Fronto-limbic abnormalities in a patient with compulsive hoarding: a 99mTc-ECD SPECT study


  • Hiromasa Ohtsuchi MD,

    1. Division of Neuropsychiatry, Department of Neuroscience, Yamaguchi University Graduate School of Medicine
    2. Yamaguchi Prefecture Mental Health Medical Center, Yamaguchi, Japan
    Search for more papers by this author
  • Koji Matsuo MD, PhD,

    Corresponding author
    1. Division of Neuropsychiatry, Department of Neuroscience, Yamaguchi University Graduate School of Medicine
    Search for more papers by this author
  • Takashi Akimoto MD, PhD,

    1. Division of Neuropsychiatry, Department of Neuroscience, Yamaguchi University Graduate School of Medicine
    Search for more papers by this author
  • Yoshifumi Watanabe MD, PhD

    1. Division of Neuropsychiatry, Department of Neuroscience, Yamaguchi University Graduate School of Medicine
    Search for more papers by this author

Koji Matsuo, MD, PhD, Division of Neuropsychiatry, Department of Neuroscience, Yamaguchi University Graduate School of Medicine, 1-1-1 Minamikogushi, Ube, Yamaguchi 755-8505, Japan. Email:


Little is known about the neuronal mechanism underpinning the pathophysiology of compulsive hoarding. We report the cerebral blood flow changes in an obsessive-convulsive patient with severe hoarding. The patient showed hyperperfusion of the fronto-temporal region and hypoperfusion of the striatal, the middle cingulate and the medial temporal regions during the stage with severe symptoms. Following improvement from the hoarding behaviors, the extent of hypoperfusion was expanded in the bilateral striatum, the anterior and middle cingulate gyrus. The result may substantiate evidence of the fronto-limbic abnormality involved in the pathophysiology of compulsive hoarding.

PATIENTS WITH OBSESSIVE-COMPULSIVE disorder (OCD) who have compulsive hoarding as their most prominent and distressing symptom (CH-OCD) account for 10–15% of all OCD patients.1 Compulsive hoarders acquire items that appear worthless to other people and have an inability to discard these items.2 It has been proposed that compulsive hoarding syndrome should be classified as a subgroup or a variant of OCD because its clinical characteristics are different from those of other OCD syndromes.3 However, the mechanisms underpinning the pathophysiology of CH-OCD remain unclear and neuroimaging studies of compulsive hoarders have not been fully investigated.4–6 Here, we report the cerebral blood flow (CBF) changes in a CH-OCD patient with severe hoarding symptoms and following her improvement from hoarding.


Case report

The patient was a 23-year-old female. The age of onset was around 12 years of age. At that time she exhibited obsessive checking, such as ‘Did I lose important items?’. When she was 16 years old, she was hospitalized for the treatment of OCD. During that period, her symptoms deteriorated due to psychosocial stress events, but her room was not littered. However, in the second grade of college, her obsessive symptom of checking her bag and desk worsened. The newspapers, magazines, wrappers of confectionaries and foods that she hoarded cluttered the rooms of her home. She could not leave her room and had little time to eat because her symptoms consumed the majority of her time. Finally, she was admitted to our hospital.

The patient met the DSM-IV-TR criteria for OCD on admission. She had hoarding symptoms but not any other obsessive behaviors. She also had comorbid social anxiety disorder. She had poor insight into her symptoms of hoarding and OCD and refused to receive treatment. Following the treatment proposed by Saxena et al.,7 we educated her as to how her compulsive hoarding and the acquired clutter impacted upon her daily life. We also encouraged her to understand her symptoms and to be motivated to participate in the treatment. Paroxetine (50 mg/day) and exposure behavioral therapy were applied. The Yale-Brown Obsessive Compulsive Scale (Y-BOCS)8 was used for assessment of the OCD severity. The 10 Y-BOCS items are each scored on a four-point scale from 0 = ‘no symptoms’ to 4 = ‘extreme symptoms’. The sum of the first five items is a severity index for obsessions, and the sum of the last five an index for compulsions. A total score indicates overall severity. Her Y-BOCS was rated 37 for total, 19 for the obsessive thoughts subcategory and 18 for the compulsive behaviors subcategory. She had 72 for state anxiety and 77 for trait anxiety on the State-Trait Anxiety Inventory (STAI).9 For exposure therapy, she was asked to bring the items she saved in her home to the hospital and to discard them in front of a psychiatrist (H.O.). It took around 1 h to discard a piece of paper in the first session. She obsessively checked the garbage that she discarded. Olanzapine administered 10 mg/day as an augmentation therapy reduced the symptoms. Following this combined therapy for 3 months, she was finally able to discard her hoarded items more easily (e.g. she cast about 100 pieces of paper and some magazines in 1 h). She recognized that important items (e.g. keys for her room and car, health insurance and credit cards) should be saved, but that a piece of newspaper, scratch books and a milk carton that had been on a kitchen table for 6 months should be thrown away. Her Y-BOCS and STAI scores at the second scan decreased compared to those at the first scan; Y-BOCS scores were 25 for total, 14 for the obsessive thoughts and 11 for the compulsive behaviors; STAI score was 53 for state anxiety and 74 for trait anxiety at the time of the second scan. Since the adverse effects of olanzapine, including overeating and weight gain, were observed we replaced olanzapine with risperidone (3 mg/day) without worsening of her symptoms. One year later, she was discharged from the hospital. She attended a clinic to treat her minor symptoms, got a part-time job and was integrated back into society. The Institutional Review Board of Yamaguchi University Hospital approved this study. Written informed consent was obtained from the patient.

