Advances in paediatric cancer treatment

Abstract

Four out of five children diagnosed with cancer can be cured with contemporary cancer therapy. This represents a dramatic improvement since 50 years ago when the cure rate of childhood cancer was <25% in the pre-chemotherapy era. Over the past ten years, while improvement in overall survival (OS) has been marginal, progress in pediatric oncology lies with adopting risk-adapted therapeutic approach. This has been made possible through identifying clinical and biologic prognostic factors with rigorous research and stratifying patients using these risk factors, and subsequently modifying therapy according to risk group assignment. This review provides a perspective for eight distinct pediatric malignancies, in which significant advances in treatment were made in the last decade and are leading to changes in standard of care. This includes four hematologic malignancies [acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), non-Hodgkin lymphoma (NHL) and Hodgkin lymphoma (HL)] and four solid tumors [medulloblastoma (MB), low grade glioma (LGG), neuroblastoma (NB) and Ewing sarcoma (ES)]. Together, they comprise 60% of childhood cancer. Improved patient outcome is not limited to better survival, but encompasses reducing both short and long-term treatment-related complications which is as important as cure, given the majority of childhood cancer patients will become long-term survivors. Risk-adapted approach allows treatment intensification in the high-risk cohort while therapy can be de-escalated in the low-risk to minimize toxicity and late sequelae without compromising survival. Advances in medical research technology have also led to a rapid increase in the understanding of the genetics of childhood cancer in the last decade, facilitating identification of molecular targets that can potentially be exploited for therapeutic benefits. As we move into the era of targeted therapeutics, searching for novel agents that target specific genetic lesions becomes a major research focus. We provide an overview of seven novel agents (bevacizumab, bortezomib, vorinostat, sorafenib, tipifarnib, erlotinib and mTOR inhibitors), which have been most frequently pursued in childhood cancers in the last decade, as well as reporting the progress of clinical trials involving these agents.

Introduction

Cancer in children is rare with an incidence of 140-155 per million per year (age <15 years) (1,2) (Table 1). This translates to ~1 in 7,000 children is diagnosed with cancer each year. For the 15 to 19 years age group, the incidence is 210 per million with a different distribution of cancer diagnosis. Despite the rarity of cancer, malignant neoplasm is the most common cause of death after accidents in children aged 5 to 14 years, accounting for 23% of mortality (7).

Table 1. Frequency, incidence, and survival of childhood cancer.

Diagnosisa	Frequencyb (% of all cancer)	Incidencec per million	Peak age (incidence per million)	5-year survival (%)d
				Year of diagnosis
				1975-1979	1992-1996	2002-2006
Age 0-14 years
All cancers	100	155		62	79	83
Acute lymphoblastic leukemia (Ib)	25.8	40.0	2-3 (95.3)	63	86	90
Acute myeloid leukemia (IIb)	5.1	7.8	0-1 (19.4)	24	50	69
Non-Hodgkin lymphoma (IIb/c)	5.8	9.0	12-14 (10.5)	50	82	89
Hodgkin lymphoma (IIa)	3.7	5.8	13-14 (16.7)	84	94	97
Central nervous system (III)	21.0	32.8	1-4 (41.8)	60	71	73
Medulloblastoma (IIIc.1)	2.8	4.5	1-7 (5.9)	50	63	70
Low grade gliomae	~8.4	~13.1		87	92	94
Neuroblastoma (IVa)	6.7	10.2	0-1 (42.7)	50	70	76
<1 year	2.2			86	90	91
1-14 years	4.4			37	57	69
Ewings’ sarcoma (VIIIc, IXd.1/d.2)	2.0	3.1	12-14 (5.5)	53	74	76
Osteosarcoma (VIIIa)	2.6	4.0	13-14 (9.3)	42	71	72
Rhabdomyosarcoma (IXa)	3.4	5.3	1-5 (7.6)	55	73	68
Wilm’s tumor (VIa.1)	4.6	7.2	1-3 (19.1)	77	93	91
Age 15-19 years
All cancers	100	211		70	79	83
Acute lymphoblastic leukemia (Ib)	8.3	17.5		34	57	70
Acute myeloid leukemia (IIb)	4.6	9.6		23	40	52
Non-Hodgkin lymphoma (IIb/c)	8.2	17.4		47	77	80
Hodgkin lymphoma (IIa)	14.4	30.3		91	95	96
Central nervous system (III)	9.6	20.3		61	76	77
Medulloblastoma (IIIc.1)	0.8	1.6		na	na	71
Low grade gliomae	~3.8	~8.1		73	88	97
Neuroblastoma (IVa)	0.2	0.4		na	na	na
Ewings’ sarcoma (VIIIc, IXd.1/d.2)	2.7	5.7		24	51	61
Osteosarcoma (VIIIa)	4.0	8.4		54	64	62
Rhabdomyosarcoma (IXa)	1.6	3.4		na	46	48
Wilm’s tumor (VIa.1)	0.1	0.3		na	na	na

^a, the Roman numerals in parentheses represent the International Classification of Childhood Cancer (ICCC) site recode extended ICD-0-3/WHO 2008; ^b, frequency: SEER 18 registry (3); year of diagnosis: 2001-2010; % of all cancer within the age group; ^c, age-adjusted incidence: SEER 18 Registry (1); ^d, survival: SEER 18 Registry (3); na, not applicable as cohort too small (n<40); ^e, low grade glioma: frequency/incidence based on 40% of all CNS tumor (4-6); survival estimated from astrocytoma (IIIb) and other gliomas (IIId) of WHO grade I or II.

The introduction of chemotherapy in the 1960s has allowed the use of multi-modality approach, i.e., in conjunction with surgery and radiation, in the treatment of cancer. As a result, survival from childhood cancers, many of which was fatal in the pre-chemotherapy era, has increased dramatically from 20-30% in the 1960s (8) to 62% in the 1970s (mortality of 4.5/100,000) (3,9) (Tables 1 and 2). The current mortality is 2.6/100,000 and survival is 83%, meaning modern medicine can cure four out of five children with cancer. This is the end result of the tremendous effort dedicated to pediatric cancer research in the last 50 years.

Table 2. Cancer deaths by age group.

Diagnosisa	0-14 years			15-19 years
	Cancer deaths per 100,000b	%		Cancer deaths per 100,000	%
All cancers	2.6	100		3.5	100
Acute lymphoblastic leukemia (Ia)	0.44	16.9		0.55	15.8
Acute myeloid leukemia (Ib)	0.26	10.1		0.40	11.4
Non-Hodgkin lymphoma (IIb/c)	0.09	3.6		0.28	7.9
Hodgkin lymphoma (IIa)	0.01	0.6		0.08	2.2
Medulloblastoma (IIIc.1)	0.15	5.7		0.05	1.5
Low grade gliomac	0.02	0.8		0.02	0.7
Neuroblastoma (IVa)	0.25	9.7		0.02	0.6
Ewings’ sarcoma (VIIIc, IXd.1/d.2)	0.05	2.1		0.24	6.7

^a, the Roman numerals in parentheses represent the International Classification of Childhood Cancer (ICCC) site recode extended ICD-0-3/WHO 2008; ^b, SEER 18 Registry (9); year of death: 2005-2009; ^c, LGG: astrocytoma (IIIb) and other gliomas (IIId) of WHO grade I or II.

