Targeting paediatric acute lymphoblastic leukaemia: novel therapies currently in development

Summary

Modifications to the treatment of acute lymphoblastic leukaemia (ALL) in children have led to a dramatic increase in survival in the past 40 years. Despite this success, a significant subset of paediatric leukaemia patients either relapse or fail to ever achieve a complete remission. Additionally, some patients necessitate treatment with intensified chemotherapy regimens due to clinical or laboratory findings which identify them as high risk. These patients are unlikely to respond to further minor adjustments to the dosing or timing of administration of the same chemotherapy medications. Many novel targeted therapies for the treatment of childhood ALL provide potential mechanisms to further improve cure rates, and provide the possibility of minimizing toxicity to non-malignant cells, given their specificity to malignant cell phenotypes. This article explores many of the potential targeted therapies in varying stages of development, from those currently in clinical trials to those still being refined in the research laboratory.

Paediatric cancer is the leading cause of disease-related mortality in children, and acute lymphoblastic leukaemia (ALL) is the most common childhood malignancy diagnosed. Over the past 40 years, alterations in paediatric ALL therapy have led to significant improvements in survival, increasing overall survival rates from less than 10% in the 1960s to over 85% in the 2000s (O’Leary et al, 2008). This dramatic improvement in survival also reflects the multi-centre collaboration of paediatric cancer networks, which collectively evaluate treatment options in large national and international trials. Survival improvements have largely resulted from changes to the administration timing, dosing, and combinations of a standard group of chemotherapeutics, as essentially no new frontline chemotherapeutics in the treatment of paediatric ALL (aside from use of imatinib in a small subset of patients) have been developed in over 30 years.

Despite vastly improved overall survival, the prognosis remains poor for some patients with ALL. Patients with certain cytogenetic abnormalities, such as the presence of Philadelphia chromosome (Ph+), severe hypodiploidy, and infants with MLL rearrangement, are known to have increased rates of induction failure and/or relapse, and event-free survival (EFS) rates less than 40% (Salzer et al, 2009; Schultz et al, 2009). Additionally, there is a subset of patients who fail to respond or who relapse following current therapies despite having apparently standard risk disease. In a recent study of paediatric patients with relapsed ALL, EFS for all subtypes combined was near 30%; relapsed T-cell ALL was noted to have a particularly grim prognosis, with overall survival approximately 15% (Einsiedel et al, 2005).

Although the majority of patients with ALL respond well to current therapies, traditional anti-neoplastic medications are known to have a multitude of short- and long-term toxic side effects, which may lead to significant morbidity. Approximately two of every three paediatric cancer survivors will have a late effect, including neurocognitive sequelae, auditory complications, cardiovascular dysfunction, gastrointestinal/ hepatic dysfunction, gonadal dysfunction, and decline in growth; one in four will experience a late effect that is severe (Landier & Bhatia, 2008). Additionally, long-term toxicities of these anti-neoplastics include the possibility of future secondary malignancy, even after remission is attained. Alternative approaches to improve cure rates, while minimizing toxicity, is the ultimate goal of modern paediatric leukaemia therapy.

Innovative approaches to treating ALL focus on selectively killing the malignant cells, while only minimally affecting non-malignant cells. Though the majority of newly developed targeted therapies are being clinically evaluated in relapsed/ refractory leukaemias, if proven effective, use of many of these agents in frontline therapy may eventually allow us to minimize conventional pan-cytotoxic anti-neoplastics and the associated toxicities, while maximizing survival and minimizing relapse. New targeted therapies in ALL are being developed with these goals in mind, and many of these therapies have been found in vitro to replace or reduce the use of traditional cytotoxic chemotherapeutics. New therapies in research and development use a variety of approaches to selectively target cancer cells, by altering intracellular signalling pathways, regulating gene expression, or targeting unique cell surface receptors.

A number of biochemical processes are involved in normal lymphocyte growth and proliferation. Dysregulation of any portion of these complex processes provides potential for uncontrolled malignant growth. In lymphoid leukaemias, the malignant cell relies on certain signalling pathways that are pathologically altered to provide a continuous survival and proliferation advantage. Intracellular pathways known to be pathologically up-regulated in ALL include PI3K (phosphoinositide 3-kinase), AKT, MAPK/ERK and mTOR (mammalian target of rapamycin) signalling, among others (See Fig 1), which the cell uses to perpetuate proliferation and survival, and evade apoptosis. Though these pathways are often utilized by non-malignant cells, it is the reliance of the leukaemia cell on these pathways for survival that make them excellent targets for cancer treatment.

Fig 1.

Cellular pathways under investigation as potential therapeutic targets in paediatric ALL. With the exception of Aurora Kinase, which is required for mitotic spindle assembly, all other kinases included here are purple and activate the PI3K-Akt and/or Ras-MAPK pro-survival pathways. Cluster of differentiation (CD) surface marker proteins are pink. Proteins in the apoptotic pathway are yellow. Molecules involved in protein degradation are green. Transcription factors are blue. Compounds in development as targeted therapeutics are described in black boxes.

The initial model of success with targeted therapy in chronic myeloid leukaemia (CML) is based on the selective inhibition of the ABL1 tyrosine kinase by imatinib. Use of imatinib showed over 75% complete cytogenetic response, compared to less than 15% with previous therapy (O’Brien et al, 2003). While tyrosine kinases are still popular targets for new therapeutics, current targeted therapy research has expanded to include a wide variety of dysregulated biochemical intracellular pathways, abnormal gene expression patterns and cell surface markers which might contribute to a cell’s malignant phenotype. With current high throughput sequencing and array technology being used to screen paediatric ALL patient samples, it is likely that additional potential targets will be discovered in the near future. In this review, we will discuss current novel approaches to targeting paediatric ALL in preclinical and clinical stages of development.

