Infective endocarditis detected by ¹⁸F‐fluoro‐2‐deoxy‐d‐glucose positron emission tomography/computed tomography in a patient with occult infection

Abstract

Integrated ¹⁸F‐fluoro‐2‐deoxy‐d‐glucose positron emission tomography/computed tomography (FDG PET/CT) has been clinically used to detect infectious lesions. We present a case with pyrexia and bacteremia of unknown origin. Whole body FDG PET/CT was arranged to look for an occult source of infection and it revealed a focal lesion with increased FDG uptake in the mitral valve area. Under suspicion of infective endocarditis, transthoracic echocardiography was repeated and then the presence of linear vegetation over the calcified mitral annulus was confirmed. Ultimately, definite infective endocarditis was diagnosed according to the Duke criteria. The patient recovered after the antibiotic therapy. In our case, FDG PET/CT can help to localize the exact site of occult infection, thereby guiding additional testing and facilitating timely definitive diagnosis and therapy.

Introduction

Early diagnosis or exclusion of infection and inflammation is very important for the optimal management of patients with fever of unknown origin (FUO). However, in most settings just like this case herein, clinicians encounter substantial challenges in detecting and localizing the exact sites of infection. Integrated ¹⁸F‐fluoro‐2‐deoxy‐d‐glucose (FDG) positron emission tomography/computed tomography (PET/CT) has demonstrated a potential role in this aspect. Herein, we report on a patient with occult infection in whom infective endocarditis was detected by FDG PET/CT.

Case presentation

The patient was an 87‐year‐old female admitted for sudden onset of high fever. Associated symptoms included general malaise, anorexia, nausea, dyspnea, and dizziness. There was no complaint of cough, expectoration, dysuria, sore throat, vomiting, diarrhea, or abdominal pain. She denied recent history of traveling and contact with animals. She received colonoscopy for health checkup 1 week ago. Her past medical history included old tuberculosis, osteoporosis, hypertension, gouty arthritis, and peptic ulcer disease. On admission, pertinent findings on physical examination included a temperature of 39.6°C and mild rales over bilateral lower lungs revealed by chest auscultation. Laboratory tests revealed a white blood cell count of 12,600/μL with elevated neutrophils (88%), microcytic anemia with a hemoglobin level of 10.7 g/dL, an elevated erythrocyte sedimentation rate (65 mm/hr), and a raised C‐reactive protein (13.2 mg/L). A chest X‐ray film revealed increased infiltration over the right lower lobe (RLL). A clinical impression of RLL pneumonia was initially diagnosed. However, serial sputum cultures were sterile. Results from serial acid‐fast stain were negative. Several separate blood cultures during admission were persistently positive for gram‐negative bacilli that were later identified as Pseudomonas aeruginosa. Transthoracic echocardiography (TTE) revealed marked calcification of mitral valve annulus with mild mitral regurgitation, but no obvious evidence of valvular vegetation was found to suggest endocarditis. Owing to P aeruginosa bacteremia with undetermined primary focus, a whole body FDG PET/CT scan was arranged to localize the source of occult infection. Before the day of performing FDG PET/CT, antibiotic treatments given to the patient were as follows: levofloxacin for 2 days and then shifted to ciprofloxacin for 1 day, ceftazidime for 2 days, and isepamicin for 1 day. Blood culture and drug sensitivity test showed that the pathogen was sensitive to these antibiotics.

After 8 hours of fasting (serum glucose 94 mg/dL), the patient was injected with 12 mCi (444 MBq) of FDG intravenously. After a routine 1‐hour waiting period, images were obtained using a GE Discovery LS (Waukesha, WI, USA) PET/CT hybrid imaging scanner. In comparison with previous FDG PET/CT scan, the current FDG PET/CT scan images showed a new focus of increased intracardiac activity corresponding to the calcified mitral valve area, highly suspicious for infective endocarditis. Besides, diffusely increased FDG uptake in the spleen was noted, which could be immunologic change related to infection (1, 2). Moreover, there was no abnormal FDG uptake in the right lower lung. This image finding did not support the initial clinical impression of RLL pneumonia. Because of the FDG PET/CT finding, TTE was performed again by an experienced echocardiographer and the presence of a tiny oscillating vegetation attached to the heavily calcified mitral annulus was confirmed (Fig. 3). Ultimately, definite infective endocarditis was diagnosed according to the Duke criteria. The patient became afebrile and the blood culture specimens were sterile within 2 weeks of initiation of the antibiotic therapy and continued antibiotic therapy for 6 weeks.

Figure 1.

Maximum intensity projection positron emission tomography images. (A) No definite evidence of abnormal ¹⁸F‐fluoro‐2‐deoxy‐d‐glucose (FDG) uptake in the intracardiac region for cancer screening 1 year ago. (B) It revealed focally increased FDG activity in the intracardiac region (arrow) and diffusely increased FDG activity in the spleen (arrowhead). There was no abnormal FDG uptake in the corresponding region, although increased infiltration over the right lower lobe was noted by previous chest X‐ray image.

Figure 2.

Selected transaxial slices of ¹⁸F‐fluoro‐2‐deoxy‐d‐glucose positron emission tomography/computed tomography (FDG PET/CT) scan are shown with CT (left), PET (middle), and PET/CT fusion (right) images. (Row A) Thoracic transaxial images show a focal intracardiac lesion of increased FDG uptake corresponding to the calcified mitral valve area (arrow). (Row B) Abdominal transaxial images show diffusely increased FDG uptake in the spleen (arrowhead).

