The pre-surgical factors that determine the decision to proceed to resection in children diagnosed with high-risk neuroblastoma in a resource limited setting.

Surgical control has prognostic value in neuroblastoma (NB). Advanced NB is common at diagnosis in South Africa. We investigated the pre-surgery factors that influenced decisions to perform surgical resections. We included 204 patients with high-risk NB from a national retrospective study, who completed induction chemotherapy between 2000 and 2016.The median age was 32.4 months (IQR 15.1 - 53.5 months). Primary tumor resection was achieved in 76.9% of patients between 0-18 months of age, 51.8% between 18-60 months and 51.7% older than 60 months (p < 0.001). Only 43.2% of patients with distant metastatic disease had surgery done (p < 0.001). LDH was >750 U/L in 46.8% and ferritin >120 g/dL in 53.1% of those who had surgery (p = 0.005). The majority (80.4%), who had achieved post-induction metastatic complete remission (mCR), were operated, while 28.7% without mCR had surgery (p < 0.001). The long-term overall survival in patients with mCR and primary tumor resection was 36.5% compared to those with mCR without primary tumor resection (25.4%) and without mCR (≤3.0%)(p < 0.001). Age (p < 0.001), stage (p < 0.001), mCR (p < 0.001) and treatment setting (p < 0.001) were of prognostic significance. The tumor site and MYCN-amplification did not significantly predict resection rates. Post-induction mCR and stage were associated with surgical resection and five-year OS (p < 0.001) on multivariate analysis.Patients with high-risk NB who achieved mCR and had primary tumor resections are curable in limited resourced settings. Stage and post-induction mCR were significant variables that led to surgery. These variables should be included as indications in the management of metastatic NB in resource limited settings.

High-risk neuroblastoma that achieved post-induction chemotherapy metastatic remission and have undergone resection, is curable, even in limited resource settings.Achieving metastatic complete remission was the only factor that significantly predicated if surgery was done.The age at diagnosis, stage and hospitals with expertise in neuroblastoma surgery were of prognostic significance in South Africa.If a patient with high-risk neuroblastoma achieves metastatic complete remission in a resource limited setting, it should be an indication for resection of the primary tumor.

The implementation of a national paediatric oncology protocol for neuroblastoma in South Africa.

PURPOSE: The aim of the World Health Organization-International Paediatric Oncology Society is to improve childhood cancer survival in low- and middle-income countries to 60% by 2030. This can be achieved using standardised evidence-based national treatment protocols for common childhood cancers. The aim of the study was to describe the development and implementation of the SACCSG NB-2017 neuroblastoma (NB) treatment protocol as part of the treatment harmonisation process of the South African Children's Cancer Study Group. METHODS: The Consolidated Framework for Implementation Research was used to identify factors that could influence the implementation of the national NB protocol as a health care intervention. The evaluation was done according to five interactive domains for implementation: intervention characteristics, inner setting, outer setting, individual or team characteristics and the implementation process. RESULTS: The protocol was developed over 26 months by 26 physicians involved in childhood cancer management. The process included an organisational phase, a resource identification phase, a development phase and a research ethics approval phase. Challenges included nationalised inertia, variable research ethical approval procedures with delays and uncoordinated clinical trial implementation. CONCLUSION: The implementation of the national NB protocol demonstrated the complexity of the implementation of a national childhood cancer treatment protocol. However, standardised paediatric cancer treatment protocols based on local expertise and resources in limited settings are feasible.

Age at diagnosis as a prognostic factor in South African children with neuroblastoma.

