Clinical features and outcome of SIL/TAL1-positive T-cell acute lymphoblastic leukemia in children and adolescents: a 10-year experience of the AIEOP group

T-cell acute lymphoblastic leukemia (T-ALL) accounts for approximately 15% and 25% of childhood and adult cases of ALL, respectively,21 and is associated with a less favorable outcome.1 In recent years, however, more intensive and risk-adapted treatment has significantly improved the outcome of patients with T-ALL, leading to cure rates approaching 70% in children and adolescents, and 30%-40% in adults.42

Whether classical risk factors, such as age and white blood cell (WBC) count at diagnosis, are relevant in T-ALL patients is still a matter of debate. At the moment, treatment response, measured at different time points, is considered the most important factor for risk stratification. The identification of new prognostically relevant diagnostic markers, such as genetic abnormalities, could, however, be instrumental in refining risk-adapted treatment stratification in T-ALL.5

Mutation of TAL1 (1p32) is a non-random genetic defect often present in childhood T-ALL. This gene encodes for a protein with a basic helix-loop-helix motif, a DNA binding domain common to several transcriptional regulatory factors. Disruption of TAL1 has been reported in up to 30% of T-ALL cases,1 and is frequently associated with a submicroscopic interstitial deletion (90 Kb) between the 5′ untranslated region (UTR) of the TAL1 and the SIL genes (9%–26% of cases, depending on the different studies). The SIL/TAL1 fusion product gives rise to the inappropriate expression of TAL1, which, in turn, may promote T-cell leukemogenesis.861 The clinical relevance and the prognostic value of this rearrangement remain unclear.1291 Different studies have reported either a favorable or unfavorable outcome trend for SIL/TAL1 positive patients.1281

We evaluated the frequency and prognostic value of SIL/TAL1 in a large cohort of children (age 1–17 years) with newly diagnosed T-ALL and treated at AIEOP centers according to the AIEOP-BFM ALL 2000 (September 2000-July 2006) and the subsequent AIEOP ALL R2006 (August 2006-December 2009) protocols. Treatment outlines and differences between the two protocols are summarized in the Online Supplementary Table S1. The study is registered at the US National Institutes of Health website (http://clinicaltrials.gov; clinicaltrials.gov identifier: 00613457).

Diagnosis of ALL was performed using cytomorphology and cytochemistry on bone marrow cells. T-cell origin of ALL was defined by a positive CD3 (either on cell surface or cytoplasmic) and a negative CD19 immunophenotype. Further subclassification (early, cortical or mature) was performed according to definition of the European Group for the Immunological Characterization of Leukemias (EGIL).13 Early T-cell precursor (ETP) subtype was not routinely screened. All diagnoses were centrally reviewed and confirmed by a reference laboratory.

Patients were stratified as standard risk (SR), intermediate risk (IR) or high risk (HR) mainly on the basis of PCR-minimal residual disease (MRD) at day 33 [time point 1 (TP1)] and day 78 [time point 2 (TP2)], as previously published.14 In addition, and regardless of the MRD results, patients with either prednisone (PDN) poor-response (PPR; ≥1000 circulating blasts/μL on day 8) or failure to achieve complete remission (CR) after induction phase IA were allocated to the HR arm.

Genomic DNA samples at diagnosis were retrospectively screened for the type 1 and type 2 TAL1 deletions by PCR amplification using the BIOMED-1 primer sets and PCR conditions.15

The cycling protocol for TAL1 deletion type 1 amplification consisted of 90 s at 94°C of initial denaturation, then 60 s at 60°C and 90 s at 72°C for the first cycle, followed by 35 cycles of 30 s at 94°C, 30 s at 60°C, 90 s at 72°C, and a final extension phase of 10 min at 72°C. The cycling protocol for TAL1 deletion type 2 amplification consisted of 3 min of initial denaturation at 92°C, followed by 40 cycles of 45 s at 92°C, 90 s at 60°C, 2 min at 72°C, and a final extension phase of 10 min at 72°C. PCR products were examined by 1% agarose gel electrophoresis and then sequenced by the Sanger method.

