Investigating the Influence of Germline ATM Variants on Susceptibility to Chronic Lymphocytic Leukemia and Other Malignancies

Introduction: Chronic lymphocytic leukemia (CLL) patients (pts) face an increased susceptibility to secondary cancers. Furthermore, the hereditary predisposition to CLL is influenced by factors that have yet to be fully elucidated by GWAS-identified common variants. Among the hereditary cancer risk mutations observed, germline Ataxia-telangiectasia mutated ( ATM) aberrations are among the most prevalent. Rare germline ATM variants are detected in 24% of CLL pts, surpassing the frequency observed in other hematologic neoplasms and the general population. In CLL, loss of ATM via 11q23 deletion is linked to an unfavorable prognosis and a shortened progression-free survival. The extent to which germline ATM variants predispose CLL pts to other malignancies remains unclear; to examine this potential risk we conducted a retrospective study to assess the influence of germline ATM variants on the predisposition to secondary neoplasms in CLL pts and their relatives.

Methods: Five hundred and eighty-five pts seen at Dana-Farber Cancer Institute and who had NGS performed to evaluate germline ATM status, either through direct germline sequencing of saliva or by inference according to the hierarchical algorithm we have previously published (Lampson, 2022), were mailed a questionnaire that investigated: demographics; personal and family history of any cancer; non-medical radiation and Agent Orange exposure; and Ataxia-Telangiectasia syndrome (AT). Of these, 333 replied (57%). Information collected from our CLL database included FISH, IGHV status and treatment status. European ancestry was categorized based on the United Nations geoscheme (UNSD, 2019). Pt characteristics are in Table 1. Pts were stratified into 4 groups, based on ATM mutational status and del(11q): Group-1, ATM germline variants alone (N=63); Group-2, ATM germline with somatic ATM mutation and/or del(11q)(N=22); Group-3, ATM somatic aberration alone (including del(11q))(N=41); and Group-4, no ATM aberration (N=207).

Results: Of the 333 pts in our cohort, baseline demographic and clinical characteristics among the 4 groups were balanced, except for unmutated IGHV status. Nineteen (86%) and 36 (86%) pts of groups 2 and 3, carrying somatic ATM aberrations, had unmutated IGHV, compared to 21 (33%) and 76 (37%) of groups 1 and 4, without somatic ATM aberrations(p<0.01). Regarding treatment, 60% of our cohort had received at least one line of therapy (median: 1 (1-7)). Among the 164 pts (49%) with history of an additional non-CLL neoplasm, 221 cancers were reported. Excluding pts with only non-melanoma skin cancers, the incidence is reduced to 31% (104 pts). The median number of second cancers per pt is 1 (1-4). Of the 134 pts with available age of onset data for the secondary malignancy, 80 (60%) were diagnosed with the secondary cancer after CLL. Comparing pts with germline ATM variants (N=85) to the ones without germline ATM variants (N=248), no significant difference in the incidence of secondary cancer was found (p=0.73). The median age of secondary cancer onset was also similar between the groups (median: 63 and 64 years in germline ATM present vs absent, p=0.28).

Of the whole cohort, 283 pts (85%) reported at least one relative with a cancer history (median 2 relatives affected (1-17)). The incidence of B-cell lymphoproliferative disorders was significantly higher (p=0.02) in pts with germline ATM variants (32%) compared to those without germline ATM variants (21%), including familial CLL (25% vs18%) (p=0.04). Familial CLL was also strongly associated with Ashkenazi Jewish ethnicity (39% incidence in Ashkenazi pts vs 16% in non-Ashkenazi, p=0.004). No difference in the incidence of other hematologic or solid malignancies was found between the relatives in the two groups.

No pts reported a familial or personal history of AT. Compared to pts without any ATM aberration, time to first treatment (TTFT) was shorter in pts harboring somatic ATM events, while no difference was observed between pts with germline ATM variants (median TTFT: 82, 59, 52, 90 months for group 1, 2, 3 and 4 respectively, p=0.01) (Figure 1).

Conclusions: Our results suggest a higher incidence of B-cell lymphoproliferative disorders, including familial CLL, in the relatives of CLL pts carrying germline ATM variants. The presence of these germline variants did not impact TTFT, compared to pts harboring somatic ATM mutations.