JAK1 mutations are not frequent events in adult T-ALL: a GRAALL study

Authors


Activating somatic mutations in Tyrosine Kinase (TK) genes are frequent events in haematological malignancies, when they can provide promising therapeutic targets (Loriaux et al, 2008; Scholl et al, 2008). Somatic mutations of JAK1, a Janus Kinase (JAK) family member which plays important roles in normal and neoplastic haematopoiesis, were reported in 2/94 adult de-novo acute myeloid leukaemia (AML) (Xiang et al, 2008). Flex et al (2008) reported JAK1 mutations in acute lymphoblastic leukaemia (ALL), particularly in adult T lineage cases, where the mutation rate was 8/38 (21%). Mutations were found within the SH2 (exon 10), Pseudo-Tyr-Kinase (exons 13 and 15) and the Tyr-Kinase (exon 18) domains of JAK1, with 6/8 mutations occurring in exons 15 and 18. The structural and functional consequences of these mutations was variable in molecular modelling and in-vitro assays (Flex et al, 2008). Mutated cases were suggested to be associated with older patients and a poor outcome (Flex et al, 2008), despite the fact that all JAK1 mutated cases also demonstrated NOTCH1 mutations, which are known to be of favourable prognosis (Breit et al, 2006; Asnafi et al, 2009). Jeong et al identified JAK1 mutations in exons 13 and 16 in 3 of 11 adult T-ALL, but did not study exons 15 and 18 (Jeong et al, 2008). We therefore assessed JAK1 in adults with T-ALL included in the French prospective multicentre Group for Research on Adult Acute Lymphoblastic Leukaemia (GRAALL) study (Huguet et al, 2009).

We screened for all of these mutations by direct sequencing of genomic DNA at diagnosis in 108 cases. Overall, JAK1 mutations were found in only four cases (Table I). Two of these mutations were previously reported and the two new mutations were absent from complete remission samples. The frequency of JAK1 mutation was significantly lower than expected and notably no mutation was observed within the TK domain. Flex et al (2008) reported that patients with mutated JAK1 allele are relatively old, consistent with the lower prevalence of mutation in childhood T-ALL. The overall median age in the present series was 30·5 years, identical to the JAK1 mutated cases (median 30 years, range 26–33 years). All JAK1 mutations were associated with NOTCH1 mutations (3/4 with PEST domain mutations associated or not with HD mutation) but this was not statistically significant because 60% of GRAALL T-ALL cases have NOTCH1 mutations (Asnafi et al, 2009). Two of the JAK1 mutated cases expressed TLX1 and a third expressed TLX3, but once again this did not have statistical value because 10 TLX3+ and 19 TLX1+ cases included in this study were JAK1 germline. Regarding the prognostic value of JAK1 mutations, all 4 cases achieved complete remission and 3/4 were cortico- and chemo-sensitive and were alive in first complete remission (CR1) at 54, 28 and 8 months from diagnosis. One patient was chemo-resistant and underwent bone marrow transplant-allograft in CR1 but died off-therapy from infectious complication. These data are in contradiction to the two previous reports (Flex et al, 2008; Jeong et al, 2008), and suggest that JAK1 mutations are rare in adult T-ALL from the GRAALL study, occurring in only 3·7% of patients, nor do they appear to be associated with a particularly poor outcome.

Table I. JAK1 mutations reported in adult T-ALL.
 Adult T-ALLExon 10Exon 13Exon 15Exon 16Exon 18Incidence (%)
SH2Tyr-kinase-likeTyr-kinase-likeTyr-kinase-likeTyr-kinase
  1. ND, not done.

Flex et al (2008)381 (S512L)1 (A634D)3 (R724H)ND3 (R879S/C/H)21·00
Jeong et al (2008)11ND2 (V658F)ND1 (L783F)ND27·30
Present report1081 (S512L)1 (Y652H)2 (R724H/Q)003·70

These results merit confirmation in order to determine whether these differences are simply due to statistical variations, small patient numbers or whether factors, such as ethnic group, impact on the incidence of JAK1 mutations in adult T-ALL.

Acknowledgements

We thank all clinicians and biologists from the GRAALL group for providing data and material and Chantal Persiaux and Caroline Caudron for technical assistance. This work was supported by the Laurette Fugain Association.

Ancillary