18F-FDG brain and whole body (WB) PET/CT in intracranial leptomeningeal carcinomatosis and in paraneoplastic neurologic syndromes

Background-Aim: Leptomeningeal carcinomatosis (LC) is usually diagnosed by brain magnetic resonance (MRI) and repeated cerebrospinal fluid (CSF) analyses. Paraneoplastic neurologic syndromes (PNS) are heterogeneous neurological disorders due to abnormal immune system response to unknown tumor and are diagnosed by serum specific autoantibodies (SSA), CSF, MRI and total body (WB) computed tomography (CT). Sometimes, in both these conditions the diagnosis is difficult. We evaluated 18F-FDG brain and WB PET/CT usefulness in patients with clinical signs suspected for LC and PNS, but with inconclusive or discordant CSF, SSA, MRI or CT. Methods: We retrospectively evaluated nine consecutive patients, five with suspected clinically signs of LC (Group A) and four with suspected PNS (Group B). In Group A, two cases (A-B) had been treated for breast carcinoma, two for lung adenocarcinomas (C-D) and the remaining one for rib chondrosarcoma (E). Initial CSF cytology was not conclusive in all patients; MRI was uncertain in A, B, C, D cases while in case E it was not practicable because of pacemaker. In Group B, one case had negative CSF and CT and positive SSA and MRI (F), one had positive CSF but negative SSA, MRI and CT (G), one had positive CFS and SSA but negative MRI and CT (H) and the remaining case had negative CSF, SSA, MRI, and CT (I). All patients underwent both WB and brain 18F-FDG PET/CT (Discovery 710, GE Healthcare) after i.v. 370 MBq dose. Results: In Group A, brain PET/CT evidenced numerous hypermetabolic foci in cerebral structures suggestive for LC in four cases (A, C, D, E). MRI remained unclear in repeated exams for intracranial lesions, while CSF only after repeated lumbar punctures ascertained cells with atypical mitosis suggestive for neoplastic cell dissemination (A, C, D, E). Moreover, in case A, WB PET/CT detected an elevated metabolic activity focus in medullary channel also evidenced by MRI and in case C a lung focus corresponding to adenocarcinoma already identified at diagnostic CT. However, both brain and WB PET/CT were negative in patients B, excluding LC, as confirmed by repeated CSF, despite unclear MRI. In Group B, brain PET/CT evidenced hypermetabolic areas in different cortical cerebral regions in three cases (F, G, H) which were classified as affected by PNS whose symptoms improved after treatment excluding metastatic lesions; however, during follow up, case H developed lung adenocarcinoma ascertained by WB PET/CT. Brain PET/CT was negative in the remaining one case I in whom further MRI evidenced demilinizing plaques in the bulb and spinal cord classifying him as affected by demyelinating disease and treated with immunosuppressant therapy. Conclusions: In the present study, both brain and WB PET/CT proved useful complementary diagnostic tools to ascertain LC and PNS, and the latter also for ascertaining extra cerebral neoplastic lesions. In particular, PET/CT is useful when MRI and CT are uncertain or not practicable or CSF and SSA are initially inconclusive. These data need to be confirmed by a larger number of case.