European Journal of Nuclear Medicine and Molecular Imaging

, Volume 35, Issue 8, pp 1421–1423

Optimizing interventions for preventing uptake in the brown adipose tissue in FDG-PET


  • Sandip Basu
    • Division of Nuclear MedicineHospital of University of Pennsylvania
    • Radiation Medicine Centre (BARC)Tata Memorial Hospital Annex
    • Division of Nuclear MedicineHospital of University of Pennsylvania

DOI: 10.1007/s00259-008-0720-6

Cite this article as:
Basu, S. & Alavi, A. Eur J Nucl Med Mol Imaging (2008) 35: 1421. doi:10.1007/s00259-008-0720-6


Brown adipose tissueFDGPETPropranololBrown fatDiazepamFentanylPET-CT

With the increasing use of 18-fluoro-deoxy-glucose-positron emission tomography (FDG-PET) in the practice of medicine, enhanced glycolysis in brown adipose tissue (BAT) has posed a major concern for accurate interpretation of the images [1, 2]. Brown fat is a sub-type of adipose tissue and regulates body temperature by non-shivering thermogenesis. This has been termed “USA Fat” by some investigators [2]. Histologically, it is characterized by high vascularity, dense mitochondria in the cells, and abundant sympathetic noradrenergic innervation. The usual distribution of BAT includes neck and shoulder region, axillae, mediastinum, retrocrural and paravertebral sites [14]. FDG uptake in BAT has been noted to occur more frequently in cold months [5]. Recent data suggests that FDG uptake in BAT occurs more often as a result of acute response to cold weather rather than the prolonged periods of average cold weather [6]. Significant localization of FDG in BAT has been particularly problematic for accurate assessment of suspected malignant lymph nodes in the neck region. Also, there have been reports [7, 8] of atypical as well as asymmetrical distribution of BAT. Therefore, it is important to recognize these variants so that they are not misinterpreted as disease sites.

In addition, the serendipitous discovery of brown fat by FDG-PET has provided a unique opportunity to study the physiology of this complex adipose tissue [9, 10]. In fact, there has been significant interest in identifying and studying the physiology of active brown adipose tissue in the recent years by in vivo functional imaging techniques [914]. It is now clear that a sizeable proportion of adult humans possess active brown adipose tissue [10] and the distribution of BAT in the human is slightly different from those in the rodents, being primarily located in the supraclavicular and the neck regions. Such studies are essential for optimal understanding of its importance in regulating human metabolism and for its potential role in the genesis and treatment of obesity. Hence, we strongly believe the therapeutic interventions that may influence BAT activity will also shed some light on this type of investigation and can provide valuable insights into the antiobesity research.

Several types of interventions [1521] have been proposed to reduce the metabolic activity of BAT, but there has not been any clear consensus or a set protocol to stop FDG uptake in BAT. The methods adopted can be broadly classified into three groups: (a) pharmacological interventions such as premedication with benzodiazepine, propranolol, reserpine, and fentanyl [1518] and (b) warming maneuvers and temperature control during the PET procedure [19, 20]. In addition, (c) precise localization with fusion PET/CT scan [1, 2, 21, 22] has been proposed as a means to reduce misinterpretation of the results generated. Obviously, comparison with the corresponding CT image is of great importance in making the optimal diagnosis in this setting. The initial results with oral diazepam have been reproduced with limited success. In one study [18], FDG uptake in BAT was reduced by intravenous fentanyl premedication, which appeared to be an effective alternative to moderate-dose (0.10 mg/kg) oral diazepam administration while, low-dose diazepam (0.06 mg/kg) was found to be ineffective. Among the various pharmacological interventions that have been tested to date, the administration of oral propranolol appears to be the most promising.

The knowledge of enhanced brown fat activity due to sympathetic overactivity has stemmed from two types of observation in recent years: (1) increased glycolysis after the administration of ß-adrenergic agonist drugs in animal studies [12], and (2) enhanced BAT metabolism in patients with pheochromocytoma, who are known to have elevated levels of circulating catecholamines [13, 14]. It has been postulated that norepinephrine released from sympathetic nerve terminals binds to ß3-adrenergic receptors on the surface of BAT and leads to heat production and enhances BAT glycolysis mainly through its effect on glucose transporter-1 receptors [23]. It has been noted in animal studies that uptake of FDG into BAT is increased by intraperitoneal injections of nicotine and ephedrine, and this effect is further enhanced when both these drugs are administered in combination [12]. These observations have suggested that a non-selective ß-blocker such as propranolol might minimize BAT-FDG activity. Tatsumi et al. [15] reported the efficacy of this approach in female Lewis rats. In this rodent study, anesthetized rats were administered propranolol, reserpine, or diazepam intraperitoneally before FDG injection to determine whether these agents reduced FDG uptake in BAT. While the control and diazepam-administered groups exhibited intense FDG uptake in BAT, those that received propranolol or reserpine showed only faint to mild FDG uptake in BAT. They concluded that propranolol or reserpine treatment can remarkably reduce the high FDG uptake in BAT. Sonderlund et al. [16] tested this proposition in 11 individuals with strong FDG uptake in the baseline scan. Reexamination of these subjects was conducted at a mean interval of 5 days (range 2–8 days), after 80 mg of propranolol was given orally 2 h before FDG administration. SUV assessments of the uptake in brown fat, lung, heart, liver, spleen, and bone marrow were also determined in addition to visual evaluation. All patients showed complete or near total resolution of the brown fat activity on the second examination (p < 0.001), both upon visual evaluation and by comparing SUVs from the two studies. There was no alteration of FDG activity in the other organs or the malignant lesions except for a mild decrease in the myocardial uptake. A similar prospective study with lower dose oral propranolol was carried out by Parysow et al. [17], who orally administered 20 mg propranolol per 60 min before FDG administration. These investigators examined 26 patients with cancer who had increased BAT FDG uptake in the baseline study. In their cohort, mean basal BAT SUVmax was 5.52 ± 2.3 and mean postpropranolol SUVmax was 1.39 ± 0.42 (P < 0.0001). Image interpretation was significantly improved in nine patients due to resolution of mediastinal FDG uptake in the postpropranolol scan. These three studies have provided us with a relatively simple solution to this problem. Obviously, it is desirable that a protocol be designed that would employ the minimum dose of propranolol and at the same time not affect the accuracy of the test. Hence, it will be worthwhile to design a prospective study where different doses of propranolol are administered to determine the optimal dose to be employed to prevent this undesirable uptake in FDG-PET examinations. This will assist the PET community to enhance the role of this modality in the assessment of patients with cancer and other disorders.

Another important aspect is to study the change in the degree and pattern of FDG uptake in BAT on dual time point scanning and to determine if this technique can help in differentiating brown fat from malignant lesions. In a retrospective study of 32 subjects [24], the SUVmax in the BAT showed wide variability (range 0.9 to 10, mean SUV was 3.8 ± 1.6) and, on dual time point imaging, interestingly, 92% of the brown fat spots demonstrated an increase in SUVmax ranging from 12 to 192%. In some instances, there was an increase in the number of brown fat spots. While more data require to be accrued in this area, it appears from this preliminary data, that this technique may not be useful to differentiate between BAT uptakes from those of the malignant lesions.


This work was supported in part by the International Union against Cancer (UICC), Geneva, Switzerland, under the ACSBI fellowship.

Copyright information

© Springer-Verlag 2008