SIOP-PODC adapted risk stratification and treatment guidelines: Recommendations for neuroblastoma in low- and middle-income settings.

Neuroblastoma is the most common extracranial solid tumor in childhood in high-income countries (HIC), where consistent treatment approaches based on clinical and tumor biological risk stratification have steadily improved outcomes. However, in low- and middle- income countries (LMIC), suboptimal diagnosis, risk stratification, and treatment may occur due to limited resources and unavailable infrastructure. The clinical practice guidelines outlined in this manuscript are based on current published evidence and expert opinions. Standard risk stratification and treatment explicitly adapted to graduated resource settings can improve outcomes for children with neuroblastoma by reducing preventable toxic death and relapse.

INTRODUCTION

Neuroblastoma is the most common extracranial solid tumor in childhood in high‐income countries (HIC), where it accounts for 10% of pediatric cancers 1. However, in low‐ and middle‐ income countries (LMIC) with population‐based registries, it accounts for only 1–3% of cancers, and in most LMIC its true incidence is unknown 2, 3, 4. The limited available data suggest that in LMIC, late diagnosis, lack of access to accurate staging, risk stratification, optimal treatment, and abandonment lead to survival rates much lower than those in HIC 3, 5, 6, 7, 8, 9. With the Pediatric Oncology in Developing Countries (PODC) committee of the International Society of Pediatric Oncology (SIOP) working group for adapted treatment regimens, we developed these recommendations for adapted management of neuroblastoma.

This manuscript focuses on neuroblastoma, since general principles of adapted therapy have been previously reported 10, 11, 12, 13. Optimizing treatment in LMIC at centers with variable capabilities and infrastructure requires adapting diagnostic methods, risk stratification, and therapy for children with neuroblastoma based on available resources 10, 11, 12, 13. In Table I, we defined each setting based on the minimal required resources to facilitate appropriate diagnosis and risk‐stratified, intensity‐adapted treatment. We have used the terminology “Setting 1” to represent centers with minimal resources, but with access to care and treatment with curative intent, while “Setting 4” represents centers with state‐of‐the‐art diagnostic and treatment capability for high‐risk neuroblastoma. Many countries, especially low‐income countries (LIC), have extreme heterogeneity in resources available where some privately‐funded centers may represent Setting 3, while nearby institutions may be Setting 1. A consistent approach to diagnosis and treatment that is adapted to local conditions and resource availability will improve the care of children in LMIC.

Table I

Resource Settings for Neuroblastoma Diagnosis, Staging, and Risk Stratification

	Setting 1	Setting 2	Setting 3	Setting 4
Diagnosis	History, Physical examination,
	Histology of small round blue cell tumor or bone marrow metastases
	Urinary catecholamines (if available)
Staging	CXR and skeletal survey, Abdominal ultrasound,	CT neck/chest/abdomen/pelvis	CT neck/chest/abdomen/pelvis	CT scan neck/chest/abdomen/pelvis
	Bilateral BM aspirate & biopsy	99mTc‐bone Scan	123I‐ MIBG or 18FDG‐PET	123I‐ MIBG or 18FDG‐PET,
		Bilateral BM aspirate & biopsy	MRI head or spine if involved	MRI head or spine if involved
			Bilateral BM aspirate & biopsy	Bilateral BM & biopsy
Laboratory	CBC, liver enzymes, LDH, ferritin, creatinine, urinalysis	CBC, liver enzymes, LDH, ferritin, creatinine, urinalysis	CBC, liver enzymes, LDH, ferritin, creatinine, urinalysis	CBC, liver enzymes, LDH, ferritin, creatinine, urinalysis
			Urine HVA/ VMA,	Urine HVA/ VMA,
		Urine HVA/VMA	Tumor lysis labs if INSS 4	Tumor lysis labs if INSS 4
			(electrolytes, Ca Mg PO4, uric acid)	(electrolytes, Ca Mg PO4, uric acid)
Pathology	H&E stain	H&E stain	H&E stain, IHC	H&E stain, IHC
		IHC	INPC classification (if available)	INPC classification
			(differentiation grade, MKI)	MYCN, DNA Ploidy
			MYCN	segmental chromosome abnormalities
Infrastructure	Nursing, Inpatient Hospital Access to RBC or whole blood	Nursing, Inpatient hospital	Nursing, Inpatient Hospital	Nursing, Inpatient Hospital
		Access to RBC & Platelets	Rapid Access to all Blood	Rapid Access to all Blood
		Pediatric Surgeon	Products Pediatric Surgeon	Products Pediatric Surgeon
		Family Housing	Family Housing	Family Housing
		Intensive Monitoring Capabilities	Pediatric ICU	Pediatric ICU
			Isolation and Transplant Facility	Isolation and Transplant Facility
Therapeutics	Antibiotics	Antibiotics	Antibiotics	Antibiotics
	Standard Chemotherapy	Standard Chemotherapy	Standard Chemotherapy	Standard Chemotherapy
		Radiation Therapy	Radiation Therapy	Radiation Therapy
			Transplant Conditioning	Transplant Conditioning
			Agents	Agents
			Isotretinoin	Isotretinoin, Anti‐GD2 antibody

CT, computerized tomography; CBC, complete blood count; LDH, lactic dehydrogenase; H&E, hematoxylin and eosin stain; IHC, immunohistochemistry; RBC, red blood cell; HVA, homovanillic acid; VMA, vannilylmandelic acid; 123I‐ MIBG, metaiodobenzylguanidine; FDG‐PET, fluorodeoxyglucose positron emission tomography; MRI, magnetic resonance imaging.

