Introduction
Neuroblastoma is the most common extracranial solid tumour of infancy and arises from the neural crest cells of the sympathetic ganglia and adrenal medulla in foetal life.Reference Thompson, Vo and London1–Reference Tolaney, Toi and Neven6 The disease accounts for about 6–10% of paediatric cancers with a median age at the time of diagnosis of about 19 months, although, in some cases, the disease can be diagnosed in utero using magnetic resonance imaging.7–Reference Shinagawa, Kitamura and Katanoda14 According to the American Cancer Society,15 there are about 800 new cases of neuroblastoma each year in the United States and it is responsible for about 15% of childhood cancer-related deaths in the country.Reference Maris, Hogarty and Bagatell16,Reference Castel, Segura and Berlanga17 It is reported that about 30% of neuroblastoma tumours arise within the adrenal medulla, approximately 60% arise from abdominal paraspinal ganglia and the remaining 10% arise from the sympathetic ganglia in the chest, head and neck, and pelvis, hence, the clinical presentation and subsequent outcome are highly variable.Reference Louis and Shohet10,Reference Schmidt, Lukens and Seeger18 Neuroblastoma is characterised by biologically and clinically heterogeneous behaviour that ranges from spontaneous regression (especially in young infants and may never present to medical attention) to cases of highly aggressive metastatic disease that could be unresponsive to standard treatment.Reference Tolaney, Toi and Neven6,Reference Louis and Shohet10,Reference Park, Bagatell and London11,Reference Maris, Hogarty and Bagatell16,Reference Brodeur19–Reference Fritsch, Kerbl and Lackner22 BrodeurReference Brodeur19 reported that most children will experience complete regression of their disease with minimal therapy, even with metastatic disease, although older children frequently have aggressive metastatic tumours, despite treatment with the most intensive multimodality approach. The likelihood of cure for neuroblastomas is reported to vary widely according to age at diagnosis, extent of the disease and tumour biology, and about half of all cases are classified as high-risk phenotype for disease relapse, with an overall survival rate of less than 40% despite intensive multimodal therapy.Reference Maris, Hogarty and Bagatell16,Reference Mosse, Laundenslager and Longo23,Reference Yáñez, Grau and Oltra24 According to Yáñez et al.Reference Yáñez, Grau and Oltra24 the prognosis for neuroblastoma depends on age at diagnosis, histology, tumour stage and avian myelocytomatosis viral oncogene homolog (MYCN) amplification status. Moreover, other molecular parameters such as tumour cell deoxyribonucleic acid (DNA) content, gain of the long arm of chromosome 17 (17q), chromosome 1 (1q) and short arm of chromosome 2 (2p) and the deletion of the short arm of chromosome 1 (1p), chromosome 3 (3p) and 4 (4p) and long arm of chromosome 11 (11q) have also been shown to be predictive of patient outcome.Reference Yáñez, Grau and Oltra24
The International Neuroblastoma Risk Group (INRG) has developed a risk-based classification system based on a combination of seven prognostic risk factors (including stage, age, histologic category, grade of tumour differentiation, status of the MYCN oncogene, chromosome 11q status and DNA ploidy) to define 16 pre-treatment groups stratified by these prognostic markers.Reference Pinto, Applebaum and Volchenboum25–Reference Ambros, Ambros and Brodeur33 Patients are therefore classified into either very low (approximately 28%), low (approximately 27%), intermediate (approximately 9%) or high-risk (approximately 36%) groups based on 5-year event-free survival rates of the 16 pre-treatment groups and the probability of advancement and relapse of the disease.Reference Pinto, Applebaum and Volchenboum25,Reference Cohn, Pearson and London27 Cohn et al.Reference Cohn, Pearson and London27 reported that high-risk neuroblastoma patients usually show an initial remission in response to treatment but followed by recurrence of aggressive tumours that may become refractory to further treatment, hence it is very important to properly risk-stratified patients to ensure they receive the best treatment modality. Therefore, in recent years, there have been several investigations into the development of various diagnostic, predictive and prognostic cancer biomarkers for patients’ stratification and towards personalised and targeted treatment.Reference Osei, Lumini and Gunasekara34,Reference Osei, Walters and Masella35 Biomarkers have the potential to measure early disease detection and diagnosis, the risk of disease development and progression, improved patient stratification for various treatment paradigms, provide accurate information of patient response to a specific treatment and inform clinicians about the likely outcome of a cancer diagnosis independent of the treatment received.Reference Tolaney, Toi and Neven6,Reference Osei, Lumini and Gunasekara34–Reference Utnes, Løkke and Flægstad36 In this paper, we report on the literature review of current and emerging clinical biomarkers used in risk assessment, screening for early detection and diagnosis, prognostication and for monitoring the response of treatment of paediatric neuroblastoma cancers.
