Outcome of children with hypodiploid ALL treated with risk-directed therapy based on MRD levels

To the editor:

Hypodiploid acute lymphoblastic leukemia (ALL) with <45 chromosomes has been associated with a dismal prognosis.^1-4  Near-haploid (24-31 chromosomes) and low-hypodiploid (32-39 chromosomes) ALL have particularly poor outcomes^5,6  and are distinct entities.⁷  Near-haploid ALL is characterized by a younger age at diagnosis⁵  and genetic alterations targeting receptor tyrosine kinase signaling, Ras signaling, and the lymphoid transcription factor gene IKZF3.⁷  Low-hypodiploid patients are older⁵  and have genetic alterations of IKZF2, RB1, and TP53 that are often inherited.⁷ 

Despite overall improved treatment outcome of childhood ALL,⁸  patients with hypodiploid ALL continued to fare poorly. In a Children’s Oncology Group study of 41 hypodiploid cases treated between 2002 and 2006, the 4-year overall survival was 54% ± 8%, and the study could not evaluate the efficacy of allogeneic transplantation.⁹  Because there are no known prognostic indicators in hypodiploid ALL, allogeneic transplantation continues to be recommended.¹⁰ 

We recently showed that minimal residual disease (MRD) levels during remission induction treatment have important prognostic and therapeutic implications in the context of MRD-guided therapy of ALL.¹¹  Notably, the adverse prognosis of pediatric Ph-like ALL can be significantly improved by this treatment approach.¹²  We therefore sought to determine if the MRD-guided treatment strategy, as applied in 2 consecutive clinical trials, could also improve the outcome of hypodiploid ALL. In parallel, we comprehensively examined genetic features and their prognostic importance.

From June 2000 to October 2007, 498 patients (1 to 18 years of age) with newly diagnosed ALL were consecutively enrolled in the St. Jude Total Therapy Study 15,¹³  and from October 2007 to June 2014, 410 patients (2 months to 18 years) in the Study 16.¹⁴  The protocols were approved by the institutional review boards and registered at www.clinicaltrials.gov as #NCT00137111 and #NCT00549848, respectively. Written informed consent was obtained from the parents or guardians, and assent from the patients, as appropriate. All hypodiploid patients received intensive chemotherapy including high-dose methotrexate, dexamethasone, vincristine, and asparaginase. Patients with MRD ≥1%, as determined by flow cytometry and/or polymerase chain reaction analysis^15,16  after completion of induction therapy were offered the option of allogeneic transplantation.

Twenty (2.2%) patients had hypodiploid ALL. Eight patients had near-haploid ALL (median age 3.6 years, range 2.5-8.3), and 12 had low-hypodiploid ALL (median age 13.9 years, range 8.2-17.5 years, P < .01). There was no significant difference in presenting leukocyte count between near-haploid and low-hypodiploid ALL: median 8.1 × 10⁹/L (range 3.0 to 52.0 × 10⁹/L) vs median 5.6 × 10⁹/L (range 1.5 to 36.9 × 10⁹/L), P = .21.

All 20 patients achieved clinical remission. Negative MRD status (<0.01% leukemic cells among bone marrow mononucleated cells) was achieved in 6 of the 8 near-haploid and 8 of the 12 low-hypodiploid patients upon completion of remission induction (P = 1.0). Of the 6 patients with detectable MRD at the end of remission induction, 5 had MRD levels between 0.01% and 0.33%, and 1 had an MRD of 4.14%. Allogeneic transplantation was performed during initial remission in 1 MRD-negative patient with near haploidy because of the preference of the primary physician, and in the low-hypodiploid patient with MRD of 4.14% based on the protocol criterion.