Targeting oncogenic kinases

BCR-ABL1 tyrosine kinase inhibition

The Philadelphia chromosome is characterized by the abnormal transposition of the q34 portion of chromosome 9 and the q11 portion of chromosome 22; the specific breakpoint found in ALL is slightly different than that found in CML (Clark et al, 1987; Kurzrock et al, 1987). The fusion gene produced in Ph+ ALL links the cytoplasmic tyrosine kinase ABL1 on chromosome 9 with the BCR gene on chromosome 22, resulting in a constitutively active kinase protein (Kurzrock et al, 2003). The dysregulated ABL1 tyrosine kinase (TK) ultimately results in decreased control of cellular proliferation via activation of phosphoinositide 3-kinase (PI3K) and the downstream pro-survival proteins AKT and mTOR, which are essential for transformation in BCR-ABL1 positive B-lineage leukaemia cell lines (Kharas et al, 2004). Activation of p21^ras (RAS) is also essential for transformation in BCR-ABL1 leukaemic cell lines (Cortez et al, 1996), which in turn activates RAF serine kinase leading to activation of ERK and JNK. The activation of RAS also has multiple other effects on the transformed BCR-ABL1 expressing cancer cell, including the signalling through the pro-survival transcription factor nuclear factor κB (NFKB1) and the increased proteasome-dependent degradation of Abi-2 (a cell cycle regulatory protein). These events downstream of RAS activation lead to cell transformation and tumourigenicity (Zou & Calame, 1999). The constitutively active BCR-ABL1 cell also has the capability to maintain the pro-apoptotic protein BAD in a phosphorylated state, which keeps it from localizing to the mitochondria and thus impedes programmed cell death (Salomoni et al, 2000). A small molecule, imatinib mesylate, was developed to treat Ph+ CML by competing for the BCR-ABL1 tyrosine kinase ATP binding site, stabilizing the inactive conformation (Vajpai et al, 2008). In clinical trials, imatinib successfully halted the aberrant TK constitutive activity, which led to sustained clinical remissions (Druker et al, 1996). Following the initial success of imatinib with CML, it has since been studied in a small subset of paediatric patients with Ph+ ALL. Imatinib had insufficient sustained effect on Ph+ ALL as a single agent (Ottmann et al, 2002), though was found to have excellent initial response rates in combination with chemotherapy (Schultz et al, 2009). Results of clinical studies with imatinib in paediatric Ph+ ALL found a significantly improved three year EFS of 80%, more than twice that of historical controls (35%). This dramatic improvement was achieved without any additional toxicities (Schultz et al, 2009), though long-term relapse and survival data are not yet known.

Despite initial successes with imatinib, variations in the breakpoint of the BCR-ABL1 fusion gene and other de novo mutations of the TK gene have allowed some Ph+ leukaemias to be resistant to this first generation ABL1 inhibitor. Newer generations of ABL1 TK inhibitors have been developed in an attempt to overcome this resistance, and many have the ability to target additional pro-survival kinase families, including SRC TKs. These newer therapeutics provide multiple points of inhibition against the leukemic cell pro-survival machinery. Dasatinib, a second-generation ABL1 TK inhibitor, has been found to bind to the ABL1 kinase in both inactive (like imatinib) and activated conformations (Vajpai et al, 2008), thus it has a two log increased potency compared to imatinib (Shah et al, 2004). It is known to inhibit at least four additional tyrosine kinases including the SRC family, c-KIT, EPHA2 and PDGFβ receptors. Dasatinib has also been found to have in vitro activity against 14 of 15 imatinib-resistant leukemic cell lines (Shah et al, 2004), and is currently being tested in several phase I-III studies of paediatric Ph+ ALL (see Table I). Additionally, a multi-national paediatric phase III study comparing the effectiveness of imatinib versus dasatinib is currently being designed.

Table I.

Current targeted therapies for the treatment of pediatric ALL.

Class	Name	Target(s)	Form	Studies in pediatric ALL
ABL1 Inhibitors	Imatinib (STI571)	ABL1, cKIT	SM; oral	FHCRC 2223·00 phase I/II pilot study of imatinib and/or nilotinib after HSCT in Ph+ ALL or CML (not including T315I+)
				EUDRACT 2004-001647-30 Phase II/III in Ph+ ALL
	Dasatinib (BMS-354825)	ABL1, SRC, cKIT, EPHA2, PDGFb, etc.	SM; oral	COG AALL0622 phase III in Ph+ ALL
				BMS CA180-226 phase II study in new diagnosis Ph+ imatinib-resistant/intolerant ALL or AML
				KNSMH KC09MIMS0255 phase II in new diagnosis Ph+ ALL ≥15 years
				MDACC 2009-0521 phase I study In combination for relapsed/refractory malignancy (including ALL)
FLT3 Inhibitors	Nilotinib (AMN107)	ABL1, cKIT, etc.	SM; oral	See Imatinib above regarding FHCRC 2223·00
	Lestaurtanib (CEP-701)	FLT-3	SM; oral	COG AALL0631 phase III study in infant MLL-R ALL
	Midostaurin (PKC-412)	FLT-3	SM; oral	Novartis CPKC412A2114/EudraCT 2008-006931-11 phase I/II study in refractory/relapsed
				MLL-R ALL or FLT3 + AML
mTOR Inhibitors	Rapamycin (Sirolimus)	mTOR	SM; oral	Dana-Faber 06-249 in relapsed ALL
				WCIEU Sirolimus phase I study in relapsed/refractory ALL
				CHOP CHH-755 phase I study of relapsed/refractory ALL or NHL
	Temsirolimus (CCI-779)	mTOR	SM; IV	MDACC 2008-0384 phase I study in patient ≥12 years in advanced/metastatic malignancies (including ALL)
				MDACC 2008-0425 phase I study in advanced/metastatic malignancies (including ALL)
Aurora Inhibitors	MLN8237	Aurora A, ABL1 (3rd generation)	SM; oral	COG ADVL0812 phase I/II study in relapse/refractory leukaemias or solid tumours (including ALL) – temporarily closed to accrual
Multi-Kinase Inhibitors	Sorafenib (BAY439006)	RAF, FLT-3, VEGFR, PDGFR, cKIT, p38α	SM; oral	COG ADVL0413 phase I/II study in relapse/refractory leukaemias or solid tumours (including ALL) – temporarily closed to accrual
				SJCRH RELHEM phase I of relapsed/refractory ALL or AML
Proteasome Inhibitors	Bortezomib (PS-341)	26S ubiquitin proteasome	SM; IV	COG-AALL07P1 phase II pilot study in relapsed/refractory ALL or LL
				TACL T2005-003 phase I/II study in relapsed ALL
				MDACC 2008-0425 phase I study in advanced/metastatic malignancies (including ALL)
BCL2 Antagonists	Obatoclax (GX 15-070)	Pan-Anti-Apoptotic BCL2	SM; IV	COG ADVL0816 phase I study in relapsed/refractory leukaemia or solid tumours (including ALL)
Histone Deacetylase Inhibitors	Vorinostat (SAHA)	HDAC	SM; oral	UMN-2008LS112 phase II study in combination with decitabine for relapsed/refractory ALL or LL
DNA Methyltransferase Inhibitors	Decitabine	DNA Methyltransferase	SM; IV	EORTC protocol 06893 for patients ≥14 years in relapsed ALL or AML
				UMN 2008LS112 phase II study with vorinostat for relapsed/refractory ALL or LL
				MDACC 2005-0895 phase I in relapsed/refractory ALL
CD Marker Antibodies	Rituximab	CD20	mAb; IV	Kinderklinik B-NHL-BFM-Rituximab phase II with CD20 + ALL or NHL
	Epratuzimab	CD22	mAb, IV	COG ADVL04P2 phase II pilot study of relapsed CD22 + ALL
	Alemtuzumab	CD52	mAb, IV	UCSDCC 090331 phase I/II study of relapsed/refractory CD52 + ALL ≥16 years
				APHP P051003 phase I/II in refractory CD52 + ALL >15 years
Conjugated CD Marker Antibodies	CAT8015 (HA22)	CD22	mAb + PSA-E; IV	MedImmune CAT-8015-1004 phase I study of CD22 + ALL or NHL
	DT2219ARL	CD19, CD22	mAb + diphtheria toxin	SWCI S-WHITE-81714 phase I relapsed/refractory CD19 + & CD22 + ALL or LL ≥12 years
	BU-12	CD19	mAb + yttrium Y 90; IV	UMN-2006LS057 phase I study of relapsed/refractory CD19 + ALL or CLL