Liquid Biomarkers for Management of Paediatric Neuroblastoma
Biomarkers are any substance, structure or process that can be measured in the body or its products and are capable of influencing or predicting the incidence or outcome of diseases.Reference Goossens, Nakagawa and Sun37,38 Cancer biomarkers can be either (1) diagnostic (used to identify whether a patient has a specific disease condition), (2) predictive (used to predict response to specific therapeutic interventions) or (3) prognostic (used to inform physicians regarding the risk of clinical outcomes such as cancer recurrence or disease progression).Reference Osei, Lumini and Gunasekara34,Reference Osei, Walters and Masella35,Reference Sawyers39–Reference Singal and Agarwala42 According to Osei et al.,Reference Osei, Lumini and Gunasekara34,Reference Osei, Walters and Masella35 understanding the molecular mechanisms and signalling pathways of biomarkers can potentially lead to personalised and targeted treatment via therapeutic targeting of specific genetic aberrant pathways which play key roles in malignant tumour formation and may provide optimal treatment choices in clinical practice. Biomarkers commonly used for diagnosis, prognostication and therapeutic monitoring can be obtained through either liquid or tissue biopsies. The liquid biomarkers can be found in several patient body fluids, including blood serum or plasma, pleural fluid, saliva, exhaled breath condensate, cerebrospinal fluid and urine.Reference Crowley, Di Nicolantonio and Loupakis43–Reference Siravegna, Marsoni and Siena46 Conventional tissue biopsies are typically surgical biopsy, radiologically guided biopsy or endoscopic biopsy and are usually invasive, expensive, painful, difficult to obtain, time-consuming, need to be consistently evaluated by expert pathologists, limited repeatability, limited by patient age and comorbidity and may introduce clinical risks to patient.Reference Siravegna, Marsoni and Siena46–Reference Sausen, Parpart and Diaz49 Furthermore, for paediatrics patients, sometimes the risks of tissue biopsy may outweigh the benefits, thus constituting a significant barrier for easy and frequent monitoring and management of paediatric cancer progressions.Reference Siravegna, Marsoni and Siena46–Reference Abbou, Shulman and DuBois50 Whereas, liquid biopsy offers a new and unique opportunity to identify potential tumour biomarkers and has a significant amount of advantages such as less invasive, easily obtainable and repeated, relatively less expensive and potential for real-time monitoring compared to tissue biopsy.Reference Hao and Zhang47–Reference Sausen, Parpart and Diaz49,Reference Perakis and Speicher51 Therefore over the past few years, there has been a significant interest in research towards liquid biopsies for biomarkers that can easily be obtained through less invasive approach.Reference Verly, van Kuilenburg and Abeling2–Reference Wang, Wang and Su4,Reference Tolaney, Toi and Neven6,Reference Monsaingeon, Perel and Simonnet12,Reference Yáñez, Grau and Oltra24,Reference Cohn, Pearson and London27,Reference Goossens, Nakagawa and Sun37–Reference Ma, Jiang and Kang147 A liquid biopsy workflow is a minimally invasive alternative to surgical tumour sampling and involves the collection and analysis of bodily fluids, including urine (which is a source of catecholamines metabolites), peripheral blood [which is a source of several circulating analytes, including cell-free DNA, cell-free ribonucleic acid (RNA) (i.e., messenger RNA (mRNA) and microRNA (miRNA)), extracellular vesicles (exosomes and microvesicles), tumour-educated platelets and circulating tumour cells (CTCs)] and other body fluids which are sources of other analytes with high-potential clinical applications.Reference Tolaney, Toi and Neven6,Reference Siravegna, Marsoni and Siena46,Reference Jahr, Hentze and Englisch52
Serum Lactate Dehydrogenase (LDH)
Lactate dehydrogenase (LDH) is a metabolic enzyme widely expressed in all living tissues, especially in skeletal muscle, heart, liver, kidneys, brain, lungs and red blood cells and is detectable in serum.Reference Liu, Cao and Gao53 LDH catalyses the interconversion of lactate and pyruvate during glycolysis and gluconeogenesis and exists in five distinct enzyme classes or isoenzymes, namely LDH-1, LDH-2, LDH-3, LDH-4 and LDH-5, which are distinguished by slight differences in their structure.Reference Liu, Cao and Gao53–Reference Wulaningsih, Holmberg and Garmo55 According to Liu et al.,Reference Liu, Cao and Gao53 several human cancers have higher levels of serum LDH than normal tissues, although the levels can also be elevated in non-malignant conditions such as heart failure, kidney disease, anaemia and hypothyroidism.Reference Petrelli, Cabiddu and Coinu54,Reference Wulaningsih, Holmberg and Garmo55 They reported that serum LDH levels are correlated with tumour burden and indicative of tumour growth and invasive potential. Furthermore, higher LDH levels are associated with shorter survival in various types of cancer and that LDH has the potential as a marker of tumour burden for advanced cancer patients.Reference Liu, Cao and Gao53 Wulaningsih et al.Reference Wulaningsih, Holmberg and Garmo55 indicated that higher pre-diagnostic LDH levels correlate with lower overall and cancer-specific survival following cancer diagnosis, and the associations between LDH and both all-cause and overall cancer mortality are stronger when LDH is measured closer to cancer diagnosis. Cohn et al.Reference Cohn, Pearson and London27 and Morgenstern et al.Reference Morgenstern, London and Stephens56 reported that serum LDH levels may be used in real-time monitoring of the progress of tumour, since levels are usually low in the early stages of the disease but show significant increase as the rate of cell turnover increases and it is relatively easy and non-invasive to obtain samples for analysis. According to Liu et al.,Reference Liu, Cao and Gao53 LDH plays an important role in tumour progression and maintenance, and its inhibition can inhibit tumour progression and thus has been considered for the therapeutic targeting of cancer energy metabolism.