Figure 3.

Transthoracic echocardiography reveals a tiny oscillating vegetation (long arrow) attached to the heavily calcified mitral annulus (short arrow).

Discussion

FDG PET has been an established diagnostic tool in oncology and the indications for FDG PET are expanding rapidly. FDG accumulates in all tissues with a high rate of glycolysis, which does not occur exclusively in neoplastic cells. Increased uptake and retention of FDG has been shown in lesions with a high concentration of inflammatory cells, such as granulocytes and activated macrophages, enabling imaging of infectious or inflammatory processes [[1], [2]]. Therefore, in the appropriate settings, FDG PET imaging can be effectively used to detect and characterize infectious and inflammatory processes. Over the past three decades, several scintigraphic techniques were introduced for examining patients with suspected infection or inflammation. Gallium‐67 is still the most widely used agent because of its low cost and wide availability. Like FDG, gallium is not specific, which not only accumulates in malignancies but also in inflammatory or infectious processes. FDG PET/CT has many advantages over conventional radionuclide imaging modalities for the diagnosis of infectious and inflammatory diseases. These advantages include the feasibility of securing diagnostic results within 1.5–2 hours, high and early target‐to‐background contrast, better spatial resolution of PET in comparison with SPECT, and accurate anatomical localization of sites of abnormality. An obvious disadvantage of PET is the relatively high cost and the lesser availability [3]. The percentage of FDG PET scans that are helpful in the diagnostic process in patients with FUO, as reported in the literature, varies from 41% to 69%, which is very high compared with radiological techniques and gallium‐67 scintigraphy [4]. FDG PET compares favorably with gallium imaging and, because it can be performed rapidly, could replace gallium scintigraphy as a radionuclide study for the evaluation of patients with FUO. In this case with unlocated infection, whole body FDG PET/CT scan effectively revealed a hypermetabolic focus in the mitral valve area, raising suspicion for infective endocarditis. Repeated TTE was done therefore, in spite of negative result of initial test 5 days ago, and confirmed the presence of a valvular vegetation.

Infective endocarditis can be the source of an occult infection and represents a severe event requiring early diagnosis and adequate therapy. The diagnosis of infective endocarditis requires the integration of clinical, laboratory, and echocardiographic data based on the Duke University criteria proposed in 1994. TTE is a rapid and noninvasive method showing a high specificity for the diagnosis of vegetations (up to 98%), but it is affected by an unsatisfactory sensitivity (60–70%) because of various technically limiting factors [[5], [6]]. In our case, the false‐negative result of the initial TTE might be attributed to a small vegetation in the markedly calcified valve, which was obscured by the intense echoes produced by the calcification and might escape detection.

The characteristic lesion, a vegetation made up of fibrins, platelets, microorganisms, and inflammatory cells, involves both native and prosthetic heart valves most frequently. Increased FDG uptakes may thus appear in the vegetations. To date, reports related to this particular application have been limited. A previous pilot study suggests that, in spite of the normal myocardial FDG uptake, FDG PET demonstrated a satisfactory accuracy, comparable with that obtained by echocardiography, in detecting infective endocarditis [7].

It is well known that FDG distribution patterns in the normal myocardium are variable and can be classified into three types: (1) fully retained FDG activity with uniform intense uptake throughout the myocardium; (2) heterogeneous FDG activity with some focal uptake mainly in the base of the heart; and (3) full change to fatty metabolism with uniform low or no uptake, as in this presented case [8]. Moreover, it has been observed that abnormal cardiac or mediastinal FDG uptake occasionally occurs in some benign lesions, such as lipomatous hypertrophy of the interatrial septum, sarcoidosis, lymphadenitis, thymic hyperplasia, and various benign mediastinal tumors [[9], [10]]. High physiological and other benign FDG uptakes perhaps interfere with the detection of endocarditis and limit the specificity of FDG PET. Coregistration of PET and CT in integrated PET/CT offers significant advantages, over the stand‐alone PET, in the localization of focal uptake and in the differentiation of pathological uptake from the normal physiological uptake [11]. In this case, valvular calcification shown by coregistered CT helps for exact localization of abnormal FDG uptake.

In addition to the focal intracardiac activity, FDG PET/CT in this case also shows diffuse splenic uptake of FDG. As an integral part of the body's immune system, the spleen executes multiple tasks, including removal of encapsulated bacteria, production of inflammatory substances and immunoglobulins, and phagocytosis of infectious agents, in addition to serving as a reservoir of cellular elements, including leukocytes. Presumably, the diffusely increased splenic activity reflects increased glucose metabolism of this organ in the setting of infection.

In conclusion, FDG PET/CT can help to localize the exact site of occult infection and effectively detects infective endocarditis in this case. This case report demonstrates the potential value of FDG PET/CT in the diagnosis of infective endocarditis, especially for those with normal echocardiograms. Furthermore, FDG PET/CT can detect hypermetabolic focus associated with infective endocarditis even after effective antibiotic treatment for a short period (up to 3 days). Although FDG PET/CT dose not directly provide a definite diagnosis (i.e. a histological or a microbiological diagnosis), it often provides anatomic localization where a hypermetabolic process is ongoing and, with the help of other techniques, facilitates timely definitive diagnosis and therapy.