FDG PET/CT in diagnostic evaluation and response assessment of Acute myeloid leukemia with chloroma

Objectives: Bone marrow biopsy is typically used for diagnosis as well as assessment of treatment in AML but it cannot identify extramedullary involvement. Conventional imaging such as CT and MRI, help detect additional lesions over clinical examination, but cannot detect small occult sites. PET/CT by combining metabolism with anatomy, should help detect such sites. In the present study, we evaluated the role of 18F-FDG PET/CT in identifying the sites of extramedullary involvement while simultaneously evaluating the intramedullary disease. We also, evaluated its role in treatment response assessment of AML with chloroma. Materials and Methods: A total of 22 AML patients with clinically detected EMD were prospectively evaluated with 18F-FDG PET/CT for primary staging, disease extent and response assessment. After fasting for 6 hours, they were injected with 5-10 MBq/kg IV F18 Fluorodeoxyglucose (FDG) and whole body PET/CT was obtained 40-60 minutes after injection. The scans were analyzed by 2 experienced nuclear medicine physicians and all the lesions were documented. SUVmax, SUVavg, metabolic tumour volume (MTV), tumour to back ground ratio (TBR) of all the identified lesions were measured using iso-contour ellipsoid ROI with threshold 2.5 SUVmax. Findings of FDG PET/CT were compared with corresponding CT images. Follow up 18-FDG PET/CT was performed after first cycle of chemotherapy.

Results: A total of 22 patients (18 males and 4 females with mean age -9.64years± 4.82) among which 19 were diagnosed with AML for the first time (mean age -9.58years ± 4.97) and 3 were known cases of relapse (mean age -10.00 years ±4.58) were included in the study. Male to female ratio of newly diagnosed cases of AML was 3:16, where as in relapse cases it was 1:2. A total of 125 lesions were detected in baseline 18F-FDG PET/CT for 22 patients (range: 1 lesion/patient - 20 lesions/patient). CT could only detect 78 lesions (range: 1 lesion/patient - 13 lesions/patient) out of 125 lesions detected by PET/CT. Thus, PET/CT detected 47 more lesions than CT. Splenic involvement was accurately detected by PET/CT which was missed by CT. Our data indicated that extramedullary AML predominantly involving certain sites such as spleen, soft tissue, lymph nodes and CNS are frequently missed by CT. Among the skeletal lesions CT was able to localize only 3 lesions whereas, PET/CT detected 5 lesions. Out of 22 patients, follow up PET/CT was done in 14 patients. In CT 13 patients were responders (Complete Response (CR)+ Partial Response (PR)) and 1 was non-responder (Progressive Disease(PD)) whereas, in PET 10 patients were responders (CR+PR) and 4 were non-responders (Stable Disease (SD)+PD). Five parameters (size, SUVmax, SUVavg, MTV and TBR) were evaluated and compared between metabolic responders and non-responders. It was found that out of five parameters only TBR was significantly higher in metabolic responders (4.57±4.34) than in metabolic non-responders (1.75±0.995) (p<0.05). Also, TBR was found to be significantly different between CR (3.66±3.41) and PR (9.08±5.71) (p<0.01). In ROC analysis the optimum cut-off for TBR to predict CR and PR was 1.83 with AUROC 0.86, sensitivity 89.8% and specificity 75%. None of the parameters were found to be significantly different in partial response vs stable disease and stable disease vs progressive disease.

Conclusions: 18F-FDG PET/CT has significant role in detecting extramedullary disease in AML patients which is often missed by clinical examination as well as by other conventional modalities like CT. PET combined with CT (PET/CT) is a highly sensitive imaging modality in the diagnostic evaluation and detection of additional extramedullary sites. Also, metabolic parameter i.e. TBR evaluated from 18F-FDG PET/CT is found to have significant role in response prediction in these patients.