ALL, acute lymphocytic leukaemia; AML, acute myelocytic leukaemia; APHP, Assistance Publique Hopitaux de Paris; BMS, Bristol-Myers Squibb; CHOP, Children’s Hospital of Philadelphia; CHUR, Centre Hospitalier Universitaire de Rennes; CLL, chronic lymphocytic leukaemia; dg-A, deglycosylated ricin-A; EORTC, European Organization for Research and Treatment of Cancer; HSCT, hemato poietic stem cell transplant; IV, intravenous; KNSMH, Kang Nam Saint Mary’s Hospital; LL, lymphocytic lymphoma; mAb, Monoclonal antibody; MDACC, M. D. Anderson Cancer Center at University of Texas; MLL-R, MLL-rearrangement positive; NHL, non-Hodgkin lymphoma; Ph+, Philadelphia chromosome positive; PSA-E, pseudomonas endotoxin; SJRH, St. Jude Children’s Research Hospital; SM, small molecule; SWCI, Scott and White Cancer Institute; TACL, Therapeutic Advances in Childhood Leukemia Consortium; UCSDCC, Rebecca and John Moores UCSD Cancer Center; UMN, Masonic Cancer Center at University of Minnesota; WCIEU, Winship Cancer Institute of Emory University.

The mechanism by which many patients become resistant to imatinib is via a mutation in the kinase domain, and of particular concern is the T315I mutation, which confers resistance to early generation ABL1 kinase inhibitors. This resistant mutation is found more frequently and has a much more rapid emergence rate in Ph+ ALL compared to CML (Jones et al, 2008; Nicolini et al, 2009). Epidemiological studies have not yet clarified the percentage of T315I mutations in newly diagnosed Ph+ ALL, but it is clear that this mutation is more commonly associated with recurrent disease after previous TK inhibitor therapy (Jones et al, 2008) and is associated with lower overall and progression-free survival (Nicolini et al, 2009). The T315I mutation has been found to be resistant to dasatinib, and other ABL1 TK inhibitors including nilotinib (AMN107), INNO-406 (NS- 187). Recently, however, AP24534 has been introduced as an effective inhibitor of T315I and other BCR-ABL1 mutants in preclinical studies (O’Hare et al, 2009).

FLT3 receptor tyrosine kinase inhibition

Point mutations or internal tandem duplications (ITD) of the Fms-like Tyrosine Kinase (FLT3) are found in MLL-rearranged ALL, some T-cell ALLs, and high hyperdiploid ALL, as well as acute myeloid leukaemia (AML) and other malignancies. In infants with ALL, presence of MLL rearrangement and/or FLT3 aberrancies are associated with a pro-survival phenotype, resistance to multiple chemotherapeutic agents, and poor prognosis (Pui et al, 1994; Behm et al, 1996; Pieters et al, 1998; Armstrong et al, 2004; Ramakers-van Woerden et al, 2004; Stam et al, 2007). These mutations allow for a constitutively active FLT3 tyrosine kinase, which in turn activates the RAS/ RAF/ERK, PI3K/AKT/mTOR and signal transducer and activator of transcription-5 (STAT5) pro-survival pathways, resulting in uncontrolled cell proliferation and loss of normal apoptotic control. Because mutant FLT3 is present on some very highly chemoresistant leukemic cells, it is considered an excellent potential target. FLT3 inhibitors, such as lestaurtinib (CEP-701), an oral small molecule inhibitor of constitutively active FLT3, have been found to decrease activation of downstream pro-survival proteins such as ERK (Levis et al, 2002). FLT3 inhibitors were initially studied in AML; unfortunately they showed variable efficacy in newly diagnosed patients, though had more consistent results in relapsed patient samples and those with higher mutant allelic burden (Pratz et al, 2010). In vitro studies have shown that lestaurtinib could induce apoptosis in paediatric ALL cell lines that express high levels of FLT3 (Brown et al, 2005), and interacted synergistically with multiple chemotherapeutic agents (Brown et al, 2006). Because MLL-rearranged infant ALL is known to have increased FLT3 expression, current phase III clinical trials of lestaurtinib are being conducted on this subset (see Table I). Successful completion of these studies will provide proof of concept that FLT3 inhibitors can be reasonably administered and are well tolerated; future projects could potentially expand on the use of this target. Additional FLT3 inhibitors, including midostaurin (PKC-412), are currently being examined in an effort to improve selectivity and efficacy of targeting this receptor.

