Important advances in the understanding of CLL pathogenesis include the discovery that NOTCH1 mutations are present in ∼28% of pts harboring +12. There is a need for improved understanding of the clinical outcomes of CLL patients (pts) with +12 on a population-level, as this subgroup is rapidly becoming the focus of biologic studies evaluating pathogenesis of disease and clinical trials investigating novel targeted therapies. In the province of BC, population 4.5 million, CLL pts receive uniform evaluation and therapy based on centrally derived protocols with FISH testing implemented since 2004. We sought to characterize the clinical outcomes of +12 in this large unselected population-based cohort of CLL pts.
Clinical and laboratory data on all pts referred for CLL FISH testing at 1 of 3 BC cytogenetic labs from 2004–2011 were entered into the BC Provincial CLL Database and included in this analysis. Pts without a confirmed diagnostic date were excluded. Baseline features of pts with and without +12 were compared using Fisher's exact test for categorical and Wilcoxon rank sum test for continuous (cnts) variables. Primary and secondary endpoints were OS and TFS (defined as time from diagnosis [dx] to first therapy). Percent of abnormal (%abn) nuclei harboring +12 was evaluated for association with OS/TFS. Cox proportional hazard (PH) models were constructed to determine predictors of OS/TFS for the +12 cohort, including age at dx, sex, Rai stage (0, 1–2, 3–4), WBC at dx, CD38 positivity and concomitant 17p-, 11q- or deletion 13q (13q-). Cox PH models were also constructed to determine effect of +12 on TFS/OS for the entire cohort.
As of Dec. 2011, 882 pts had CLL FISH testing in BC of which 164 (19%) had +12 on their 1st FISH test: 8 (5%) with concomitant 17p-; 14 (9%) with 11q-; 142 (86%) without either 17p- or 11q-, of which 43 (30%) had 13q-; 16/124 tested (13%) had an IGH translocation [t(IGH)]. Of the 164 +12 pts, median age at dx was 60 yrs (range 35–93), 70% were male, 10% had Rai stage 3–4. At median follow-up of 4.5 yrs (range 0–19), 95 pts (59%) received treatment, 31 (19%) died. For the +12 cohort, median OS was 14.7 yrs (95% CI 9.8–19.0) and median TFS 3.7 yrs (95% CI 2.7–5.4).
Of the 658 non +12 CLL pts (N12CPs), prevalence of recurrent cytogenetic abnormalities (RCA) were: 17p-, 10%; 11q-, 11%; 13q-, 60%; t(IGH) 7%. Significant differences between +12 and N12CPs included more CD38+ pts (66% vs 28%, P<0.001), higher t(IGH) incidence (13% vs 7%, P=0.04) and fewer 17p- (5% vs 10%, P=0.03) or 13q- (26% vs 60%, P<0.001) abn among +12 pts. When pts were grouped by hierarchical FISH abn, +12 pts retained an intermediate OS (median 15.9 yrs) and TFS (median 4.2 yrs) when compared to other RCAs (Fig 1A). Multivariate analysis (MVA) for the whole cohort (n=822) demonstrated no significant effect of +12 on OS (HR 0.72, 95% CI 0.36–1.43, P=.35) or TFS (HR 0.86, 95% CI 0.69–1.36, P=.86) after adjustment for covariates.
For the +12 cohort (n=162), univariate analysis demonstrated shorter OS associated with age (P=.001), Rai stage (P=.01) and 17p- (P=.07). A longer OS was associated with presence of 13q- (median OS 11.6 vs 18.7 yrs, P=.04), Fig 1B. Shorter TFS was associated with Rai stage (P<.001), WBC at dx (P=.01) and 17p- (P=.04). %abn nuclei harboring +12 was not predictive of OS (P=.33) or TFS (P=.25) as a cnts variable; however those with <20% vs ≥20% abn had a significant improvement in OS (P=.02). MVA for the +12 cohort demonstrated Rai stage (HR 3.26, 95% CI 1.23– 8.63, P=.02) and 11q- (HR 9.07, 95% CI 1.44–57.02, P=.02) as independent risk factors for OS, while 13q- did not retain its protective effect (P=.98). For TFS, MVA found Rai stage (HR 2.92, 95% CI 1.78–4.78, P<.001) and 17p- (HR 5.44, 95% CI 1.52–19.43, P=.01) as negative predictors while 13q- (HR 2.01, 95% CI 1.08–3.75, P=.03) again had a positive effect.
We report the largest, population-based cohort of CLL pts with FISH testing and confirm that +12 occurs in 19% of CLL pts and in the absence of 17p- or 11q-, confers an intermediate prognosis. The presence of 13q- had a protective effect on TFS and a trend towards improved OS, thus improving the prognosis of a subset of +12 pts. This finding is consistent with recent observations that NOTCH1 mutations and 13q- are mutually exclusive in +12 pts and may explain the clinical heterogeneity seen in this subgroup. Further research into these distinct subsets of +12 pts is warranted.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.