METHODS

A Global Neuroblastoma Network (GNN) subcommittee of pediatric oncologists from LMIC and HIC reviewed published evidence about treatment strategies for low‐risk, intermediate‐risk, and high‐risk neuroblastoma and applied these strategies during weekly online Global Neuroblastoma Network (GNN) case discussions held via www.Cure4Kids.org. The subcommittee consensus recommendations are based on review of published evidence (using PubMed ®) for neuroblastoma diagnostics, risk assessments, prognostic markers, and treatment strategies combined with expert opinion about neuroblastoma and healthcare delivery in LMIC. The draft was reviewed by peers in HIC and LMIC and presented to a broad group of pediatric solid tumor experts via web conferences of the SIOP PODC working group on adapted treatment regimens, the annual SIOP meeting, and the GNN weekly online multidisciplinary conferences 14.

DIAGNOSIS

The diagnosis of neuroblastoma is based on pathology with immunohistochemistry (IHC) complemented by urine catecholamines. In all settings, physical examination, basic laboratory, radiographic, and pathologic evaluation should be performed (Table I). Neuroblastoma should be suspected in children with cervical, mediastinal, paraspinal, abdominal or pelvic masses, and the diagnosis made by either biopsy of the primary tumor or by identification of characteristic tumor cells in the bone marrow. The presence of hypertension may suggest neuroblastoma or Wilms tumor. Elevated urine catecholamines, including urine homovanillic acid (HVA), vannilylmandelic acid (VMA), strongly suggest the diagnosis, and is helpful to distinguish among small round blue cell tumors 15. Conversely, if urine catecholamines are not available or are non‐diagnostic, neuroblastoma can be differentiated from other small round blue cell tumors using IHC staining for tyrosine hydroxylase, CD56, and synaptophysin. Alternative diagnosis should be considered for IHC positive staining for leukocyte common antigen CD45 in lymphoma, CD99 in Ewing sarcoma, and desmin and myogenin in rhabdomyosarcoma 16. In Setting 1, IHC is usually not available, and the treating doctor will have to rely on a combination of the pathologic appearance and typical clinical features (e.g., age at diagnosis, symptoms, sites of disease, sites of metastases, lack of response to glucocorticoid therapy).

STAGING AND RESPONSE EVALUATION

Accurate staging and response evaluation are essential for determining treatment and prognosis (Table I). In all settings, bilateral bone marrow aspirate and biopsy should be performed to detect bone marrow metastases, which occur in 70% of metastatic neuroblastoma patients 17. Radiologic and surgical staging depend on locally available resources: Advance imaging, including magnetic resonance imaging (MRI), 123I‐metaiodobenzylguanidine (123I‐ mIBG) scans, and 18FDG‐PET have increased the accuracy of diagnosis and staging and facilitate local control with surgery and radiotherapy, but due to high cost, are inconsistently available in LMIC 18, 19. Surgical staging of the primary tumor is according to the International Neuroblastoma Staging System (INSS‐Supplemental Appendix I) 20. The recent International Neuroblastoma Risk Group (INRG‐Supplemental Appendix I) staging system is a simpler, radiologic staging, which is not influenced by degree of resection 21.

Setting 1

Staging and response assessment depend on imaging studies and bilateral bone marrow biopsy with simple hematoxylin and eosin staining to document presence of disease (Table I). Plain radiographs and abdominal ultrasound can document the extent of thoracic and abdominal disease 22, but are suboptimal for staging and pre‐operative planning due to their lack of sensitivity to detect small metastatic lesions in lymph nodes or lungs. Evaluation for bone disease should include bone radiographs (e.g., a skeletal survey), with the understanding that plain radiographs are less sensitive (26% lesions identified) than bone scan (59% lesions identified) to detect skeletal involvement 23, 24.

Setting 2

Computerized tomography (CT) of neck, chest, abdomen, and pelvis with intravenous contrast combined with 99mTc phosphonate nuclear imaging provides greater sensitivity of anatomic location, associated radiologic risk factors for surgery, and metastasis (Table I). CT is more sensitive for bone involvement than plain radiographs. If spinal involvement is suspected, magnetic resonance imaging (MRI), if available, should be obtained to assess invasion of the neural foramina and spinal cord compression.

Setting 3

In Setting 3, the neuroblastoma staging and risk classification should include CT of neck, chest, abdomen, and pelvis with intravenous contrast; MRI of sites with suspected spinal involvement and sites where tumor extent is not well defined by CT; and 123I‐ mIBG scans, or 18FDG‐PET (Table I). 123I‐ mIBG, which is taken up via the norepinephrine transporter in 90% of patients with neuroblastoma 19 is the most sensitive and specific test for metastatic disease in bone or soft tissue. An 18FDG‐PET scan is also sensitive, but less specific than mIBG for neuroblastoma but can be used in patients whose tumors are 123I‐ mIBG negative or when 123I‐ mIBG scans are unavailable 18.

RISK STRATIFICATION

In the past two decades, risk stratification using clinical and tumor biological features has facilitated intensity‐adjusted neuroblastoma therapy with resultant improvements in outcomes and decreased complications in HIC 22, 25, 26, 27, 28. Established prognostic markers utilized over the past two decades include clinical features such as stage, age, serum lactate dehydrogenase (LDH), serum ferritin; tumor histological classification; and tumor biologic features of MYCN gene amplification status, DNA ploidy and segmental chromosomal aberrations 29, 30, 31.

Histologic classification according to International Neuroblastoma Pathology Classification (INPC) could be helpful in all settings to guide risk stratification 15. Unfortunately, INPC has been underutilized in LMIC due to shortages of trained pathologists familiar with pediatric malignancy and INPC. Fluorescent in situ hybridization (FISH) or RT‐PCR for tumor MYCN gene amplification is desirable, but are most essential in risk stratification for children with stage 4 tumor in patients <18 months of age, and for all stage 3 tumors. DNA ploidy is most helpful in infants without MYCN amplified disease, who have stage 4 or 4 S disease, as patients with diploid tumors have significantly lower survival, and may require more intensive therapy 32, 33.