mTOR kinase inhibitors

As previously discussed, the pro-survival PI3K/AKT pathway is often upregulated in paediatric leukaemias. Specific targeting of this pathway has been attempted mainly by focusing on the downstream serine/threonine kinase mammalian target of rapamycin (mTOR) and the TOR complexes (TORC1 and TORC2), due to their central role in cell cycle regulation, cell survival, and proliferation (Hidalgo & Rowinsky, 2000). Up-regulation of mTOR results in dysregulation of cell cycle controls by phosphorylation of eukaryotic initiation factor 4E binding protein 1 (4EBP1) and ribosomal S6 kinase (S6K), both of which increase translation of cell proliferation and survival proteins (Schmelzle & Hall, 2000). Rapamycin and other inhibitors of mTOR prevent the translation of these proliferative and cell cycle transition proteins, halting the cell in the G₁ phase. Inhibition of mTOR has been found, both in vitro and in vivo, to increase killing of malignant cells because of the heavy reliance on its upregulation for survival. Inhibiting mTOR has been found to be effective in several types of cancer, including paediatric ALL (Brown et al, 2003; Avellino et al, 2005). Additionally, mTOR inhibitors were found to work synergistically with several traditional chemotherapeutics in vitro (Teachey et al, 2008). Clinical application of mTOR inhibitors, such as rapamycin and the second-generation mTOR inhibitor temsirolimus (CCI-779), are currently being evaluated in phase I clinical trials of paediatric ALL (see Table I). Two later generation mTOR inhibitors, everolimus (RAD-001) and ridaforolimus (AP23573), are currently being studied in non-hematopoietic paediatric malignancies, and are potentially useful therapies for paediatric ALL. More recently, innovative approaches to targeting this pathway, such as dual TORC1/2 inhibitors, have been shown in pre-clinical research to improve leukemic cell killing (Janes et al, 2010), giving hope for targeting of this pathway as a treatment strategy in paediatric ALL.

JAK tyrosine kinase inhibition

The Janus kinase (JAK) family of tyrosine kinases is activated by cytokine binding to a Type I cytokine receptor. Activation of JAK leads to phosphorylation of STAT, and subsequent activation of both the RAS/RAF and PI3K/AKT pathways, ultimately leading to cell proliferation. In ALL cell lines, members of this TK family are abundantly expressed, though JAK2 has been noted to be expressed more frequently than JAK1 or JAK3. In addition to frequent overexpression, several activating JAK mutations have been identified (Meydan et al, 1996; Mullighan et al, 2009a). Constitutively active JAK/STAT results in uncontrolled proliferation and has been associated with poor prognosis (Mullighan et al, 2009a). Activating mutations of JAK also correlate with other gene abnormalities, including deletion or mutation of the Ikaros Family Zinc Finger 1 gene (IKZF1) and genomic rearrangement involving the Cytokine receptor-like factor 2 gene (CRLF2) which results in its overexpression, both of which confer poor prognosis (Mullighan et al, 2009a; Yoda et al, 2010). IKZF1 codes for the IKAROS transcription factor, which is required for normal lymphocyte development. Deletion of IKAROS causes a dominant-negative effect, resulting in abnormal IKAROS subcellular compartmentalization in patient samples (Sun et al, 1999) and leading to leukemogenesis in mouse models (Winandy et al, 1995). CRLF2 is a subunit of the type I cytokine receptor, which forms a heterodimer with interleukin seven receptor (IL7R); cytokine binding to this receptor is known to stimulate B-cell proliferation. Rearrangements involving CRLF2 have been found to cause constitutive dimerization with IL7R, resulting in cytokine-independent activation of JAK2 and STAT5, subsequent B-cell proliferation, and possibly cell transformation, especially in the presence of a constitutively activated JAK mutation (Mullighan et al, 2009b). JAK inhibitors are currently in varying stages of development; clinically they have only been used in adult trials, and have not yet been assessed in paediatric patients. By targeting cells with activated JAK mutations, it may be possible to improve prognosis for patients with IKAROS mutations and CRLF2 overexpression because of the known high-frequency association of these abnormalities. JAK inhibitors may be especially useful in certain populations of pre-B ALL in which CRLF2 rearrangements are significantly more prevalent, for example patients with Down syndrome and Latino/Hispanic patients (Mullighan et al, 2009b; Harvey et al, 2010).

Aurora kinase inhibitors

Aurora serine/threonine kinases A and B have been found to regulate cell proliferation through control of mitosis. Aurora kinase A is known to regulate centrosome function and spindle assembly (Gautschi et al, 2008), in addition to providing G₂ → M transition checkpoint control (Marumoto et al, 2002). Aurora kinase B is known to regulate the chromosomal passenger complex, which has multiple effects on mitosis including segregation of chromotids, histone modifications, and cytokinesis (Vader et al, 2006). Increased expression of these kinases have been found in several human tumours and leukaemia cell lines (Zhou et al, 1998; Houghton et al, 2008a; Hardwicke et al, 2009), and are thought to cause amplification, chromosomal instability and transformation of cancer cells (Zhou et al, 1998). Aurora kinase A and/or B inhibitors have been tested in vitro and in vivo, and found to have additional inhibitory effects on several other kinases including the mutant ABL1 TK (including those cells with T315I mutation) and FLT3 (Gautschi et al, 2008). Currently, a phase I/II study of MLN8237, an Aurora A kinase small molecule inhibitor, is being conducted on paediatric patients with relapsed or refractory leukaemias and solid tumours. This treatment modality is expected to have pro-apoptotic effects in a multitude of paediatric cancers. Currently, several Aurora A, Aurora B, and pan-Aurora inhibitors are being investigated in adult trials, and may also have efficacy in paediatric patients, especially those compounds with activity against mutated BCR-ABL1 and/or FLT3 (Gautschi et al, 2008).

Other multi-kinase inhibitors

Some inhibitors target a broad range of kinases, thus preventing multiple pro-survival or anti-apoptotic signalling cascades. Sorafenib (BAY 43-9006) is a multi-kinase inhibitor that has been found to inhibit several receptor tyrosine kinases including VEGFR-2, FLT3, c-KIT, PDGFR, in addition to the serine/ threonine kinases p38α and RAF. By targeting multiple kinases, sorafenib blocks the pro-survival RAS/RAF/ERK pathway at several different points (Wilhelm & Chien, 2002), thus prevents compensatory upregulation of this pathway, which has been described as a mechanism of acquired resistance to more selective tyrosine kinase inhibitors. Additionally, sorafenib has been found to increase levels of the tumour necrosis factor-related apoptosis-inducing ligand (TRAIL), resulting in activation of caspases-3, -7, -8, and -9, and release of mitochondrial cytochrome-c (Rosato et al, 2007). Currently, phase I studies of sorafenib are underway in paediatric patients with a variety of relapsed and refractory malignancies, including ALL. It is hoped that the strategy of multi-kinase inhibition will circumvent the redundant pro-survival pathways activated in lymphoblasts, though it is unclear what long-term effects a multi-kinase inhibitor will have on normal cell function.