SPECT acquisition and analysis

We used 99mTc-ethylcysteinate dimer (ECD, 600 MBq; Fuji Film Parma Ltd, Tokyo, Japan) SPECT. The dimer was injected via the right brachial vein. The SPECT imaging scan was at resting state. A triple-headed gamma camera system (GCA-9300A/PI; Toshiba Medical, Shibaura, Japan) was used. SPECT images were spatially normalized to standardized stereotactic space and smoothed with a 12-mm Gaussian filter. The smoothed images were analyzed using the Easy Z-Score Imaging System (eZIS).10 eZIS analyzed the patient's image using an a priori template from healthy volunteers (age range: 20–39 years) on SPM2 and showed the results of the Z-score map rendered upon the standardized MRI template. The extent threshold of the cluster size was set at ≥300 voxels. The two-tailed view with the levels of Z score ≥2 and ≤−2 was represented. A Z score ≥2 indicated hyperperfusion of the brain and a Z score ≤−2 indicated hypoperfusion. She was examined with the first SPECT during the severe state of her illness (Y-BOCS 37, State of STAI 72) and following improvement from hoarding (Y-BOCS 25, State of STAI 53). The patient received the first scan when she was hospitalized due to severe hoarding symptom and took paroxetine (50 mg/day), olanzapine (10 mg/day) and hypnotics and received the exposure therapy for 8 months. After 3 months, when she was discharged, she received the second scan when she took paroxetine (50 mg/day), risperidone (1 mg/day) and hypnotics.


During the stage with severe symptoms we observed hyperperfusion in the bilateral frontal, the left lateral temporal and the bilateral occipital regions (Fig. 1a). The superior medial region showed strong hyperperfusion in the frontal cortex. There was hypoperfusion in the bilateral striatal region, the left middle cingulate and the right medial temporal gyrus (Fig. 1a). Following improvement from hoarding, the area of hyperperfusion during the severe state shrank (Fig. 1b), while the extent of hypoperfusion was expanded in the bilateral striatum, the anterior gyrus, the middle cingulate gyrus and the medial temporal gyrus (Fig. 1b).

Figure 1.

Cerebral blood flow change of fronto-limbic structures in an obsessive compulsive disorder patient who is a compulsive hoarder analyzed by eZIS. (a) The cerebral blood flow of the patient with severe compulsive hoarding; (b) the CBF of the patient following improvement. A color bar indicates the Z-score.


This is the first longitudinal study of cerebral blood flow changes in CH-OCD. The regions with hyper- and hypoperfusion are where prior neuroimaging and lesion studies have shown neural abnormalities in patients with compulsive hoarding and collecting.4–6,11,12 Frontal area of the severe state showed wider hyperperfusion compared to that of the improvement and, in contrast, the striatum, the cingulate and the medial temporal gyrus of the two states exhibited hypoperfusion. Those of the improvement were broader than those of the severe state. The results suggest that altered perfusion of these regions may be associated with hoarding symptoms and the finding is involved in the pathophysiology of compulsive hoarding.

To date, there has been neuroimaging studies of CH-OCD published.4–6 A [18F]-fluorodeoxyglucose PET study showed that CH-OCD patients had lower glucose metabolism in the posterior cingulate gyrus and cuneus compared to healthy subjects and lower glucose metabolism in the dorsal anterior cingulate gyrus compared to OCD patients who did not have compulsive hoarding as a primary symptom.6 The hoarding-related pictures induced a greater activation in the left precentral/superior frontal gyrus, left fusiform gyrus and right orbitofrontal cortex of OCD patients when compared to healthy subjects. Within the OCD patients, hoarding-related anxiety was positively correlated with activation of the precentral/superior frontal gyrus.4 Another fMRI study of a hoarding-related provocation task showed that CH-OCD patients had a greater activation of the bilateral anterior ventromedial prefrontal cortex than OCD patients and healthy subjects.5 Lesion studies demonstrated that subjects with orbitofrontal (OFC), frontopolar, caudate or anterior cingulate damage had compulsive collecting behaviors.11,12 These findings indicate that the fronto-limbic structures including the dorsolateral and ventral-medial prefrontal, cingulate, striatum and medial temporal cortex are involved in the pathophysiology of compulsive hoarding.13 Together, the finding of the present study may support the evidence of pathophysiology in compulsive hoarding. However, we did not show abnormal OFC perfusion during severe hoarding symptoms or remission. The OFC plays a role in impulsive behavior and reward regulation and consists of the neural circuits underpinning the pathophysiology of compulsive hoarding.4 Although the reason for the failure of abnormal OFC perfusion in this case is unclear, the medication taken by the patient may contribute to weaken abnormal OFC activation. Longitudinal functional neuroimaging studies of a substantial number of medication-free CH-OCD and control subjects would resolve this issue. Another limitation is that because the patient had different severity of hoarding symptoms and other factors such as acquirement of insights of treatments and different medications between the first and second scan, these other factors potentially confounded the difference of regional cerebral blood flow between the two scans. Another limitation is that it is difficult to elucidate the association between regional cerebral blood flow changes and the patient's treatment because she received both pharmacological and behavioral therapies at the time of the two scans and this study is a case report. Future well-designed controlled studies with a substantial number of OCD patients with hoarding are required to clarify the association. Although the present study is a case report, it provides an insight into the pathophysiology of CH-OCD.