One of the challenges of pediatric cancer research is the small disease population, comparing with adult cancer that is 40 times more frequent. To overcome this obstacle, multi-center clinical trials are essential to generate statistically meaningful results. For example, the Children Oncology Group (COG) represents the world’s largest organization devoted exclusively to childhood cancer research and comprises >200 member institutions, majority of which are based in the US. However, even with multi-center set up, it often takes five years to complete a phase III randomized controlled trial (RCT). Furthermore, it takes another six years to formally publish mature 5-year survival data (10-12). The CCG-1991 study enrolled 3,054 acute lymphoblastic leukemia (ALL) patients from 109 institutions between 2000 and 2005 and detected a <5% difference in survival. Results were published in 2011 (10). The HIT-SIOP PNET 4 Trial enrolled 340 children with medulloblastoma (MB) from 122 European centers between 2001 and 2006 and published the result in 2012 (12). It often takes over a decade for any progress established from clinical trials to become standard of care, and this has not taken into consideration the preceding ten years expended in preclinical and phase I/II studies.

We reviewed eight pediatric malignancies, in which significant advances in treatment were made in the last ten years, and most importantly are leading to improved standard of care. This included four hematologic malignancies and four solid tumors, comprising 60% of childhood cancer. The progresses were disseminated in published literature in the last decade and reflect clinical trials conducted between mid-1990 and mid-2000. We also explored ongoing studies with attention to trials conducted by COG, which have been developed based on knowledge gained in preclinical and early clinical studies in the last decade and discussed some of the more promising molecular targets for each of the eight cancers. Finally, we reviewed seven novel agents that have been most frequently pursued in childhood cancers.

Acute lymphoblastic leukemia (ALL)

Childhood leukemia has an incidence of 490 per million, of which 80% are ALL making it the commonest childhood cancer, peaking at 2 to 3 years old (1). With contemporary chemotherapy protocols, 5-year survival approaches 90% (Table 1). However, infant ALL with MLL/11q23 rearrangement have significantly worse outcome (EFS <30%) (13). The following is an overview of the recent advances in ALL treatment.

Minimal residual disease (MRD)

The rapidity of clearance of leukemic cells from the bone marrow is a strong prognostic factor (14,15) and is best evaluated by MRD using quantitative flow cytometry or polymerase chain reaction (PCR) assay of immunoglobulin and T-cell receptor rearrangements. These techniques are sensitive to 1×10^–5. After two decades of research, MRD is now widely used in risk stratification and provides a validated early measure of treatment response. MRD response to induction is best measured at day 15 and day 33 for pre-B ALL and at day 79 for T-cell ALL (16). Clinical trials currently evaluating treatment modifications based on MRD include the St Jude TOT XVI, AIEOP-BFM ALL 2009, DFCI-05001, COG-ALL0331, ALL-REZ BFM 2002.

Genome-wide analysis and targeted therapy

The collaborative TARGET research program has identified several key aberrations associated with high-risk ALL phenotypes, including BCR/ABL1, JAK, MLL, CRLF2 and IKZF1 (17). JAK kinases activation is present in 10% of BCR/ABL1-negative high-risk cases (18).

One of the best examples of targeted therapy is the use of tyrosine kinase inhibitors (TKI) to complement chemotherapy in BCR/ABL-positive ALL, which accounts for 3% of childhood ALL. Continuous exposure to imatinib in an intensive chemotherapy regimen yielded 3-year event-free survival (EFS) of 80%, more than twice that of historical controls (19). Second generation TKIs (dasatinib and nilotinib) with more potent BCR/ABL signaling suppression can overcome imatinib-resistance (20) and are being studied in a number of pediatric ALL studies. A COG phase III RCT of FLT inhibition with lestaurtinib (TKI) with chemotherapy is currently underway for MLL-rearranged infant ALL, which has been shown to express high levels of FLT3 mRNA (21) (NCT00557193). Lestaurtinib, which has also been shown to inhibit JAK2 (22,23), may potentially be active in JAK-mutated ALL.

Novel therapies

New formulations of the old armamentarium have been developed to improve delivery and reduce toxicities and can potentially be used in ALL therapy (Table 3). Continuous asparagine depletion has been associated with better EFS than intermittent depletion and a lower incidence of CNS relapse (28). Pegylated (PEG) asparaginase, approved for ALL treatment, has a long half-life and lowers the risk of allergic reactions and anti-asparginase antibody formation while maintaining efficacy similar to conventional E. coli asparaginase (24).

Table 3. New drug formulations for ALL therapy.

Drug	Benefits
Pegylated asparaginase	Long half-life, reduced immunogenicity (24)
Liposomal daunorubicin	Decreased cardiotoxicity (25)
Liposomal annamycin	Decreased cardiotoxicity (NCT00430443)
Sphingosomal vincristine	Decreased neuropathy, high tissue concentration, non-vesicant (26)
Liposomal cytarabine	Long half-life (27) (NCT00002704, NCT00991744)

Second generation nucleoside analogues are part of a new repertoire of drugs against ALL. Nelarabine shows specific anti-leukemic effect in T-cell ALL (29) and is currently being evaluated in newly diagnosed (phase III: NCT00408005) and refractory T-cell ALL (phase IV: NCT00866671). Clofarabine is being studied in both de novo and recurrent disease (NCT00372619, NCT01228331). Furthermore, the Interfant-06 Study Group is conducting a RCT (NCT00550992) investigating the novel addition of AML-type therapy during intensification for MLL-rearranged infant ALL.

CNS-directed treatment

Intensification of systemic and intrathecal chemotherapy abrogates the need for prophylactic cranial irradiation without compromising survival (30,31). This removes the long-term side effects of radiotherapy-induced secondary malignancies and neurocognitive harm. The isolated CNS relapse rates were <3%. Risk factors for CNS relapse included t(1;19)/TCF3-PBX1, CNS involvement at diagnosis and T-cell immunophenotype.

Immunotherapy

Monoclonal antibodies are designed to potentiate chemotherapy, particularly in the setting of relapsed leukemia (32,33). This includes rituximab (anti-CD20), alemtuzumab (anti-CD52), epratuzumab (anti-CD22), inotuzumab ozogamicin (anti-CD22) for B-cell ALL. Blinatumomab (CD19/CD3-bispecific antibody) are particularly promising for T-cell ALL (34). Chimeric antigen receptor T cell (CAR-T cell) is an investigational novel approach to relapsed/refractory leukemia. Autologous T cells, which are genetically modified ex vivo to express a chimeric receptor that recognizes a surface antigen on the patient’s own tumor cells, home to the disease sites and persist with time (35). CD19-specific CAR-T cell therapy has been used with success in adult relapsed CLL (36) and has recently been used in two children with refractory ALL, albeit with toxicities (37). Killer-cell immunoglobulin-like receptor (KIR)—mismatch natural killer cell therapy is also another novel approach (38,39).

Acute myeloid leukemia (AML)

AML represents 5% of childhood malignancy and 20% of childhood leukemia. While cure of childhood AML has tripled since the 1970’s, current survival of 70% makes it one of the less curable cancers in children. Chemotherapy intensification, response-directed therapy and better supportive care including routine anti-fungal prophylaxis have improved survival by 15% in the last decade, a substantial progress compared to other childhood cancers (Table 1).

AML can be categorized into three genetically defined prognostic groups (40). Favorable risk comprises acute promyelocytic leukemia (APL), myeloid leukemias of Down Syndrome (MLDS), core binding factor (CBF) AML with t(8;21) and inv(16), and AML with NPM1 and CEBPA mutations. APL is characterized by t(15;17)/PML-RARA and its unique sensitivity to all-trans-retinoic acid (ATRA) and arsenic trioxide, both of which target the PML-RARA fusion protein (41). Children with Down syndrome (DS) (Trisomy 21) have a 20-times increased risk of developing leukemia and are particularly susceptible to chemotherapy and treatment-related complications (42). High risk aberrations include monosomy 5 and 7, some 11q23/MLL rearrangements and FLT3/ITD. Intermediate risk is neither favorable nor high risk. CBF, APL and MLL represent 50% of pediatric AML.

