Comprehensive Genomic Profiling (CGP) Informs Molecular Diagnosis, Prognosis, and Therapeutic Decision-Making for Patients with Angioimmunoblastic T-Cell Lymphoma (AITL)

Background: AITL, under the 2016 WHO umbrella classification of nodal peripheral T cell lymphomas (PTCL) with a T follicular helper (TFH) phenotype, accounts for ~20% of all PTCL cases. Genomic analysis has revealed recurrent somatic mutations, such as RhoA ^G17V, that are hallmarks of AITL. Overlapping genomics have been observed in other PTCL subgroups, such as PTCL not otherwise specified (PTCL-NOS). We and others have reported frequent co-occurrence of RhoA ^G17V with mutations in the epigenetic regulators TET2 and IDH2^R172 in AITL (Morley Blood 2015). Since treatment for the disease remains unsatisfactory, efforts are underway to assess the clinical utility of genomic information to guide therapy. The current study aimed to comprehensively assess the genomic landscape of AITL further, including identification of diagnostically and therapeutically relevant alterations.

Methods: We explored the FoundationCORE™ database (Foundation Medicine Inc, Cambridge, MA) for cases assigned as AITL or harboring RhoA ^G17V positivity irrespective of disease ontology. This led to the inclusion of 42 cases: confirmed AITL (concordant diagnoses by internal and external pathologists, 29), suspected AITL (discordant diagnoses, 9), lymph node lymphoma NOS (PTCL, 1), or other neoplasm [pleomorphic myxofibrosarcoma (MFS), 1; marginal zone lymphoma (MZL), 1; cutaneous T-cell lymphoma (CTCL), 1]. Median cohort age was 65.5 years, 65 years for the AITL-specific subset; range for both, 34-88y. Pathology reports indicated Epstein-Barr virus positivity in 16/38 of AITL cases and prior autologous stem cell transplant in 4/38. DNA and RNA were extracted from formalin-fixed paraffin embedded samples or peripheral blood, and CGP performed using the CLIA-certified, NYS-approved, CAP-accredited FoundationOne^® Heme assay (He Blood 2016).

Results: RhoA^G17V was seen in 57% (24/42) of cases in the cohort, including 53% (20/38) of AITL samples and in 1 case each of MFS, MZL, CTCL, and PTCL. Confirmed AITL harbored RhoA^G17V in 52% (15/29), and suspected AITL in 56% (5/9), of samples. TET2 mutations (m TET2) were found in 89% of AITLs (26/29, confirmed; 8/9, suspected), and in 1/1 PTCL; as reported previously, m TET2 was not identified in the MFS, MZL, or CTCL. IDH2 mutations (m IDH2), all occurring at R172, were seen in 16% (6/38) of AITL cases, with all 6 co-occurring with RhoA^G17V plus m TET2 . DNMT3A mutations (m DNMT3A) were seen in 32% (12/38) of AITLs (9/29, confirmed; 3/9, suspected) and in the single PTCL. RhoA^G17V co-occurred with m TET2 (but notm DNMT3A orm IDH2) in 26% (10/38), m TET2 plus m DNMT3A in 7.9% (3/38), and m TET2, mDNMT3A, plus m IDH2 in 7.9% (3/38) of cases; all RhoA^G17V cases(20/20) co-occurred with m TET2 . TP53 mutation was seen in 21% (8/38) of AITL cases. Other recurrent somatic mutations included BRCA1 or BRCA2 in 7.9% (3/38), CTNNB1 in 7.9% (3/38), and KRAS, ATM or STAT3 in 2.6% (1/38), of AITL samples. Assessment of tumor mutational burden (TMB), a biomarker for checkpoint inhibitor efficacy in solid tumors, revealed low TMB (<6 mutations/Mb) in 8/10 AITL cases and intermediate TMB (TMB-Intermediate, 6-19 mutations/Mb) in 2/10.

Conclusion: This study utilized CGP to explore alterations of diagnostic and therapeutic potential in AITL. Confirming prior studies, in a significant proportion of AITL cases RhoA^G17V co-occurred with m TET2 and m IDH2 ; all AITL cases positive for RhoA^G17V had concurrent m TET2. We also found RhoA^G17V, but not m TET2 or m IDH2, in solitary cases of MZL, MCL, and CTCL.Taken together, co-occurrence of RhoA^G17V and m TET2 may provide greater specificity for AITL diagnosis than RhoA^G17V alone. We also detected putative actionable targets in AITL. While predictors of response to hypomethylating agents (e.g. m TET2, m IDH2) were seen in most cases, a subset were found to have mutations in BRCA1/2 (3 pts), KRAS (1 pt), CTNNB1 (3 pts), or STAT3 (1 pt). These findings suggest alternative therapeutic opportunities in AITL (e.g. inhibitors of PARP, MEK, mTOR, and STAT3). We further report that mutational burden in AITL is generally low; however, a subset of cases (2/10) were TMB-Intermediate. Although the predictive power of TMB in lymphoma is unclear, in some solid tumors TMB-Intermediate status predicts response to immune checkpoint inhibition. In total, our findings highlight the utility of CGP to identify alterations that may influence the clinical management of AITL.