The prognostic potential of serum LDH in oncology has been long recognised in several investigations.Reference Cohn, Pearson and London27,Reference He, Yan and Tang41,Reference Liu, Cao and Gao53–Reference Morgenstern, Potschger and Moreno59 Cangemi et al.Reference Cangemi58 investigated the prognostic value of LDH, commonly evaluated at diagnosis in 279 children diagnosed with neuroblastoma. They observed that high LDH levels are positively associated with a worse prognosis in neuroblastoma patients with localised and metastatic disease and are also associated with poor prognosis independent of disease stage in patients with MYCN amplification. They reported that patients with localised disease and LDH levels >1,300 IU/mL had 12·9 times greater chance of relapse, whereas LDH levels <1,300 IU/mL were found to be independent predictors of good prognosis in patients with localised neuroblastoma without MYCN amplification.Reference Cangemi58 Furthermore, they observed LDH as an independent prognostic factor in patients with stage IV neuroblastoma without MYCN amplification and as an independent predictor of poor event-free and overall survival in patients older than 18 months at diagnosis with metastatic disease. They concluded that LDH and catecholamine metabolites measurement at diagnosis may be helpful for an easy, cost-effective, patient risk stratification.Reference Cangemi58 Cohn et al.Reference Cohn, Pearson and London27 investigated the statistical and clinical significance of 13 potential prognostic factors including LDH in children diagnosed with neuroblastoma and reported that a single LDH cut-off level of 587 IU/mL could potentially stratify patients into good and poor prognosis. Morgenstern et al.Reference Morgenstern, Potschger and Moreno59 analysed 1,053 patients data extracted from the database of the European High-Risk Neuroblastoma 1 (HRNBL1) – International Society for Paediatric Oncology European Neuroblastoma trail to develop a model and score for risk stratification for patient aged ≥18 months with metastatic disease. They observed that patients aged >5 years with elevated serum LDH above a median level of 1,250 U/L at diagnosis had significantly poorer event-free and overall survival.
Serum Ferritin
Ferritin is an intracellular protein that stores iron and releases it in a controlled manner thereby acting as a buffer against iron deficiency and iron overload in humans and are present in both glycosylated form (which is the form that results from active secretion by cells) and non-glycosylated form (which is a form that is directly released by damaged cells).Reference Hann, Levy and Evans60–Reference Hann, Stahlhut and Evans63 Ferritin is found in most tissues as a cytosolic protein and small amounts circulate in human serum in normal and pathological states where it functions as an iron carrier.Reference Hann, Levy and Evans60 In general, normal ferritin levels range from 12 to 300 ng/mL for males and 12 to 150 ng/mL for females. However, elevated serum levels of ferritin are frequently present at diagnosis in children with actively growing neuroblastoma, although the levels will usually return to normal with clinical remission.Reference Hann, Levy and Evans60–Reference Hann, Stahlhut and Evans63 Hann et al.Reference Hann, Evans and Cohen61 reported that ferritin is present in tumours tested directly from patients and observed that in vitro neuroblastoma cells and nude mice bearing human neuroblastomas have detectable levels of human ferritin in their sera, thus suggesting that the increased ferritin in the serum of patients with neuroblastoma is probably derived from the tumour. They indicated that the elevated ferritin in sera of these patients could be derived from three sources namely: (1) leakage of intracellular ferritin from damaged normal cells due to tumour invasion; (2) leakage of tumour cell ferritin during tumour necrosis; or (3) augmented synthesis of intracellular ferritin by tumour cells and subsequent secretion of ferritin into plasma.Reference Hann, Evans and Cohen61 StudiesReference He, Yan and Tang41,Reference Meany, London and Ambros62,Reference Hann, Evans and Siegel64 have demonstrated the detection of serum ferritin at different levels in patients at different stages of neuroblastoma observed elevated ferritin levels in patients with stage III or IV neuroblastoma but infrequently elevated in sera in patients with stages I and II disease and observed usually low or normal in patients with stage IV-S tumours. Thus ferritin levels in serum can serve as a poor prognostic biomarker independent of age at diagnosis and disease stage.
