Prognostic Significance Of JAK3 Activating Mutations In Natural-Killer/T-Cell Lymphomas

Background We have shown that the JAK/STAT pathway is dysregulated in Natural-killer/T-cell lymphomas (NKTL) and that the JAK3A572 and JAK3A573 activating mutations were found in 35% of NKTL. JAK3 mutant (MT) NKTL cell lines demonstrated IL-2 independent growth and constitutive activation of downstream JAK/STAT signaling molecules. When these cell lines were treated with JAK3 siRNA, downstream JAK/STAT signaling is inhibited. Reciprocally, transient expression of in JAK3A572 in wild-type (WT) NKTL resulted in IL-2 independent growth and constitutive activation of JAK/STAT signaling. The main aim of this study was to determine the prognostic significance of JAK3 mutation in NKTL. Methods Patients with NKTL treated at the National Cancer Centre Singapore and Singapore General Hospital were identified through clinical and pathology databases. All cases of NKTL were centrally reviewed by hematopathologists to confirm the diagnosis of NKTL based on the 2008 WHO classification. Whenever possible, FFPE tumor samples were examined for JAK3A572 and JAK3A573 mutations using Sanger sequencing and results confirmed with high-resolution melt. Patient demographics, treatment details and survival outcomes were obtained from medical records. Data for hemophagocytic lymphohistiocytosis (HLH) and site of disease involvement were specifically captured. Treatment was stratified according to curative or palliative intents and whether radiotherapy (RT) or stem cell transplant (SCT) was administered. The primary endpoint was to compare survival; OS and PFS, between JAK3 MT and WT patients. Secondary endpoints were to compare clinical characteristics between JAK3 MT and WT patients and identify poor prognostic features amongst NKTL patients who were treated with curative intent. Results Between 1997 and 2011, 98 patients were diagnosed with NKTL. Of these, JAK3 mutation testing was successful in 58 tumor samples: 37.9% (22/58) were JAK3 MT and 62.1% were JAK3 WT. There were 10% with HLH. The 2-year PFS for JAK3 MT and WT patients were 17% and 30% respectively (p=0.91) and the 2-year OS for JAK3 MT and WT patients were 31% and 40% respectively (p=0.45). There was no difference in the baseline clinical characteristics between JAK3 MT and WT patients. (Table 1) In our cohort of NKTL patients, the median OS for patients for stage I/II was 74.2 months (95% CI, 21.7-NR) and the median OS for stage III/IV was 5.2 months (95% CI, 2.93 - 7.86). Amongst the 73 patients treated with curative intent, on univariate analysis, age >60 years, stage III/IV disease, ≥2 extra-nodal sites, non-nasal extranodal involvement, B symptoms, higher IPI scores, HLH and the absence of RT were associated with a poorer OS. On multivariate analysis, age >60years, stage III/IV disease and presence of HLH were independently associated with a poorer survival. Discussion This study showed that JAK3 mutation was not prognostic for poorer OS or PFS and JAK3 MT and WT patients are clinically indistinguishable. Since in vitro, treatment of both JAK3 MT and WT cell lines with a pan-JAK inhibitor resulted in a reduction in cell viability and current treatment strategies against NKTL are poor, future clinical trials should focus on targeting the JAK/STAT pathway in NKTL patients. We also report that HLH was found in 10% of NKTL patients and it was independently prognostic of a poorer OS. The association of HLH and NKT, in addition to them being associated with EBV may point towards a shared etiology. Disclosures: No relevant conflicts of interest to declare.