Author Affiliations: Division of Hematology-oncology, Children's Hospital, Zhejiang University School of Medicine, Key Laboratory of Reproductive Genetics (Zhejiang University), Ministry of Education, Hangzhou 310003, China (Jia M, Wang ZJ, Zhao HZ, Shen HP, Cheng YP, Luo ZB, Tang YM)

 

Corresponding Author: Yong-Min Tang, Division of Hematology-oncology, Children's Hospital, Zhejiang University School of Medicine, Key Laboratory of Reproductive Genetics (Zhejiang University), Ministry of Education, Hangzhou 310003, China (Tel: +86-571-88873450; Fax: +86-571-87033296; Email: Y_M_TANG@zju.edu.cn)

 

doi: 10.1007/s12519-015-0019-1

 

 

 

 

 

 

 

 

Acute lymphoblastic leukemia (ALL) is the most common childhood malignancy and one of the major causes of cancer-related mortality in children. Survival rates for childhood ALL have increased dramatically over the last four decades as a result of risk-adjusted treatment, which implements risk factors such as an early treatment response and cytogenetic abnormalities, as well as rationally designed phases in the treatment backbone including remission induction, central nervous system prophylaxis, and intensification and maintenance phases of treatment.^[1] However, despite these major improvements, ALL therapy still fails in approximately 20% of children, and surviving patients often experience serious toxicities.^[2,3] This underscores the need to identify specific biomarkers that could serve as prognostic factors for ALL and stratify different risk groups for appropriate treatment.

The deregulated expression of cytokine receptor-like factor 2 (CRLF2), also known as thymic stromal lymphopoietin receptor (TSLPR), in B-cell precursor ALL was first described by Russell et al^[4] in 2009. Two primary CRLF2 genomic lesions have been identified: a cryptic chromosomal translocation that juxtaposes CRLF2, located in the pseudoautosomal region (PAR1) of chromosome X or Y, to the immunoglobulin heavy chain locus (IGH@); and a PAR1 deletion upstream (centromeric) of CRLF2 that juxtaposes the first noncoding exon of P2RY8 to the entire coding region of CRLF2, resulting in a P2RY8-CRLF2 fusion. Recent studies^[4-6] have also revealed that over-expression of CRLF2 is driven by its juxtaposition to either the IGH@ enhancer (IGH@-CRLF2) or the P2RY8 promoter (P2RY8-CRLF2). High expression levels of CRLF2 are thought to contribute to leukemogenesis by activation of the JAK-STAT pathway, leading to increased cell proliferation.

To date, several studies^[1,6-17] have examined the prognostic impact of elevated CRLF2 expression or CRLF2 deregulation (CRLF2-d) in childhood ALL; however, whether CRLF2 represents a prognostic biomarker remains controversial. Because limited sample sizes in some studies might cause variations in the clinical significance of the results, we performed a systematic review and a meta-analysis to estimate the prognostic importance of CRLF2 alterations in ALL. We assessed the association of better clinical decision-making with the improvement of patient survival in childhood ALL.

 

 

The electronic databases PubMed, Embase, Web of Science, the Cochrane Library, ClinicalTrials.gov, China National Knowledge Infrastructure, and Wanfang Data were searched for studies to be included in the present meta-analysis. The last search was updated on March 2014, and no lower date limit was used. Search terms included "acute lymphoblastic leukemia" or "lymphoblastic lymphoma" or "precursor cell lymphoblastic leukemia-lymphoma [Mesh]", "CRLF2" or "TSLPR", "variant" or "deregulation" and "prognosis" or "outcome". The references cited by the identified studies were also screened for any other potential studies.

