SET-NUP214 is a recurrent γδ lineage-specific fusion transcript associated with corticosteroid/chemotherapy resistance in adult T-ALL

T-cell acute lymphoblastic leukemia (T-ALL) accounts for 25% of adult and 10% of pediatric ALL.^1,2  Over the last decade, great progress has been made toward the identification of molecular genetic abnormalities in T-ALL.³  Proto-oncogene activation by promoter/enhancer substitution (TLX1/3, TAL1/2, LMO1/2, MYB, and CCND2) or gene mutations (NOTCH1, FBXW7, WT1, and RAS) are frequent events in T-ALL,³  whereas chromosomal translocations leading to fusion genes are relatively rare.⁴ PICALM-MLLT10 (also known as CALM-AF10) and MLL rearrangements are the most frequent fusion proteins in T-ALL, in which they are specific to the T-cell receptor γδ (TCRγδ) lineage and lead to overexpression of HOXA genes.^5-7  The SET-NUP214 (TAF1/CAN) fusion gene resulting from either cryptic t(9;9)(q34;q34) or del(9)(q34.11q34.13) was first described in a patient with acute undifferentiated leukemia,⁸  then in 1 patient with acute myeloid leukemia (AML),⁹  and also in a very limited number of pediatric^10-12  and adult T-ALLs.^10,13-16 SET-NUP214, similar to MLL and CALM-AF10 rearrangements, contributes to T-ALL pathogenesis, at least in part by transcriptional activation of HOXA genes.^12,14  However, since few SET-NUP214–positive T-ALL patients have been reported to date, their clinicobiologic features have not been fully determined. These patients were suggested to have a poor prognosis, but this is based on sporadic reports and has not been evaluated within the context of prospective clinical trials. We therefore undertook to evaluate the frequency, clinical, and biologic characteristics of patients with SET-NUP214–positive T-ALL and the prognostic significance of SET-NUP214–positive T-ALL in a consecutive series of 196 adult patients with T-ALL enrolled in the Group for Research on Adult Acute Lymphoblastic Leukemia (GRAALL) 2003 and 2005 trials.

Between November 2003 and May 2010, 264 adult T-ALL patients were randomized in the consecutive GRAALL 2003 and 2005 (NCT00327678) trials.^1,17  This study concerns 196 (74%) of these patients, for whom diagnostic complementary DNA was available for SET-NUP214 screening by reverse-transcriptase polymerase chain reaction (RT-PCR), as described.¹²  These 196 patients were representative of the overall GRAALL T-ALL population, with a 3-year overall survival (OS) of 66% vs 68% (P = .84; supplemental Table 1). This study was approved by local and multicenter research ethical committees and by the institutional review board of the French Regulatory Agency. This study was conducted in accordance with the Declaration of Helsinki.

Within GRAALL studies, corticosteroid-resistant patients (Cr) are defined by an absolute number of circulating blasts >10⁹/L after 7 days of prednisone. Chemotherapy-resistant patients (CHr) are those with circulating blasts or >5% of bone marrow blasts after 1 week of chemotherapy. Patients who demonstrated either corticosteroid or chemotherapy resistance (Cr/CHr) were allografted in first complete remission (CR). Complete methods and GRAALL protocols are described in the supplemental Data.

SET-NUP214 was detected by RT-PCR analysis in 11 patients (6%), 5 of whom had available single nucleotide polymorphism 6.0 array data (supplemental Figure 1) showing del(9)(q34.11q34.13). This incidence (6%) is relatively similar to that previously reported.^13,14  Baseline characteristics of SET-NUP214–positive and SET-NUP214–negative T-ALL patients were not significantly different (Table1). SET-NUP214 was mutually exclusive with CALM-AF10, SIL-TAL, TLX1, or TLX3 overexpression (Table 1). NOTCH1 and/or FBXW7 mutations were seen in 36% of SET-NUP214–positive compared with 70% of SET-NUP214–negative T-ALL patients (P = .04). As expected,¹²  all SET-NUP214–positive T-ALL samples overexpressed HOXA9 transcripts (HOXA9/ABL1: median, 227%; range, 65% to 1800%).

Conventional cytogenetic data of SET-NUP214–positive T-ALL patients showed 9 abnormal karyotypes, 5 of which were complex (supplemental Table 2). Strikingly, we observed del(12p) and 5q aberrations in 4 patients, including 3 with both abnormalities. Likewise, 2 patients had concomitant del(6q), del(11q), and del(12p).

All SET-NUP214–positive T-ALLs but 1 had an immature (IM) immunophenotypic profile with no cytoplasmic TCRβ (cTCRβ) or surface CD3/TCR (sCD3/TCR) expression (supplemental Table 3). They showed TCRδ only (IMδ; n = 2) and/or TCRγ (IMγ; n = 7) rearrangements but no complete variable diversity joining (VDJ) TCRβ by genomic PCR. One patient harbored a VDJ TCRβ rearrangement and thus was classified as IMβ. When compared with SET-NUP214–negative T-ALL patients, SET-NUP214–positive patients were strikingly associated with an IMδ or IMγ (IMδ/γ) genotype (82% vs 17%; P < .0001; Table 1). All 10 patients without TCR rearrangements (IM0) were SET-NUP214–negative. The single mature sCD3⁺SET-NUP214–positive T-ALL expressed TCRγδ, whereas none of the TCRαβ-positive or pre-αβ T-ALL patients were SET-NUP214–positive. We therefore undertook to determine whether IMδ/γ SET-NUP214–positive could represent TCRγδ precursor T-ALLs. We previously showed that a CD5⁺CD2^– phenotype is particularly common in TCRγδ-positive T-ALL patients (53%) compared with only 7% of TCRαβ-lineage T-ALL patients.⁵  We also showed that CALM-AF10 in T-ALL is specific to the TCRγδ lineage and that the CD5⁺CD2^– phenotype was much more frequent in IMδ/γ CALM-AF10–positive compared with IMδ/γ CALM-AF10–negative patients, suggesting that this phenotype may identify TCRγδ precursors.⁵  Interestingly, 7 of 9 IMδ/γ SET-NUP214–positive patients in this series were CD5⁺CD2^–. Secondly, TCRδ rearrangements in IMδ/γ SET-NUP214–positive patients, like IMδ/γ CALM-AF10–positive patients, were predominantly Dδ2-Jδ1, indicative of T-lymphoid lineage restriction.^18,19  In addition, TCRγ rearrangements involved the functional VγfI/Vγ9 segments in the majority of IMδ/γ SET-NUP214–positive patients. These TCR profiles reinforce the immunophenotypic evidence of a TCRγδ lineage origin for immature SET-NUP–positive T-ALLs.