Morgenstern et al.Reference Morgenstern, London and Stephens56 analysed data from the INRG database for 2,250 patients with stage IV disease and reported a median ferritin level of 156 ng/mL for metastatic site index of 1 versus 247 ng/mL for metastatic site index > 1. They observed that increased metastatic site index correlated with serum ferritin level for patients of all ages and was also associated with a worse outcome and higher baseline ferritin level. Furthermore, for patients aged ≥18 months at diagnosis, there was a high correlation of worse outcome with increasing metastatic site index, thus, reflecting greater disease burden and more widespread metastatic dissemination at diagnosis. Hann et al.Reference Hann, Levy and Evans60 screened serum ferritin levels in 58 children with neuroblastoma and reported that increased ferritin levels in serum ≥400 ng/mL correlated well with the presence of active disease; however, ferritin levels returned to normal with clinical remission, suggesting that the tumour was the origin of the elevated ferritin levels. Therefore, they suggested that serum ferritin levels could be used as an indicator of disease activity and as a guide to therapy. They observed that primary neuroblastoma tumours and cells from neuroblastoma cell lines contain ferritins with the electrophoretic characteristics different from normal liver ferritin. Hann et al.Reference Hann, Evans and Cohen61 studied 17 patients with stage IV neuroblastoma who were treated with vincristine, cyclophosphamide and imidazole carboxamide and 13 patients with stage IV-S disease who were not treated and reported that serum ferritin levels were elevated in 88% (15 of 17) of children with stage IV, but not in the 13 with stage IV-S. They observed that neuroblastoma stage IV and IV-S diseases with extensive disease burdens but with significantly different prognosis can be differentiated by serum ferritin levels. In another study, Hann et al.Reference Hann, Evans and Siegel64 measured ferritin levels in sera obtained at diagnosis from 241 patients with neuroblastoma to determine both the incidence of elevated ferritin at diagnosis in relationship to stage of disease and the relationship between ferritin level and subsets of patients with stages III and IV disease with different prognoses. They reported that ferritin level was rarely elevated in sera in patients with stages I and II disease but was highly elevated in patients with stages III 37% (89 of 241) and IV 54% (130 of 241) neuroblastomas. Furthermore, they reported that the progression-free survival at 24 months of follow-up for patients with stage III disease with normal ferritin level was 76% and with elevated ferritin was 23%, whereas the progression-free survival for patients with stage IV disease was 27 and 3% with normal and elevated ferritin levels, respectively. They reported that elevated ferritin was associated with a significantly poorer prognosis than normal ferritin levels for patients with stages III and IV disease and the correlation was independent of stage and age at diagnosis. They concluded that the determination of the level of ferritin in serum at diagnosis is useful for selecting appropriate therapy for patients with stage III neuroblastoma, and ferritin level is capable of defining subgroups of patients with stages III and IV neuroblastoma with different outcomes at diagnosis. Hann et al.Reference Hann, Stahlhut and Evans63 examined in sera of 36 children with neuroblastoma and from 16 normal healthy children to determine whether the increased amounts of ferritin at diagnosis are of glycosylated or non-glycosylated ferritin and observed that elevated ferritin in sera of patients with neuroblastoma seen at diagnosis is the result of secretion of ferritin by the tumour. Meany et al.Reference Meany, London and Ambros62 analysed 1,483 patients’ data from the INRG database to define patient and tumour characteristics that are predictive of outcome. They reported that serum ferritin ≥96 ng/mL was associated with poor event-free survival and overall survival in older children, patients with serum ferritin <96 ng/mL have excellent prognosis, tumour grade and serum ferritin were prognostic of outcome for patients >547 days of age, and ferritin may prove to be an important factor for patient risk stratification.
