Background:The somatic hypermutation (SHM) of immunoglobulin heavy chain variable (IGHV) region genes classifies chronic lymphocytic leukemia (CLL) patients into two subgroups: unmutated CLL (UM‐CLL, ≥98%) associated with an adverse prognosis and mutated CLL (M‐CLL, <98%) showing a good outcome. In cases with a % of IGHV identity close to the 98% cut‐off, defined as “borderline” (BL‐CLL), the prognosis remains controversial.Aims:To investigate the clinico‐biologic features and prognosis of BL‐IGHV CLL in view of their undefined outcome.Methods:We retrospectively analyzed 811 untreated CLL patients, including 327 at diagnosis, sequenced at our Institution. BL‐CLL were defined as cases with a 97–97.9% IGHV identity, according to the 2017 European Research Initiative on CLL (ERIC) recommendations. IGHV were amplified using family‐specific VH primers (framework region 1 [FR1] or leader primers) and sequenced using the 3500 Series Genetic Analyzer. Following the ERIC recommendations, we updated the IGHV analysis with two different approaches in order to accurately determine the SHM rate: 1) re‐sequencing of 82 old FR1 cases with leader primers; 2) realignment of all cases to the updated IMGT/V‐QUEST tool. From the re‐analysis with leader primers of 82 old FR1 sequenced cases, none shifted the UM/M IGHV category, while 6 (7.3%) were reclassified as follows: 4 M‐CLL as BL‐CLL and 2 BL‐CLL as UM‐CLL. No case shifted across the three categories after the realignment to the updated IMGT/V‐QUEST tool. Major stereotyped subset analysis was determined according to the ARResT/AssignSubsets tool (bat.infspire.org/arrest/ericll.org/pages/services/tool). Time‐to‐first treatment (TFT) was evaluated as a clinical endpoint. Two further independent cohorts of 308 and 465 newly diagnosed CLL patients were used to validate the data.Results:In the 811 untreated patients, 418 (51.5%) UM‐CLL, 39 (5%) BL‐CLL and 354 (43.5%) M‐CLL were identified, with a higher frequency of UM‐CLL due to the enrichment of cases enrolled in first line clinical trials. Sixteen cases (2%) belonged to stereotyped subset #2. BL‐CLL showed a similar proportion of TP53 deletion and mutation compared to UM‐CLL, but a significant lower frequency of other unfavorable prognostic factors (stage, CD38/ZAP70 expression) (Table 1). Moreover, BL‐CLL were enriched in subset #2 (5/39, 13%), in agreement with the higher incidence of subset #2 among these cases. Applying the conventional 98% IGHV cut‐off, TFT was significantly shorter in UM‐CLL than M‐CLL (p < 0.0001), as well as in subset #2 CLL vs M‐CLL (p = 0.017), as expected. When analyzing the 97–97.9% IGHV subgroup, BL‐CLL showed a significantly longer TFT than UM‐CLL (p < 0.0001), similar to that of M‐CLL (Fig. 1a). The TFT of BL‐CLL cases remained comparable to that of M‐CLL, also when the analysis was restricted to the 327 CLL patients at diagnosis (80% Binet stage A) (125 UM‐CLL, 19 BL‐CLL and 183 M‐CLL) (Fig. 1b), as well as excluding subset #2 cases. These findings were validated in two independent cohorts of CLL at diagnosis (105 UM‐CLL, 16 BL‐CLL, 187 M‐CLL; TFT BL vs UM‐CLL: p < 0.0001 and 143 UM‐CLL, 13 BL‐CLL, 309 M‐CLL; p = 0.024), that included 719 Binet stage A CLL.Summary/Conclusion:Our data suggest that the prognosis of BL‐CLL is better than that of UM‐CLL and similar to that of M‐CLL, thus supporting the 98% cut‐off for clinical purposes. The issue of BL‐CLL remains open to further investigations.image
Read full abstract