Prognosis

In HIC, children with localized disease and favorable biology tumors (INSS 1, 2, and 4 S or INRG L1, and MS) have an overall survival (OS) of 96% with surgery alone, or in some cases observation only 20, 21, 22, 31, 34, 35. In the intermediate risk group, comprising of infants with metastatic disease or infants and children with unresectable (INSS 3 or INRG L2) disease without tumor MYCN amplification, the OS remains greater than 90% when treated with 3–6 months of outpatient chemotherapy of moderate intensity 26. Yet in LMIC, survival rates even for localized disease is usually lower 5, 6, 7 perhaps due to inaccurate staging or risk stratification. Lack of access to biologic risk classification may lead to over‐treatment and higher risk of death from aggressive surgical, toxicity, or treatment abandonment. Treating high‐risk disease, comprised mainly of children >18 months with stage 3 and 4 disease and any stage with MYCN amplified tumors is challenging even in HIC, with OS of 40% using modern protocols in settings with excellent supportive care 36, 37, 38, 39, 40. Due to economic, social, and health system factors, high‐risk patients typically form a greater proportion of patients diagnosed with neuroblastoma in LMIC than HIC. In LMIC, especially in Setting 1 centers, high risk patient are often not treated with curative intent, since expected survival rates would be <10% 5, 6.

Adapted Risk Stratification

Guidelines for risk stratification are shown in Table II. Age and stage continue to be the strongest clinical prognostic factors for neuroblastoma. Age <18 months implies favorable disease, even in the presence of metastases, provided there is no tumor MYCN gene amplification 31, 41. In Setting 3, where INPC classification may be performed, toddlers age 12–18 months with stage 4 disease with Unfavorable Histology (if available) and/or diploidy/hypodiploidy should be considered to have high‐risk disease even without MYCN amplification. However, toddlers with stage 3 disease without MYCN amplification should still be treated on the intermediate risk protocol, as the benefit of transplant is not clearly defined 42. Children ≥18 months of age with stage 4 disease should be treated as high‐risk, without further need for biologic or clinical markers 31.

Table II

Adapted Risk Stratification and Treatment Assignment for Neuroblastoma in LMIC

INSS	Initial Status	Risk Group	Age (yr)	LDH	Ferritin	MYCN	Rx S‐1	Rx S‐2	Rx S‐3
1	Resection a	Low	0.5‐21	any	any	any	Resection	Resection	Resection
1	Observation	VLR	<0.5	any	any	any	Observation	Observation	Observation
2A/2B	Resection ≥50%, asymptomatic	Low	any	any b	any b	NA/Unknown b	Observation	Observation	Observation
2A/2B	Resection ≥50%, symptomatic	Intermediate	any	any b	any b	NA/Unk b	IR c x 4 cycles‐80% c	IR c x 4 cycles‐100%	IR c x 4 cycles‐100%
2A/2B	Resection <50%	Intermediate	any	any b	any b	NA/Unk b	IR c x 4 cycles‐80%	IR c x 4 cycles‐100%	IR c x 4 cycles‐100%
2A/2B	Any resection	High	any	any	any	A	HR‐1 or IR c x 8 cycles‐80%	HR‐2	HR‐3
3		Intermediate	<1.5	<750	<120	NA/Unk	IR c x 4 cycles‐80%	IR c x 4 cycles‐100%	IR c x 4 cycles‐100%
3		Intermediate	≥1.5	<750	<120	NA/Unk	IR c x 8 cycles‐80%	IR c x 8 cycles‐100%	IR c x 8 cycles‐100%
3		High	any	≥750 b	≥120 b	A/Unk	HR ‐1 or IR c x 8 cycles‐80%	HR‐2	HR‐3
4		High	<1.5	≥750 b	≥120 b	A/Unk	HR ‐1 or IR c x 8 cycles‐80%	HR‐2	HR‐3 f
4		Intermediate d	<1.5	<750	<120	NA/Unk	IR c x 4 cycles‐80%	IR c x 4 cycles‐100%	IR c x 4 cycles‐100%
4		High	≥1.5	any	any	any	HR ‐1 or IR c ‐8 cycles 80%	HR‐2	HR‐3
4S	Asymptomatic	Low	<1	<750	<120	NA/Unk	Observation	Observation	Observation
4S	Symptomatic e	Intermediate	<1	<750	<120	NA/Unk	IR c x 4 cycles‐80%	IR c x 4cycles‐100%	IR c x 4 cycles‐100%
4S	Asymptomatic or symptomatic	High	<1	any	any	A	IR c x 8 cycles‐80%	HR‐2	HR‐3

VLR, very low risk; LR, low risk; IR, intermediate risk; HR, high risk; A, Amplified; NA, Non‐Amplified; Unk, Unknown.

in low risk patients (INSS Stage 1 and 2), residual tumor (< 50%) is acceptable, as the outcome remains excellent.

If MYCN is unavailable for low risk patients, would assume that they are low risk. LDH and ferritin have not been established as prognostic markers in INSS 2, Unavailability of MYCN with LDH ≥750 units/L and/or ferritin ≥120 ng/μl in stage 3 and 4 disease, would upstage to high risk approach.

IR therapy entails chemotherapy + surgery to achieve tumor response >CR/VGPR/PR. IR will entail a minimum of 4 cycles and upto 8 cycles of intermediate risk therapy per A3961, modified dosing based on respective setting designation.