Neuron-Specific Enolase (NSE)
Neuron-specific enolase (NSE) is a cell-specific isoenzyme of the glycolytic enzyme enolase and is a highly specific marker for neurons and peripheral neuroendocrine cells.Reference Isgro, Bottoni and Scatena65–Reference Massaron, Seregni and Luksch68 There are three isozymes of enolase which are expressed by different genes in mammals, namely alpha enolase which is ubiquitous, beta enolase which is muscle-specific and gamma enolase which is neuron-specific, and the three isoforms are characterised by different tissue distributions and by distinct biochemical and immunological properties.Reference Isgro, Bottoni and Scatena65,Reference Tsuchida, Honna and Iwanaka67,Reference Massaron, Seregni and Luksch68 The human NSE is a major brain protein that constitutes between 0·4 and 2·2% of the total soluble protein of brain, depending on the region and in some neurons accounts for 3–4% of the total soluble protein.Reference Isgro, Bottoni and Scatena65,Reference Kintzel, Sonntag and Strauss69 It has been reported that NSE appears during neurogenesis and its amount increases during early neuronal differentiation and is observed in specific tissues under normal conditions; therefore, it has been hypothesised that increased levels of NSE in body fluids (spinal fluid, serum or urine) may occur with malignant tumours and thus can be of value in diagnosis, staging and treatment of related neuroendocrine tumours.Reference Isgro, Bottoni and Scatena65,Reference Zeltzer, Marangos and Parma66 Furthermore, the specific localisation of NSE in neurons and neuroendocrine cells suggests that assays of NSE levels in biological fluids, such as spinal fluid, serum or urine, might provide useful diagnostic information regarding disorders due to altered metabolism or turnover of these cell types.Reference Isgro, Bottoni and Scatena65,Reference Zeltzer, Marangos and Parma66,Reference Massaron, Seregni and Luksch68
Several applications of NSE determination in medical oncology have been investigated including the determination of NSE content in tissue biopsies; serum measurements of NSE as a marker of tumour diagnosis, disease extent and response to therapy; and determination of cerebrospinal fluid NSE as an indicator of cranial and CNS metastases.Reference Singal and Agarwala42,Reference Zeltzer, Marangos and Parma66,Reference Kintzel, Sonntag and Strauss69–Reference Berthold, Engelhardt-Fahrner and Schneider72 These studies reported that serum NSE levels are correlated with response to therapy, and whereas low level of NSE is a predictor of good outcome, high levels are associated with bad prognosis. Riley et al.Reference Riley, Heney and Jones70 conducted a systematic review and a meta-analysis of molecular and biological tumour markers in neuroblastoma to establish an evidence-based perspective on their clinical significance for screening, diagnosis, prognosis and monitoring of patients. They observed that patients with elevated serum levels of NSE have a significantly worse outcome in terms of disease-free survival and overall survival, and they identified NSE biomarker as a potentially important prognostic tool in neuroblastoma. Zeltzer et al.Reference Zeltzer, Marangos and Parma66 investigated the serum levels of NSE measured at diagnosis in 122 children with clinical stage IV metastatic neuroblastoma. They reported a mean serum NSE level of 207 ng/mL, whereas level in normal age-matched children was 7·5 ng/mL. They observed that the level of NSE at diagnosis >100 ng/mL was associated with poor outcome and highly significant in children aged <1year at diagnosis. They concluded that serum NSE may be a useful disease marker and a prognostic indicator in children with metastatic neuroblastoma. In another study, Zeltzer et al.Reference Zeltzer, Marangos and Evans71 analysed serum NSE measured at diagnosis in 61 children at all stages of neuroblastoma and reported median levels for stages I, II, III, IV and IV-S of 13, 23, 40, 214 and 40 ng/mL, respectively. They observed a disease-free survival rate of 79% (27 of 34) for patients with NSE levels <100 ng/mL compared to 10% for patients with higher levels, and elevated serum NSE level in patients with stages III and IV disease was associated with poor outcome. Furthermore, serum NSE levels were significantly lower in patients with stage IV-S disease than patients in stage IV, despite their extensive tumour burden. They concluded that serum NSE estimation may confirm stage IV-S status and may suggest a more benign clinical course. Massarona et al.Reference Massaron, Seregni and Luksch68 evaluated NSE levels in 80 paediatric patients with neuroblastoma and reported NSE median levels of 9·9, 45·1, 49, 93·9 and 53·4 ng/mL for stages I, II, III, IV and IV-S, respectively, and the levels were correlated with prognosis. They observed a good correlation between NSE levels and disease course and concluded that NSE is a good marker for neuroblastoma and its quantitative determination in serum is valuable in the management of these patients to confirm diagnosis, monitor effect of treatment and detect recurrent disease. Kintzel et al.Reference Kintzel, Sonntag and Strauss69 measured NSE from cord blood samples taken from 192 healthy term newborns by enzyme immunoassay and reported median concentration of 8·0 μg/L in the reference group with no differences between male and female newborns as a proposed reference values for NSE in cord blood of healthy term newborns. They reported that measurement of NSE in cord blood in comparison with their reference values may offer the earliest postnatal possibility of confirming the diagnosis of neuroblastoma in cases of suspicion raised by foetal sonography.