MLL-rearranged pediatric AML generally has an unfavourable outcome [overall 5-year EFS 44%; overall survival (OS) 56%] (43). However, it is a heterogeneous disease with the MLL gene having over 60 different translocation partners (44) and 5-year EFS ranging from 11% to 92% for different translocation partners (43). t(9;11) (p22;q23)/MLL-AF9 is the most common rearrangement (50%) and has a 5-year EFS of 50%. The 3% of patients with t(1;11) (q21;q23)/MLL-AF1q have the best outcome (5-year EFS 93%). In contrast, t(6;11) (q27;q23)/MLL-AF6 (5% of cases) has the lowest survival (5-year EFS 11%) (43,45).

Internal tandem duplication (ITD) of the receptor tyrosine kinase FLT3 results in activating mutations. FLT3-ITD-positive patients have significantly inferior progression-free survival (PFS) than FLT wild-type patients (4-year PFS 31% vs. 55%; P<0.001) (46). However, prognosis is influenced by ITD allelic ratio (AR) (mutant to wild-type ratio) and NPM1 status. Patients with high ITD-AR (>0.4) were reported to have significantly worse outcome than those with low AR (PFS 16% vs. 72%) (46). In contrast, NPM1 mutations were found to improve survival of patients with FLT-ITD. In the UK MRC AML 10 and 12 trials, young adults who were FLT-ITD-positive had significantly better outcome when concurrent NMP1 mutations were present (5-year DFS 31% vs. 15%) (47).

Advances in AML treatment in the last decade include:

• In APL, addition of ATRA to anthracycline monotherapy or multi-agent chemotherapy containing idarubicin and high-dose cytarabine increased survival to 85-90% (48,49). Arsenic trioxide has been shown to be very effective when incorporated into adult APL therapy (50,51). COG is currently conducting a phase III trial of introducing arsenic to a multi-agent regimen containing ATRA, idarubicin, and cytarabine in newly diagnosed APL (NCT00866918) (52);

• Reducing chemotherapy intensity did not compromise outcome in MLDS with EFS ranging from 79 to 83% and OS 84 to 91% in three different studies (53-55);

• Risk-adapted therapy utilizing cytogenetic risk stratification improved outcome, yielding EFS from 56 to 61% and OS from 66 to 75% in a number of large studies (56-59);

• MRD-positivity by flow cytometry after first course of chemotherapy predicts survival. Relapse-free survival (RFS) was found to be significantly worse in MRD-positive (RFS 14-43%) than in MRD-negative patients (RFS 65-85%) in three large studies (57,60,61). AML trials evaluating response-guided therapy based on MRD include NOPHO-DBH AML 2012 (NCT 1828489) and COG-AAML-1031 (NCT01371981).

Molecular target

FLT3-ITD reported in 10% of pediatric AML (40) is a candidate for targeted therapy with TKI. In a phase I study, sorafenib (multikinase inhibitor) in combination with clofarabine and cytarabine induced complete remission in relapsed AML patients with FLT3-ITD (62). In a current COG phase III trial for newly diagnosed AML (NCT01371981), FLT3-ITD-positive patients receive sorafenib.

Targeting CD33 using gemtuzumab ozogamicin (GO) (Mylotarg^®), a calicheamicin-conjugated CD33 antibody, has promising activity in AML. In pediatric AML, GO induced complete remission in 35% of refractory disease (63,64) and was safe in de novo AML when combined with chemotherapy (65). In adults, GO improved survival in newly diagnosed AML in some but not all large studies (65-70). Pediatric trials of GO in de novo AML are ongoing (NCT00372593, NCT00476541).

Non-Hodgkin lymphoma (NHL)

NHL constitutes 6% of childhood cancer and is most common in the second decade of life. NHL can be classified according to phenotype (B-cell vs. T-cell) and differentiation (71). Unlike adults NHL that is generally low/intermediate grade, pediatric NHL is frequently high grade. It falls into three categories: (I) mature B-cell NHL including Burkitt/Burkitt-like lymphoma and diffuse large B-cell lymphoma (DLBCL); (II) lymphoblastic lymphoma (LL) (mostly precursor T-cell); and (III) anaplastic large cell lymphoma (ALCL) (mature T-cell or null-cell). Burkitt lymphoma (BL) is most common, accounting for one-third of pediatric NHL.

Currently, cure from childhood NHL is 90%. The steady increase in survival over the last 40 years was due to the initial recognition that LL was best treated with two years of ALL therapy (72,73) and later on due to increasingly effective risk-stratified therapies evolving from large multi-centre clinical trials including BFM-NHL90 (74), NHL-BFM95 (75) and FAB/LMB96 (76-78). Survival is excellent for low-stage LL (>90%) (74), BL (>85%) and DLBCL (90%) (79). However, survival is inferior for BL with CNS involvement (75%) (76), primary mediastinal B-cell lymphoma in DLBCL (73%) (80) and ALCL (70-85%) (81,82).

Advances in the treatment of childhood NHL include:

• Reducing multi-agent chemotherapy to only two courses and omitting intrathecal chemotherapy while maintaining an excellent cure rate of 99% in completely resected localized BL (75,77);

• Reducing treatment in early responding patients with intermediate-risk B-NHL (78);

• Use of multi-agent chemotherapy tailored to disease burden and initial response in B-NHL (79);

• Substituting cranial irradiation with high-dose methotrexate for CNS prophylaxis that is essential for T-cell LL, thus avoiding the long-term side-effects of cranial irradiation (83-85).

Molecular therapy

BL and DLBCL, both mature B-cell phenotype, express high levels of CD20. Rituximab (CD20 antibody) in combination with chemotherapy improved survival in adult DLBCL (86). It is an approved drug for the treatment of DLBCL. In pediatric BL, rituximab has activity as a single agent (87) and can be safely combined with chemotherapy (88). Based on these studies, an international collaborative study INT-B-NHL ritux 2010 (NCT01516580) is currently evaluating rituximab with chemotherapy for high-risk BL and DLBCL in children.

Rituximab may also be useful as a second-line therapy in post-transplant lymphoproliferative disorder (PTLD) or BL secondary to Epstein-Barr virus reactivation with immunosuppression after solid organ or stem cell transplantation. The benefit of rituximab with low dose chemotherapy has been described in a small number of pediatric refractory PTLD (89,90).

Advanced-stage disease is common in ALCL and has less favourable outcome. ALCL expresses CD30 and frequently harbors the t(2;5)/NPM-ALK aberration. Brentuximab vedotin (Bv) (SGN-35, Adcetris^®), an antibody-drug conjugate of chimeric CD30 antibody and monomethylauristatin E, resulted in an objective response of 86% and complete remission of 57% in an adult phase II trial of relapsed systemic ALCL (91). Crizotinib (ALK inhibitor) elicited response in eight of nine ALCL in a phase I pediatric study (92). COG is currently conducting a phase I study of crizotinib in combination with chemotherapy for relapsed ALCL (NCT01606878) and a phase II trial where newly diagnosed ALCL patients are randomized to either crizotinib or Bv with chemotherapy (NCT01979536).

Hodgkin lymphoma (HL)

HL is the most common cancer in the 15 to 19 years age group and is four to five times more frequent than in the <15 years age group. It is characterized by the Reed-Sternberg multinucleated giant cell or its variants and histologic subtypes are defined by the number of Reed-Sternberg cells, characteristic inflammatory milieu and degree of fibrosis. HL is categorized into classical and nodular lymphocyte predominant HL.

HL was fatal until the 1960s when the MOPP nitrogen mustard containing chemotherapy regimen was introduced (93). The cure rate of HL in children has been >90% in the last two decades and is one of the most curable childhood cancers. Unfortunately, survivors of childhood HL are at significant risk of long-term treatment-related morbidity and mortality. In a study of ~2,700 childhood HL survivor, 23% of deaths were from secondary malignant neoplasms and 14% from cerebrovascular and heart disease. The 30-year cumulative incidence of secondary malignant neoplasms was significantly higher in females than in males (26% vs. 11%) due to the high incidence of invasive breast cancer in female survivors treated with radiation (94). Hence, to reduce long-term side effects while maintaining excellent survival, pediatric oncologists have adopted a risk-adapted approach for HL and attempted to reduce or omit radiation. However, risk assignment and definition of response has not been uniform in clinical trials and has made comparison of outcomes across trials challenging. Study results should therefore be interpreted in the context of risk stratification strategy and chemotherapy regimen, as well as response criteria and timing.