Seven (78%) of our 9 IMδ/γ SET-NUP214–positive T-ALL patients expressed stem cell and myeloid markers (supplemental Table 3). In keeping with our data, all 6 previously reported adult SET-NUP214–positive T-ALL patients with an available phenotype were CD34⁺, CD13⁺, and/or CD33⁺.^10,13,15  Coustan-Smith et al²⁰  defined a very high-risk subgroup of pediatric T-ALLs as early T-cell precursor acute lymphoblastic leukemia (ETP-ALL) on the basis of its associated distinctive immunophenotype (CD1a⁻, CD8⁻, CD5^weak with stem cell and/or myeloid markers). Five of 11 of our SET-NUP214–positive T-ALLs met these ETP-ALL criteria²⁰  (Table 1 and supplemental Table 3), and SET-NUP214–positive T-ALLs were more often ETP-ALL (45%) than SET-NUP214–negative T-ALLs (19%; Table1). The poor prognosis of ETP phenotype, however, is increasingly contested in adult²¹  and even pediatric T-ALL.²² 

We performed SET-NUP214 RT-PCR screening in 22 AML overexpressing HOXA genes, including 5 minimally differentiated AML (French-American-British AML-M0) patients. One AML-M0 patient was SET-NUP214–positive. Importantly, this patient was CD7⁺ (but cCD3^–) and showed TCRδ and TCRγ rearrangements (supplemental Figure 2). Taken together, these data demonstrate that SET-NUP214 is permissive of a certain degree of TCRγδ lineage differentiation while maintaining myeloid features.

When compared with SET-NUP214–negative patients, SET-NUP214–positive patients showed a significantly higher rate of Cr (91% vs 44%; P = .003) and CHr (100% vs 44%; P = .0001; Table 1). This strong association of SET-NUP214 with Cr/CHr remained statistically significant (odds ratio, 2.287; 95% confidence interval, 1.37 to 6.85; P = .003) after adjusting for white blood cell count (>100 × 10⁹/L) and age (≥ 35 years).

All SET-NUP214–positive patients but 1 achieved CR, and 9 were allografted (Table 2). Despite their very poor initial sensitivity to induction treatment, the event-free survival (EFS) and OS at 3 years of SET-NUP214–positive patients were not significantly different from those of SET-NUP214–negative patients (45% vs 59%; P = .52 for EFS and 73% vs 68%; P = .86 for OS) (Table 1 and supplemental Figure 3). Similarly, EFS and OS following transplantation were comparable for SET-NUP214–positive and SET-NUP214–negative patients (Table 1). Furthermore, the survival of allografted SET-NUP214–positive T-ALL patients was similar to that of allografted SET-NUP214–negative Cr/CHr patients and of SET-NUP214–negative chemotherapy- and corticosteroid-sensitive patients, while nonallografted SET-NUP214–negative Cr/CHr patients had a significantly inferior outcome (supplemental Figure 4). In keeping with this, we considered the fact that SET-NUP214–positive and SET-NUP214–negative T-ALL patients had a similar outcome is most likely due to the benefit of allografting Cr/CHr patients. This is particularly pertinent in the light of evolving evidence suggesting that appropriately targeted allografting in first CR is the best available option for cure in younger adult high-risk ALL patients.²³  Emerging data also suggest that allogeneic transplantation can be effective in ETP-ALL.^24,25 

In conclusion, our results demonstrate that SET-NUP214 is a recurrent oncogenic fusion transcript in adult patients with T-ALL and is specific to the TCRγδ lineage. SET-NUP214 is strongly associated with corticosteroid and chemotherapy resistance but does not negatively influence clinical outcome after allogeneic transplantation.

We thank all participants in the GRAALL 2003 and 2005 study groups for collecting and providing data and samples. We are grateful to V. Lheritier for her help in collecting clinical data.

This work was supported by grants from the Association Laurette Fugain (12/09), and an Institut National du Cancer Translational Research grant (caractéristiques oléculaire et épigénétique des LAM de l'enfant [CARAMELE]). The GRAALL studies were supported by grants P0200701 and P030425/AOM03081 from Le Programme Hospitalier de Recherche Clinique, Ministère de l’Emploi et de la Solidarité (France), and by the federal government in Switzerland.

Contribution: R.B.A., A. Roggy, E.M., and V.A. wrote the manuscript; R.B.A., A. Roggy, A.C., A.T., H.M., and V.A. performed and/or analyzed molecular and cellular data; T.L., A. Renneville, A.B., E.R., D.C., B.L., A.D., C.P., N.I., and H.D. contributed to the sample collection or provided patient data; and V.A. designed and oversaw conceptual development of the project.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Vahid Asnafi, Hôpital Necker Enfants Malades, Laboratoire d’hématologie, 149 rue de Sèvres, 75015 Paris, France; e-mail: vahid.asnafi@nck.aphp.fr.