The Role of 18F-Fluorodeoxyglucose Positron Emission Tomography and CT (PET-CT) for the Diagnosis of Infections in Hemato-Oncological Patients with Persistent Neutropenia and Fever.

Abstract 3121

Poster Board III-58

Hemato-oncological patients with prolonged neutropenia and persistent fever are at high risk for bacterial as well as invasive fungal infections (IFIs). Due to limitations in diagnosing infections in these patients at an early stage, these patients suffer sometimes from under and over diagnosis and treatment. 18-Fluoro-2-deoxy-D-glucose positron emission tomography is an emerging imaging technique for the diagnosis of infections. We aimed to evaluate the role of PET-CT for detection of bacterial and invasive fungal infections among high-risk hemato-oncological patients with persistent neutropenia and fever as compared with the final clinical diagnosis. We included hemato-oncological patients undergoing chemotherapy, allogeneic or autologous hematopoietic stem cell transplantation (HSCT) at our center. All consecutive, consenting patients with neutropenia (<500/mm³) and persistent fever (>4 days) despite broad spectrum antibiotic treatment underwent a PET-CT examination instead of a routine chest and sinus CT, which is our common practice. Data were prospectively collected from the day of recruitment until 30 days following neutropenia resolution (or death), including patient characteristics, underlying malignancy and treatment, characteristics of the febrile episode, neutropenia duration, antibiotic and antifungal management and outcomes. Each PET- CT examination was reviewed independently by a radiologist who analyzed the CT part alone and a specialist in nuclear medicine who analyzed the results of the full PET- CT scan. Both specialists received the clinical information available prior to the test and were blinded to the interpretation of the comparator test. The reference standard was the final clinical diagnosis, as determined by an expert panel consisting of a hematologist and an infectious diseases expert, 30 days following neutropenia resolution. We assumed that infections will be apparent or can be definitely ruled out by that time. Sensitivity of the PET-CT was compared to that of chest and sinus CT by means of the McNemar test. Results were available to clinicians in real time. Between January 2008 and July 2009, 35 PET-CT examinations were performed for different episodes of persistent febrile neutropenia in 32 patients. The cohort included: 24 patients with acute leukemia undergoing chemotherapy (27 episodes: 19 episodes during induction therapy, 2 episodes during consolidation, 6 episodes during salvage), 5 patients undergoing allogeneic HSCT (4 acute leukemia and 1 myelofibrosis), and 3 patients undergoing autologous HSCT (all lymphoma).

The final clinical infectious diagnoses were: IFI (8), IFI and bacteremia (5), clinically-documented infection (10), microbiologically documented bacterial infection/ bacteremia (6) and fever of unknown origin (6). Chest and sinus CT demonstrated findings in addition to the routine workup for persistent neutropenia and fever in 23 patients, and PET- CT demonstrated findings in 29 of 35 patients. The findings on PET- CT in addition to chest and sinus CT included abdominal infections as appendicitis (2 patients), typhlitis, pseudomembraneous colitis, diverticulitis and pyelonephritis. All these infections were visualized on abdominal CT, and emphasized by FDG. 4 of these 6 patients were asymptomatic at the time PET-CT was performed. Diagnosis by PET-CT preceded subsequent symptoms by 2-4 days. In addition cellulitis, abscess and an endovascular infection (infected right atrial thrombus originating from a central venous catheter) were diagnosed by the PET component of the test. PET-CT changed patient management in 8 cases (23%) (change of antibiotics in 2, appendectomy in 2, tonsillar abscess drainage in 1, perianal abscess drainage in 1, tooth abscess extraction in 1, central line removal in 1).

The diagnosis according to CT alone was consistent with the final clinical diagnosis in 23 cases (sensitivity 66%), while according to PET-CT it was consistent with final diagnosis in 28 cases (sensitivity 80%). The difference was without statistical significance given the small sample size.

PET- CT has a potential role for the diagnosis of infections in neutropenic patients with persistent fever. It seems that the addition of abdominal CT to the routine chest and sinus CT adds valuable information. PET-CT accentuates these findings and helps in diagnosing abdominal infections and systemic candidiasis.