TAM tyrosine kinase inhibitors

The Tyro-3, Axl and Mer (TAM) family of receptor tyrosine kinases is aberrantly expressed in a variety of cancers (Linger et al, 2008). Specifically, Mer is expressed at high levels in T-and B-cell leukaemias (Graham et al, 2006; Linger et al, 2009), and not found on normal lymphocytes. Presence ofMer is associated with increased phosphorylation of AKT and ERK, and subsequent prevention of apoptosis (Guttridge et al, 2002). Up-regulation and overexpression of Mer in T- and B-cell ALL leads to increased survival pathway activation, through AKT and ERK (Keating et al, 2006; Linger et al, 2009). Both in vitro and in vivo studies have demonstrated that inhibition of Mer reduces survival of leukaemia cells. Furthermore, Mer inhibition interacts synergistically with chemotherapeutic agents in vitro (Linger et al, 2009). Currently, targeting Mer RTK is being investigated in preclinical studies of paediatric ALL cell lines, using both monoclonal antibody and small molecule TK inhibitors.

Other oncogenic targets

Proteasome inhibitors

Proteasomes work to degrade multiply ubiquitinated intracellular proteins, many of which are involved in cell maintenance, transcription, cell cycle control and apoptosis. In particular, proteasomes are known to degrade short-lived cell cycle regulatory cyclin proteins and the tumour suppressor protein p53 (Adams, 2004). Additionally, proteasomal degradation of IκB allows the transcription factor NFKB1 to translocate into the nucleus (Li et al, 1995) and activate the promoter region of several anti-apoptotic proteins (e.g. BCL2, XIAP, JNK) and inactivate the pro-apoptotic caspase-8 (Wang et al, 1998). NFKB1 is constitutively active in paediatric ALL patient samples, and is thought to promote survival by suppressing apoptosis (Kordes et al, 2000). Blocking proteasome function may reconstitute normal cell cycle regulation and apoptosis pathways by restoring levels of proteins, such as IκB and p53. Bortezomib (PS341) is a specific proteasome inhibitor targeting the 26S ubiquitin proteasome pathway, which performs the degradation of IκB, p53 and several other cell cycle control proteins. This agent works to increase cellular sensitivity to apoptosis and has demonstrated clinical efficacy in several adult haematopoietic cancers. More recently, it has been investigated as a potential targeted therapy for paediatric ALL both in vitro and in mouse xenograft models, which demonstrate its substantial anti-leukaemic effect both as a single agent and in conjunction with traditional chemotherapy agents (Horton et al, 2006; Houghton et al, 2008b). Currently, bortezomib is being assessed clinically in phase I/II studies of paediatric patients with relapsed ALL. Several subsequent generations of proteasome inhibitors have been developed and are currently being assessed in adult malignancy trials.

Farnesyltransferase inhibitors

As previously mentioned, the intracellular protein RAS is associated with pro-survival cytokine receptor signalling via RAF, MEK and ERK 1/2, and RAS activation has been observed in several paediatric leukaemias. Post-translational modification of RAS, including addition of a farnesyl isoprene group by farnesyltransferase, is required for localization of RAS to the cellular membrane and subsequent cell transformation (Willumsen et al, 1984; Schaber et al, 1990). Farnesyltransferase inhibitors are therapies in development that aim to decrease farnesylation of multiple intercellular proteins, including RAS. Blocking this enzyme is thought to decrease RAS activity, and its subsequent pro-survival downstream effects. Tipifarnib (R115777) is a farnesyltransferase inhibitor that was studied in a paediatric phase I trial of relapsed and refractory leukaemias, but only moderately decreased farnesyltransferase activity in ALL blasts, by 67% in comparison to 81% in AML (Widemann et al, 2003). In vitro studies suggest that T-cell leukaemias are much more sensitive to this medication than precursor B-cell leukaemias (Goemans et al, 2005). Though farnesyltransferase inhibitors are not currently being studied in clinical trials of paediatric ALL, novel pursuits in targeting RAS are likely in the near future because of its frequent overactivation in paediatric ALL.

Heat shock protein inhibitors

Heat shock proteins (HSP) are molecular chaperones for many intracellular proteins, and serve to properly refold proteins whose structure has been distorted (i.e. misfolded) due to a cellular stress event. If not properly refolded, these proteins become ubiquitinated, degraded, and can trigger apoptosis in the affected cell. Upregulation of HSPs occurs transiently in normal stressed cells, but proliferating leukaemia cells exhibit sustained elevation of HSP levels which may inhibit normal apoptotic stimuli (Strahler et al, 1991; Yufu et al, 1992). In addition to blocking the refolding function of HSPs, studies suggest that HSP inhibitors down-regulate the AKT and RAF (pro-survival) pathways (Jia et al, 2003; Nimmanapalli et al, 2003). In addition to their preclinical effectiveness in paediatric acute leukaemia cell lines (Hawkins et al, 2005), HSP90 inhibitors also have marked in vitro effects on both FLT3 and BCR-ABL1 positive leukaemias, including those with the imatinib-resistant T315I mutation (Yao et al, 2003; Peng et al, 2007). The HSP90 inhibitor tanespimycin (17-AAG) has been evaluated in two phase I trials of paediatric solid tumours. Unfortunately, soluble preparation requires substantial dimethyl sulphoxide, which limits it maximum dosing and thus its potential efficacy (Bagatell et al, 2007; Weigel et al, 2007). Alvespimycin (17-DMAG), a more potent and more water-soluble HSP90 inhibitor, has been developed (Smith et al, 2005) but was demonstrated to have only moderate anti-leukaemic activity in preclinical trials (Smith et al, 2008). Future generations of HSP90 inhibitors, which are currently being evaluated in adult trials, are thought to be more useful in combination with other therapies (e.g. proteasome inhibitors, tyrosine kinase inhibitors, etc.), rather than as monotherapy. While only HSP90 inhibitors have been evaluated in clinical trials to date, it is likely that HSP27 and/or HSP70 inhibitors will see further development as potential therapies in ALL, as they are found to be overexpressed in many refractory or relapsed ALL cells lines (Thomas et al, 2005).