Urine Catecholamines
Catecholamines are hormones produced by the adrenal glands and their levels in the body can easily be measured quantitatively in urine and blood, although they are most commonly measured with urine test.Reference Schilling, Spix and Berthold73 There are three main catecholamines, namely dopamine, norepinephrine (noradrenaline) and epinephrine (adrenaline), and several studies have reported that elevated urinary excretion of these adrenal hormones and/or their metabolites [vanillylmandelic acid or 3-methoxy-4-hydroxymandelic acid (VMA), homovanillic acid or 3-methoxy-4-hydroxyphenylacetic acid (HVA), 3-methoxytyramine (3-MT), vanillactic acid (VLA), vanilglycol, normetanephrine, metanephrine, methyldopamine] are associated with ganglioneuroma, ganglioneuroblastoma and neuroblastoma and are thus very useful for the diagnosis of these diseases.Reference Zambrano and Reyes-Múgica3,Reference Monsaingeon, Perel and Simonnet12,Reference Schilling, Spix and Berthold73–Reference Tsubono and Hisamichi77 Increased urinary excretion of dopamine, VMA and HVA by neuroblastoma patients is well established, and the detection of elevated urinary levels of these compounds has become a useful diagnostic test and has been used in clinical diagnosis as well as for early detection in screening programmes.Reference Monsaingeon, Perel and Simonnet12,Reference Strenger, Kerbl and Dornbusch76–Reference Woods, Gao and Shuster79 The correlation of neuroblastoma with elevated urinary levels of VMA and HVA is reported to be related to the stage of the disease and it is less likely to have elevated levels in low-stage tumours.Reference Zambrano and Reyes-Múgica3,Reference Berthold, Hunneman and Harms80 The serial determinations of urinary catecholamines and their metabolites are also useful in the follow-up of neuroblastoma patients, since reappearance or increase in these compounds in the patient’s urine is often associated with recurrence and/or progression of the disease.Reference Zambrano and Reyes-Múgica3,Reference LaBrosse, Com-Nougue and Zucker75,Reference Laug, Siegel and Shaw78 According to Monsaingeon et al.Reference Monsaingeon, Perel and Simonnet12 and Verly et al.,Reference Verly, van Kuilenburg and Abeling2 assays of the commonly analysed catecholamines in urine and their metabolites (e.g., dopamine,VMA, HVA) establishes a biochemical diagnosis of neuroblastoma in over 80–90% of patients. Furthermore, the determination of the two major catecholamine metabolites, VMA and HVA, in urine has been widely recognised as a valuable tool for diagnosis and evaluation of response to treatmentReference Berthold, Hunneman and Harms80 or diagnosis of relapse.Reference Zambrano and Reyes-Múgica3
Schilling et al.Reference Schilling, Spix and Berthold73 screened urine samples of about 2,581,188 children (at approximately 1 year of age) with neuroblastoma to compare the incidence of disseminated disease and mortality. They reported two different types of neuroblastoma from their study: a type with a favourable prognosis that may regress spontaneously and a more clinically aggressive type with an unfavourable prognosis. Strenger et al.Reference Strenger, Kerbl and Dornbusch76 investigated urinary catecholamine, their metabolites patterns and their correlation to stage, biological features, and outcome in 114 clinically diagnosed neuroblastoma patients with median age at diagnosis of 9·25 months (0–239 months) and reported measurement sensitivities of 80·7, 71·9 and 61·3% for VMA, HVA and dopamine, respectively. They observed that high VMA levels were associated with favourable biological features and high dopamine levels were predominantly found in biologically unfavourable disease. They reported that dopamine-to-VMA ratio was capable of separating stage IV and stage IV-S patients with metastatic disease and may serve as a tool for ‘“biological grading”’ in metastatic disease and potentially be helpful for the decision of ‘wait and see’ strategy in selected neuroblastoma patients with localised disease. LaBrosse et al.Reference LaBrosse, Com-Nougue and Zucker75 analysed VMA and HVA in initial urine samples of 288 patients with neuroblastoma and related tumours of neural crest origin and examined the interrelationships between excretion of these metabolites, age at diagnosis, stage of the disease, site of the primary tumour and prognosis. They reported that patients in stages IV and IV-S excreted VMA at levels significantly higher than patients in stages I–III, and that the initial urinary excretion levels of VMA and HVA are of significant value as an aid in diagnosis of patients with neuroblastoma. Verly et al.Reference Verly, van Kuilenburg and Abeling2 retrospectively studied a panel of eight urinary catecholamine metabolites (VMA, HVA, 