Studies eligible for inclusion in the meta-analysis had to meet the following criteria: 1) case-control or cohort study; 2) describing the prognostic value of CRLF2 over-expression or deregulation in childhood patients with ALL; 3) CRLF2 dichotomized as "high" and "low" in a CRLF2 over-expression study group, and as "CRLF2-normal" and "CRLF2-d" status in a CRLF2-d group; 4) investigating the association between CRLF2 alterations and patient prognosis relapse-free survival (RFS), and/or event-free survival (EFS), and/or overall survival (OS); 5) providing sufficient information about survival data; and 6) containing a 95% confidence interval (CI) for the hazard ratio (HR), or sufficient data to calculate these numbers. Two authors (Ming Jia and Yong-Min Tang) independently determined the study eligibility, and any disagreements were resolved by consensus. The following studies were excluded: 1) review articles, editorials, correspondences, case reports, meeting abstracts, and animal studies; 2) studies with insufficient data to calculate a HR; and 3) multiple articles based on the same population and published by the same research team. In such cases, only the latest or the largest population study was included.

 

Data retrieved from the selected studies included first author name, publication year, country, recruitment time, sample size, gender distribution, age range, ethnicity, National Cancer Institute (NCI) risk group, treatment, follow-up period, detection method of CRLF2 alterations, cut-off value, CRLF2 over-expression (%), subtype of CRLF2-d (IGH@-CRLF2 or P2RY8-CRLF2), cases with CRLF2-d, and HR and its 95% CI of relevant outcomes (EFS, RFS or OS) (Tables 1 and 2). Items were treated as "not specified" when data from any of the above categories were not reported in the primary study. We also contacted authors of some meeting abstracts for additional or unreported information. We did not use pre-specified quality-related inclusion or exclusion criteria, and did not weigh each study by a quality score because this has not received general agreement for use in a meta-analysis, especially in observational studies.^[18]

 

Included studies were divided into two groups for analysis: those with data on CRLF2 over-expression and those regarding CRLF2-d. The main outcomes were RFS, EFS, and OS, comparing ALL patients with high CRLF2 expression to those with low expression or measuring the impact of CRLF2-d status on prognosis between the two survival distributions of ALL patients. HRs and corresponding 95% CIs were combined across the studies. When these variables were not available in an article, we estimated them from given data using methods reported by Tierney et al.^[19] Some of the following data were collected to summarize the HR: HR rate, variance, patients in each experimental arm, observed events, expected events, total events, follow-up time, Kaplan-Meier curves, and P value of the log-rank test. Additionally, we used Engauge Digitizer 4.1 software to read survival curve data and calculated survival data using the HR calculation sheet previously published.^[19]

Heterogeneity between trials was checked by the Chi-square test according to Peto's method.^[20] Additionally, I² statistics was calculated for heterogeneity tests. Heterogeneity was defined as P<0.10 or I²>50%. If no heterogeneity was identified, the fixed-effects model was applied. Otherwise, a random-effects model was used. By convention, an observed HR >1 implied worse survival for the analysis group with corresponding factors. For CRLF2-d, subgroup analysis was performed to investigate the prognostic impact of P2RY8-CRLF2 fusion on ALL patients. Two parallel analyses were performed, one focusing on univariate HRs and the other on multivariate HRs, to determine whether CRLF2 is an independent prognostic factor of different survival endpoints.

Finally, publication bias was assessed by Begg's funnel plot^[21] and Egger's test.^[22] P values <0.05 were regarded as statistically significant. Statistical analyses were performed by STATA 12.0 software (StataCorp, College Station, TX).

 

 

The process of identifying eligible studies is shown in Fig. 1. Following stepwise selection, twelve studies^[1,6-16] published in the period of 2010-2013 were eligible for meta-analysis. The total number of patients included was 5945, ranging from 53 to 1061 patients per study (median, 486). Seven studies reported the prognostic value of CRLF2 over-expression for survival in childhood ALL (Table 1). These studies were conducted in seven countries: the Netherlands, the United States, Germany, China, Italy, Japan, and Israel. Of these seven studies, two were performed in Asian populations, and the remaining five focused on non-Asian patients. The treatment options in the seven studies differed from each other.

Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) was used to assess CRLF2 expression in these seven studies. However, the "high" CRLF2 expression group was defined by using different cut-off values between the studies. The proportion of patients exhibiting CRLF2 over-expression ranged from 4.7% to 62.0% in individual studies.

The main characteristics of the six eligible publications that evaluated the prognostic value of CRLF2 deregulation for survival in childhood ALL are summarized in Table 2. These studies were conducted in five countries: England, Austria, Germany, Italy, and the Netherlands, and involved Caucasian patients. Fluorescence in situ hybridization analyses were used in two studies to assess CRLF2-d, while qRT-PCR/RT-PCR was used in the other three studies. Different treatment options were used in all six studies.

