The Prognostic Importance of Various 13q14 Deletions In CLL

Abstract 3587

Recurrent chromosomal deletions centered on the chromosomal marker D13S319 (known collectively as 13q14 deletions) are present in ~50% of all CLL patients. The clinically used CLL FISH panel measures the presence of 13q14 deletions, and singular 13q14 deletions often indicate indolent CLL. Recently, substantial anatomic heterogeneity in 13q14 deletions in CLL has been recognized. This includes short deletions of ~0.8-1.0 Mb length that involve either one or both chromosomes and that do not involve RB (type I) as well as much longer deletions that extend over various distances on 13q resulting in the removal of one copy of RB (type II) and a variable number of other genes. Importantly, the clinically used FISH test cannot discriminate short from long 13q14 deletions and is unreliable in discriminating heterozygous from homozygous short deletions.

To further our understanding of the anatomy and clinical importance of 13q14 deletions in CLL, we have reviewed clinical FISH data and SNP 6.0 array genomic profiling data of DNA from sorted CD19+/CD3+ cells from 256 CLL patients, of which 201 were untreated and 55 relapsed at study enrollment. The boundaries of 13q14 deletions were mapped and lesions grouped based on the exclusion (type I, short, N=72) or inclusion of RB (type II, long, N=50). Next, we estimated TTFT and OS for patient groups with either lesion type. While the TTFT was not significantly different between the 13q14 type I and II cohorts, the hazard ratio was 3.32 (1.16-9.49) and 5.93 (2.1-16.79) for short OS (type II versus type I) for the data analyzed based on diagnosis date and enrollment date, respectively. Thus, post-therapy, the presence of type II 13q14 deletions was associated with a more aggressive disease course.

Subsequently, three CLL groups were analyzed for outcome; single type I 13q14 deletions, single type II 13q14 deletions or complex 13q14 deletions of types I or II combined (defined as the presence of any additional FISH lesion in addition to a 13q14 deletion). In this cohort 10/72=14% (type I) and 17/50=34% (type II) of 13q14 deletions were FISH complex. Complex FISH inclusive of 13q14 deletions was a very strong marker for aggressive CLL, with short OS (34 months; CI=15-53 months) from the date of study enrollment when compared with single 13q14 deletions of either type I or II (p<0.01). Interestingly, after exclusion of all cases with 13q14 deletions in the setting of complex FISH, singular type II 13q14 deletions were also adverse for OS (date of enrollment to date of death) when compared with singular type I deletions (HR=6.1, p=0.03).

Next we sequenced all coding exons of RB in all 50 CLL cases with type II 13q14 deletions and identified 2 cases with mutations resulting in truncated Rb proteins, thus identifying Rb null states in a small fraction of CLL. Work is in progress attempting to clarify the possible prognostic impact of varying 13q14-resident miR15 and miR16 expression levels (as measured though normalized Q-PCR) on CLL outcome as well the significance of homozygous versus heterozygous type I deletions (as measured through array-based copy number estimates).

In conclusion, our data suggest that CLL with 13q14 deletions in the setting of additional FISH abnormalities is aggressive. Furthermore, the identification of singular non-complex type I versus type II 13q14 deletions may be clinically important, as the latter may delineate patients with shortened survival post-therapy. Such identification could be readily accomplished if the currently used FISH tests included a probe centered on RB and ultimately could be accomplished through use of array-based genomic analyses in CLL.