Neuroblastoma in Africa: A Survey by the Franco-African Pediatric Oncology Group.

PURPOSE: Neuroblastoma is a sympathoadrenal lineage neural crest-derived tumor. It is the third most common childhood malignancy in the Western world. Studies from the United States show that black patients with neuroblastoma have a greater proportion of high-risk neuroblastoma with poorer prognosis compared with white patients. In Africa, there are few published data on the epidemiology and management of neuroblastoma. The primary aim of this study was to assess the diagnostic and therapeutic resources available for the management of neuroblastoma within the Franco-African Pediatric Oncology Group (GFAOP).
METHODS: A survey was conducted in the pediatric oncology centers of the GFAOP. Participating GFAOP centers were Abidjan, Algiers, Bamako, Dakar, Lubumbashi, Lomé, Ouagadougou, Rabat, Tananarive Antananarivo, and Tunis. Questionnaires were sent out by e-mail to the principal investigators at each participating GFAOP center in December 2013.
RESULTS: Ten (62%) of 16 GFAOP centers responded to the questionnaire. Neuroblastoma represented only 3% to 5% of childhood cancers in the sub-Saharan African centers, with the exception of Antananarivo, where it represented 7.5%. In contrast, in the northern African centers of Tunis, Rabat, and Algiers, neuroblastoma accounted for 30%,10%, and 7% of childhood cancer, respectively. At initial diagnosis, 50% to 80% of patients had metastatic neuroblastoma in eight of 10 centers.
CONCLUSION: Based on this survey, neuroblastoma seems to be less common in sub-Saharan Africa. The proportion of patients with metastatic neuroblastoma seems to be higher than reported in Western countries.

INTRODUCTION

Neuroblastoma is a family of malignant embryonal tumors of young children derived from the neural crest that affects the sympathetic chain and/or the adrenal medulla.[1,2] It accounts for 7% of childhood cancer and is the third most common childhood malignancy after acute lymphoblastic leukemia and brain tumors in developed countries.[3] On the basis of prognostic factors (age, stage, ploidy, MYCN status, and Shimada histologic classification), the Children’s Oncology Group stratifies neuroblastoma into low, intermediate, and high risk. Although intermediate- and low-risk patients have excellent survival rates of 85% to 100%, less than 50% of high-risk patients achieve long-term survival despite intensive multimodal therapy including chemotherapy, surgery, autologous stem-cell transplantation, radiation, and immunotherapy.[4]

The incidence of neuroblastoma is lower in black populations compared with other ethnic groups.[5,6] However, black patients seem to have a higher proportion of high-risk disease associated with sperm-associated antigen 16 single nucleotide polymorphism, which carries a poorer prognosis.[7-9]

The Franco-African Pediatric Oncology Group (GFAOP) is a consortium made up of pediatric oncology centers in francophone African countries with the primary aim of supporting the management of childhood cancer in Africa. The treatment of childhood cancer in Africa has made significant progress thanks to the efforts of the GFAOP.[10,11] Pediatric oncology units have been developed in participating countries by this consortium, but as a result of limited resources, attention has mostly focused on the most common pediatric oncology pathologies that are less resource intensive and have generally good prognosis, such as acute lymphoblastic leukemia, Burkitt lymphoma, and Wilms tumor.

In Africa, and particularly in sub-Saharan Africa, there is hardly any epidemiologic or survival data on neuroblastoma.[12,13] We describe the results of a survey initiated by the GFAOP to assess the number of new patients with neuroblastoma, as well as available resources for the management of neuroblastoma in member institutions. The study also assessed the feasibility of a multicenter protocol to harmonize the management of neuroblastoma within the GFAOP.

METHODS

This survey was carried out in December 2013. Questionnaires (Data Supplement) were sent out by e-mail to the principal investigators of member institutions of the GFAOP. The questions focused on epidemiology, diagnostic imaging, tumor biology, histopathology, therapeutic options, and health care finance. The data generated were analyzed using Epi Info 7 (Centers for Disease Control and Prevention, Atlanta, GA).

RESULTS

Ten of 16 centers responded to the questionnaire, a response rate of 62%. Participating centers include Abidjan, Algiers, Bamako, Dakar, Lubumbashi, Lomé, Ouagadougou, Rabat, Tunis, and Antananarivo. The numbers of new patients treated at each center per year were as follows: Abidjan, Lubumbashi, Lomé, and Ouagadougou, fewer than five patients; Bamako, Algiers, and Dakar, five to 20 patients; Tunis and Antananarivo, 20 to 30 patients; and Rabat, 30 to 40 patients. Neuroblastoma represented only 3% to 5% of childhood cancers in the sub-Saharan African centers, with the exception of Antananarivo, where it represented 7.5%. In contrast, in the northern African centers of Tunis, Rabat, and Algiers, neuroblastoma accounted for 30%, 10%, and 7% of childhood cancer, respectively (Table 1). The median age at diagnosis was 4 years (range, 8 months to 6 years), with a male-to-female sex ratio of 2:1; the median duration of symptoms before diagnosis was 3 months. At initial diagnosis, 50% to 80% of patients had metastatic neuroblastoma at all centers except Ouagadougou and Rabat, where 20% to 50% of patients had metastatic disease. Plain radiography and ultrasonography were available at all centers. Although all centers had either magnetic resonance imaging or computed tomography scans, not all patients received either modality in their initial treatment evaluation. None of the sub-Saharan centers had metaiodobenzylguanidine (MIBG) scans. Algiers, Tunis, and Rabat centers had MIBG scans, but only in Algiers did all patients receive evaluation with MIBG scans because the cost is borne entirely by the government. None of the sub-Saharan centers surveyed had the ability to measure urinary catecholamines locally, and therefore, they shipped samples overseas, which resulted in long turnaround times ranging from 35 to 45 days. However, all the northern African centers (Tunis, Rabat, and Algiers) run urine catecholamines locally.

