¹Department of Haematology, Royal Liverpool University Hospital, Liverpool, UK

Correspondence: Dr PD Sherrington, Department of Haematology, Royal Liverpool University Hospital, Liverpool L7 8XP, UK. Fax:+44 0 1517065810; E-mail: pds@liv.ac.uk

Received 12 October 2004; Accepted 3 February 2005; Published online 17 March 2005.

Abstract

Some cellular and molecular features of chronic lymphocytic leukaemia (CLL) cells that are associated with prognosis may reflect the context within which their progenitors encountered antigen. It follows that the nature of antigen drive in CLL could influence the clinical course and we were prompted to assess the impact, if any, of affinity maturation (an antigen-driven process) on prognosis. Statistical models for assessing affinity maturation status are typically applied to V_H gene sequence data analysed using a web-based resource like IMGT or VBASE. Since these resources differ with respect to some key relevant features, we evaluated a cohort of CLL cases by applying statistical models to V_H data derived from both IMGT and VBASE. Important differences between the resulting data sets became apparent. These resulted from database variance and because IMGT and VBASE define complementarity-determining and framework regions (CDRs, FRs) in different ways. Thus, the numbers of mutations identified and their distribution between CDRs/FRs varied between the data sets for the majority of clones. Consequently, two different but overlapping sets of cases with evidence of affinity maturation were defined. Notwithstanding their differences, no significant associations of affinity maturation status with CD38 expression, p53 functional status or survival were identifiable in either data set.