Two studies focused on CRLF2-d, based on IGH@-CRLF2 and P2RY8-CRLF2, while the other four studies focused on the prognostic significance of subtype P2RY8-CRLF2. The positive ratio of CRLF2-d in individual studies ranged from 3.89% to 54.7%. Five studies measured HRs and 95% CIs derived directly from multivariate analyses, and only one study reported HRs and 95% CIs derived from survival curves with different outcomes (EFS/OS).

 

A total of seven studies including 3586 cases were evaluated for the impact of CRLF2 over-expression on the prognosis of childhood ALL. A fixed-effects model was used to combine HR values. The results of the meta-analysis are shown in Fig. 2A. Three studies were evaluated for the impact of CRLF2 over-expression on the RFS of ALL patients. The pooled HR value was 1.70 (95% CI=1.28-2.24, P=0.000), without evidence of heterogeneity (P=0.645, I²=0.0%), suggesting that CRLF2 over-expression is associated with decreased RFS in ALL patients. The combined data on EFS showed that patients with CRLF2 over-expression had a shorter EFS (HR=1.78, 95% CI=1.05-3.01, P=0.032) than those without CRLF2 over-expression, with no evidence of data heterogeneity (P=0.686, I²=0.0%). The pooled HR of CRLF2 over-expression on ALL patient survival for OS was 2.28 (95% CI=1.42-3.65, P=0.001) without evidence of heterogeneity (P=0.598, I²=0.0%).

We then conducted two parallel analyses on univariate HRs and multivariate HRs. The combined HR for three eligible studies evaluating CRLF2 over-expression on RFS by multivariate analysis was 1.70 (95% CI=1.28-2.24), with no significant heterogeneity (P=0.645, I²=0.0%). The fixed-effects model was adopted (Fig. 2B). These results suggested that CRLF2 over-expression is an independent prognostic factor of poor RFS for childhood ALL. However, a pooled analysis according to data type could not be performed for EFS and OS because of the lack of original data.

No evidence for significant publication bias was identified for RFS (Begg's test, P=0.296; Egger's test, P=0.437), EFS (Begg's test, P=1.000; Egger's test, P=0.627), or OS (Begg's test, P=1.000; Egger's test P value could not be calculated) in the whole group analysis.

 

In six studies of 2429 cases, the survival data (EFS, RFS, and OS) were assessed for the impact of CRLF2-d on the prognosis of childhood ALL. Fig. 3A shows the forest plot of the association between CRLF2-d and prognosis in childhood ALL. A fixed-effects model was used to combine the HR values. The pooled HR for EFS showed that CRLF2-d carriers had a worse EFS than non-carriers (HR=1.95, 95% CI=1.46-2.61, P=0.000). No heterogeneity was observed (P=0.447, I²=0.0%). Combined RFS data showed that patients with CRLF2-d had a shorter RFS than those without CRLF2-d (HR=2.20, 95% CI=1.53-3.18, P=0.000), and that no significant heterogeneity was observed in the data (P=0.173, I²=39.7%). We identified three studies assessing the OS of CRLF2-d carriers and the pooled HR of the CRLF2-d effect on ALL patient survival for OS was 1.89 (95% CI=1.24-2.87, P=0.003). There was no evidence of heterogeneity (P=0.999, I²=0.0%).

Fig. 3B shows the results of the subgroup analysis according to data type. The fixed-effects model was adopted to combine the HR values. In studies reporting HRs based on multivariate analysis, CRLF2-d was significantly associated with poor EFS in ALL patients (HR=1.85, 95% CI=1.34-2.57, P=0.000). The same result was observed for OS (HR=1.89, 95% CI=1.20-2.97, P=0.006). No significant heterogeneity was observed in the studies on EFS (P=0.358, I²=7.0%) and OS (P=0.965, I²=0.0%). These results suggest that CRLF2-d is an independent prognostic factor of poor prognosis for childhood ALL.

