Screening Cerebrospinal Fluid for Aggressive B-Cell Lymphoma: Significance of Clonal Populations Found with 8-Color Flow Cytometry

Abstract

Background: Secondary central nervous system (CNS) lymphoma confers a poor prognosis, and prompt identification is critical for guiding management. Over the last decade, we have employed cerebrospinal fluid (CSF) analysis with 8-color flow cytometry (FCM) in screening patients deemed at risk of CNS relapse. However, there is little data to guide interpretation and management of positive results, and concurrent clinical and radiographic features of FCM positive (FCM+) patients, remain areas of uncertainty. Methods: We analyzed newly diagnosed (ND) and relapsed/refractory (R/R) diffuse large B cell lymphoma (DLBCL) and high grade B cell lymphoma (HGBCL) patients who underwent a screening lumbar puncture (LP) from 2011-2020. Clinical, imaging, and CSF findings (differential, cytology, FCM) were collected. Patients with prior CNS involvement or FCM showing CLL/low-grade B-NHL were excluded. “FCM-only” patients were defined as FCM+ but without any other evidence of CNS spread (negative cell count/cytology, MRI, and no symptoms), whereas others were deemed to have overt CNS disease. Results: There were 415 LPs among 361 patients (243 ND, 172 R/R). 21 (5%) patients were FCM+, 10 ND and 11 R/R patients (Table 1). Four had FCM-only involvement (2 ND, 2 R/R). Among these, the clonal population was 1.63% of WBC (range, 0.3-2.7%), and 3 had concurrent bone marrow (BM) or peripheral blood (PB) involvement. All 4 received intrathecal (IT) therapy. At median follow up of 23 months, 3 (75%) have died (2 from CNS disease progression, 1 unknown due to loss in follow up), and 1 remains alive in remission 44 months later following IT methotrexate (MTX) and CAR T-cell therapy. 17 had overt CNS disease (8 ND, 9 R/R). The clonal population was 9% (range, 0.2-98%). Eight (47%) had concurrent BM or PB involvement. Twelve (70%) received IV high dose MTX, and the remainder received only IT therapy. At median follow up of 42 months, 9 (53%) have relapsed (2 CNS only, 1 systemic, 6 both), and 9 have died. Median time to CNS relapse was 5 months. Conclusions: CSF 8 color FCM screening of DLBCL/HGBCL patients yields a low proportion of positive results, mostly in the setting of overt CNS disease. However, even FCM-only patients had poor outcomes (2/4 with CNS progression), suggesting that blood contamination does not explain CSF findings, and that these patients should be approached as secondary CNS involvement. Better pt selection for CSF screening is needed to reduce unnecessary LPs. Intensified or novel approaches for all FCM+ patients are needed.