Advances in the treatment of pediatric HL in the last decade include (Table 4):

Table 4. Clinical trials in Hodgkin lymphoma using risk-adapted and response approach.

Study	Risk	Chemotherapy	Intervention	IFRT	Outcome
Reducing chemotherapy in RER maintained survival
P9426 (95)	Low	ABVE	RER: 2 cycles; non-RER: 4 cycles	Yes	RER (n=112) vs. non-RER (n=135); 8-year EFS 87% vs. 85%
P9425 (96)	Intermediate & high	ABVE-PC	RER: 3 cycles; non-RER: 5 cycles	Yes	RER (n=132) vs. non-RER (n=84); 8-year EFS 86% vs. 83%
Omitting IFRT in RER did not compromise survival
HOD99 (97)	Low	VAMP ×4	RER—no IFRT; non-RER—IFRT	+/–	RER (n=47) vs. non-RER (n=41); 2-year EFS 89% vs. 92%
GPOH-HD95 (98)	Low	OPPA ×2 (F); OEPA ×2 (M)	RER—no IFRT; non-RER—IFRT	+/–	RER (n=273) vs. non-RER (n=56); 10-year PFS 97% vs. 92%
GPOH-HD2002 (99)	Low	OPPA ×2 (F); OEPA ×2 (M)	RER—no IFRT; non-RER—IFRT	+/–	RER (n=62) vs. non-RER (n=126); 5-year EFS 93% vs. 92%
Modification based on gender and response
GPOH-HD2002 (99)	Intermediate & high	OPPA-COPP (F); OEPA-COPDAC (M)	Procarbazine replaced with dacarbazine & etopside in male	Yes	Male (n=183) vs. female (n=195); 5-year EFS 90% vs. 85%
C5942 (100)	High	RER (F): BEACOPP-COP/ABV (8 cycles); RER (M): BEACOPP-ABVD (6 cycles)/IFRT; non-RER: BEACOPP ×8/IFRT		F: +/–; M: +	5-year EFS 94% (n=98)

RER, rapid early responder after two cycles of chemotherapy; F, female; M, male; IFRT, involved field radiation therapy; ABVE (doxorubicin, bleomycin, vincristine, etoposide); ABVE-PC (prednisone and cyclophosphamide); VAMP (vinblastine, doxorubicin, methotrexate, prednisone); OPPA (vincristine, procarbazine, prednisone, doxorubicin); OEPA (vincristine, etoposide, prednisone, doxorubicin); OEPA-COPDAC (cyclophosphamide, vincristine, prednisone, dacarbazine); OPPA-COPP (cyclophosphamide, vincristine, prednisone, procarbazine); BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine, prednisone).

• Reduction in chemotherapy exposure in patients who are in complete response (CR) after two cycles of chemotherapy, referred to as rapid early responder (RER) (95,96);

• Omission of involved-field radiation therapy (IFRT) in low-risk RER (97-99);

• Modification of chemotherapy in RER according to gender, based on the principle that less gonadal toxic alkylating therapy in males would reduce the risk of infertility and avoiding IFRT in female would reduce the risk of breast cancer (99,100).

The rapidity of early response is prognostic in HL (96,97,99). Positron emission tomography (PET) has been evaluated as an interim imaging modality in adult HL and was more superior than other response assessments (101). COG has recently completed four HL trials treating ~2,400 patients of different risks while addressing the utility of PET. Mature results are yet to be published. There are also ongoing clinical trials in North America and Europe addressing the utility of PET for rapid early response and the omission of IFRT in RER [NCT01858922, NCT00846742, 2006-000995-33 (102)]. All these trials should provide valuable data on the value of PET to predict outcome and whether EFS can be maintained after adjusting therapy based on early PET response.

Molecular therapy targeting CD30 expressed by the Hodgkin/Reed-Sternberg cells may be effective in classical HL. In addition to ALCL, Bv is currently being studied in a number of phase I and II trials for pediatric relapsed HL (NCT01780662, NCT01393717, NCT01492088, NCT01508312) and in newly diagnosed unfavourable risk HL patients (NCT01920932).

Medulloblastoma (MB)

MB is a malignant embryonal neuroectodermal tumor in the cerebellum and comprises 15% of childhood brain tumors. Clinical staging stratified MB into standard (60%) and high-risk (40%) (metastases, residual volume ≥1.5 cm², large cell/anaplastic histology) (103). Current survival is 80% in standard-risk (104,105), 70% in high-risk (106-108) and 50% in children <3 years with MB (109). Contemporary MB therapy includes maximal surgery, adjuvant radiation [cranial spinal irradiation (CSI) with posterior fossa boost] and chemotherapy. Therapeutic approach in young children focuses on delaying or avoiding radiation to minimize the detrimental effects on the immature CNS through the use of multi-agent chemotherapy.

Advances in the last decade and current clinical trials include:

1. Standard-risk MB and age >3 years:

   • Post-radiation adjuvant chemotherapy allowed lowering CSI dose from 36 to 23.4 Gy while maintaining 5-year EFS at 83% (n=86) (104) and 81% (n=379) (105);

   • Current COG phase III RCT (NCT00085735) asks whether EFS can be maintained with 18 Gy CSI and conformal tumor site radiation with concomitant chemotherapy.

2. High-risk MB and age >3 years treated with CSI (36-39.6 Gy):

   • Chemoradiotherapy using carboplatin and vincristine (CV) followed by maintenance chemotherapy in a phase I study resulted in 5-year PFS of 71% (n=55) (107);

   • Post-operative chemotherapy for eight weeks followed by hyperfractionated accelerated radiotherapy +/– consolidation with myeloablative chemotherapy produced 5-year EFS of 70% (n=33) (106);

   • Four courses of post-radiation intensive chemotherapy with hemapoietic stem cell support in SJ-MB-96 study achieved 5-year EFS of 70% (n=48) (108);

   • Current COG phase III RCT (NCT00392327) evaluates the addition of carboplatin to chemoradiotherapy and isotretinoin to maintenance chemotherapy.

3. MB in young children:

   • Histology is an important prognostic factor. In a systematic review of five studies, desmoplastic/nodular (DN) or MB with extensive nodularity (n=108) have significantly better outcome (8-year EFS 55%; OS 76%) than classical MB (n=145; 8-year EFS 27%; OS 42%) in children <5 years (110);

   • Thiotepa-based myeloablative therapy eliminated the need of CSI in 50% of young children (<3 years) with non-metastatic MB. 5-year EFS and OS were 52% and 70% (n=21) (111);

   • Current COG phase III RCT (NCT00336024) assesses the addition of high-dose methotrexate to thiotepa-based myeloablative therapy in high-risk MB.

4. Molecular profiling performed over the last decade has allowed segregation of MB into four distinct groups: WNT, sonic hedgehog (SHH), Group 3 and Group 4 (112-115) (Table 5):

   • The WNT tumors, of which only 5% to 10% are metastatic, are found in children >3 years with classical MB, and have the best prognosis (OS 90%) (116). Patients may be candidate for reduced intensity therapy;

   • The SHH tumors are bimodal in age distribution (<3 years and adult) and have intermediate prognosis (OS 60-80%). DN MB is exclusive to this group (115). Smoothened (SMO) receptor is a target for SHH (117) and vismodegib (GDC-0449), a SMO inhibitor, is a potential therapeutic (118,119). Phase II trials of vismodegib in refractory (NCT01239316) and newly diagnosed MB (NCT01878617) are ongoing;

   • Group 3 comprises non-WNT and non-SHH tumors that are MYC-driven. It has the highest incidence of large cell/anaplasia (LCA) histology and metastases and the worst survival (50%);

   • Group 4 tumors, of which 35% are metastatic, are largely classical MB with intermediate outcome (OS 75%).

