The Association of Clinical Characteristics and Tumour Markers With Image-Defined Risk Factors in the Management of Neuroblastoma in South Africa.

AIMS: Image-defined risk factors (IDRFs) in neuroblastoma predict surgical complications and management outcomes. As there is a lack of data regarding the association of IDRFs with clinical and pathological factors, this study evaluated the prognostic value of IDRFs to predict neuroblastoma survival outcomes. MATERIALS AND METHODS: This was a retrospective study including 345 patients and reviewed diagnostic imaging for 20 IDRFs, pleural effusions and ascites. The IDRFs were grouped into five 'primary IDRFs' cohorts with vascular encasement, involvement of multiple body compartments, organ infiltration, airway obstruction and intraspinal extension. The association between clinical, histopathological and biological characteristics of neuroblastoma and management was evaluated. RESULTS: More patients without IDRFs had operations compared with patients with IDRFs, with a trend towards significance (64.4% versus 35.6%, P = 0.082). Patients with multiple compartment tumour involvement (P = 0.003) and organ infiltration (P < 0.001) had a higher risk of surgical complications. The 5-year overall survival of the group with more than one IDRF was 0.0% and those with pleural effusions or ascites 6.7%, associated with the worst outcome (P = 0.005). The total number of IDRFs was not predictive of the metastatic remission rate (P = 0.585) or overall survival (P = 0.142), with no conclusive association found between IDRF groups and clinical or biological markers. CONCLUSIONS: Patients with more than one IDRF had the shortest survival time, whereas those with pleural effusions and ascites at diagnosis had a poor outcome. Standardised reporting of IDRFs is crucial for predicting prognosis.

The Association between Tumour Markers and Meta-iodobenzylguanidine Scans in South African Children with High-risk Neuroblastoma.

AIMS: Diagnostic and post-induction 123I-meta-iodobenzylguanidine (123I-mIBG) scans have prognostic significance in the treatment of neuroblastoma, but data from low- and middle-income countries are limited due to resource constraints. The aim of this study was to determine the association between neuroblastoma-associated tumour markers (lactate dehydrogenase [LDH], ferritin and MYCN amplification) and 123I-mIBG scans (modified Curie scores and metastatic disease patterns) in predicting complete metastatic response rates (mCR) and overall survival. MATERIALS AND METHODS: Two hundred and ninety patients diagnosed with high-risk neuroblastoma in South Africa between January 2000 and May 2018 and a subanalysis of 78 patients with diagnostic 123I-mIBG scans were included. Data collection included LDH, ferritin and MYCN amplification at diagnosis. Two nuclear physicians independently determined the modified Curie scores and pattern of distribution for each diagnostic and post-induction 123I-mIBG scans with high inter-rater agreement (r = 0.952) and reliability (K = 0.805). The cut-off values for the diagnostic and post-induction modified Curie scores of ≥7.0 (P = 0.026) and 3 (P = 0.009), respectively, were generated. The association between the tumour markers and the modified Curie score of the 123I-mIBG scans was determined using post-induction mCR and 2-year overall survival. RESULTS: Diagnostic LDH (P < 0.001), ferritin (P < 0.001) and the diagnostic modified Curie scores (P = 0.019) significantly predicted mCR. Only ferritin correlated with diagnostic modified Curie scores (P = 0.003) but had a low correlation coefficient of 0.353. On multivariable analysis, the only significant covariate for 2-year overall survival at diagnosis was LDH <750 U/l (P = 0.024). A post-induction chemotherapy modified Curie score ≤3.0 had a 2-year overall survival of 46.2% compared with 30.8% for a score >3.0 (P = 0.484). CONCLUSION: LDH, ferritin and the diagnostic 123I-mIBG scans significantly predicted mCR, but only LDH predicted 2-year overall survival. Ferritin and the modified Curie scores correlated with each other. MYCN amplification neither correlated with any aspect of the 123I-mIBG scans nor significantly predicted mCR or 2-year overall survival. LDH and ferritin are therefore appropriate neuroblastoma tumour markers to be used in low- and middle-income countries with limited or no access to mIBG scans and/or MYCN amplification studies.