Abstract

Isolated mono-allelic deletion of chromosome region 13q14 is seen in 40–60% of patients with typical B-CLL and is associated with good risk disease. Bi-allelic deletion of 13q14 region has an overall incidence of 4–5% in typical B-CLL. It is not known if loss of both allelic regions impacts on disease progression and survival. FISH analysis was performed on 995 patients with CLL. 451 cases were assessed using 13q14 PAC probes 933-e9 and 273–2 (Dr Stilgenbauer, Ulm University, Germany) and 540 cases were assessed with the B-CLL probe set (α12/13q14/13q34 and p53/ATM probe sets, Abbott/Vysis).

Normal results for 13q14 were obtained in 501 (50.4%) cases, 450 (45.2%) had a mono-allelic deletion of 13q14 and 44 (4.4%) cases had a bi-allelic deletion of 13q14. Control probes (13q34 region) had normal patterns in all cases tested. Detailed analysis was carried out to test the hypothesis that loss of both allelic regions conferred a poorer prognosis by comparing the 44 cases with bi-allelic deletion of to 162 cases with mono-allelic deletion good risk CLL. Cases with bi-allelic deletion have similar clinical characteristics compared to cases with a mono-allelic deletion. There was no significant difference in Binet stage between patients with a mono-allelic (84% A, 7% B, 9% C) or a bi-allelic deletion of 13q14 (73% A, 13% B, 13% C). However, Binet stage was significantly predicative of outcome independent of the cytogenetic abnormality. Immunophenotypes were compared using the dChip program. Unsupervised hierarchical cluster analysis showed that mono-allelic and bi-allelic deletion 13q14 CLL cases have different protein expression profiles. CD5 and CD38 were more strongly expressed (1.23 fold increase, p = .06 and 2.36 fold increase, p = .02, repsectively), whereas CD22 and CD23 had lower levels of expression (1.61 fold decrease, p = <.001 and 1.75 fold decrease, p = <.001 respectively) in cases with bi-allelic deletion compared to mono-allelic deletion. 12% of cases with bi-allelic deletion also have sub-populations of cells with mono-allelic 13q14 deletions. 29% of cases with bi-allelic deletion (5.8% p53; 11.7% ATM; 13.5% +12) had additional cytogenetic abnormalities compared to just 6.5% of cases with mono-allelic deletion (4.5% p53; 1.1% ATM; 0.6% +12)(p = <.001). The overall survival between patients with a bi-allelic deletion compared to a mono-allelic deletion was not significant but median follow-up is relatively short (25 months). Patients with a bi-allelic deletion were significantly more likely to have disease progression and require treatment (median progression free survival (PFS) 12 months) (p = <.001) compared to patients with a mono-allelic deletion (median PFS 26 months) (progression was defined as time to first theraputic intervention; 58% of cases with mono-allelic deletion have not required treatment and 40% of cases with bi-allelic deletion have not required treatment to date). PFS remained significant when cases with deletions of ATM &p53 and +12 were excluded from analysis suggesting that the poor PFS is due to the bi-allelic deletion of 13q14.

In conclusion, we have demonstrated that bi-allelic 13q14 deletion occurs in approximately 5% of B-CLL patients and that it is associated with an inferior progression free survival. The altered immunophenotypic profile associated with bi-allelic 13q14 deletion is intrigueing and further study of this group is warranted.

Author notes

Corresponding author