Gamma secretase inhibitors

NOTCH is a transmembrane heterodimeric receptor that, when sequentially cleaved by an ADAM metalloproteinase, then γ-secretase, releases the intracellular domain Notch1 (Six et al, 2003). Activated Notch1 translocates to the nucleus and associates with DNA-binding proteins, such as the CSL family of transcription factors. This complex functions as a transcription activator that regulates T-cell development in normal cells (Radtke et al, 1999), and has been shown to activate transcription of genes such as MYC and NFKB1 (Palomero et al, 2006; Vilimas et al, 2007). Mutations in the NOTCH receptor have been found in T-cell ALL and can result in ligand-independent cleavage of Notch1 (Ellisen et al, 1991). One study reported that NOTCH mutations were found in over 50% of T-cell ALL cases (Weng et al, 2004), and an in vivo study showed that mice transplanted with cells carrying defective NOTCH genes developed T-cell ALL (Pear et al, 1996). Inhibition of γ-secretase is thought to prevent release of Notch1 from the transmembrane receptor, thereby decreasing the viability of T-cell ALL. Currently, γ-secretase inhibitors are under development, though in animal models and phase I clinical trials the use of these inhibitors has been limited by severe gastrointestinal toxicity (Milano et al, 2004; Deangelo et al, 2006). There is evidence that this toxicity may be eliminated by concomitant use of glucocorticoids, which also seem to increase the anti-leukaemic effects of γ-secretase inhibitors (Real et al, 2009). Second generation γ-secretase inhibitors, which exhibit decreased toxicity, are currently being evaluated in clinical trials of adult malignancies.

Securin

Securin, also known as pituitary tumour-transforming factor 1 (PTTG1), is a protein involved in the process of cell division. Securin regulates sister chromatid segregation; abnormal expression of securin is thought to cause irregular chromosomal separation and result in chromosomal loss or gain, and subsequent cell transformation (Zou et al, 1999). Up-regulation of securin has been found in a number of human malignancies, including ALL (Saez et al, 2002). In a screening study of paediatric ALL patient samples, securin was found to be significantly up-regulated at the time of relapse compared to at diagnosis (Bhojwani et al, 2006) and thus targeting securin would probably serve as a much needed treatment for relapsed paediatric ALL. Inhibition of securin is currently in preclinical stages of development, and is hoped to have clinical applicability in paediatric relapsed ALL.

Targeting apoptotic pathways

BCL2 antagonists

Another approach to targeting the apoptotic pathway of the leukemic cell is to block certain members of the BCL2 family, which are central to apoptosis prevention. Pro-apoptotic molecules (e.g. Bax, Bak, etc.) are known to induce permeabilization of the mitochondrial outer membrane causing cytochrome-c release and subsequent initiation of apoptosis via the apoptosome and caspase-9. Several members of the BCL2 family of proteins serve to protect the mitochondrial membrane from the pro-apoptotic actions of Bax and Bak, thereby preventing apoptosis. Up-regulation of BCL2 proteins has been observed in leukemic cells compared to normal B-cell precursors, which contributes to delayed apoptosis (Campana et al, 1993) and correlates with chemoresistance in relapsed paediatric leukaemia (Haarman et al, 1999). Inhibition of BCL2 has been approached in a variety of ways, including direct binding of a small molecule inhibitor to the BH3 binding pocket of the molecule, or by preventing its expression with an anti-sense mRNA binding oligonucleotide. In preclinical trials of paediatric ALL cell lines, BCL2 small molecule inhibitors were effective initiators of apoptosis (Lock et al, 2008). Currently the pan-BCL2 family small molecule inhibitor Obatoclax (GX15-070) is undergoing a phase I study on paediatric patients with relapsed or refractory leukaemia. G3139, the anti-sense oligonucleotide which binds to the first six codons of the human BCL2 mRNA, is currently undergoing phase I paediatric trials in patients with solid tumours. G3139 decreases BCL2 expression in vitro resulting in increased apoptosis in ALL cell lines (Szegedi et al, 2008), though clinical efficacy has not yet been assessed.

TRAIL receptor agonists

A different approach to targeted killing of malignant cells uses the apoptotic cellular pathways to trigger cell death. Activation of Death Receptors 4 (DR4 or TRAIL-R1) and 5 (DR5 or TRAIL-R2) by ligand binding has been found to induce apoptosis via caspase-8 and -10 activation (Kischkel et al, 2001), which leads to activation of caspase-3 and resultant degradation of cellular contents (Stennicke et al, 1998). The TRAIL (or Apo2L) family of cytokines binds to DR4 and DR5. Interestingly, exposure to TRAIL has been found to cause increased apoptosis of malignant cells, without causing significant toxicity to normal cells (Ashkenazi et al, 1999). Recombinant TRAIL, TRAIL receptor monoclonal antibody (mAb) agonist, and TRAIL small molecule agonists have been developed. The TRAIL mAb has been found to have a longer half-life compared to native or recombinant TRAIL (Ashkenazi et al, 1999; Pukac et al, 2005). This difference in stability is probably because of the specificity of the mAb to the DR4 and DR5 receptors, whereas native and recombinant TRAIL are known to bind to several decoy DR receptors. Recently, a specific DR4 antibody, mapatumumab (HGS-ETR1), was studied preclinically, though it showed limited efficacy in ALL (Smith et al, 2010). A TRAIL mAb specific to the DR5 receptor, lexatumumab (ETR2-ST01), is currently being tested in paediatric solid tumours and lymphoma. Additional compounds which target this pathway are currently in laboratory development, and may be more efficacious in treating paediatric ALL.

Survivin inhibitors

Survivin is a member of the inhibitor of apoptosis protein (IAP) family. It is expressed in fetal tissues, but in normal somatic cells is expressed only transiently during the G2 → M phase of the cell cycle (Adida et al, 1998; Li et al, 1998). Survivin binds to microtubules of the mitotic spindle, which results in inhibition of caspases-3 and -7, and prevents apoptosis (Li et al, 1998; Tamm et al, 1998; Shin et al, 2001). Survivin is overexpressed in many malignancies, including paediatric pre-B cell ALL, where it has been found to be an independent risk factor for early relapse (Troeger et al, 2007). Overexpression of survivin is thought to override apoptotic G2 → M checkpoint control, and allow for aberrant progression of transformed cells through mitosis. Additionally, elevated levels of survivin have been found to decrease apoptotic effects of chemotherapy in vitro (Tamm et al, 1998). Inhibition of survivin by use of small interfering RNA (siRNA) has been found to increase apoptosis in ALL cell lines (Zhu et al, 2008). This approach has also been used in vitro with several other malignancies that express high levels of survivin, but clinical applicability of siRNA inhibitors may be limited due to concerns of abnormal viral insertion into the human genome. Currently, small molecule inhibitors of survivin are in early phase clinical studies of several adult malignancies, but have not yet been tested in paediatrics.