To assess whether P2RY8-CRLF2 is associated with the prognosis of childhood ALL, we pooled the data into a subgroup analysis (Fig. 4). This stratified analysis indicated that P2RY8-CRLF2 was associated with a poor ALL prognosis in terms of both EFS and RFS, with a pooled HR of 2.51 (95% CI=1.66-3.78, P=0.000) and 3.42 (95% CI=1.98-5.93, P=0.000), respectively, using a fixed-effects model with no significant heterogeneity (I²=0.0%, P=0.684 and I²=0.0%, P=0.746, respectively).

Begg's funnel plot and Egger's test of the whole group analysis indicated that there was no risk of publication bias in this meta-analysis (EFS: Begg's test P=0.462, Egger's test P=0.195; RFS: Begg's test P=1.000, Egger's test P=0.958; OS: Begg's test P=1.000, Egger's test P=0.878).

 

 

ALL is a markedly heterogeneous disease with a complex spectrum of genetic abnormalities, including recurring chromosomal translocations, and genomic deletions and amplifications.^[23-26] The characterization of molecular and cytogenetic abnormalities has not only provided insights into the mechanisms of leukemogenesis, but also led to the establishment of new treatment strategies targeting these abnormalities.

To date, several studies^[1,6-17] have examined the prognostic impact on CRLF2 alterations in ALL, however, the use of CRLF2 as a prognostic biomarker remains controversial. To our knowledge, this meta-analysis is the first study to systematically evaluate the prognostic role of CRLF2 alterations in ALL. Our results show that both CRLF2 over-expression and deregulation are associated with poor prognosis in childhood ALL, while subgroup analyses according to multivariate HRs reveal that both CRLF2 over-expression and CRLF2-d are independent prognostic factors of poor prognosis for childhood ALL. Therefore, the detection of CRLF2 alterations may be useful to stratify patients at high risk, guide their clinical surveillance, and to select the most appropriate chemotherapy.

Children with Down syndrome (DS) have a higher rate of associated ALL. DS-ALL has been shown to be a highly heterogeneous disease that is not definable as a unique entity.^[27] Recently, CRLF2 abnormalities have been described in approximately 60% of DS-ALL patients. In our study, the conclusion didn't change substantially despite we included two studies^[6,16] focused primarily on the prognosis of DS-ALL based on CRLF2 aberrations.

The present meta-analysis has a number of limitations. First, the original studies might be different in patient characteristics such as age, sex distribution, ethnicity, NCI risk group, and treatment received. These may contribute to between-study heterogeneity. A comparative analysis of Italian association of pediatric hematology and oncology-Berlin Frankfurt Münster study group and the UK national cancer research institute-children's cancer and leukemia group^[28] revealed that P2RY8-CRLF2+ patients had an intermediate protocol-independent outcome. However, it remains to be determined whether different treatment protocols affect the outcome of patients with CRLF2 alterations. Second, the method of detecting CRLF2 alterations was not standardized between the studies, and the application of different cut-off values for determining high CRLF2 levels is a potential source of bias. Thus, we propose that well-defined standardized methods should be used in the future to reproducibly evaluate biological markers. Third, the method of HR and 95% CI extrapolation is another potential source of bias. In our study, if multivariate survival analysis data were unavailable, we extrapolated them from the survival curves, which could have caused heterogeneity. Fourth, as there was a limited number of prospectively designed prognostic studies concerning biomarkers and our meta-analysis was only concerned with retrospective studies, the results should be confirmed by well-designed prospective studies. Additionally, our findings should be interpreted with caution because of the small number of studies included in each group analysis. Further high-quality and large-sample studies are therefore needed to confirm our results. Finally, publication bias is inevitable, even though not evident from statistical testing, because negative or controversial results might not be reported. To minimize publication bias, we attempted to retrieve all relevant data that are unavailable from the published reports, but it is likely that some data are still missing.

In conclusion, CRLF2 over-expression and deregulation could both be used as biomarkers to predict the prognosis of patients with childhood ALL. Based on the current findings, an assessment of CRLF2 alterations would provide improved prognostic information and could be used as a novel therapeutic target for patients with ALL. To strengthen our findings, well-designed prospective studies with a standardized assessment of prognostic markers should be conducted to explore the relationship between CRLF2 and survival of pediatric patients with ALL.