PURPOSE: Low- and middle-income countries (LMICs) reported a higher median age at diagnosis of neuroblastoma (NB) compared to high-income countries. The aim was to determine if the optimal age at diagnosis, which maximizes the difference in overall survival between younger versus older patients in the South African population was similar to the internationally validated 18 months age cut-point. METHODS: Four hundred sixty NB patients diagnosed between 2000 and 2016 were included. Receiver operating characteristic (ROC) curves were used to predict potential age cut-point values for overall survival in all risk group classifications. Risk ratios, sensitivity, specificity, and positive and negative predictive values at the specific cut-points were estimated with 95% confidence intervals, and time to mortality by age at the specific cut-points was shown with Kaplan-Meier curves and compared using log-rank tests. RESULTS: The median age at diagnosis for the total cohort was 31.9 months (range 0.2-204.7). For high-risk (HR), intermediate-risk, low-risk, and very low-risk patients, the median age at diagnosis was, respectively, 36 months (range 0.4-204.7), 16.8 months (range 0.7-145.1), 14.2 months (range 2.0-143.5), and 8.7 months (range 0.2-75.6). The ROC curves for the total NB cohort (area under the curve [AUC] 0.696; P < .001) and HR (AUC 0.682; P < .001) were analyzed further. The optimal cut-point value for the total cohort was at 19.1 months (sensitivity 59%; specificity 78%). The HR cohort had potential cut-point values identified at 18.4 months age at diagnosis (sensitivity 45%; specificity 87%) and 31.1 months (sensitivity 67%; specificity 62%). The 19.1 months cut-point value in the total cohort and the 18.4 months cut-point value in HR were as useful in predicting overall survival as 18 months age at diagnosis. CONCLUSION: The 18 months cut-point value appears to be the appropriate age for prognostic determination, despite the higher median age at diagnosis in South Africa.

Prognostic factors affecting survival in children and adolescents with HIV and Hodgkin lymphoma in South Africa.

South African children with Hodgkin lymphoma (HL) and human immunodeficiency virus (HIV) have low 5-year overall survival (OS) rates. In this retrospective multicenter study, 271 South African pediatric patients with HL were studied to determine OS and prognostic factors in those with HIV and HL. Univariate risk factor analysis was performed to analyze prognostic factors. The 29 HIV-infected patients were younger (p = .021), more likely to present with wasting (0.0573), stunting (0.0332), and Stage IV disease (p = .000) than HIV-uninfected patients. The 5- and 10-year OS of HIV-infected patients of 49% and 45% versus 84% and 79%, respectively for HIV-uninfected patients (p = .0001) appeared to be associated with hypoalbuminemia (<20 g/dL) and CD4 percentage of <15%. Causes of death in the HIV-infected group included disease progression (6/14), infection (4/14), unknown (3/14), and second malignancy (1/14). HIV-infected pediatric patients with HL experience increased mortality due to post-therapy opportunistic and nosocomial infections.

The evaluation of induction chemotherapy regimens for high-risk neuroblastoma in South African children.

Achieving remission after induction therapy in high-risk neuroblastoma (HR-NB) is of significant prognostic importance. This study investigated remission after induction-chemotherapy using three standard neuroblastoma protocols in the South African (SA) setting. Retrospective data of 261 patients with HR-NB diagnosed between January 2000 and December 2016, who completed induction chemotherapy with standard treatment protocols were evaluated. The treatment protocols were either OPEC/OJEC or the St Jude NB84 protocol (NB84) or rapid COJEC (rCOJEC). The postinduction metastatic complete remission (mCR) rate, 2-year overall survival (OS) and 2-year event free survival (EFS) were determined as comparative denominators. The majority (48.3%; n = 126) received OPEC/OJEC, while 70 patients received (26.8%) rCOJEC and 65 (24.9%) NB84. Treatment with NB84 had the best mCR rate (36.9%), followed by OPEC/OJEC (32.5%) and rCOJEC (21.4%). The 2-year OS of treatment with NB84 was 41% compared to OPEC/OJEC (35%) and rCOJEC (24%) (p = 0.010). The 2-year EFS of treatment with NB84 was 37% compared to OPEC/OJEC (35%) and rCOJEC (18%) (p = 0.008). OPEC/OJEC had the least treatment-related deaths (1.6%) compared to rCOJEC (7.1%) and NB84 (7.5%) (p = 0.037). On multivariate analysis LDH (p = 0.023), ferritin (p = 0.002) and INSS stage (p = 0.006) were identified as significant prognostic factors for OS. The induction chemotherapy was not significant for OS (p = 0.18), but significant for EFS (p = 0.08) Treatment with NB84 achieved better mCR, OS and EFS, while OPEC/OJEC had the least treatment-related deaths. In resource-constrained settings, OPEC/OJEC is advised as induction chemotherapy in HR-NB due to less toxicity as reflected in less treatment-related deaths.

