Atypical Chronic Lymphocytic Leukemia with CD200 Expression Shares Similar Outcomes to Classical Chronic Lymphocytic Leukemia

Introduction

Chronic lymphocytic leukemia (CLL) diagnosis relies on a characteristic immunophenotype of B-cell lymphocytes. The modified Matutes scoring system is based on five key membrane markers: CD5, CD23, CD79b, FMC7 and surface membrane immunoglobulin (smIg) (Matutes et al., Leukemia 1994; modified by Moreau et al., Am J Clin Pathol 1997). CLL typically demonstrates dim staining for smIg, low or absent expression of CD22, CD79b and FMC7 and strong expression of CD5 and CD23. However, up to 7% of B-cell clonal lymphoproliferative disorder (CLPD) cannot be classified as CLL, nor as any other well described CLPD, namely: mantle cell lymphoma, marginal zone lymphoma, follicular lymphoma, hairy cell leukemia, diffuse large B-cell lymphoma, B-cell prolymphocytic leukemia or Waldenstrom's macroglobulinemia and are classified as atypical CLL (A-CLL) (Boucher et al., ASH 2013). Furthermore, leukemic phase of mantle cell lymphoma may mimic CLL (Nelson et al., Mod Pathol 2002), leading to suboptimal management. The expression of CD200 emerged as a powerful marker to distinguish MCL from CLL and therefore better recognize A-CLL (Spacek et al., ASH 2014). Clinical outcomes of patients with A-CLL with CD200 expression have not been compared to patients with classical CLL.

Methods

Patients diagnosed with CLL or A-CLL expressing CD200 on flow cytometry analysis performed at Maisonneuve-Rosemont Hospital (Montreal, Canada) between July 2014 and July 2015 were reviewed. Multiparameter flow cytometric immunophenotyping (8 colors) was performed on 1 x 10⁶cells using the Euroflow panel of CLPD (van Dongen et al., Leukemia 2012). Data were acquired using FACS CantoII flow cytometer and FACSDiva software program (Becton Dickinson Bioscience, San Jose, CA) and were analysed with Infinicyt software (Cytognos SL, Salamanca, Spain). Diagnostic criteria for A-CLL were negative or ≤ intermediate expression of CD23, ≤ intermediate expression of CD5 or strong expression of CD20. We retrospectively collected data on clinical presentation, flow cytometry and cytogenetics analysis, therapies and outcomes. Treatment response was evaluated according to Hallek (Blood 2008). Comparison between outcomes were evaluated through time to first line therapy (TT1T), time to second line therapy (TT2T) and overall survival (OS) with Kaplan-Meier curves and log-rank tests and response to first line therapy (R1T) was compared with Fisher's test. Treatment was according to physician choice. Our study was approved by local research and ethic committees.

Results

We retrieved 25 CLL and 10 A-CLL patients with a median age of 67 (46-93) and 67 (51-81) respectively at diagnosis. All these 35 CLL expressed CD200 as per inclusion criteria. Among the 10 A-CLL: 4 did not express CD23 (2 were negative for t(11;14) by FISH analysis and 2 had negative immunohistochemistry staining for cyclin D1 on bone marrow biopsy), 4 had intermediate expression of CD23, 1 had intermediate expression of CD5 and CD23 and 1 had intermediate expression of CD5. One of the A-CLL not expressing CD23 had strong expression of CD20. At diagnosis, 3/25 (12%) CLL presented with Rai 3-4 stage compared to 1/9 (11%) of A-CLL. Deletion 17p was detected in none of the 7 CLL and 8 A-CLL patients tested and deletion 11q was detected in 2/9 (22%) CLL and 1/8 (13%) A-CLL patients tested. At time of analysis, 7 patients with CLL and 4 patients with A-CLL received treatment. Chlorambucil was administered to 4 CLL and 2 A-CLL patients, fludarabine, cyclophosphamide and rituximab to 2 CLL and 2 A-CLL patients and rituximab only to 1 CLL patient for refractory immune thrombocytopenia. After a median follow up of 18 months, no significant difference in outcomes was observed. Median TT1T was 26 months (m) for CLL and not reached (NR) for A-CLL (p=0.53). Complete remission was achieved for 2 CLL and 2 A-CLL and partial remission for 5 CLL and 2 A-CLL patients respectively (p=0.58). During follow up, 3 CLL and 3 A-CLL patients required second-line therapy. Median TT2T was 111 m for CLL and NR for A-CLL. No Richter transformation was observed and no difference in OS could be detected (only one death in the entire cohort).

Conclusion

Outcomes of atypical CLL expressing CD200 appear similar to CLL using CLL management strategies in this cohort. A longer follow-up of a larger cohort may allow us to confirm the value of this marker in guiding A-CLL management.