 

 

Funding: This study was supported in part by grants from the National Natural Science Foundation of China (No: 30971283, No: 81170502), the Zhejiang Provincial Natural Science Foundation of China (No: LZ12H08001), and the Health and Family Planning Commission of Zhejiang Province.

Ethical approval: Not needed.

Competing interest: All the authors declare no conflict of interests.

Contributors: Jia M is the first author of this article who was responsible for study design, data analysis, manuscript writing and revision. Zhao HZ and Shen HP reviewed and revised the manuscript. Cheng YP, Luo ZB assisted in completing the statistical analysis. Wang ZJ is responsible for English checking. Tang YM is the guarantor. Jia M and Wang ZJ contributed equally to this paper.

 

 

1   van der Veer A, Waanders E, Pieters R, Willemse ME, Van Reijmersdal SV, Russell LJ, et al. Independent prognostic value of BCR-ABL1-like signature and IKZF1 deletion, but not high CRLF2 expression, in children with B-cell precursor ALL. Blood 2013;122:2622-2629.

2   Schrappe M, Reiter A, Zimmermann M, Harbott J, Ludwig WD, Henze G, et al. Long-term results of four consecutive trials in childhood ALL performed by the ALL-BFM study group from 1981 to 1995. Berlin-Frankfurt-Münster. Leukemia 2000;14:2205-2222.

3   Möricke A, Reiter A, Zimmermann M, Gadner H, Stanulla M, Dördelmann M, et al. Risk-adjusted therapy of acute lymphoblastic leukemia can decrease treatment burden and improve survival: treatment results of 2169 unselected pediatric and adolescent patients enrolled in the trial ALL-BFM 95. Blood 2008;111:4477-4489.

4   Russell LJ, Capasso M, Vater I, Akasaka T, Bernard OA, Calasanz MJ, et al. Deregulated expression of cytokine receptor gene, CRLF2, is involved in lymphoid transformation in B-cell precursor acute lymphoblastic leukemia. Blood 2009;114:2688-2698.

5   Mullighan CG, Collins-Underwood JR, Phillips LA, Loudin MG, Liu W, Zhang J, et al. Rearrangement of CRLF2 in B-progenitor- and Down syndrome-associated acute lymphoblastic leukemia. Nat Genet 2009;41:1243-1246.

6   Hertzberg L, Vendramini E, Ganmore I, Cazzaniga G, Schmitz M, Chalker J, et al. Down syndrome acute lymphoblastic leukemia, a highly heterogeneous disease in which aberrant expression of CRLF2 is associated with mutated JAK2: a report from the International BFM Study Group. Blood 2010;115:1006-1017.

7   Chen IM, Harvey RC, Mullighan CG, Gastier-Foster J, Wharton W, Kang H, et al. Outcome modeling with CRLF2, IKZF1, JAK, and minimal residual disease in pediatric acute lymphoblastic leukemia: a Children's Oncology Group study. Blood 2012;119:3512-3522.

8   Mi JQ, Wang X, Yao Y, Lu HJ, Jiang XX, Zhou JF, et al. Newly diagnosed acute lymphoblastic leukemia in China (II): prognosis related to genetic abnormalities in a series of 1091 cases. Leukemia 2012;26:1507-1516.

9   Cario G, Zimmermann M, Romey R, Gesk S, Vater I, Harbott J, et al. Presence of the P2RY8-CRLF2 rearrangement is associated with a poor prognosis in non-high-risk precursor B-cell acute lymphoblastic leukemia in children treated according to the ALL-BFM 2000 protocol. Blood 2010;115:5393-5397.

10 Palmi C, Vendramini E, Silvestri D, Longinotti G, Frison D, Cario G, et al. Poor prognosis for P2RY8-CRLF2 fusion but not for CRLF2 over-expression in children with intermediate risk B-cell precursor acute lymphoblastic leukemia. Leukemia 2012;26:2245-2253.