In Setting 3, toddlers age 12–<18 months with stage 4 disease without MYCN amplification should be considered high risk if the tumor is unfavorable by INPC, if it is diploid, or if LDH and/or ferritin elevated.

Stage 4 S symptomatic patients, inability to obtain biology, consider treatment until symptom resolution, with up to 8 cycles.

In a setting (i.e., Setting 3), where the INPC determination may be feasible, it would be helpful to be guided with combination of data for adequate risk stratification.

One significant limitation in LMIC is the inability to assess MYCN amplification in some centers, despite the fact that it is present in approximately 25% of patients 43 and is critical to differentiate between intermediate and high‐risk disease for patients <18 months with stage 4 disease, and for all patients with stage 3 disease 31. Although MYCN amplification in stage 1, 2, and 4 S tumors could potentially change the risk group, it is uncommon in this group 25. When MYCN assessment is not available, we would consider use of serum LDH and serum ferritin as surrogate markers for MYCN amplification in INSS stage 3 and 4 disease 44. In Children's Cancer Group Study 3891, 70 eligible newly diagnosed patients with Evans stage 3 disease were evaluated for ferritin as a prognostic marker, demonstrating 5‐year EFS 38 ± 9% for patients with ferritin >143 ng/ml versus 78 ± 8% for those with ferritin <143 ng/ml 42, 45. A review of INRG database identified 1483 INSS stage 3 patients, of which ferritin was known in 666 patients, and LDH in 959 patients. Five‐year EFS associated with ferritin <96 ng/ml versus ≥96 ng/ml was 77 ± 3% versus 61 ± 4%, respectively (P = <0.0001), while the 5‐year EFS for patients with LDH <580 U/L versus ≥580 U/L was 78 ± 2% versus 67 ± 3%, respectively (P = <0.0001) 42. Therefore, any patient with stage 3 disease and those <18 months with stage 4 disease whose LDH and/or ferritin exceed this threshold may be stratified as high‐risk 33, 42, 45, 46, 47. If we had to determine a single marker, LDH would be arbitrarily used, as that has the most literature support. The reported threshold at which elevated LDH serves as a prognostic marker varies from >580 IU/L to >1000 IU/L 42, 48; therefore, we recommend the intermediate level of 750 IU/L to intensify therapy in LMIC when MYCN status is unknown. The reported threshold at which elevated ferritin has been shown to affect outcome has varied, from >92 ng/ml to >142 ng/ml 31, 48, 49; therefore we recommend considering a pre‐treatment ferritin ≥120 ng/ml to intensify therapy in LMIC when MYCN status is unknown. For both ferritin and LDH, we chose to use an intermediate rather than the highest value as the proposed threshold, acknowledging differences in laboratory ranges while noting that the largest INRG analysis used the lower value. LDH and ferritin assessments are of uncertain prognostic significance in patients with INSS 1, 2, or 4 S tumors, and therefore not recommended for changing the risk category in those patients 22, 25, 28, 31, 34.

TREATMENT GUIDELINES

Treatment guidelines according to risk group and setting are outlined in Table II.

Low Risk Disease

In patients with INSS 1 and 2 tumors, the outcome is excellent, with >95% 3 year OS 27, 50. Regardless of setting, surgical resection is sufficient for asymptomatic patients whose tumor is largely (>50%) resected, even if some residual tumor remains. Patients with tumor resection can be monitored without further therapy, with ultrasound or CT every 3 months for the 1st year and then every 6 months until 2 years after diagnosis 27. Perinatal adrenal neuroblastoma (Very Low Risk—VLR) patients, with small adrenal masses detected before birth or within the first six post‐natal months, have an especially favorable outcome with >95% 4 year OS, without any surgical or medical intervention 35. We recommend monitoring these patients with ultrasonography and physical exam every 6–8 weeks, until the mass resolves, or until significant growth warrants surgical resection. Occasionally, these patients will progress to INSS 4 S, but they can be safely observed as long as they remain asymptomatic 51, 52.

Infants with stage 4 S asymptomatic disease have an overall 5 years survival 100% for those patients that did not require chemotherapy 51. For patients with symptomatic stage 4 S neuroblastoma with hepatic dysfunction, coagulopathy, respiratory compromise, or renal impairment, we recommend treatment with intermediate risk therapy. Infants younger than 3 months with hepatomegaly even without symptoms deserve prompt intermediate risk treatment until symptoms abate due to the higher risk of death at this age 33, 51.

Intermediate Risk Disease

There have been significant advances in the therapies for children with intermediate risk neuroblastoma, allowing de‐intensification of therapy while maintaining survival. Children with stage 3 intermediate risk neuroblastoma or infants with stage 4 were treated in the 1990s with cisplatin, etoposide, doxorubicin, and cyclophosphamide over 9 months, with a 4‐year EFS of 100% for favorable stage 3 53 and 3‐year EFS >93% for infants with INSS 4, MYCN‐Non amplified disease 54. The poor prognostic indicators identified in the infant group were diploid and/or MYCN‐amplified tumors 32, 55. A French cooperative trial showed that INSS 3 infants without MYCN gene amplification had a 94 ± 5% EFS with moderate‐intensity chemotherapy 56. A large INRG report showed that patients with INSS stage 3 tumors without MYCN amplification had an 81% 5‐year EFS 42.

Therapy has since been modified to shorten the course of treatment, substitute carboplatin for cisplatin, and decrease the cumulative doses of active agents by 60% in an effort to decrease toxicities, hospital stay, and costs, while maintaining survival outcomes, with 88% 3‐year EFS and 96% OS 26. In settings where carboplatin is unavailable or is cost‐prohibitive, cisplatin can be an alternative. Table III outlines a well‐tolerated intermediate‐risk therapy suitable for each setting 26. We recommend dose reduction as outlined in Table II according to availability of resources given the typically favorable outcome and potential need to conserve blood products.

