The changing definition of CLL

In this issue of Blood, Shanafelt and colleagues provide a clearly written, analytical critique of the ALC in the diagnosis of CLL in daily practice and the role of prognostic factors.

The recent International Workshop on Chronic Lymphocytic Leukemia update of the National Cancer Institute 1996 guidelines for chronic lymphocytic leukemia (CLL) by Hallek et al changed the definition of CLL by requiring an absolute B-cell count (BALC) of 5000 cells/μL rather than the previous absolute lymphocyte count (ALC) of 5000 cells/μL.¹  This created a considerable amount of controversy for both clinicians and patients. BALCs of 5000 or more cells/μL were designated B-cell monoclonal lymphocytosis (MBL), causing many former Rai stage 0 and I patients to be reclassified as MBL or SU (small lymphocytic lymphoma). MBL is well recognized in both the United States and Western Europe in population studies, blood bank donors, in aging individuals, and in unaffected first-degree relatives with familial CLL.²  It is now thought to be the precursor in CLL.^3,4  If the ALC of 5000 was arbitrarily selected, there is the notion that the selection of a BALC of 5000 was likewise arbitrarily selected for “consistency” and was not based on objective clinical outcome data. This situation was further aggravated by the requisite change in diagnosis of CLL to MBL in 40% of patients previously diagnosed with Rai stage 0 CLL⁵  and there is no standardized method to determine the BALC.⁶  Given the seriousness of the diagnosis of leukemia and an understanding of the evolution of the diagnostic criteria for CLL, Shanafelt et al undertook an evaluation of the ALC, B-cell count and conventional prognostic markers.⁷ 

The Mayo Clinic CLL Database was used to identified 459 consecutive patients diagnosed with Rai stage 0 CLL over a 7-year period. The database is ongoing, and it allows for a sophisticated statistical analysis of presenting clinical data including ALC and flow cytometry and the prognostic factors of CD38, fluorescence in situ hybridization (FISH), zeta-chain associated protein 70 (ZAP70), and immunoglobulin gene variable reagent heavy chain (IGVH) mutational status. The objective clinical outcome measurements were treatment-free survival (TFS) and overall survival (OS). Both the BALC and the ALC predicted TFS and OS. The B-cell threshold that best predicted both OS and TFS was 11 000 cells/μL. The authors noted that the current recommended B-cell count value of 5000 did predict TFS but did not predict OS. Interestingly, the ALC of 5000 did not predict either TFS or OS! What is of even greater interest is that the ALC becomes predictive of OS and TFS at 12 000, which suggests that at this numerical value, the percentage of B cells in such a patient sample will probably be 75% or greater. At that value, the ALC and BALC become interchangeable. The predictability of the prognostic factors and absolute B-cell counts were then determined. B-cell count, IGHV mutational status, and FISH appeared to be the best predictors of OS while ZAP70 and IGHV mutation status were the stronger predictors of TFS. Interestingly, the predictive value of the B-cell count was similar to or slightly better than FISH and CD38. According to Shanafelt et al, when analyzed together, the combination of B-cell count and ZAP-70 was the best predictor of TFS, while the combination of B-cell count and FISH was the best predictor of OS. The size of the B-cell clone is stressed as it relates to disease outcome. It would now appear that since an absolute B-cell clone size of 11 000 cells/μL is an independent prognostic indicator and or a surrogate biomarker, even without a revision in the definition of the diagnosis of CLL (clone size 5000 vs 11 000), the ALC can be conveniently used to determine the second flow cytometric analysis for a more refined B-cell lymphocyte prognostic immunophenotype.

Shanafelt et al have presented the most sophisticated analysis of this single clinical feature to date. The size of the B-cell clone is at the heart of the matter. I would be surprised if these findings are not confirmed on a second cohort. In fact, whereever other CLL databases exist, it should be easy to confirm these findings in subjects who have had multiple B-cell count determinations. From a historical viewpoint, in 1973 Hansen defined CLL as an ALC of 10 000 cells/μL with 60% mature lymphocytes.⁸  Regardless of the definition used, the role of prognostic indicators has been enhanced and remain important in the management of CLL. Pre-MBL can now be explored in terms of intraclonal heterogeneity and chronic antigen stimulation for MBL/CLL/SLL.⁹