The 5-year event-free survival was 73.6% (95% CI, 47.7-88.1) for all 20 patients (Figure 1) and was 68.2% (95% CI, 39.6-85.4) for the 16 treated in Study 15 as compared with 89.9% (86.5-92.5) for the other 409 B-cell ALL patients treated in Study 15 (P < .001). The result was not significantly different between the 8 patients with near-haploid and the 12 patients with low-hypodiploid ALL (72.9% [27.6-92.5] vs 75.0% [40.8-91.2], P = .80). However, it was significantly better for the 14 patients who achieved negative MRD status at the end of remission induction as compared with the 6 patients with detectable MRD: 85.1% (52.3-96.1) vs 44.4% (6.6-78.5), P = .03 (Figure 1). There were no other features associated with treatment outcome in this cohort. Twelve of the 13 hypodiploid patients with negative MRD status at the end of remission induction treated with chemotherapy only are alive in continuous complete remission for 0.7+ to 12.3+ years (median 7.0+ years); the 5-year event-free survival for this group was 91.7% (53.9-98.8). The near-haploid patient with negative MRD at the end of induction who was transplanted during initial remission died of transplant-related toxicity. Among the 6 patients with detectable MRD at the end of induction, the one with MRD of 4.14% remains in remission for 9.2+ years after transplantation, and 4 of the other 5 with lower levels of MRD (0.01% to 0.33%) treated with chemotherapy alone died of relapse with only 1 alive in remission for 1.4+ years.

Sixteen hypodiploid patients had suitable tumor samples, and all patients had remission samples for genomic analysis and sequencing, performed as previously described (Table 1).^17-21  The spectrum of genetic alterations was similar to those previously reported.⁷  Somatic Ras pathway alterations were identified in 7 of 8 near-haploid patients, with a higher proportion of loss-of-function alterations of NF1 than in our previous study (6 of 7 cases, with a KRAS mutation in 1 case). Four of 8 low-hypodiploid patients harbored TP53 mutations, with 1 identified in the germ line. We also observed alterations of IKZF2 in 2 low-hypodiploid patients and mutations of the histone acetyl-transferase genes CREBBP and EP300 in near-haploid ALL.

Five patients had hematologic relapse, including 2 with near haploidy (nos. 2 and 7) and 3 with low hypodiploidy (nos. 15, 18, and 19). Both near-haploid patients harbored Ras-activating mutations, 1 NF1 and 1 KRAS, and 1 also acquired IKZF1 and PAG1 deletions at relapse. Of the 3 low-hypodiploid patients that experienced relapse, 1 (no. 15) had alterations of IKZF2 and TP53, the second (no. 18) had a sequence mutation of RB1, and the third (no. 19) had deletions of PAX5 and CDKN2A/B. An additional 2 low-hypodiploid patients (nos. 9 and 16) had detectable MRD at the end of induction but did not experience relapse. Patient 16 had a somatic TP53 mutation, and patient 9 lacked tumor material for analysis but did not have a germ-line TP53 mutation. Three patients (nos. 6 and 7 with near haploidy and no. 18 with low hypodiploidy) also exhibited a Ph-like gene expression profile,²¹  2 of whom (nos. 7 and 18) relapsed. No associations between individual genetic alterations and outcome were observed.

In summary, our study demonstrates for the first time that MRD is the most important prognostic indicator for childhood hypodiploid ALL and that the outcome can be substantially improved by MRD-guided therapy. Importantly, hypodiploid patients with negative MRD status at the end of 6-week remission induction are highly curable with intensive chemotherapy alone. Studies are needed to determine if transplantation in first remission will improve the outcome of those with positive MRD at the end of remission induction. Genomic analysis confirmed the distinct genetic alterations characteristic of hypodiploid ALL⁷  but, because of the small number of cases, did not identify any genomic features associated with treatment response or outcome. However, 1 near-haploid patient acquired a deletion of PAG1 (also known as CBP, Csk-binding protein) at relapse. PAG1 encodes a putative modulator of Ras signaling, and PAG1 alterations were previously associated with treatment failure in hypodiploid ALL.⁷  Additional studies are needed to develop effective targeted therapy for hypodiplod ALL. In this regard, inhibitors of phosphatidylinositol-3-kinase signaling were identified as potentially effective drugs in our previous study⁷  and are being explored in preclinical studies as a new therapeutic strategy.