Overall survival for neuroblastoma in South Africa between 2000 and 2014.

BACKGROUND: Outcome data for neuroblastoma in sub-Saharan Africa are minimal, whereas poor outcome is reported in low- and middle-income countries. A multi-institutional retrospective study across South Africa was undertaken to determine outcome. METHODS: Patients treated between January 2000 and December 2014 in nine South African pediatric oncology units were included. Kaplan-Meier curves and Cox regression models were employed to determine two-year survival rates and to identify prognostic factors. RESULTS: Data from 390 patients were analyzed. The median age was 39.9 months (range, 0-201 months). The majority presented with stage 4 disease (70%). The main chemotherapy regimens were OPEC/OJEC (44.8%), St Jude NB84 protocol (28.96%), and Rapid COJEC (22.17%). Only 44.4% had surgery across all risk groups, whereas only 16.5% of high-risk patients received radiotherapy. The two-year overall survival (OS) for the whole cohort was 37.6%: 94.1%, 81.6%, and 66.7%, respectively, for the very-low-risk, low-risk, and intermediate-risk groups and 27.6% for the high-risk group (P < 0.001, 95% CI). The median survival time for the whole group was 13 months (mean, 41.9 months; range, 0.1-209 months). MYCN-nonamplified patients had a superior two-year OS of 51.3% in comparison with MYCN-amplified patients at 37.3% (P = 0.002, 95% CI). CONCLUSIONS: Limited disease had an OS comparable with high-income countries, but advanced disease had a poor OS. South Africa should focus on early diagnosis and implementation of a national protocol with equitable access to treatment.

Treatment outcomes of children with Hodgkin lymphoma between 2000 and 2010: First report by the South African Children's Cancer Study Group.

BACKGROUND: Children with Hodgkin lymphoma (HL) have excellent survival rates in high-income countries, but there are minimal outcome data in South African patients. Differing approaches to treatment are used in centres across South Africa, and the South African Children's Cancer Study Group (SACCSG) embarked on a programme to audit outcomes to improve survival rates. PATIENTS AND METHODS: A multicentre study was conducted to analyse outcomes and prognostic factors of children with HL in South Africa. Ten dedicated South African paediatric oncology units participated in a retrospective data review. All patients with HL treated consecutively between January 2000 and December 2010 were included. Kaplan-Meier curves and Cox regression model were employed to determine survival rates and prognostic factors. RESULTS: Two hundred and ninety-four patients were eligible for inclusion. The median age at presentation was 9.6 years (range 2.9-18.8); 55.4% of the patients presented with Stage III and IV disease and 9.9% were human immunodeficiency virus (HIV) positive. First-line therapy consisted of adriamycin, bleomycin, vinblastine and dacarbazine (ABVD) in 158 patients, vincristine, procarbazine/etoposide, prednisone and doxorubicin in 97 and adriamycin, bleomycin, vincristine and dacarbazine-chlorambucil, vinblastine, prednisone and procarbazine in 23 patients. The 5-year overall survival (OS) was 79% (95% confidence interval 73-84%). Multivariate analysis demonstrated that HIV infection (P = 0.018) and Ann Arbor Stage III and IV disease (P = 0.006) conferred a poor prognosis, while treatment with ABVD was associated with higher survival rates (P = 0.028). CONCLUSION: OS rates are encouraging for a middle-income country, although economic disparities continue to impact negatively on outcomes. Study results will form the basis for the development of national protocol and continued advocacy to rectify disparities.