Childhood cancer is the leading cause of death due to disease from the first year of life through adolescence. Acute lymphoblastic leukaemia (ALL) is the most frequent type of cancer in children.1 The mortality of ALL has decreased substantially in recent year, with 5-year overall survival rates increasing to up to 90%.2 However, the presence of genetic changes such as the t (9;22) translocation, found in 2–4% of paediatric patients with ALL, is associated with an increased risk of refractory disease and relapse.3

The chromosome that results from the aforementioned translocation, t (9;22) (q34;q11), is known as the Philadelphia chromosome. Nowell and Hungerford first described it in 1960 in adult patients with chronic myeloid leukaemia,4 and it has since been associated with the pathogenesis of some forms of acute myeloid leukaemia, ALL and mixed-phenotype acute leukaemia. The t (9;22) translocation of the tyrosine-protein kinase encoding gene ABL1 (chromosome 9) to the breakpoint cluster region BCR gene (chromosome 22) results in the fusion gene BCR-ABL1, which encodes a protein with constitutive tyrosine kinase activity.5 This results in permanent activation of cell proliferation pathways in haematopoietic stem cells, promoting pathogenicity through the inhibition of differentiation pathways and resistance to conventional treatment.6

Traditionally, a diagnosis of Philadelphia chromosome-positive ALL (Ph+ ALL) has been considered high-risk and to have a poor prognosis. Allogeneic haematopoietic stem cell transplantation (HSCT) has substantially improved survival in this disease compared to the exclusive use of chemotherapy.7,8

The addition of tyrosine kinase inhibitors (TKIs) to conventional chemotherapy has significantly improved the prognosis of Ph+ ALL, with the 5-year event-free survival increasing to approximately 70%,9,10 although still remaining lower compared to paediatric patients with Philadelphia chromosome-negative ALL.11

Imatinib is a TKI that acts by interacting with ABL, inhibiting its phosphorylation and therefore the activation of intracellular signal transduction cascades involved in proliferation. It also promotes apoptosis.12–14 Its combination with intensive chemotherapy followed by HSCT has been found to increase survival.15–18 However, it has not proven effective as monotherapy due to the development of drug resistance.11

The early response to treatment, determined by the measurement of minimal residual disease (MRD) through immunophenotyping or molecular techniques such as quantitative polymerase chain reaction (qPCR), is a very important prognostic factor in ALL, with more favourable outcomes in patients that have negative results for MRD.10,19–21

Patients treated with conventional chemotherapy and imatinib that achieve complete remission (CR) with negative MRD are considered low-risk. In this subset of patients, survival outcomes are not inferior to patients that also undergo HSCT. However, in the case of patients with positive MRD results after induction therapy, who are considered high-risk, survival can improve with performance of HSCT.9,15,22

At present, paediatric patients with a diagnosis of Ph+ ALL receive chemotherapy according to the protocol of the Sociedad Española de Hemato-Oncología Pediátrica (Spanish Society of Paediatric Haematology and Oncology, SEHOP) and the Spanish Haematology Treatment Programme (Spanish acronym, PETHEMA) 2013 update for the group of high-risk patients, by which imatinib is added from day 15 of conventional induction chemotherapy.

The primary objective of our study was to assess the clinical impact of HSCT in our region in paediatric patients with a diagnosis of Ph+ ALL by analysing the historical case series of the Spanish Group of Bone Marrow Transplantation in Children (Spanish acronym, GETMON) and the Spanish Group of Haematopoietic Stem Cell Transplantation (Spanish acronym, GETH).

Fifty-two patients (74.28%) underwent HSCT in first CR, 9 (12.86%) in second CR and 9 (12.86%) in subsequent remissions. In 36 patients (51.4%), the results of MRD were negative before HSCT.

The source of haematopoietic stem cells was bone marrow in 32 patients (45.7%), in 13 cases (18.57%) from HLA-identical related donors, umbilical cord blood in 21 patients (30%), in 1 case from an HLA-identical related donor, and mobilised peripheral blood in 17 patients (24.3%), in 10 cases (14.28%) from HLA-identical related donors.

The conditioning regimen was myeloablative in 55 of the HSCTs (78.57%) and included TBI in 41 (58.57%). Conditioning included administration of ATG in 41 patients (58.57%). All patients received cyclosporin, alone or combined with another agent, for post-transplant immunosuppression (Table 2).

Two (2.86%) patients experienced graft failure. The most frequent post-transplant complication was GvHD, in the form of acute GvHD in 38 patients (54.3%) and chronic GvHD in 17 (24.3%).

Of the patients that developed acute GvHD, 11 (28.9%) had mild presentations, while 27 (71.1%) developed moderate to severe disease. Infectious complications developed in 24 patients (34.3%) and were the second most frequent type of complication.

A total of 15 patients (21.43%) experienced relapse. There were 32 deaths (45.72%): 6 (18.75%) due to relapse and 26 (81.25%) transplant-related (Table 2).

The median duration of follow-up in the cohort was 6.92 years (IQR, 12.25). The 6-year OS was 56.9% (SD, 5.9%) and the 6-year EFS 48.4% (SD, 6%). The 6-year PR was 23.7% (SD, 6.1%) and the 6-year TRM was 35.7% (SD, 6%) (Fig. 1).

Figure 1.