Table III

Intermediate Risk Regimen for Setting 1, 2, and for High Risk Setting 2

Risk Group	Cycle a	Chemotherapy b	Dose c and Administration
IR 1	1	Carboplatin (D1)	560 mg/m2 (18 mg/kg) in 125 mL/m2 D5 ½ NS IV over 1 hour day 1
		Etoposide (D1‐3)	120 mg/m2 (4 mg/kg) in 300 mL/m2 D5 ½ NS IV over 2 hours (maximum concentration 0.4mg/mL) Days 1, 2, 3 immediately after Carboplatin infusion is completed
	2	Carboplatin (D1)	560 mg/m2 (18 mg/kg) in 125 mL/m2 D5 ½ NS IV over 1 hour day 1
		Cyclophosphamide (D1)	1,000 mg/m2 (33 mg/kg) in 125 mL/m2 D5 ½ NS IV over 1 hour day 1 after Carboplatin
		Doxorubicin (D1)	30 mg/m2 (1 mg/kg) in 125 mL/m2 D5 ½ NS IV over 15‐60 minutes day 1 after cyclophosphamide
	3	Cyclophosphamide (D1)	1,000 mg/m2 (33 mg/kg) in 125 mL/m2 D5 ½ NS IV over 1 hour day 1
		Etoposide (D1‐3)	120 mg/m2 (4 mg/kg) in 300 mL/m2 D5 ½ NS over 2 hours daily days 1,2,3 after cyclophosphamide
	4	Carboplatin (D1)	560 mg/m2 (18 mg/kg) in 125 mL/m2 D5 ½ NS IV over 1 hour day 1
		Etoposide (D1‐3)	120 mg/m2 (4 mg/kg) in 300 mL/m2 D5 ½ NS over 2 hours daily days 1,2,3 after Carboplatin
		Doxorubicin (D1)	30 mg/m2 (1 mg/kg) in 125 mL/m2 D5 ½ NS IV over 15‐60 minutes day 1 after Etoposide
	Surgery d
	If feasible
IR 2	5	Carboplatin (D1)	560 mg/m2 (18 mg/kg) in 125 mL/m2 D5 ½ NS IV over 1 hour day 1
		Etoposide (D1‐3)	120 mg/m2 (4 mg/kg) in 300 mL/m2 D5 ½ NS over 2 hours daily days 1,2,3 after Carboplatin
	6	Carboplatin (D1)	560 mg/m2 (18 mg/kg) in 125 mL/m2 D5 ½ NS IV over 1 hour day 1
		Cyclophosphamide (D1)	1,000 mg/m2 (33 mg/kg) in 125 mL/m2 D5 ½ NS IV over 1 hour day 1 after Carboplatin
		Doxorubicin (D1)	30 mg/m2 (1 mg/kg) in 125 mL/m2 D5 ½ NS IV over 15‐60 minutes day 1 after Cyclophosphamide
	7	Carboplatin (D1)	560 mg/m2 (18 mg/kg) in 125 mL/m2 D5 ½ NS IV over 1 hour day 1
		Etoposide (D1‐3)	120 mg/m2 (4 mg/kg) in 300 mL/m2 D5 ½ NS over 2 hours daily days 1,2,3 after Carboplatin
	8	Cyclophosphamide (D1)	1,000 mg/m2 (33 mg/kg) in 125 mL/m2 D5 ½ NS IV over 1 hour day 1
		Doxorubicin (D1)	30 mg/m2 (1 mg/kg) in 125 mL/m2 D5 ½ NS IV over 15‐60 minutes day 1 after Cyclophosphamide
	Surgery End of treatment
Follow up q 3 months x 4, q 6 months x 4, then yearly x 2 with physical exam, urinalysis, VMA/HVA, ultrasound

NS, Normal Saline; IV, intravenous; IR, intermediate risk, Additional Notes: IR1 will be minimum of 4 cycles of intermediate risk therapy per A3961, modified dosing based on respective setting designation. IR2 will be up to 8 cycles of intermediate risk therapy per A3961, modified dosing based on respective setting designation.

Cycles are administered at minimum of 3 week intervals and start after absolute neutrophil count is >1000/μl and platelet count is >100,000/μl. Prior to each cycle, CBC, creatinine, electrolytes, ALT, bilirubin, urinalysis. End of therapy CBC, creatinine, ALT, bilirubin, echocardiogram, imaging (based on setting), CT, mIBG (or bone scan).

Recommended to prehydrate with D5 ½ normal saline at 125 ml/m2/hr × 2 hr prior to cyclophosphamide cycles and achieve Urine Specific gravity ≤1.010 prior to starting cyclophosphamide; post hydration with D5 ½ normal saline at 125 ml/m2/hr after chemotherapy with 2 hr hydration after carboplatin cycles and 4 hr of hydration after cyclophosphamide containing cycles.

Doses are adjusted to mg/kg for infants <10 kg. For Setting 1, it is recommended to reduce the dose so that only 80% of indicated dose is given, to reduce the likelihood of fever and neutropenia or platelet requirements.

After 4 cycles, evaluate for surgery, obtain CBC, creatinine, urinalysis CT, mIBG. The goal of therapy is to achieve resolution of metastases and primary tumor response of >50% with chemotherapy and surgery.