Table 5. Four distinct molecular/genetic subgroups in medulloblastoma.

	WNT	SHH	Group 3	Group 4
Frequency (%)	10	30	25	35
Age (years)	>3	<3 and >16	<16	all
Male:female	1:1	1:1	2:1	2:1
Metastatic (%)	5-10	15-20	45	35
Histology	Classical	Classical; DN (50%)	Classical; LCA (35%)	Classical
Survival (%)	>90	60-80	50	75
Distinct genetic aberrations	CTNNB1	PTCH1	MYC amplification
Chromosomal abnormality	6- (~100%)			i17q (65%); X- (80% female)

SHH, sonic hedgehog; DN, desmoplastic/nodular; LCA, large cell/anaplasia; WNT, wingless/integrated.

Low grade glioma (LGG)

Glioma is a CNS tumor of glial cell origin and can arise anywhere within the CNS. It consists of astrocytomas, oligodendroglioma and mixed gliomas, and can be classified into low grade (WHO grade I and II) and high grade (WHO III and IV). LGG, most frequently located in the cerebellum, is the most common type of brain tumor (30-50%) (4-6). Pilocytic (WHO grade 1) and diffuse fibrillary astrocytoma (WHO grade 2) are the predominant histology in childhood LGG. Children with neurofibromatosis type 1 (NF1) have a high incidence (20%) of optic pathway LGG (OPG) (120) and contributes to 60% of OPG (121). The natural history of LGG is not always predictable.

Surgery remains the first-line treatment for LGG and is considered curative in areas of the brain amenable to complete resection. The 8-year PFS and OS are 90% and 99% following gross total resection (122). Incomplete resection and young age are both poor prognostic factors. Partially resected LGG has a10-year EFS of 18% and OS of 87% (123). Infants with diencephalic syndrome from hypothalamic lesion have the worst outcome (5-year PFS 11%; 10-year OS 43%) (124). Patients <1 year old have EFS of 19%, compared with patients aged >5 years with EFS of 64% (123,125). Radiotherapy can provide long-term control of LGG (123,126), but with significant late cognitive and endocrine sequelae (126,127), especially in young children who are more prone to the late effects of radiation (128). In NF1-associated OPG, the risk of second tumor was found to be three times higher in those who received radiation (129).

In the last decade, the growing trend to manage unresectable, progressive, or recurrent LGG with chemotherapy has successfully deferred or avoided radiation in young children. The two most widely used chemotherapy regimen are CV (130,131) or thioguanine, procarbazine, lomustine and vincristine (TPCV) (132,133). In the German study of CV (n=213), 32% experienced partial response (PR) and 75% did not require subsequent radiation. 10-year PSF and OS was 44% and 88%, respectively (123). However, it remains unclear whether chemotherapy improves visual outcome in OPG (134).

COG conducted a phase III RCT of CV versus TPCV (125). Initial response was similar for both groups (1/3 PR, 1/3 stable disease (SD), 1/3 progressive disease). While 2-year EFS was similar at 60% for both regimens, there were increasingly more events in the CV group after two years, resulting in 5-year EFS of 39% for CV and 52% for TPCV. Due to non-proportionality, log rank test for EFS was not significant (P=0.1); however, by cure model analysis, TPCV was more superior to CV (P=0.007). 5-year OS was equivalent (87% vs. 86%). Toxicity profile was different with more CNS toxicity for TPCV. Allergic reaction was only reported in CV due to carboplatin allergy. CV is the regimen of choice for NF1 patients due to potential increased risk of second malignancy from alkylating agents in TPCV. Other agents that has been tested in phase I/II studies for LGG include vinblastine (135,136), temozolomide (137) and bevacizumab and irinotecan (138).

Neuroblastoma (NB)

NB, an embryonal tumor of the sympathetic nervous system, is the most common extracranial solid tumor in children. Majority of NB is found in children younger than five years. It is well known for its heterogeneity ranging from spontaneous regression, differentiation to benign ganglioneuroma to aggressive metastatic disease (139).

NB is risk-stratified based on age, disease stage and biologic features (140), with survival >90% for low and intermediate risk (141,142) and 50% for high-risk NB (143). MYCN amplification is a well-established molecular prognostic factor and places patients into the high-risk category irrespective of age and stage. A pre-treatment risk stratification system (International Neuroblastoma Risk Group classification system) was developed in 2008 using 13 potential prognostic factors in a cohort of 8,800 children diagnosed with NB worldwide with an aim to facilitate comparison of clinical trials and development of international collaborative studies (140). Standard therapy for high-risk NB includes intensive chemotherapy, surgery, myeloablative chemotherapy with stem cells rescue, radiation and biological therapy. Low or intermediate NB is managed with surgery with or without chemotherapy, with an aim to appropriately minimize therapy.

The following is an overview of advances in NB treatment in the last decade (Table 6).

Table 6. Progress in neuroblastoma.

Discovery	Outcome	Ref.
Localized NB with normal MYCN
For resectable stage 1 and 2 NB, surgery without adjuvant chemotherapy is adequate	Stage 1: 5-yr EFS 93%, OS 99%; Stage 2: 5-yr EFS 85%, OS 95%	(142,144)
Infant localized NB with normal MYCN
Observation does not compromise survival in those with resectable NB	Infant <12 months old: 3-yr OS 99% (n=93); subsequent local or metastatic progression in 30% & 11%. Infant <12 months old & small adrenal mass: 3-yr OS 100% (n=87); 19% required subsequent surgical intervention	(145,146)
Omission of anthracyclines in upfront chemotherapy does not compromise survival in those with unresectable NB	5-yr EFS 90%; OS 99% (n=120); subsequent anthracyclines due to inadequate response in 38%	(147)
Intermediate-risk NB
For favorable biology tumor, reducing chemotherapy from 8 to 4 cycles maintained excellent survival	Favorable biology tumor (4 cycles): 3-yr EFS 90%; OS 98% (n=323). Unfavorable biology tumor (8 cycles): 3-year EFS 83%; OS 93% (n=141)	(141)
High-risk NB
Immunotherapy with ch14.18 (chimeric antibody against GD2) provided the best outcome for post-consolidation biologic therapy	ch14.18 antibody/GM-CSF/IL-2/ isotretinoin: 2-year EFS 66%; OS 86% (n=113). Isotretinoin alone: 2-year EFS 46%; OS 75% (n=113) (P=0.01)	(148)
11q loss of heterozygosity (LOH)
Poor prognostic factor in stage 2, 3, and 4s tumors without MYCN amplification	<18 months old (stage 2 & 3): 11q LOH, 5-yr EFS 60%; OS 84% (n=176); normal 11q, 5-yr EFS 83%; OS 98% (n=19). ≥18 months old (stage 2 & 3): 1q LOH, 5-yr EFS 61%; OS 73% (n=18); normal 11q, 5-yr EFS 80%; OS 100% (n=49). Stage 4s: 11q LOH, EFS 38% ;OS 63% (n=8); normal 11q, EFS 87%; OS 97% (n=62)	(140)

IL-2, interleukin-2; OS, overall survival.