Six-year survival curves (A) Overall survival curves (OS). (B) Event-free survival (EFS). (C) HSCT-related mortality (TRM). (D) Probability of relapse (PR).

(0.17MB).

Fig. 2 presents the statistically significant differences in OS, EFS, PR and TRM in patients that underwent transplantation in first CR.

Figure 2.

Six-year survival curves (A) Overall survival curves (OS). (B) Event-free survival (EFS). (C) HSCT-related mortality (TRM). (D) Probability of relapse (PR).

CI, confidence interval; 1CR, first complete remission; ≥2CR, second or subsequent complete remission.

(0.26MB).

The use of imatinib increased significantly from 2004. In addition, the lowest proportion of relapse occurred in the group of patients treated with imatinib exclusively before HSCT (10%; P = .006), compared to patients who did not receive imatinib, who received it both before and after HSCT or who received it only after HSCT (Table 3).

Our study summarises the experience in Spain with HSCT in paediatric patients with Ph+ ALL, an infrequent disease with a poor prognosis, over a long 22-year period.

The data evinces that the use of imatinib before HSCT has a relevant role, with a substantial impact in the probability of relapse. This was consistent with the findings of previous studies.9,25,26 However, it is important to note that some authors question the impact of imatinib in terms of increasing survival.27,28

Overall, the data showed progressive improvement in HSCT outcomes. This was a significant trend evinced in increases in OS and EFS and decreases in TRM performed from 2004. This inflection point may be explained by the increased cumulative experience, the optimization of donor selection, conditioning strategies, advances in immunosuppressive therapy and adequate prophylaxis.29

Another significant factor was disease status before HSCT. In patients that underwent HSCT when they were in first CR, the OS and EFS were higher and the TRM lower. On the other hand, the absence of detectable MRD was associated with a lower PR. In the univariate analysis, the age of the patient at the time of transplantation appeared to be an important factor, as TRM was lowest in patients aged less than 10 years.26

The increased survival observed in patients treated with chemotherapy and TKI suggests that not all patients with Ph+ ALL may require treatment with HSCT after achieving first CR. The COG-AALL0031 study found that patients treated with chemotherapy combined with imatinib who had not undergone HSCT had a 5-year EFS of 70% (SD, 12%), without significant differences compared to patients that received a sibling donor bone marrow transplant (65% ± 11%) or an unrelated donor transplant (59% ± 15%).10 In the EsPhALL2010 study, treatment consisted of imatinib and chemotherapy. Haematopoietic stem cell transplantation was performed in high-risk patients and in low-risk patients with access to an HLA-identical donor. The 5-year EFS was 62.7% in low-risk patients (95% CI, 52–71.6) and 46.3% in high-risk patients (95% CI, 32.3–59.2). There was a significant increase in EFS in low-risk patients treated with chemotherapy and imatinib compared to patients that also underwent HSCT during the first CR.9,26

Thus, the option of HSCT should be contemplated in patients that do not achieve an undetectable MRD at the end of induction, and not routinely, due to the favourable outcomes achieved with chemotherapy combined with TKIs in low-risk patients. In our study, the use of imatinib was associated with significant improvements in HSCT outcomes from year 2004, with a decreased frequency of relapse. However, there are contradictory results in the literature as regards patients treated with TKIs.25,26,30

In the cohort under study, there was a significant number of patients who developed GvHD, both in its acute and chronic forms, but it did not ultimately have a negative impact on survival. This has been observed in the past.27 Moreover, we ought to highlight that patients who developed acute GvHD had a higher EFS and lower PR in the univariate analysis. This could be explained by the graft-versus-leukaemia effect associated with GvHD. Other studies have associated the development of chronic GvHD with a lower relapse rate, suggesting that it may be a protective factor.31

Treatment with ATG as part of the conditioning regimen had a positive impact on OS and EFS. It was associated with a decreased TRM, so it acts as a protector against the development of severe forms of GvHD, as described in other studies.32–34

The use of TBI in children is currently questioned due to its important late side effects.35 A recent study36 did not find a significant association of TBI with survival or PR in children that underwent TBI prior to HSCT, and instead found an association with the development of GvHD. The findings of our retrospective study were similar.

As we already mentioned, in our study we found the strongest association of undetectable MRD with a decrease in PR. However, the data available for study did not allow us to determine which of the techniques used to quantify MRD (flow cytometry vs molecular methods) is more appropriate for the purpose.36

Compared to previous studies, we did not find differences between patients that received the transplant from an HLA-identical related donor versus an HLA-identical unrelated donor.8,37 This is quite relevant, as an HLA-identical related donor will not be available in most cases.

In light of the above, we think it is relevant to share our findings, especially given the current importance of CAR T-cell therapies in B-cell ALL. However, it is important to keep in mind that the indication of CAR T-cell therapy in the case of Ph+ ALL is restricted to patients who cannot tolerate or do not respond to 2 TKIs or in whom the use of TKIs is contraindicated. In spite of these, favourable outcomes have been reported with the use of CAR T cells.38–40

There were limitations to this study, chiefly related to its retrospective design and the prolonged period under study.

Imatinib has a beneficial impact on disease status before transplantation, promoting complete remission and MRD negativity and therefore decreasing the probability of relapse.

In patients in first complete remission that received ATG, the OS and EFS were higher, so the use HSCT should be considered in these patients.

The authors have no conflicts of interest to declare.