3-methoxytyramine, dopamine, epinephrine, metanephrine, norepinephrine and normetanephrine) and their correlation with clinical factors from 301 neuroblastoma patients at diagnosis. They observed that elevated catecholamine metabolites correlated with 75% (9 out of 12) neuroblastoma characteristics such as tumour stage, age at diagnosis, MYCN amplification, loss of heterozygosity for 1p and bone marrow invasion. They reported diagnostic sensitivities of 89, 82, 66, 84 and 95% for normetanephrine, HVA, VMA, combination of VMA and HVA and combination of all eight metabolites, respectively. Furthermore, they indicated that analysis of urinary catecholamine metabolite panel, comprising eight metabolites, ensures the highest sensitivity to diagnose neuroblastoma patients, specifically patients that are much more difficult to diagnose and could also be linked to prognosis and monitoring response to therapy.Reference Verly, van Kuilenburg and Abeling2 BondReference Bond74 studied the pattern of urinary excretion of VMA in 50 children (33 boys and 17 girls) with neuroblastoma and observed that the urinary VMA level was elevated at diagnosis in eight of the ten disease-free survivors but rapidly returned to normal after treatment. Furthermore, three out of four children with residual tumour had their VMA level remained persistently elevated, and 28 of the 36 children who died had their VMA level elevated at diagnosis and remained elevated throughout the course of their disease. They reported that measurement of urinary VMA is a simple diagnostic test capable of confirming the presence of neuroblastomas, and the prompt return to normal levels in the disease-free survivors confirms the importance of serial VMA determination in assessing response to treatment and prognosis.
Zambrano and Reyes-MugicaReference Zambrano and Reyes-Múgica3 examined the correlation between urinary levels of catecholamines (dopamine and norepinephrine) and their metabolites (VMA and HVA) with age at diagnosis, clinical tumour stage and histopathological classification of 33 patients with neuroblastoma. They found a direct correlation between advanced tumour stages (stages III and IV) and increased urinary dopamine, norepinephrine, VMA and HVA levels. They reported that 88% (29 of 33) of patients with tumour stages III and IV and 8% (3 of 33) of patients with tumour stages I and II had urinary dopamine levels that were higher than twice the upper limit of normal levels. Furthermore, urinary VMA was significantly higher (more than 10 times the upper limit of normal) in 61% of patients with tumour stages III and IV, whereas only 7% of patients with tumour stages I and II had urinary VMA at that level. Similarly, 44% of patients with tumour stages III and IV had urinary norepinephrine levels that were higher than 3 times the upper limit of normal, in contrast to the stages I–II group, in which no patients had urinary norepinephrine at that level. Finally, 65% of patients with stages III–IV had urinary HVA levels 3·5 times higher than the upper limit of normal, but no patients in stages I and II had urinary HVA levels in that range. They suggested that the increased net amount of catecholamines and their metabolites in patients with advanced stage disease reflects greater absolute tumour burden. They also reported that increased urinary dopamine-to-norepinephrine and dopamine-to-VMA ratios are associated with poor prognosis; low dopamine-to-norepinephrine ratios was associated with longer disease-free survival and low VMA-to-HVA ratio was associated with poorly differentiated tumours and poor prognosis. They concluded that aggressive neuroblastomas are associated with higher urinary levels of dopamine, norepinephrine, VMA and HVA, and that a subset of neuroblastomas with particularly unfavourable features (tumour stage IV, age older than 12 months, unfavourable histology and MYCN amplification) shows a blockade in the metabolism of dopamine. Furthermore, determination of urinary dopamine, VMA, HVA, and the dopamine-to-VMA and dopamine-to-HVA ratios could be used as a valuable and rapid initial assessment of prognosis in patients with neuroblastoma.Reference Zambrano and Reyes-Múgica3 Laug et al.Reference Laug, Siegel and Shaw78 examined the initial urinary catecholamine metabolite excretion patterns in 54 children with neuroblastoma to determine if the levels of those compounds in initial urine samples are predictive of the disease outcome. They reported that initial excretion patterns of urinary catecholamine metabolites are predictive of prognosis, and that prognosis in disseminated disease was correlated directly with the urinary VMA-to-HVA ratio but not with the absolute levels of HVA, and the presence of VLA and increased amounts of cystathionine and/or low levels of VMA indicated poor prognosis.