Low and intermediate-risk NB

Surgery alone was found to be adequate in localized resectable (stage 1 and 2) NB (142,144). In infants <12 months old with resectable tumor (145) or in young infants <6 months old with small adrenal mass (146), observation approach at diagnosis did not compromise survival while upfront low dose chemotherapy without anthracyclines maintained excellent outcome in infants with localized unresectable tumor (147). In intermediate-risk NB, reducing chemotherapy to four cycles in tumors with favorable biologic features did not adversely affect survival (141). Furthermore, chromosome 11q loss of heterozygosity (LOH) was found to be a poor prognostic factor (140).

High-risk NB

Immunotherapy has provided the best outcome in the setting of MRD. GD2 is a disialoganglioside highly expressed by NB. Immunotherapy with ch14.18 (chimeric antibody against GD2) in conjunction with cytokines and isotretinoin was evaluated in a phase III RCT (n=226). In patients who have achieved at least a PR after myeloablative therapy, immunotherapy was superior to isotretinoin alone (2-year EFS 66% vs. 46%, P=0.01; OS 86% vs. 75%, P=0.02) (148). Immunocytokine therapy using the humanized 14.18-IL-2 fusion protein has been tested in phase I (149) and II (150) studies in recurrent NB.

Metaiodobenzylguanidine (MIBG) scan is routinely used to diagnose NB. Metastatic response assessed by MIBG scan was found to predict survival (151,152). A semiquantitative MIBG scoring system (Curie score) showed that patients with a score >2 after induction therapy had a significantly worse outcome than those with scores ≤2 (3-year EFS: 15% vs. 44%; P<0.001; n=237) (153).

Targeted radiotherapy

Radiolabeled 131-I-MIBG selectively targets radiation to catecholamine-producing NB cells. Phase I studies of 131-I-MIBG in combination with radiosensitizer (154) or with myeloablative therapy and stem cell rescue (155,156) has demonstrated efficacy (25-65% response) in refractory NB. Studies of 131-I-MIBG followed by myeloablative therapy in newly diagnosed NB are ongoing (NCT01175356, NCT00798148).

NB genome

Whole genome sequencing studies have shown paucity of somatic mutations in the high-risk NB genome and identified ALK and ATRX aberrations in ~20% of high-risk NB (157-159). Crizotinib, an ALK inhibitor, is currently being evaluated alone or with chemotherapy in recurrent NB (NCT00939770, NCT01606878).

Ewing sarcoma (ES)

The incidence of ES is highest in the second decade of life. It most commonly presents as an undifferentiated bone tumor and less frequently as a soft tissue mass (extraosseous ES). The pathognomonic t(11;22)(q24;q12)/EWS-FLI translocation is also found in peripheral primitive neuroectodermal tumor (PNET), a more differentiated tumor of bone or soft tissue. These tumors are collectively referred to as the ES family of tumors (ESFT) (160).

In addition to surgery +/– radiotherapy for local control, doxorubicin-containing chemotherapy regimen is essential for effective eradication of micro-metastases in ES (161). Improvement in survival of localized disease in the 1990’s was attributed to the addition of ifosfamide and etoposide (IE) to the standard three drugs [vincristine, doxorubicin, cyclophosphamide (VDC)] in the US (162,163). However, progress has been marginal in the last 20 years and outcome of metastatic ES remains poor. Currently, survival for localized and metastatic ES is 80% (164) and <35%, respectively (165,166).

For localized ES, one advance of note in the last decade is the effective use of “interval compression” of chemotherapy (as opposed to dose intensification). The COG AEWS001 phase III RCT in 587 patients demonstrated 5-year EFS of 73% and OS 83% in the intensified arm (alternating VDC and IE every two weeks), compared with 5-year EFS of 65% and OS 73% in the standard arm (3-week cycle) (P=0.048) with no worsening toxicity (164).

Topotecan and irinotecan, both topoisomerase I inhibitors, have shown activity in recurrent ES. Response was observed in a third of recurrent ES treated with topotecan and cyclophosphamide in two prospective studies (167,168) and in 60% of patients receiving irinotecan and temozolomide in a retrospective study (169). Based on this evidence, the current COG phase III RCT (NCT01231906) examines the efficacy of adding CVT (cyclophosphamide, vincristine, topotecan) to the intensively timed 5-drug (VDC/IE) regimen in newly diagnosed localized ES.

Large tumours, pelvic site and poor histologic response to induction chemotherapy are known poor prognostic factors in localized disease (170,171). The French EW93 study demonstrated an improved 5-year EFS of 45% in a small number of high-risk patients treated with high-dose busulphan/melphalan and hemapoietic stem cell support (172). Furthermore, high-dose chemotherapy may be beneficial in the setting of isolated pulmonary metastases, which has a better outcome than other types of metastases (170,173). Hence, both the recently completed multi-center EuroEwing 99 (NCT00020566) study and the current EuroEwing 2012 (ISRCTN92192408) (174) phase III RCT examine therapy intensification with high-dose busulphan and melphalan in patients with isolated lung metastases and in those with large (>200 mL volume) or poor histologic response (>10% viability) tumors.

Novel targeted therapy is needed for metastatic and recurrent ES. Inhibition of insulin growth factor-1 receptor (IGF1R) using monoclonal antibody has shown efficacy in relapsed ES. Anti-IGF1R antibody that has been studied in phase I/II studies included figitumumab (CP-751,871) (175,176), ganitumab (AMG-479) (177) and cixutumumab (IMC-A12), either alone or with temsirolimus (mTOR inhibitor) (178,179). These studies reported SD and response between 25-50%. COG is developing a phase II RCT (AEWS1221) to assess the feasibility of adding ganitumab to the interval-compressed regimen (180).

Novel therapeutics in pediatric cancer

Our increased understanding of the molecular basis of childhood cancer in the last decade has allowed researcher to define molecular targets and evaluate available therapeutics that are in clinical development for adult cancers. We selected seven novel agents or class of drugs that are most frequently studied in clinical trials involving childhood cancer. In most instances, the agent has a track record in adult oncology and has been subjected to safety scrutiny in adult phase I/II trials, in which a maximum tolerated dose was established. Tables 7 and 8 provide a list of completed and ongoing clinical studies that involve pediatric patients. All but one drug have approval from the US Food and Drug Administration (FDA) for specific cancers; however, none are pediatric cancers.

Table 7. Completed clinical trials of the seven novel agents.

