from a tree and shooting other children.  In addition, for 9 months prior to evaluation he was negative, irritable, and moody.  His family doctor recommended psychotherapy, which was ineffective.  When he was brought to our clinic for evaluation he had a brain SPECT study, which revealed no perfusion in the region of the left temporal lobe.  A MRI revealed an arachnoid cyst, the size of a gold ball, occupying his left middle cranial fossa.  When the cyst was drained, the bad thoughts went away, his mood and behavior significantly improved.

J, 16, had serious problems with aggression.  He had been expelled from 5 schools for fighting.  He was also oppositional, argumentative, and underachieved in school. There was a family history of violence, with his father and uncle spending time in prison for aggressive behavior.  His SPECT study was significantly abnormal, showing marked hyperperfusion in the anterior cingulate gyrus, and significant hypoperfusion in the prefrontal cortex and temporal lobes.  He was placed sequentially on a combination of an anticonvulsant to stabilize his temporal lobes (our experience tells us to do this first), an anti-obsessive antidepressant to calm anterior cingulate gyrus hyperperfusion, and a psychostimulant to enhance prefrontal cortex perfusion.

Substance Abuse

In my experience, SPECT can add valuable information to the prevention, evaluation, and treatment of substance abusers.  Three-dimensional SPECT images demonstrating brain damage from substance abuse may have a positive educational effect on deterring children and teenagers from drug abuse.  SPECT studies may also help to break though the denial that often accompanies substance abuse.  When patients are faced with abnormal cerebral perfusion in their own brain, it is hard for them to remain in denial.  SPECT might also help evaluate the presence of other neuropsychiatric illnesses (such as brain trauma) that contribute to drug abuse.
Substance abusers have demonstrated a number of cerebral perfusion abnormalities in brain areas related to behavior, especially in the frontal and temporal lobes (28).  Over time, many amphetamine and cocaine abusers display multiple cerebral perfusion defects on SPECT, even after 6 months of abstinence (29).  Alcohol abuse is also associated with cerebral perfusion abnormalities, especially found in the frontal and temporal regions of the brain.  Nicholas et. al. (30) performed SPECT studies on 40 chronic alcoholic patients and found that, compared to a normal control group, alcohol abusers showed significantly decreased perfusion in all lobes of the brain.  Perfusion defects were most frequently seen in the frontal lobes (65%) and correlated with cognitive deficits on neuropsychological testing.  Chronic marijuana smoking, as well as opiate abuse, seems to reduce cerebral blood flow, particularly in the frontal and temporal lobes (31, 32).  These functional abnormalities are often found in tissues that are otherwise structurally intact.  Fortunately, some authors report that the reductions in cerebral blood flow associated with chronic use of ethanol, nicotine, inhalants, and solvents is at least partially reversible upon abstinence. 

Here are two examples of how SPECT can be helpful in the clinical management of substance abusers.

G, 37, was a corporate executive using alcohol and cocaine on a daily basis.  His work performance was poor and his family relationships strained, even though he did not recognize the problem.  His psychiatrist referred him for a SPECT study to confront him with the biological evidence of potential brain impairment.  His SPECT study showed marked hypoperfusion in multiple areas of his brain, especially in the prefrontal cortex and temporal lobes.  The scan helped convince him that his drug abuse was a problem and he entered into a treatment program.  He was scanned a year drug free, which showed an overall improved scan (although still not normal).

Conclusions

When used properly SPECT can be a useful tool in evaluating brain dysfunction in patients with brain trauma, violent behavior, or substance abuse.  A number of factors limit the large scale use of SPECT, including the sophistication of imaging equipment (single headed cameras are inferior to the multi-headed cameras) and the experience of the physicians interpreting the scans and utilizing the data.  Like any medical test, SPECT should only be used in the context of a patient’s clinical history and examination.  SPECT is not a “doctor in a box” and findings need clinical correlation.