Aggressive surgery may cause significant morbidity in patients with intermediate risk neuroblastoma. Of 235 children with intermediate risk group on the COG A3961 trial, 28% had surgical complications including major hemorrhage, vascular and renal injuries, and intra‐operative need for organ resection 26. Subsequent pilot investigations showed that intermediate risk patients will often mature without progression despite residual unresected tumor 57. Given the high survival rate for the intermediate group regardless of resection extent 49, we recommend chemotherapy for 4–8 cycles depending on control of metastases and tumor reduction with conservative surgery performed if needed to achieve >50% primary tumor response. Radiation therapy is not recommended even in the setting of residual disease. Patients can be monitored with ultrasound or CT every 3‐months for the 1st year and then every 6 months until 2 years after diagnosis.

High Risk Disease

Despite numerous advances and multimodal therapies in the treatment of children with neuroblastoma, those with high‐risk disease continue to be our greatest challenge. The incorporation of three distinct phases of therapy has improved outcome for high‐risk neuroblastoma: intensive induction treatment, myeloablative chemotherapy with autologous hematopoietic stem cell rescue, and treatment of MRD 58, 59, 60. The goal of induction therapy is to achieve maximum reduction of tumor burden, including reduction of metastatic bone marrow, and bone disease within a timeframe that will minimize the risk of developing resistant tumor clones and clinical progression. Subsequently, potentially resistant residual tumor is treated with high‐dose myeloablative therapy followed by autologous hematopoietic stem cell transplantation (ASCT), which has been shown to significantly improve EFS 58. The relapse rate of greater than 40% even after such therapies has led to post‐transplant treatment for MRD‐positive patients using isotretinoin and monoclonal anti‐GD2 antibodies 39, 58, 59.

Due to the lack of transplant expertise, reduced access to blood products and pheresis, and difficulty supporting patients through the period of myelosuppression, few Setting 1 centers in LMIC have attempted curative treatment of high‐risk neuroblastoma, and have commonly prescribed palliative care. Furthermore, the cost of isotretinoin and lack of access to monoclonal anti‐GD2 antibody impede effective MRD treatment in some LMIC. However, a coordinated treatment approach based on resource capabilities can improve outcome for these children progressively, and will eventually lead to overall improvement in supportive care and therapy of childhood cancer. We propose a graduated intensity plan for high‐risk patients in different LMIC settings (Table IV).

Table IV

Induction Therapy for High‐Risk Neuroblastoma in Setting 3

Regimen: POG 9341‐Modified	Chemotherapy	Dose and Administration
Course 1	Cisplatin	40 mg/m2/dose (1.33 mg/kg) a in 125 ml/m2 NS containing 3 g/m2 mannitol, IV over 1 hr, day 1‐5
	Etoposide b	100 mg/m2/dose (3.3 mg/kg) a BID 250 ml/m2 NS IV over 2 hr, day 1‐3
	Hydration	D5 1/2NS + KCl 30 mEq/L + MgSO4 500 mg/L + Ca Gluconate 250 mg/L at 200 mL/m2/hr for 2 hours prior and 6 hours after each dose of Cisplatin
Course 2	Vincristine	1.5 mg/m2/dose (.05 mg/kg) a IV push days 1, 8, 15
	Cyclophosphamide a	1 g/m2/dose (33 mg/kg) a IV in 125 mL/m2 D5 ½ NS IV over 1 hour days 1and 2.
	Doxorubicin	60 mg/m2 /dose (2 mg/kg) IV over 15 min day 1
	Hydration b	D5 ½ NS IV at 125 ml/m2/hr for 2 hours prior to and 4 hours after each dose of cyclophosphamide
Course 3	Ifosfamide a	2 g/m2/day (66.6 mg/kg) a IV over 1 hour Days 1‐5 with MESNA 400 mg/m2 (13.3 mg/kg) in 200 ml/m2 D5 1/2 NS.
	Etoposide	75 mg/m2 (2.5 mg/kg) a IV/1 hour in 200 ml/m2 D5 1/2 NS days 1‐5
	Hydration b	Prehydration: D5 ½ NS IV at 125 ml/m2/hr for 2 hours and Post hydration: D5 1/2 NS at 150 ml/m2/hr and MESNA over 15 min at hr 4, 7, 10 on days 1‐5.
PBSC collection
Course 4	Carboplatin	500 mg/ m2/day (16.7 mg/kg) a IV days 1,2
	Etoposide	75 mg/m2 (2.5 mg/kg) a BID IV/1 hour in 200 ml/m2 D5 1/2 NS
	Hydration b	Post hydration with D5 1/2 NS at 125 ml/m2/hr x 2 hours
Course 5	Cisplatin	40 mg/m2/dose (1.33 mg/kg) a in 125 ml/m2 NS containing 3 g/m2 mannitol, IV over 1 hr, day 1‐5
	Etoposide	100 mg/m2/dose (3.3 mg/kg) a BID 250 ml/m2 NS IV over 2 hr, day 1‐3
	Hydration b	D5 1/2NS + KCl 30 mEq/L + MgSO4 500 mg/L + Ca Gluconate 250 mg/L at 200 mL/m2/hr for 2 hours prior and 6 hours after each dose of Cisplatin
Surgery if ≤ 5 metastatic sites remaining

NS, Normal Saline; IV, intravenous.

Doses are adjusted to mg/kg for infants <10 kg.

Achieve Urine Specific gravity ≤1.010 prior to starting cyclophosphamide or ifosfamide. Granulocyte colony stimulating factor should be considered 24 hr after completion of chemotherapy and continued until post‐nadir ANC >1,500.

Zage PE, Kletzel M, Murray K, et al. Outcomes of the POG 9340/9341/9342 trials for children with high‐risk neuroblastoma: A report from the Children's Oncology Group. Pediatr Blood Cancer 2008:51(6):747‐753.