Agent	Phase	Other agents	Cancer	DLT or AE (grade) [%]	Efficacy [%]	Trial ID (52)	Ref.
Bevacizumab	I	–	Solid	0/18 [0]	OR 0/18 [0]	NCT00085111	(181)
	II	–	EPM	1/13 [8] (Gr 4)	SD 2/13 [15]	–	(182)
	II	IRN	HGG/BSG	2/31 [7] (Gr 4)	SD 13/31 [42]	–	(183)
	Pilot	VCR/IRN/TMZ	Solid	0/13 [0]	PR 2/13 [15] SD 1/13 [8]	–	(184)
	I	IRN	Solid	2/9 [22]	PR 1/11 [9] SD 3/11 [27]	–	(185)
	I	VCR/IRN/TMZ	Solid	2/12 [17]	OR 5/13 [38]	NCT00993044	(186)
	I†Φ	Sorafenib, CPM	Solid	5/19 [26]	PR 5/17 [29] SD 5/17 [29]	NCT00665990	(187)
	II†	IRN	LGG	3/35 [9] (Gr ≥3)	PR 2/35 [6]	NCT00381797	(188)
	I	IRN/TMZ	CNS	–	–	NCT00876993	np
Bortezomib	I	–	Solid	2/5 [40]	OR 0/11 [0]	NCT00021216	(189)
	I/II	–	ALL	1/10 [10]	CR 7/9 [78]	NCT00440726	(190)
	I	–	Leuk	2/12 [17]	SD 2/12 [17]	NCT00077467	(191)
	IΦ	Vorinostat	Solid	2/17 [12]	OR 0/23 [0]	NCT00994500	(192)
	I	Chemotherapy	ALL	21/22 [95] (Gr ≥3)	CR 16/22 [73]	NCT00440726	(193)
	IΦ	Vorinostat	Solid	–	–	NCT01132911	np
	II	–	AML	–	–	NCT00666588	np
	II	–	HL	–	–	NCT00381940	np
Vorinostat	I/II†	–	Solid/leuk	–	–	NCT01422499	(194)
	I	TMZ	CNS	4/18 [22]	SD 3/16 [19]	NCT01076530	(195)
	I	Isotretinoin	Solid	14/53 [26]	PR 1/14 [7] SD 2/14 [14]	NCT00217412	(196)
	II	Decitabine	ALL	–	–	NCT00882206	np
Sorafenib	I	–	Solid/AML	4/60 [7]	SD 16/60 [27]	NCT01445080	(197)
	I	–	NF1	4/9 [44]	–	NCT00727233	(198)
	I†	clofarabine cytarabine	AML	3/12 [25]	CR 8/12 [66] PR 1/12 [8]	NCT00908167	(62)
	II	–	OS	–	–	NCT00889057	np
Erlotinib	I	radiotherapy	CNS	14/50 [28] (Gr ≥3)	SD 11/47 [23]	NCT00418327	(199)
	I/II†	radiotherapy	HGG	16/23 [70] (Gr ≥3)	SD 16/23 [70]	NCT00124657	(200)
	IΦ	sirolimus	LGG	2/19 [10] (Gr ≥3)	PR 1/6 [17] SD 1/6 [17]	–	(201)
	I	TMZ	Solid	3/36 [8]	SD 17/43 [40]	NCT00077454	(202)
	II	–	EPM	–	–	NCT01032070	np
	II	–	EPM	–	–	NCT01247922	np
mTor inhibitors: Sirolimus (S), Temsirolimus (T), Everolimus (E), Deforolimus (D)	II (T)	–	HGG/NB RMS	23/52 [44] (Gr ≥3)	PR 1/52 [2] SD 18/52 [35]	NCT00106353	(203)
	I (T)	–	Solid	1/19 [5]	CR 1/19 [5] SD 5/19 [25]	–	(204)
	I (E)	–	Solid	5/25 [20]	SD 6/25 [24]	NCT00187174	(205)
	I (D)	–	Solid	0/15 [0]	SD 6/15 [40]	NCT00704054	(206)
	II (S)	–	NF–1	–	–	NCT00652990	np
	I (S)	Vinblastine	Solid	–	–	NCT01135563	np
	I (T)	IRN/TMZ	Solid	–	–	NCT01141244	np
	I (T)	Cixutumumab	Solid	–	–	NCT00880282	np
	II (E)	–	LGG	–	–	NCT00782626	np
	I (D)	–	Solid	–	–	NCT01431534	np
	I (D)	Dalotuzumab	Solid	–	–	NCT01431547	np
Tipifarnib	I	Radiotherapy	DIPG	5/17 [29]	1Y–PFS [9.4]	NCT00079339	(207)
	II	Radiotherapy	DIPG	4/40 [10] (Gr ≥3)	PR 7/40 [18]	NCT00079339	(208)
	II	–	CNS	31/97 [32] (Gr 4)	PR 1/31 HGG PR 1/35 BSG	NCT00070525	(209)
	II	–	Leuk	5/18 [28]	OR 0/23 [0]	NCT00022451	(210)
	I	–	Solid/NF1	12/40 [30]	OR 0/40 [0]	–	(211)
	II	–	NF1	–	–	NCT00021541	np
	II	–	NF1	–	–	NCT00076102	np
	II	–	JML	–	–	NCT00025038	np

AE, adverse event; BSG, brain stem glioma; CPM, cyclophosphamide; CR, complete response; DIPG, diffuse intrinsic pontine glioma; DLT, dose-limiting toxicity; EPM, eppendymoma; HGG, high-grade glioma; IRN, irinotecan; leuk, leukemia; NF1, neurofibromatosis type-1-related plexiform neurofibroma; np, no published results; OR, objective response; OS, osteosarcoma; PR, partial response; RMS, rhabdomyosarcoma; SD, stable disease; TMZ, temozolomide; VCR, vincristine; †, ongoing clinical trial; Φ, trial involving two or more of the seven novel agents.

Table 8. Ongoing clinical trials of the seven novel agents.

Agent		Phase	Other agents		Tumor type	Trial ID
Bevacizumab		I	Cediranib		Solid	NCT00458731
		I †	Everolimus		Solid	NCT00756340
		I/II	VCR/IRN/TMZ		Solid	NCT00786669
		I †	Sorafenib/CPM		Solid/leuk	NCT00665990
		II	Chemo		Sarcomas	NCT00643565
		II †	Temsirolimus, vinorelbine/CPM		RMS	NCT01222715
		II	CHEMO/RT		RMS	NCT01871766
		II	–		NF–2	NCT01767792
		II	–		NF–2	NCT01207687
		I	–		NB	NCT00450827
		I	CPM/zoledronic acid		NB	NCT00885326
		II	IRN/TMZ		NB	NCT01114555
		II	CPM/topotecan		ES/NB	NCT01492673
		III	VCR/topotecan/CPM		ES	NCT00516295
		I †	Sorafenib, temsirolimus		ES/DSRCT	NCT01946529
		Pilot	IRN/TMZ/chemo		DSRCT	NCT01189643
		II	Chemo		OS	NCT00667342
		II	Chemo		GCT	NCT00936936
		II	IRN		CNS	NCT00381797
		II	Chemo		MB	NCT01356290
		II	Chemo		MB	NCT01356290
		II	IRN/TMZ		MB/PNET	NCT01217437
		II	TMZ/RT		HGG	NCT01390948
		II	Valproic acid/RT		HGG	NCT00879437
		II/III †	TMZ/vorinostat/RT		HGG	NCT01236560
		Pilot	IRN/TMZ		HGG, DIPG	NCT00890786
		II †	Erlotinib/TMZ/RT		DIPG	NCT01182350
		I/II	Cetuximab		Glioma	NCT01884740
		II	Lapatinib		EPM	NCT00883688
Bortezomib		I	IRN		NB	NCT00644696
		II	–		ALL	NCT00873093
		II	–		ALL	NCT00440726
		II	Arsenic		APL	NCT01950611
		III †	Sorafenib		AML	NCT01371981
Vorinostat		I/II	–		Solid	NCT01294670
		I	Isotretinoin		MB/PNET	NCT00867178
		I	Isotretinoin		NB	NCT01208454
		I	131-i-MIBG		NB	NCT01019850
		I/II	Decitabine		ALL	NCT01483690
Sorafenib	I		Topotecan	Solid			NCT01683149
	I		IRN	Solid			NCT01518413
	II		–	RMS/kidney liver/thyroid			NCT01502410
	II		–	LGG			NCT01338857
	II		Chemo/RT	RMS			NCT01871766
Erlotinib	I		Pralatrexate	Advanced			NCT01532011
	I		Radiotherapy	CNS			NCT00360854
	I †		Sirolimus	LGG			NCT00901849
	II †		Sirolimus	GCT			NCT01962896
	III		Chemo	CNS			NCT00602667
mTOR inhibitors: sirolimus (S), temsirolimus (T), everolimus (E)	I (S)		–	ALL			NCT01658007
	I (S)		–	Leuk/NHL			NCT00068302
	I (S)		–	Solid			NCT01331135
	I (S)		Topotecan	Solid			NCT01670175
	I (S)		IRN	Solid			NCT01282697
	II (S)		–	NF-1			NCT00634270
	II (S)		Dasatinib, IRN/TMZ	NB			NCT01467986
	I (T)		–	ALL/NHL			NCT01403415
	I (T)		Etoposide/CPM	ALL/NHL			NCT01614197
	I (T)		Perifosine	Solid			NCT01049841
	I (T)		Cixutumumab	Sarcoma			NCT01614795
	II (T)		IRN/TMZ/ch14.18	NB			NCT01767194
	I (E)		–	ALL			NCT01523977
	I (E)		–	NF-2			NCT01419639
	I(E)		–	LGG			NCT01158651
	II (E)		–	LGG			NCT01734512
	II (E)		–	OS			NCT01216826
	II (E)		–	Sarcomas			NCT01216839

Chemo, chemotherapy; CPM, cyclophosphamide; DIPG, diffuse intrinsic pontine glioma; DSRCT, desmoplastic small round cell tumor; EPM, eppendymoma; GCT, germ cell tumor; HGG, high-grade glioma; IRN, irinotecan; leuk, leukemia; NF1, neurofibromatosis type-1-related plexiform neurofibroma; NF2, neurofibromatosis type-2-related vestibular schwannomas; OS, osteosarcoma; RMS, rhabdomyosarcoma; RT, radiotherapy; TMZ, temozolomide; VCR, vincristine; †, trial involving 2 or more of the 6 novel agents.

