Screening for Leptomeningeal Disease by High-Sensitivity Flow Cytometry in High Risk Patients with Aggressive Non-Hodgkin’s Lymphoma.

Background: Central nervous system (CNS) involvement by non-Hodgkin’s lymphoma (NHL) portends a very poor prognosis. There is no consensus in the literature on the “high- risk” features that predict for leptomeningeal disease, and no standardized clinical guidelines exist regarding CNS surveillance, prophylaxis or treatment for patients at increased risk. 2–4 colour flow cytometry (FCM) has been reported to be more sensitive than standard cytology in detecting occult leptomeningeal disease (Blood 2005,105:496). The current study evaluates the utility of a high-sensitivity (5-colour) flow cytometry technique for detecting occult lymphoma cells in the cerebrospinal fluid (CSF) of high-risk patients with NHL.

Method: Patients with a new diagnosis of histologically aggressive B or T cell NHL were included in this study if they displayed one or more “high-risk” features for CNS involvement. Patients suspected of CNS relapse of NHL were also eligible for participation. Patients underwent routine staging investigations, with the addition of a diagnostic lumbar puncture (LP) during initial assessment. CSF was tested by standard cytology, cell count and biochemistry, and an additional 5 ml was obtained for analysis by high-sensitivity FCM on a Beckman Coulter FC500. The antibody panel (5 antibodies per tube) was customized according to the phenotype of the lymphoma. The key markers for B cell lymphoma were CD19/kappa/lambda with CD5 or CD10. CD45 was used to identify all white blood cells in the sample.

Results: Seventeen patients (8M/9F) with a median age of 59 (range 36–85) have been tested. Patients displayed anywhere from 2–6 “high-risk” features for CNS involvement. These included: HIV positivity (2), primary mediastinal B-cell lymphoma (4), bone marrow (5), multifocal bone (2), paraspinal (1), nasopharyngeal (2) or orbital (1) involvement, elevated serum LDH (12), multiple extranodal sites of disease (5), poor performance status (2), high IPI (3), B-symptoms (9), stage IV disease (11), and otherwise unexplained neurological symptoms (3). 14 patients underwent CSF analysis at time of initial diagnosis, one of whom had cranial nerve palsies secondary to a nasopharyngeal mass extending to the skull base. The other 3 were tested at relapse, transformation, and suspected CNS relapse ultimately diagnosed as a stroke. Despite the presence of these features, CSF analysis was negative for lymphoma cells by both cytology and FCM in all but one of the patients tested. However this patient had very high numbers of circulating lymphoma cells in the peripheral blood (PB), and the positive result was felt to be due to PB contamination of the CSF during a “bloody tap.” One patient with vague neurological symptoms had a negative LP at diagnosis, and later developed frank CNS involvement by lymphoma, but was too unwell to undergo a repeat LP.

Conclusions: Given the limited number of patients enrolled thus far and the low prevalence of patients with NHL and CNS involvement (2/17), it is difficult to fully assess the utility of high-sensitivity FCM in the diagnosis of occult leptomeningeal disease. It is of interest that CSF analysis was negative even in the patient with cranial nerve palsies and in the patient who later developed multiple CNS lesions secondary to lymphoma, suggesting that this technique may have limited sensitivity in diagnosing leptomeningeal disease. The systematic screening of high-risk patients cannot yet be recommended as standard clinical practice.