Overall, 359 of 382 children with T-ALL (92%) were screened for SIL/TAL1 deletion type 1 and 52 of them (14.5%) were positive; 290 of 359 cases with biological material available were subsequently tested for deletion type 2 and only 4 of them (1.4%) were positive, as expected from the literature.7 Since SIL/TAL1 type 2 deletion is quite rare, the 69 SIL/TAL1 type 1 negative patients who were not screened for type 2 deletion were included in the negative cohort. Thus a total of 56 cases (16%) were positive for SIL/TAL1. The characteristics of the SIL/TAL1 positive and negative patients and data on their early response to treatment are reported in Table 1. Male gender was highly prevalent in the SIL/TAL1 patients (P=0.03). There was no significant difference in age distribution between the two groups (P=0.31), with a tendency for more adolescents in the SIL/TAL1 positive group. There was a statistical difference in WBC count at presentation between the two subgroups (P<0.001) with SIL/TAL1 positive and negative patients having respectively a median WBC count of 174×10/L versus 67×10/L and a WBC count of 100×10/L or more in 68% versus 39% of patients. Detailed immunophenotype was available in 341 patients: 53 SIL/TAL1 positive and 288 negative. The immunophenotype distribution was similar in the two subgroups (P=0.28). There was a significant difference in response to the PDN pre-phase between the two groups: 54.5% of SIL/TAL1 positive children were PPR versus 31% of those who are negative (P<0.001). Data on morphological response at the end of phase IA were available for 351 patients; the rate of patients not in CR was 3.6% and 7.4% in SIL/TAL1 positive versus negative cases (P=0.31). There was no significant difference in MRD distribution between the two subgroups (P=0.77). The percentage of HR patients was higher in the SIL/TAL1 positive group compared to the group of negative patients mainly due to a higher frequency of PPR. The final risk stratification was as follows: 7% SR, 36% IR and 57% HR in SIL/TAL1 positive versus 11%, 49% and 39% in SIL/TAL1 negative patients (P=0.05).

Table 1.Clinical characteristics of patients according to genetic group.

The 5-year event-free survival (EFS) was 64.1% (SE 6.8) versus 71.2% (2.6), and the overall survival (OS) 78.6% (5.5) versus 76.3% (2.5) (P=0.34; P=0.82) for patients SIL/TAL1 positive or negative, respectively. At a median follow up of 6.9 years, a total of 86 relapses were registered: 14 (25.0%) in the SIL/TAL1 positive and 72 (23.8%) in the negative groups, respectively (Table 2). There was no significant difference in cumulative incidence of relapse (CIR) between the two genetic groups [5-year CIR of 26.1% (6.1) versus 23.8% (2.5) in SIL/TAL1 positive or negative patients; P=0.72]. SIL/TAL1 positive patients did, however, have a higher frequency of extramedullary relapses (8 of 14, 57% versus 19 of 72, 26%; P=0.01). There was no significant difference in EFS and CIR curves within HR and no HR patients (Figure 1A–D). In analysis according to age (1–9 years versus 10–17 years), the EFS for SIL/TAL1 positive and negative patients, was 67.9% (9.9) versus 71.1% (3.4) (P=0.96) and 59.3% (9.5) versus 71.3 (4.2) (P=0.17), respectively. SIL/TAL1 positivity did not have a significant prognostic value also when analyzed in a Cox regression model, after adjusting for factors known to potentially influence outcome (risk group, age and WBC count at diagnosis).

Figure 1.(A). Event-free survival in patients SIL/TAL1 positive versus SIL/TAL1 negative in the HR group. (B). Cumulative incidence of relapse in patients SIL/TAL1 positive versus SIL/TAL1 negative in the HR group. (C). Event-free survival in patients SIL/TAL1 positive versus SIL/TAL1 negative in no HR group. (D). Cumulative incidence of relapse in patients SIL/TAL1 positive versus SIL/TAL1 negative in the no HR group.

Table 2.Events according to genetic group.

In this relatively large study, 16% of the children screened showed the presence of SIL/TAL1 with a preponderance of breakpoint db1. As already reported,11 no difference was observed in the two cohorts of children with regard to immunophenotypic features of the leukemic cells. The increase in SIL/TAL1 frequency with age observed here, although not statistically significant, is in keeping with the results of Cavè et al. which suggest a modal distribution, with a higher incidence of this alteration in adolescents and young adults.81 SIL/TAL1 deletion was also significantly more frequent in males, strongly associated with higher initial WBC count and PPR, but, interestingly, no difference according to MRD response or response to phase IA was seen. Although in our study the cumulative incidence of relapses was similar for SIL/TAL1 positive and negative patients (both overall and also separately in non-HR or HR subgroups), a higher frequency of extramedullary relapses was observed in the group of positive patients. This original finding, not reported previously, suggests that this gene alteration might play a role in the migration process of leukemia T cells. The median time of extramedullary relapses in SIL/TAL1 positive patients was very similar to that of relapses involving the bone marrow (0.9 versus 1.1 years from diagnosis).

In conclusion, our study shows an association of SIL/TAL1 deletion with adolescence, higher WBC count at diagnosis, and PPR. No major differences in overall outcome were seen. EFS was slightly inferior in the SIL/TAL1 positive subgroup, but there was no difference in survival. This finding may be due to the relapse sites. In fact, the excess of relapses in the SIL/TAL1 positive subgroup is largely represented by patients suffering extramedullary relapse who have a higher probability of being subsequently rescued. This study, like many others aimed at assessing the role of single genetic lesions, was not able to detect a prognostic value in childhood T-ALL, suggesting that further investigations should take into account multiple genetic alterations rather than focus on single ones.