Bevacizumab (Avastin^®)

Vascular endothelial growth factor (VEGF) is a regulator of tumor angiogenesis and is a prerequisite for cancer growth. Bevacizumab, a humanized anti-VEGF monoclonal antibody, sequestrates VEGF (212). Bevacizumab has FDA approval for adult malignancies including colorectal, lung and prostate cancer (213). In the pediatric population, bevacizumab was most effective when combined with irinotecan +/– temozolomide (184-186). Numerous Phase II clinical trials are ongoing to investigate bevacizumab mostly in combination with other agents in a variety of childhood malignancies.

Bortezomib (Velcade^®)

The ubiquitin-proteasome pathway regulates the stability of proteins and deregulated proteolysis has been reported in many tumor types. Proteasome inhibition has been found to selectively induce pro-apoptotic proteins in cancer cells (214). Bortezomib is the first FDA approved proteasome inhibitor for the treatment of multiple myeloma and mantle cell lymphoma (215,216). Multiple preclinical and early clinical trials demonstrated its safety and significant anticancer activity towards haematological malignancies in both adults (217) and children (190,193). Bortezomib is currently being evaluated in a COG phase III study in de novo AML and in a phase II study with arsenic in APL.

Vorinostat (Zolinza^®)

Histone acetylation is a reversible process where histone acetyltransferases transfer the acetyl moiety from acetyl co-enzyme A to a lysine while histone deacetylases (HDACs) remove this acetyl groups. HDAC inhibition results in accumulation of acetylated proteins and induces growth arrest, apoptosis and reactive oxygen species-mediated cell death (218). Vorinostat is the first HDAC inhibitor approved for advanced cutaneous T-cell lymphoma (219). It has also been tested in phase I trials in pediatric solid tumors with predominantly disease stabilization (195,196).

Sorafenib (Nexavar^®)

It is a multi-kinase inhibitor that targets tyrosine kinases VEGFR, PDGFR and FLT3, as well as the RAF/MEK/ERK pathway (220). It is currently approved for the treatment of metastatic renal cell carcinoma and hepatocellular carcinoma (221,222). In pediatric early clinical trials, sorafenib combination therapy resulted in 30% response in solid tumors (187) and 75% response in AML (62). Furthermore, next generation TKIs that have been developed to provide increased target specificity and inhibition have been evaluated in adult clinical trials. Pazopanib (223), axitinib (224) and tivozanib (225) are potent VEGFR inhibitors whereas quizartinib and crenolanib are potent FLT3 inhibitors (226).

Erlotinib (Tarceva^®)

Targeting the epidermal growth factor receptor (EGFR) is well established in the treatment of a number of adult malignancies. Erlotinib, a potent EGFR TKI (227), has been approved for the treatment of advanced NSCLC and pancreatic cancer (228,229). In early clinical trials of erlotinib combined with radiation, prolonged SD was observed in pediatric high-grade gliomas (HGG) and relapsed/refractory brain tumors (199,200).

Tipifarnib (Zarnestra^®)

Farnesylation is necessary for Ras activation, which triggers activation of the PI3-kinase/AKT and RAF/MEK/ERK pathway, and is implicated in the pathogenesis of solid and hematologic malignancies. Tipifarnib (farnesyl transferase inhibitor) was shown to have anti-leukemic activity in adult phase I/II trials (230,231). There are no ongoing paediatric trials of tipifarnib, most likely due to lack of activity as a single agent in a number of pediatric studies (207-211).

The mammalian target of rapamycin (mTOR) inhibitors

The mammalian target of rapamycin (mTOR) is a serine/threonine protein kinase and the PI3-kinase/AKT/mTOR pathway plays an important role in the regulation of cell growth, proliferation, motility, survival and transcription (232). mTOR inhibitors include sirolimus (Rapamune^®), its analogue deforolimus, and its derivates temsirolimus (Torisel^®) and everolimus (Afinitor^®). The latter two are approved for the treatment of renal cell carcinoma, subependymal giant cell astrocytoma, pancreatic and breast cancer (233). There are quite a number of ongoing pediatric phase I/II trials of mTOR inhibitors in both hematologic and solid malignancies.

Discussion

Advances in research that evolve into improved standard of care and outcome in childhood cancer is the result of the remarkable effort invested in childhood cancer research. The current 5-year OS of 83% in the US (Table 1) (3) and 79% in Europe (2) represents only a marginal increase over a decade. While cancers that already have an excellent outcome did not show much improvement, 5-year survival of cancers including AML, NB in >1 year, and ALL and ES in 15-19 years old has increased by 10% to 20% in the last ten years (Table 1). Furthermore, progress in oncology is not limited to better survival. Reducing both short and long-term treatment-related complications is as important, given majority of childhood cancer patients will become long-term survivors. This is best achieved by risk-adapted therapeutic approach that has been made possible through identifying clinical and biologic prognostic factors with rigorous research, stratifying patients using these risk factors and modifying therapy according to risk group assignment.

Optimizing delivery of conventional therapeutics has been the driving force behind continuous improvements in pediatric cancer survival in the last 40 years. However, the pediatric oncology field acknowledges that further escalation of conventional therapy is unlikely going to yield improvement in cancers that currently have unacceptably low cure rates. This include AML, high-risk ALL and NB, high-grade brain tumors, and metastatic bone tumors and sarcomas. Advances in medical research technology have led to a rapid increase in our understanding of the genetics of childhood cancer in the last decade and will continue to facilitate identification of molecular targets that can potentially be exploited for therapeutic benefits. As we move into the era of targeted therapeutics, searching for novel agents that target specific genetic lesions in this group of poor prognosis cancer becomes both a priority and a challenge.

Unlike conventional cytotoxic chemotherapies, the premise of effective targeted therapy involves “hitting” the intended target resulting in disruption of specific signalling pathways. Hence clinical trials will need to focus on biologically defined patient subsets, meaning even smaller patient population. It will no longer be feasible to conduct the standard phase III RCT that requires hundreds of patients. The key challenge will be to design trials that can clearly delineate the effect of the new agent under study. Options include single arm study, of which results will be compared with historical control, and phase II randomized trial comparing an active but non-curative cytotoxic chemotherapeutic regimen with or without the new agent. National and international collaborative studies will be required to attain sufficient patients and complete trials in a timely manner. Furthermore, new endpoints that utilize functional imaging or molecular biomarkers can be incorporated into clinical trials as a measure of response and MRD. However, these endpoints need to be properly validated to ensure they accurately reflect clinical benefits.

As we continue into the 21st century, our increased understanding of the molecular and genetic basis of childhood cancer will facilitate further refinement of risk-adapted therapy that utilizes molecular and genetic signatures for risk stratification. The ultimate goal is to cure childhood cancer with the best quality of long-term survivorship.