11 Yamashita Y, Shimada A, Yamada T, Yamaji K, Hori T, Tsurusawa M, et al. IKZF1 and CRLF2 gene alterations correlate with poor prognosis in Japanese BCR-ABL1-negative high-risk B-cell precursor acute lymphoblastic leukemia. Pediatr Blood Cancer 2013;60:1587-1592.

12 Ensor HM, Schwab C, Russell LJ, Richards SM, Morrison H, Masic D, et al. Demographic, clinical, and outcome features of children with acute lymphoblastic leukemia and CRLF2 deregulation: results from the MRC ALL97 clinical trial. Blood 2011;117:2129-2136.

13 Morak M, Attarbaschi A, Fischer S, Nassimbeni C, Grausenburger R, Bastelberger S, et al. Small sizes and indolent evolutionary dynamics challenge the potential role of P2RY8-CRLF2-harboring clones as main relapse-driving force in childhood ALL. Blood 2012;120:5134-5142.

14 Palmi C, Valsecchi MG, Longinotti G, Silvestri D, Carrino V, Conter V, et al. What is the relevance of Ikaros gene deletions as a prognostic marker in pediatric Philadelphia-negative B-cell precursor acute lymphoblastic leukemia? Haematologica 2013;98:1226-1231.

15 Dörge P, Meissner B, Zimmermann M, Möricke A, Schrauder A, Bouquin JP, et al. IKZF1 deletion is an independent predictor of outcome in pediatric acute lymphoblastic leukemia treated according to the ALL-BFM 2000 protocol. Haematologica 2013;98:428-432.

16 Buitenkamp TD, Pieters R, Gallimore NE, van der Veer A, Meijerink JP, Beverloo HB, et al. Outcome in children with Down's syndrome and acute lymphoblastic leukemia: role of IKZF1 deletions and CRLF2 aberrations. Leukemia 2012;26:2204-2211.

17 Haslam K, Kelly J, Morris T, Connaghan G, Gilligan O, Browne P, et al. Screening for CRLF2 overexpression in adult acute lymphoblastic leukemia. Int J Lab Hematol 2011;33:e17-e19.

18 Altman DG. Systematic reviews of evaluations of prognostic variables. BMJ 2001;323:224-228.

19 Tierney JF, Stewart LA, Ghersi D, Burdett S, Sydes MR. Practical methods for incorporating summary time-to-event data into meta-analysis. Trials 2007; 8:16.

20 Yusuf S, Peto R, Lewis J, Collins R, Sleight P. Beta blockade during and after myocardial infarction: an overview of the randomized trials. Prog Cardiovasc Dis 1985;27:335-371.

21 Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics 1994;50:1088-1101.

22 Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629-634.

23 Pui CH, Behm FG, Downing JR, Hancock ML, Shurtleff SA, Ribeiro RC, et al. 11q23/MLL rearrangement confers a poor prognosis in infants with acute lymphoblastic leukemia. J Clin Oncol 1994;12:909-915.

24 Pui CH, Pei D, Sandlund JT, Campana D, Ribeiro RC, Razzouk BI, et al. Risk of adverse events after completion of therapy for childhood acute lymphoblastic leukemia. J Clin Oncol 2005;23:7936-7941.

25 Teitell MA, Pandolfi PP. Molecular genetics of acute lymphoblastic leukemia. Annu Rev Pathol 2009;4:175-198.

26 Uckun FM, Nachman JB, Sather HN, Sensel MG, Kraft P, Steinherz PG, et al. Clinical significance of Philadelphia chromosome positive pediatric acute lymphoblastic leukemia in the context of contemporary intensive therapies: a report from the Children's Cancer Group. Cancer 1998;83:2030-2039.

27 Forestier E, Izraeli S, Beverloo B, Haas O, Pession A, Michalova K, et al. Cytogenetic features of acute lymphoblastic and myeloid leukemias in pediatric patients with Down syndrome: an iBFM-SG study. Blood 2008;111:1575-1583.

28 Attarbaschi A, Morak M, Cario G, Cazzaniga G, Ensor HM, te Kronnie T, et al. Treatment outcome of CRLF2-rearranged childhood acute lymphoblastic leukaemia: a comparative analysis of the AIEOP-BFM and UK NCRI-CCLG study groups. Br J Haematol 2012;158:772-777.