Introduction

Evaluation of fever of unknown origin (FUO) is a well-established indication for 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) / computed tomography (CT) scan as it sometimes might be a challenge for the clinician to determine the cause of a fever after ruling out the most common causes (1,2). Often, despite having gone through the whole diagnostic work-up, including different imaging modalities as well as extensive clinical examinations, it still remains impossible to identify the focus causing the fever. Deep venous thrombosis, sometimes resulting in thrombophlebitis, leads to systemic inflammatory reaction of the body including fever. Identifying the underlying cause of the fever is of crucial importance in the course of optimal therapeutic approach (2). In this setting, FDG-PET was shown to be one of the most reliable diagnostic modalities and therefore of value in the clinical setting of FUO. FDG-PET correlates well with the metabolic activity of a certain lesion such as inflammation (3,4,5). FDG-PET/CT is therefore not only of value in the initial localization of the underlying cause of FUO but might also serve as a surrogate marker for a positive therapy response by showing decreased metabolic (inflammatory) activity.

Case Report

A 60-year old male patient was admitted to the hospital because of FUO. Additionally, he presented with a generalized malaise, and subtle neurologic symptoms. Blood sample tests were positive for staphylococcus aureus and revealed increased CRP, leucocytes and blood sedimentation. One month before the admission to the hospital, the patient had self extracted an infected toe nail. Besides the positive blood culture, no other examinations including urine culture, repeated echocardiography of the heart, CT of the cerebrum and thorax, and fecal cultures showed any indication for a bacterial focus. He was treated with intravenous and per oral antibiotics (flucoxacillin, clindamycin) during which the fever and local complaints subsided and inflammatory and microbiological parameters in blood normalized. He was discharged but soon again admitted with shivering fever and collapse. Now, a FDG-PET/CT was performed and showed an occlusion in the left brachiocephalic vein which was very intensely PET-positive, attributable to a thrombosis with thrombophlebitis, possibly with associated infection. Additionally, several small foci of increased FDG uptake were seen in the lungs caused by probably infected emboli (Figure 1). A subsequently performed ultrasound examination of the neck revealed a thrombus in the left jugular vein and also in the left supraclavicular vein.

Based on the results of the FDG-PET/CT, low molecular heparin (in a therapeutic dose) was added to the antibiotic therapy. Clinical improvement was observed shortly thereafter. Fever disappeared about 1-2 days after the introduction of low molecular heparin. Additionally, CRP dropped to values of 8 mg/dl, from values between 18 and 40 mg/dl and did not rise again until the end of treatment. Four weeks after the first, a second FDG-PET/CT was performed before patient’s discharge from the hospital. No abnormal FDG uptake could be shown in the intravascular localization and intrathoracic foci. Clinically, all previous foci had disappeared. The patient was than discharged and successfully continued with the treatment of oral antibiotics for another 2 weeks, and anticoagulants for about one year. Clinical follow up examinations during that time course remained negative for a relapse of the infection and/or thrombosis.

Literature Review and Discussion

The salient feature of this case report is the use of a FDG-PET/CT scan for the monitoring of the success of treatment of deep septic thrombophlebitis. In our patient, FDG-PET/CT scan was also crucial for the diagnosis of a septic thrombophlebitis in the brachiocephalic vein. Some studies have reported FDG-PET/CT scanning in venous thrombotic disorders. Miceli et al. (1) reported 9 cases of septic deep venous thrombophlebitis localized in vena cava, innominate, subclavian, and internal jugular veins, showing in one case histopathology of an organizing thrombus with neutrophillic cell infiltrate in the vessel wall as the cause for increased FDG uptake. Sopov et al. (3) have made a distinction between non-metabolically active, so-called “bland” thrombi and metabolically active, FDG-positive thrombi as a result of 1. vascular wall inflammation (phlebitis), 2. systemic inflammatory disorder (vasculitis), 3. infection, 4. catheter-related, or 5. tumor tissue thrombus. Although the diagnosis in our patient was not confirmed by histological examination of the thrombus, it is most likely that in our patient central venous thrombosis and -phlebitis was caused by septicemia after septic toe nail extraction and recurrent thrombophlebitis as a result of previous infusion of intravenous antibiotics. In this situation follow-up PET-CT was very valuable to indicate reconvalescence in monitoring successful treatment with oral antibiotics and anticoagulants after previous relapse, and for the prognosis and safe hospital discharge of the patient. Other imaging modalities such as CT and ultrasonography might also be able to detect deep venous thrombosis (6,7), however they can not conclude whether the vascular occlusion is infected or not. In contrast, noninvasive diagnosis of infectious changes is a major advantage of FDG-PET, knowing that leukocytes concentrating in sites of infection, accumulate glucose for their respiratory burst - reaction used for fighting infection (8). This correlates directly with the intensity of FDG accumulation and facilitates the detection of infectious foci. The additional value of combined PET and CT imaging is based on the fact that both anatomic and metabolic information are obtained by the hybrid imaging modality, enabling exact localisation in particular vessels. Furthermore, as in our case whole-body FDG-PET/CT is able to detect potentially multiple active embolic. In the context of fever of unknown origin, such as in our patient FDG-PET/CT performed with high accuracy and a pooled sensitivity and specificity origin of 0.982 and 0.859 respectively (9). In conclusion we suggest that non-invasive therapy evaluation of deep venous thromboflebitis seems possible with FDG-PET/CT and can provide valuable information about additional infectious localisations and success of treatment.