Table 1

Distribution of Patients With Neuroblastoma at GFAOP Centers

Morphologic evaluation of the primary tumor by hematoxylin and eosin is available at all centers. However, immunohistochemical characterization is only feasible at the northern African centers of Tunis, Rabat, and Algiers. Only the Rabat and Tunis centers had the resources to run fluorescence in situ hybridization or real-time quantitative polymerase chain reaction for determination of MYCN amplification in neuroblastoma.

Most centers used either a local or some form of modified Western/international protocol for treating patients with neuroblastoma. Only three centers, all sub-Saharan, did not have a neuroblastoma protocol (Table 2).

Table 2

Treatment Protocol and Survival in Different GFAOP Pediatric Oncology Centers

In sub-Saharan Africa, families paid for all medical care expenses, in contrast to Algiers where medical care was entirely government funded. In Rabat, Tunis, and Madagascar, the cost of medical care was shared between families and the government.

DISCUSSION

On the basis of this survey, the average incidence of neuroblastoma is 10 cases per year per pediatric oncology center in sub-Saharan Africa. This figure is probably an underestimation because lack of health insurance, difficult access to health services for populations in remote areas, poverty, and the use of traditional medicine mean that many patients with cancer do not reach pediatric oncology centers.[14] The incidence of neuroblastoma was highest (30%) in Tunis. However, this may be falsely high because patients with leukemia are treated in adult units and, therefore, are not included in the number of new cancer diagnoses.

The higher proportion of metastatic neuroblastoma at presentation in sub-Saharan centers would strengthen the hypothesis of a genetic predisposition for neuroblastoma in black patients. Many of these metastatic patients have high-risk neuroblastoma and, therefore, have a poor prognosis.[15,16]

The diagnosis of neuroblastoma is based on the histology of the primary tumor or on elevated urine catecholamines (homovanillic acid and vanillylmandelic acid) and involved bone marrow by morphology.[17] Sub-Saharan African centers are unable to perform immunohistochemistry for diagnosis of neuroblastoma, so diagnosis is based on the finding of small round blue cells on hematoxylin and eosin stains and elevation of urine catecholamines. Unfortunately, urine catecholamines are unable to be run locally, and therefore, urine samples are shipped to Western countries, resulting in considerable delay in diagnosis.

MYCN amplification confers poor prognosis in patients with neuroblastoma and is an extremely important prognostic factor for risk stratification to tailor therapy.[18] Among the GFAOP centers, molecular tests for MYCN amplification are currently only feasible at the Rabat and Tunis centers. MYCN-amplified tumors are generally considered high risk and are treated intensively with induction chemotherapy, surgical resection of primary tumor, autologous stem-cell transplantation, involved-field radiation, and maintenance therapy with 13-cis-retinoic acid either alone or in combination with cytokines and anti-GD2 antibody.[19,20] As a result of the prohibitive financial burden, lack of transplantation expertise, and unavailability of adequate supportive care during the period of prolonged myelosuppression, this therapy cannot be adopted currently by most centers of the GFAOP. The use of chemotherapy agents in small or metronomic doses has gained interest in recent times.[21] In most centers in Africa, this metronomic therapeutic approach would be a reasonable palliative alternative for patients with high-risk neuroblastoma.[22,23] The treatment of localized neuroblastoma is essentially surgical in the absence of MYCN amplification, with a cure rate of greater than 90%.[24,25] Given that surgical resection of the primary tumor is not an issue in GFAOP centers, attention should focus on treatment of localized disease while a new risk stratification system and a protocol applicable to all centers are being developed by the GFAOP. The survival outcomes of various treatment protocols (local institution or modified international protocols) used in different GFAOP (Table 2) are currently unsatisfactory, making the case for development of a GFAOP protocol that takes into account the resources of member institutions.

In summary, this survey gives an estimate of the epidemiology of neuroblastoma in Africa, the resources available for management, and the generally poor outcomes. The limitation of these data is that they are based on a survey, which is subject to recall bias. The GFAOP is currently working with member institutions to collect data on patients with neuroblastoma and also to develop treatment protocols that can be used across centers.