Epigenetic targets

Histone deacetylase inhibitors

In the previous sections, we have discussed targeted therapeutic strategies that involve blocking activation of pro-survival or anti-apoptotic molecules. An alternative approach is also under investigation whereby therapeutics are designed to restore expression of tumour suppressor genes. Histone deacetylases (HDACs) are enzymes that remove acetyl groups from histones, increasing the affinity of histones for DNA, resulting in chromatin compaction, decreased access of transcription factors to promoter regions, and decreased protein production (Lane & Chabner, 2009). Inhibition of HDACs has been found to augment expression of several cell cycle control proteins such as p21^WAF1/CIP1, ultimately leading to cell cycle arrest (Sandor et al, 2000). HDAC inhibitors (HDACi) also increase transcription of apoptotic proteins, such as TRAIL and DR5, in leukemic cells, but not in normal haematopoietic progenitors (Insinga et al, 2005). Therefore, inhibition of HDACs in leukaemia cells decreases proliferation and increases apoptosis. In vitro studies of HDACi on ALL cell lines have shown anti-leukemic effects (Bernhard et al, 2001; Tsapis et al, 2007; Scuto et al, 2008). The HDACi vorinostat (SAHA) has been explored in paediatric preclinical trials, but showed no appreciable anti-leukaemia effect as a single agent (Keshelava et al, 2009). Despite this finding, HDACi are thought to have anti-leukemic effect in combination with certain traditional pan-cytoxic chemotherapeutics (Kano et al, 2007; dos Santos et al, 2009), and it has been proposed that HDACi might be best used in combination with other targeted anti-neoplastic agents, such as proteasome inhibitors, HSP inhibitors, or DNA methyltransferase inhibitors, rather than as monotherapy. In vitro studies suggest that concomitant use of HDACi with imatinib in BCR-ABL1-positive leukaemias may increase blast apoptosis (Kano et al, 2007). Currently, use of vorinostat in combination with etoposide and the DNA methyltransferase inhibitor decitabine is being evaluated in paediatric patients with refractory and relapsed ALL or lymphoblastic lymphoma.

DNA methyltransferase inhibitors

Methylation of DNA promoter CpG islands promotes increased histone binding and chromatin compaction, which inhibits binding of transcription factors to promoters. Aberrant hypermethylation of the promoter regions of several proapoptotic or cell cycle control genes has been found in ALL, and the degree of aberrant methylation may portend a poor prognosis (Roman-Gomez et al, 2007; Stumpel et al, 2009). DNA methyltransferase inhibitors prevent hypermethylation of promoter CpG islands and rescue normal expression of tumour suppressor genes. Clinical trials of the DNA methyltransferase inhibitor 5-azacytidine (azacitadine) has been assessed in adult malignancies, but not in paediatric ALL. The more potent 5-aza-2′-deoxycytidine (decitabine) is currently being evaluated in several early phase studies in paediatric ALL, and has been associated with successful remission status in a case report of a paediatric patient with multiply relapsed ALL (Yanez et al, 2009).

Non-oncogenic surface targets

Cell surface marker antibodies

Leukaemia and other malignant cells often express unique cell surface proteins that do not contribute to oncogenesis, but can be used to selectively target malignant cells. For example, a diverse group of proteins called cluster of differentiation (CD) surface markers are expressed at high levels on various types of cancer cells. The use of monoclonal antibodies (mAb) to target CD markers on malignant cells has been proven effective in the treatment of many leukaemias and lymphomas (e.g. the anti- CD20 mAb rituximab). CD22 is a marker expressed on normal mature B lymphocytes whose main function is intracellular signalling and B-cell activation (Sato et al, 1998). CD22 is also expressed in over 95% of pre-B leukaemias (Gudowius et al, 2006). Several anti-CD22 mAb (e.g. epratuzimab) have been developed that cause internalization of the CD22 receptor resulting in cytotoxicity. A pilot study of epratuzimab plus standard re-induction chemotherapy in paediatric CD22− positive ALL demonstrated a favorable early response in the majority of patients (Raetz et al, 2008). This agent is now being evaluated in a larger phase II study in paediatric patients with relapsed CD22-positive ALL.

Several other mAbs are in various stages of development that target non-oncogenic surface receptors, which are commonly seen on lymphoblastic leukaemia cells. The anti-CD52 mAb, alemtuzumab, has been assessed in phase I/II studies of CD52+ ALL, though to date its use has been limited because of its known short-term side effects including fever, allergic reaction and hypotension, and failure to appreciably induce remission as a single agent (Angiolillo et al, 2009).

Conjugated antibodies

The use of mAbs to target malignant cells expressing a certain CD marker may not be an effective treatment strategy for all types of ALL because not all leukaemic cell markers cause increased apoptosis as a result of antibody binding. Some groups have adapted this approach to deliver cytotoxic agents, such as antibiotics, bacterial exotoxin, or radioisotopes to the leukaemic cells. In the treatment of AML, this method was applied in the development of gemtuzimab ozogamicin, an anti-CD33 mAb conjugated to the antibiotic calicheamicin, which results in DNA cleavage once internalized by the targeted cell. In ALL, several similar models of calicheamicin linked to mAb exist, including anti-CD22 (CMC-544 or inotuzumab ozogamicin) and anti-CD19, which have been shown to have significant anti-leukaemic effects in mouse models (Dijoseph et al, 2007; Bernt et al, 2009). Conjugation of other toxins to CD marker mAb, are also being explored in paediatric ALL in both preclinical and early clinical stages. An example of this technique is seen with CAT 3888 (BL22), an anti-CD22 mAb linked to the Pseudomonas exotoxin PE38, which is endocytosed and blocks protein translation resulting in cell death (Kreitman & Pastan, 2006). CAT 8015 (HA22) is a second-generation anti-CD22+ PE38 mAb that works by the same mechanism of action, but exhibits greater affinity for the CD22 receptor (Bang et al, 2005); this molecule is currently under investigation in a paediatric phase I trial. Another agent with potential in paediatric ALL treatment is Combotox, a dual anti-CD22 and anti-CD19 molecule linked to the immunotoxin deglycosylated ricin-A (RFB4-dgA and HD37-dgA). In a recent phase I trial, this therapy was found to reduce tumour burden in paediatric patients with refractory ALL (Herrera et al, 2009). Additional combinations of immunotoxins or radioisotopes conjugated to single or combination mAb are being evaluated in both preclinical and early phase clinical studies.

