Comprehensive Hybrid Capture-Based Genomic Profiling of T-Cell Leukemias and Lymphomas Reveals Targetable JAK1 and JAK3 Co-Existing Mutations

Background: T-cell leukemias and lymphomas (TL&Ls) are rare when compared with other hematologic malignancies, and pose a challenge for clinical management. TL&Ls are characteristically resistant to chemotherapeutic regimens used to treat other hematologic malignancies (PMID 17369126, 22649104, 25042790). Recently, activating JAK1 and JAK3 mutations have been described in a subset of mature TL&Ls including adult T-cell leukemias/lymphoma (ATLL), NK/T-cell lymphoma, and T-cell prolymphocytic leukemia (T-PLL). We used a comprehensive hybrid capture-based sequencing assay to identify genomic alterations that might suggest benefit from targeted therapy.

Methods and Results.

Using FoundationOne® and FoundationOne® Heme assays, we performed genomic profiling on 28 consecutive TL&Ls (leukemias, n=10; lymphomas, n=18) for all alterations in 370 or 405 cancer-related genes and select introns of 19 or 31 genes commonly rearranged in cancer, or both DNA and RNA were extracted and captured using custom baits for 405 genes and 265 frequently rearranged genes, respectively. The samples were sequenced to high uniform coverage, averaging >500X for DNA and >8M total pairs for RNA (PMID 24142049). JAK3 alterations were the most common at 25% (7/28) with an enrichment in T-cell derived leukemias (2 T-PLL, 3 T-ALL, and 1 cutaneous T-cell lymphoma (CTCL)). Other altered genes in the 28 cases included CDKN2A/B (21%), TP53 (18%), and TET2 , DNMT3A , and STAT3 (11% each). A complex rearrangement involving FCHO1 and JAK3 resulting from a ~70kb genomic segment deletion was also identified in a CTCL. The most frequently co-altered genes identified in the JAK3 mutated cases were relevant to the chromatin remodeling pathway. Focal STAT3, STAT5A, and STAT5B amplifications were identified in 2 cases (CTCL and ALCL) as well as one STAT3 D661Y in a T-cell large granular lymphocytic leukemia (T-LGL). All 3 cases with STAT gene alterations were mutually exclusive with JAK3 alterations. Additionally, 2 patients with T-PLL and T-ALL respectively were both found to harbor multiple co-existing JAK1 and JAK3mutations (2/28; 7%).

One heavily pre-treated T-PLL patient harbored a major JAK1 V658F clone (40%) and a minor JAK3 M5111 clone (5%), and loss of heterozygosity for mutant ATM and TP53. Based on assay results, the patient was treated with ruxolitinib, a potent and specific inhibitor of JAK1 and JAK2. The patient had an impressive clinical response of duration of 4 months after which her T-PLL cell count rose markedly to pre-treatment levels and she developed worsening thrombocytopenia, indicating acquired resistance to therapy. The resistant T-PLL was assayed and was found to harbor an increased frequency of JAK3 M5111 (28%), decrease of JAK1 V658F (18%) and a new EZH2mutation. The patient’s disease progressed through another line of cytotoxic chemotherapy and she succumbed to her disease. Signal transduction analysis was conducted which showed in T-PLL, the JAK1 and JAK3 mutations were functionally significant as they induced constitutive phosphorylation of downstream STAT proteins. Importantly, these phosphorylations could be inhibited by JAK inhibitors. Collection of clinical outcome on additional cases is ongoing.

Conclusion. In this study, we discovered a high frequency of JAK-STAT alterations in 28 T-cell neoplasms with 25% of cases harboring JAK3 missense mutations and two cases with concordant JAK1 and JAK3 mutations. In one case, treatment with the JAK1/2 inhibitor, ruxolitinib, resulted in an impressive clinical response. Our data argue for oncogene dependence upon the JAK-STAT pathway in diverse T-cell neoplasms and underscore the importance of genomic profiling by FoundationOne Heme® of rare and difficult to treat cases by revealing alterations that may be targeted by specific inhibitors, such as ruxolitinib, and may also be predictive of treatment resistance.