Joint Program in Nuclear Medicine

Scintigraphy of Brain Abscesses

Richard C. Hom, MD, PhD
Finn Mannting, MD, PhD

March 1, 1994

Presentation

A 43 year old man with history of Grade III glioblastoma was treated with surgical resection. Post operatively, he had an infection at the surgical site which grew P. acne and S. epidermidis. He was treated with antibiotics for 6 weeks. An MRI, a brain perfusion scintigraphy with Tc-99m HMPAO, and In-111 labelled white blood cell scintigraphy were performed prior to closure of the craniotomy site.

Imaging Findings

A composite image (45k bytes) show a large region of increased T2 signal on MRI with a somewhat smaller region of decreased perfusion on the Tc-99m HMPAO brain perfusion study. The white blood cell study shows moderate increased uptake in this same region, indicating a persistent infection at the site of resection.

Discussion

It has been estimated that brain abscess accounts for approximately 1 in 10,000 general hospital admissions and 4-10 cases are seen yearly on active neurosurgical services in hospitals of developed countries. Because of the low incidence along with the presence of competing modalities such as CT and MRI scanning, the role of scintigraphic imaging of brain abscesses has been rather limited, often used when the CT and MRI results are equivocal or nonspecific such as when there is cerebritis early in the infection. The published scintigraphic techniques have been with Gallium scanning and in the past 10-15 years, with In-111 labelled white blood cells. Studies done using Tc-99mHMPAO-labelled white blood cells have been rare, probably because the tracer development came later on.

Gallium scanning has been used in the diagnosis of brain abscess. Waxman and Siemsen (1976) was able to diagnose 5/5 cases of brain abscesses with Ga-67 scintigraphy. This was, however, a very small series. Because of concerns with respect to the agent's sensitivity and specificity and the requirement that the study be performed up to 48 hours following tracer injection, later attention was given to the use of white blood cell-labelling.

Rehncrona et al (1985) published a series of 16 patients who underwent In-111 labelled white blood cell scanning for brain abscesses 24-48 hours following injection: 

          Final Diagnosis                No. with + scan


    Brain abscess (clinical course in 3/5,
                   surgical finding in 2/5)   4*/5			        
    
    Brain tumor (8/10 with astrocytoma)	      0/10
    Brain tumor (necrotic astrocytoma)	      1/1


    *Of note is that the one brain abscess with the false
     negative study had the study done 9 days following
     steroid and antibiotic therapy.
Similarly Schmidt et al (1990) retrospectively studied 28 patients in Denmark over a 5 year period for suspected brain abscess using In-111 labelled white blood cell scanning: 
    Final Diagnosis      No. with + Scans


    Brain Abscess         7/7  (all were intense; when patients were
                                        not on high dose of steroids)
    Brain Tumor/mets.     9/18 (of the 9 +s: 3 moderate; 6 weak)
    Brain infarction      0/1
    Brain hematoma        0/1
Of the 2 patients with brain abscesses who were studied twice, both had negative studies when they were on high dose steroids. One patient on high dose steroids was studied once and had a negative study. Of the 9 patients with brain tumors with positive scans, 6 were on steroids. Those with negative scans, 5/10 were on steroids.

Despite the successes of the technique of using labelled white blood cell for imaging brain abscesses in patients not on high dose steroids, periodic reports describe the false positives as occurring in the presence of brain tumor. Similar to the data reported by Schmidt et al, Bellotti et al have suggested that using the criteria that any focal uptake in excess of the physiological uptake of the base of the skull reduces the false positives from malignant cystic lesions. So far, there have been no series published with SPECT imaging or with the use of Tc-99m HMPAO for imaging brain abscesses. Whether SPECT imaging will have a significant role in increasing sensitivity of scintigraphy of brain abscesses remains to be determined.

The role of PET scanning in the diagnosis of brain abscess has been exceedingly limited due to the small number of cases. Reports have been mainly anecdotal. Most published studies have been case reports using agents which are designed for imaging tumors. One study showed that L-methyl [C-11]-methionine is taken up by a brain abscess in one patient. Another study showed that [F-18]-fluorodeoxyglucose is taken up in abscesses.

References

  1. AD Waxman and JK Siemsen. Gallium scanning in cerebral and cranial infections. Am J. Roentgenol 127:309-314 (1976).

  2. S Rehncrona, J Brismar, and S Holtas. Diagnosis of brain abscesses with Indium-111-labelled leukocytes. Neurosurgery 16:23-26 (1985).

  3. C Bellotti, MG Aragno, M Medina, AL Viglietti, et al. Differential diagnosis of CT-hypodense cranial lesions with indium-111-oxine-labelled leukocytes. J. Neurosurg. 64:750-753 (1986),

  4. KG Schmidt, JW Rasmussen, PB Frederiksen, C Kock-Jensen, and NT Pedersen. Indium-111-Granulocyte scintigraphy in brain abscess diagnosis: Limitations and Pitfalls. J. Nucl Med 31:1121-1127 (1990).

  5. S Balachandran, M. Husain, JR Adametz, JS Pallin, TL Angtuaco, and CM Boyd. Uptake of Indium-111-labeled leukocytes by brain metastasis. Neurosurgery 20:606-609 (1987).

Click here to go to Joint Program in Nuclear Medicine home page and Copyright notice.

J. Anthony Parker, MD PhD, Tony_Parker@bidmc.harvard.edu