Induction and Local Control

No single chemotherapy induction regimen has proven to be superior for induction response and EFS in high‐risk neuroblastoma. All regimens reported to date employ multi‐agent chemotherapy with a platinum agent, alkylator therapy such as cyclophosphamide, topoisomerase therapy, usually with etoposide or less frequently, doxorubicin, and vincristine. Initial reports suggested that increasing dose intensity might improve response rate, but this has not been borne out in single arm trials which all showed response rates from 70 to 80% 39, 61, 62, 63. The dose‐intensive rapid COJEC 64 did show an improvement in EFS compared in a randomized trial to OPEC–OJEC, though the 3‐year EFS of 31% was not better than that reported with other North American and European regimens 40.

Given the increased need for supportive care with more intensive regimens and without significant benefit expected for patients without the option of myeloablative therapy, we have recommended induction regimens in Setting 1 using a regimen with low dose metronomic oral cyclophosphamide 25 mg/m2 daily (max 50 mg) or oral etoposide 50 mg/m2/day for palliation of pain and improvement of quality of life, High Risk Setting 1 (HR‐1) 65, 66. An alternate approach non‐curative therapy would entail administering 1–2 cycles of intermediate risk chemotherapy and local radiation to the primary tumor bed (dosing per Tables III and V) to allow for some tumor reduction, pain relief and improved quality of life. Aggressive pain management is strongly recommended, as end‐of‐life bone pain can be significant 67.

Table V

Consolidation and MRD Therapy for High Risk Neuroblastoma

Setting	Consolidation	Local Radiation	MRD‐directed therapy
Setting 1	Cyclophosphamide 25 mg/m2/day PO or etoposide 50 mg/m2/day PO	21 Gy to primary tumor and symptomatic metastases d	none
Setting 2	Repeat cycles 1‐4 of P9341 (Table IV) a , b , c	21 Gy to Primary tumor bed and up to 5 residual bone metastases d	Isotretinoin 80 mg/m2/dose PO BID 14 days of each 28 day cycle for 6 cycles
Setting 3	ASCT with busulfan/melphalan (see supplemental Table 1)	21 Gy to primary tumor bed and up to 5 residual bone metastases d	Isotretinoin 80 mg/m2/dose PO BID 14 days of each 28 day cycle for 6 cycles
Follow up	End of therapy US or CT of the primary site, bone scan (setting 1,2) or mIBG scan (Setting 3), bone marrow biopsy, urine VMA/HVA, echocardiogram, hearing test (setting 3). After completion of therapy, monitor q 3 months for 2 years then q 6 months for 3 years with US/CT, urine VMA/HVA.

ASCT, Autologous stem cell transplant; US, Ultrasound; CT, Computed tomography; VMA, Vannillylmandelic acid; HVA Homovanillic acid; Gy, Gray.

Hydration guidelines: per Table IV.

Achieve Urine Specific gravity ≤1.010 prior to starting cyclophosphamide or ifosfamide and use hydration guidelines as outlined in Table IV.

Granulocyte colony stimulating factor should be considered 24 hr after completion of chemotherapy and continued until post‐nadir ANC >1,500.

Gross residual disease should receive additional 14.4 Gy boost at the end of consolidation therapy.

For Setting 2, we recommend either the same regimen used for intermediate risk patients, (Table III ), for 8 cycles or modified‐Pediatric Oncology Group (POG) 9341 induction regimen, (Table IV ) 26, 36. For Setting 3, we suggest the modified‐POG 9341 induction regimen, which is more intensive than prior French CAdO‐PE 68, and has fewer reported acute toxicities than N7 63 or the SIOPEN rapid COJEC 40. It optimizes platinum, uses less doxorubicin, and has shorter hospitalization (primarily given in a day hospital) and higher response rate than the CCG 3891 protocol 58. This protocol is currently being used in Morocco for high‐risk patients and has been tolerable in >60 patients without toxic death 69. An alternative option is to administer the OPEC–OJEC regimen that incorporates more carboplatin but avoids doxorubicin; although response rates are lower than those reported for POG 9341 40. Surgery to resect the primary tumor is recommended after the 4th or 5th induction cycle in patients, whose metastases have responded significantly to chemotherapy, including either a complete metastatic response or a partial response (PR) and fewer than six residual bone metastases that might be treated with radiotherapy. In a review of patients with high risk neuroblastoma on the CCG A3891 trial, a statistically significant improvement was noted in patients with stage 4 neuroblastoma achieving complete remission after surgery with a 5 year EFS 26% ± 4% versus 19% ± 3% 70 although there are conflicting evidence as to the benefit of complete resection when radiotherapy is also incorporated 71, 72. Given the minimal likely benefit, we recommend avoiding mutilating surgery likely to risk major complications in LMIC and instead using radiation therapy to control unresectable primary tumor and residual bone metastases.

The decision to proceed to a more intensive consolidation should be based on the response to induction chemotherapy. Multiple analyses of myeloablative regimens show that the outcome is significantly better for patients who have achieved at least a partial response at the end of induction 73, and preferably a good partial response in metastases as manifested by a low MIBG Curie score 74.

Consolidation

The recommended consolidation for curative intent is myeloablative therapy with ASCT, based on three randomized trials 39, 75, 76.

Preliminary results from the European SIOP‐EN HR NBL‐1 randomized phase III trial comparing busulfan and melphalan (Bu/Mel) versus carboplatin, etoposide, and melphalan (CEM) as a preparative regimen for autologous transplant, showed improved survival in patients treated on the Bu/Mel arm 77, 78, 79. Due to its lesser toxicity and apparently better EFS, we recommend that Setting 3 centers with transplant capabilities utilize either BuMel, (Supplemental Appendix II), or consider melphalan alone (180 mg/m2), as this is the active agent that has been the backbone of almost all neuroblastoma transplant regimens, and was successful in the first randomized trial 76. One caveat for those utilizing the Bu/Mel regimen is the higher risk of sinusoidal obstructive syndrome (SOS), which may be more severe in LMIC where options for potential prophylaxis or treatment such as defibrotide are not readily available 80.

