Page 115 - 2020_11-Haematologica-web

Prognostic impact of CLL stereotyped subsets
AB
P=0.009 P=0.027 P=0.058
CD
P=0.069 P=0.011 P=0.042
Figure 2. Outcomes of advanced stage chronic lymphocytic leukemia (CLL) cases according to subset classification. (A) Median time-to-first-treatment (TTFT) for the subset #1: 20.6 months, subset #2: 28.5 months, subset #4: 42.8 months, subset #8: 5.5 months, U-CLL: 17.8 months, M-CLL: 36.1 months. (B) Median time- to-next-treatment (TTNT) for the subset #1: 60.2 months, subset #2: 57.3 months, subset #4: not reached, subset #8: 43.7 months, U-CLL: 47.4 months, M-CLL: not reached. (C) Median progression-free survival (PFS) for subset #1: 41.7 months, subset #2: 33.3 months, subset #4: not reached, subset #8: 26.3 months, U-CLL: 26.5 months, M-CLL: 54.7 months. (D) Median overall survival (OS) for subset #1: 67.5 months, subset #2: not reached, subset #4: not reached, subset #8: 43.3 months, U-CLL: 78.9 months, M-CLL: not reached; Cum: cumulative.
#1, 2 and 8 (HR: 0.53, CI: 0.38-0.74, P<0.001). This same pattern was also seen when M-CLL was compared to sub- set #2 alone; however, this difference was not statistically significant (P=0.052) (Online Supplementary Table S4). U- CLL had a significantly shorter PFS compared to subsets #1, 2 and 8 (P=0.028) as opposed to M-CLL that showed a significantly longer PFS (P=0.002) (Online Supplementary Table S5). In the supplementary analysis, U-CLL also had a shorter PFS compared to subset #2 (P<0.05) (Online Supplementary Table S6). In the OS final model, we observed a significant difference between M-CLL and sub- sets #1, 2 and 8 (P=0.002). In the second analysis, we found a statistically significant difference between TTNT in m-subset #2 and m-IGHV (P=0.005) (Online Supplementary Table S7). With regards to PFS, no differ- ences were found between m-subset #2 and m-IGHV (P=0.122).
Discussion
Our study evaluated the impact of BcR IG stereotype on four relevant clinical outcomes (TTFT, TTNT, PFS, and
OS) with a sufficient follow-up period in four major CLL stereotyped subsets, i.e., subsets #1, 2, 4 and 8. Two dif- ferent clinical scenarios were considered in the present analysis: early stage patients with CLL included in the 'watch and wait' arm of the CLL1 study, and advanced stage, treatment-naïve patients with CLL treated within three clinical trials of the GCLLSG. The analysis was per- formed in patients within prospective clinical trials, where the information regarding the most relevant clinical and genetic characteristics of subsets #1, 2, 4 and 8 could be analyzed in a controlled setting with mature follow-up.
Subset #1 had a similar prognosis to U-CLL in both early stage and advanced stage CLL patients. These findings were consistent with previous publications, where this subset has been associated with a poor prognosis in terms of survival and TTFT even when compared to CLL with the same IGHV genes but a heterogenous BcR IG.14
Subset #2 patients are known to have a shorter TTFT as well as OS regardless of IGHV mutational status.15,33,34 In our study, we observed this same behavior in both early and advanced stage CLL. Furthermore, consistent with previous publications, we found that subset #2 cases had a higher rate of SF3B1 mutations in both the IGHV mutat-
haematologica | 2020; 105(11)
2603