BiTE antibodies

Building on the concept of using mAbs to target leukaemic cells, some groups have developed a bispecific mAb, in which two different single-chain mAb are covalently linked. This dual mAb is designed to bind the targeted leukaemic cell with one arm, and a T-cell receptor with the other, bringing the malignant cell and the T-cell in close proximity, ultimately resulting in T-cell mediated cytolysis of the targeted leukemic cell (Haagen et al, 1992). These constructs have been coined bispecific T-cell engager (BiTE) single-chain antibodies, and have been shown to successfully trigger the signalling cascade of the T-cell receptor complex (Schlereth et al, 2006) causing T-cell mediated malignant cell destruction (Anderson et al, 1992; Haagen et al, 1992). Of potential use in paediatric ALL, a BiTE antibody composed of an anti-CD19 antibody and an anti-CD3 antibody (i.e. blinatumomab or MT103) has been developed and is currently being evaluated in adult early phase trials. This molecule, in combination with a co-stimulatory CD28 antibody, has been tested in vitro on paediatric pre-B leukaemia bone marrow samples with noted T-cell mediated cytolysis of the malignant B-cells (Manzke et al, 1999). Initial phase I studies in adults showed that this treatment caused reversible central nervous system-related effects (e.g. confusion, speech disorders, tremors, etc.), though this has not yet been observed in the on-going adult phase II trials (Nagorsen et al, 2009). Several groups are developing further variations of this concept, which have the potential to more potently target or more directly kill malignant cells by using a variety of malignant cell and T-cell receptors.

Chimeric T-cell receptors

Using the immune system to trigger T-cell cytolytic response to leukaemia has been attempted using another approach conceptually similar to BiTE antibodies, though different in that the T-cell itself expresses the anti-tumour antibody. Virally transduced T-cells are made to express chimeric receptor constructs which have an extracellular domain (composed of the variable chains of a monoclonal antibody) specific to binding the leukaemic cell, and an intracellular cytolytic signalling domain. When these chimeric receptors are expressed in T-cells, binding of leukaemic cells to the extracellular domain causes direct cytolytic killing. An example that has been shown to cause in vitro apoptosis of pre-B cell ALL is a receptor construct in which an anti-CD19 composes the external domain, and a T-cell receptor which induces expansion and cytolytic response (e.g. CD3 or 4-1BB) composes the intracellular domain (Cooper et al, 2003; Imai et al, 2004). Similar chimeric T-cell receptors have been assessed, and are thought to have clinical potential in paediatric ALL. Specifically, these virally transduced T-cells may have optimal use in reducing drug-resistant leukaemias and/or minimal residual disease after successful induction of remission, or haematopoietic stem cell transplantation because of the graft-versus- leukaemia effect. Similarly, natural killer cells transduced with chimeric receptor constructs have been found to cause increased killing of pre-B cell ALL blasts (Imai et al, 2005). Though initial killing of CD19 expressing lymphocytes aids in tumour lysis during or around the time of therapy, prolonged anti-CD19 effects are not desirable. The addition of a suicide gene has been explored in which plasmid expression of anti-CD19 receptor constructs can be turned off (Serrano et al, 2006). With the use of gene transfer therapy, there is always the theoretical risk of inducing new malignancies or other genetic aberrations due to abnormal insertion of the viral DNA into the human genome. Further optimization will probably be needed before this therapy is used clinically in paediatric patient trials.

Conclusion

With each of these promising therapies, there are many challenges to overcome. With imatinib and other therapies, there is the issue of resistance, either intrinsic to the leukaemic cell or acquired as a result of treatment. Additionally, inhibiting a single cellular protein may not completely reverse pathological alterations to the cell’s normal auto-regulatory pathways. Cross-talk between intracellular pathways may lead to compensatory upregulation of a related pathway, causing persistence of leukaemic proliferation by promotion of other aberrant signalling receptors, growth factors, etc. Of concern with cell surface receptor targeted therapies is the likelihood of altered cell surface receptor expression after exposure to a therapy. This has previously been observed in relapsed lymphomas, which no longer express CD20 after treatment with rituximab, and relapsed AML, which no longer express FLT3 after exposure to lestaurtinib. Because of these potentially redundant oncogenic mechanisms, there is concern that targeted therapy could select for high resistance.

Though there is hope that targeted therapeutics limit effects on normal cells, many of these therapies have some level of activity against normal somatic cells, thus have their own list of potential side effects. These effects, such as those previously discussed with gamma-secretase inhibitors on the gastrointestinal tract, may be short-lived, but of more concern are the potential long-term side effects that may not be elucidated for several years after initial use. We are just beginning to understand some of the long-term side effects of therapies such as imatinib and dasatinib, including prolonged immunocompromise, growth suppression or arrest, and cardiac dysfunction. Side effects such as these may be seen in both paediatric and adult patients, though paediatric patients may have additional unique side effects related to normal development, growth and maturation, such as delayed or permanently reduced growth velocity. As time passes, we may begin to see even more long term effects that were not previously anticipated from these medications.

Despite the possible secondary effects of targeted therapies, these novel treatments have great potential. Research in the past several years has greatly increased our understanding that paediatric ALL is a heterogeneous group of cancers, containing different genotypic and phenotypic signatures. This greater understanding has allowed development of novel treatments based on the exact specifications of the disease, such as treating pre-B cell leukaemia with an anti-CD22 mAb, treating MLL-rearranged leukaemia with a FLT3 inhibitor, or treating T-cell ALL with a gamma-secretase inhibitor. With high throughput sequencing and array technology, we will continue to improve our understanding of paediatric ALL, which will permit us to systematically refine our ability to develop new therapies focused on the precise aberrancies of the leukaemic cell while bypassing effects on normal somatic cells.

The development and eventual implementation of novel targeted therapies is necessary for future improvements in outcome for paediatric ALL patients. As presented here, there are many approaches to specific targeting of leukaemia cells with minimal toxicity to normal cells, including targeting aberrantly up-regulated cellular survival pathways, modulating gene expression, or using unique proteins expressed on the cell surface to deliver cytotoxic molecules to cancer cells. As these therapies continue to undergo refinements, the eventual goal is that their specificity for leukaemia cells will allow us to reduce or eliminate the use of pan-cytotoxic chemotherapeutic agents, thereby decreasing both short- and long-term side effects of paediatric ALL treatment.