We recommend that high‐risk patients in Setting 1 continue on their same palliative chemotherapy used in induction for a total of 12 months. In Setting 2, a tolerable consolidation therapy would be 4 cycles of modified POG 9341 regimen (Table IV), the French CAdO/CE regimen for 2 cycles each 81 or cyclophosphamide/topotecan for 6 cycles at 3–4–week intervals (Supplemental Appendix III) followed by radiation therapy to the primary tumor bed and residual bone metastases 68. For Setting 3, if resources are available for ASCT, then a peripheral blood stem cell (PBSC) harvest could be performed after 2–5 cycles of induction. An earlier timing of apheresis generally leads to improved feasibility in obtaining sufficient PBSC numbers. If apheresis is not available, autologous bone marrow could be collected and stored providing that metastases have cleared. If resources to conduct ASCT are not available, then a consolidation using outpatient‐based cyclophosphamide/topotecan for 6 cycles at 3–4 week intervals should be considered (Supplemental Appendix III), as this combination was successful in obtaining response in 30% of resistant or relapsed patients 82. An alternative therapy option would be to continue induction therapy (POG 9341) for another 4 cycles. It is reasonable to continue such chemotherapy with curative intent, as there were still >20% EFS for patients treated with intensive consolidation chemotherapy on the randomized CCG protocol 39. We recommend administration of 21.6 Gy radiation to the primary tumor bed and to residual bone metastases (fewer than six sites) at the end of consolidation phase of therapy 83, 84.

Minimal Residual Disease Therapy

Although myeloablative therapy has improved the outcome, up to 40% of patients will relapse after this therapy. The CCG reported the benefit of treating children after myeloablative therapy with a differentiating agent, isotretinoin, with significant improvement in EFS for all children post consolidation 58. Isotretinoin for 6 months is recommended in all settings as post‐consolidation therapy for patients who have achieved a complete remission (Table V). It is not effective, and not recommended, for patients with gross residual disease. We would urge particular caution before prescribing isotretinoin for women of child‐bearing age to review risks and reinforce pregnancy prevention measures, as there is a serious risk of birth defects in exposed pregnancies. Currently, the anti‐GD2 antibody Ch14.18 based trials, which has been shown to further improve outcome, is not commercially available and requires intensive supportive nursing care for administration, thus limiting access in LMIC.

Recurrent/Progressive Disease and Alternative Regimens

Despite intensive upfront multimodal therapy, over 50% of high‐risk neuroblastoma patients continue to relapse. After relapse these patient have poor outcomes with 5‐year OS < 10% 85. Regimens such as cyclophosphamide combined with topotecan or irinotecan combined with temozolamide have demonstrated response rates for refractory/relapse disease of 32% and 15%, respectively 82, 86. 131I‐mIBG therapy is also effective in relapsed disease, with a 30% response rate, but is available in few LMIC 87. External beam radiation therapy will often effectively control symptomatic sites of disease 88. Oral metronomic therapies may be options for palliation for settings without access to camptothecins or temozolomide 66, 89. Localized recurrence, particularly in low and intermediate risk disease, may be salvageable with additional surgery, radiation, and chemotherapy 26, 27, 90. The determination of therapy is guided by timing of recurrence, previous therapies received, individual patient factors and preferences of the family and providers, in addition to the resources available within those centers.

SUPPORTIVE CARE

Ability to care for children with fever and neutropenia is critical in terms of life threatening complications. Therefore aggressive multi‐agent therapy should only be given at center capable of ensuring that patients have rapid access to medical care and broad spectrum antibiotic therapy either by providing affordable housing in close proximity to the medical center or at a facility close to patients' home. It is equally important to educate on fever management with handouts provided to patients and families at each visit 13. For Setting 3, a central venous catheter should deliver intensive therapy and blood products. It is also recommended in Setting 3 that a pediatric intensive care unit is located within the treating center. Ongoing education should be provided for nursing and medical assistants regarding the care of a child with cancer. Red blood cell products and antibiotics should also be readily available in all settings.

ADVANCING TREATMENT INTENSITY

The optimal intensity of therapy at each institution should be determined based on the center's own toxicity data, including grade 3 or 4 infection, sepsis, renal impairment, and death. Review of relapses may also guide the decision to adjust therapy, since reduction of intensity may produce a modest decrease in toxic death but a larger increase in relapse. HIC currently report toxic death rates below 5% for both induction and consolidation therapies 91. A toxic death rate >5% for induction or 10% for consolidation should trigger careful assessment and use of lower treatment intensity if supportive care cannot be improved. Reviewing institutional data on abandonment rates may help guide practices to promote patient adherence.

CONCLUSION

Risk classification using an adapted approach that takes into consideration limitations in diagnostic imaging and tumor biology allows risk assignment for children with neuroblastoma, regardless of their setting, and selection of feasible and appropriate treatment. This approach does not preclude active efforts to implement the full range of capabilities in every pediatric cancer unit, but can improve outcomes for those patients who must be treated with the current resources.

Additional Supporting Information may be found in the online version of this article at the publisher's web‐site.

Supplemental Appendix I: Supplemental Appendix I.International Neuroblastoma Staging System (INSS) and International Neuroblastoma Risk Group Staging System (INRG).

Click here for additional data file.

Supplemental Appendix II: Consolidation Regimen Setting 3: Bone Marrow Transplantation Conditioning.

Click here for additional data file.

Supplemental Appendix III: Alternative Consolidation Regimen Options for Setting 2 and 3.

Click here for additional data file.