Circulating Deoxyribonucleic Acid (DNA)
The main sources of circulating DNA that are non-invasively assessed and have been demonstrated to be present in patients’ blood or serum in both the early and late stages of paediatric neuroblastoma include the circulating cell-free DNA (cfDNA), cell-free tumour DNA (ctDNA) and CTCs.Reference Wang, Wang and Su4,Reference Tolaney, Toi and Neven6,Reference Sausen, Parpart and Diaz49,Reference Bettegowda, Sausen and Leary81,Reference Trigg, Martinson and Parpart-Li82 The potential applications of these serum-circulating DNA in the diagnosis and prognosis of cancer were first demonstrated in 1977, when elevated levels of circulating DNA were detected in the serum of cancer patients but dropped when the patients responded to radiotherapy.Reference Combaret, Hogarty and London83
Circulating cell-free DNA (cfDNA)
The cfDNA refers to the DNA from both malignant and non-malignant cells that are shed into the circulatory system, and their presence in circulation was first formally reported in 1948.Reference Sausen, Parpart and Diaz49,Reference Jahr, Hentze and Englisch52 Since then, several investigatorsReference Tolaney, Toi and Neven6,Reference Siravegna, Marsoni and Siena46,Reference Sausen, Parpart and Diaz49,Reference Perakis and Speicher51,Reference Chicard, Colmet-Daage and Clement84,Reference Janoueix-Lerosey, Schleiermacher and Michels85 have harness this non-invasive and easily accessible genetic information in circulation of patients for the detection of cancer cells-specific genetic aberrations or mutations. The biological half-life of cfDNA in serum is reported to range from 15 min to a few hours and is then eliminated by the liver, spleen and kidney.Reference Siravegna, Marsoni and Siena46 According to Kurihara et al.Reference Kurihara, Ueda and Onitake86 cfDNA consists of small double-stranded fragments of DNA derived from the tumour and can be found in the plasma of cancer patients. When cfDNA is isolated from plasma, it can potentially be used for the detection of gene aberrations or mutations, including MYCN amplification, chromosome 17q gain and anaplastic lymphoma kinase (ALK) mutations in neuroblastoma patients.Reference Kurihara, Ueda and Onitake86–Reference Pezzolo, Rossi and Gimelli90 Some studiesReference Wang, Wang and Su4,Reference Van Roy, Van Der Linden and Menten91,Reference Su, Wang and Wang92 have also reported that plasma cfDNA can be used to evaluate tumour burden, is correlated with treatment response and tumour size, and can predict relapse in neuroblastoma patients. Su et al.Reference Su, Wang and Wang92 investigated the potential use of plasma cfDNA as a biomarker to determine tumour burden and minimal residual disease of 58 neuroblastoma patients during chemotherapy. They observed that cfDNA level was down-regulated in neuroblastoma patients with partial remission when compared to those with stable disease. Furthermore, they reported that the dynamic change of cfDNA can be considered as a surrogate biomarker to evaluate tumour burden and minimal residual disease of neuroblastoma during early and intermediate therapy periods. According to Kurihara et al.Reference Kurihara, Ueda and Onitake86 the cfDNA detection technique is a very useful and less invasive approach for identification of genetic aberrations in childhood solid tumours.
Wang et al.Reference Wang, Wang and Su4 evaluated plasma cfDNA as a promising biomarker for tumour burden in 79 patients (median age of 36 months) with newly diagnosed neuroblastoma using quantitative polymerase chain reaction (qPCR). They reported that cfDNA levels were highly elevated in newly diagnosed neuroblastoma patients compared to control group and higher in patients with larger tumours. They concluded that the quantification of plasma cfDNA is an effective and more precise biomarker to evaluate tumour burden and minimal residual disease of neuroblastoma in a clinical setting. Kurihara et al.Reference Kurihara, Ueda and Onitake86 assessed cfDNA using next-generation sequencing in 44 childhood solid tumour patients who underwent surgical intervention and analysed the relationships with clinical parameters. They reported that cfDNA concentrations were significantly associated with disease stage with distant metastasis and were capable of identifying 15 mutations or deletions in solid tumours, including ALK, RET mutations and MYCN high copy numbers in neuroblastoma. Chicard et al.Reference Chicard, Boyault and Daage87 evaluated the feasibility of genomic copy number profiling of cfDNA isolated from plasma of 70 neuroblastoma patients at diagnosis. They demonstrated the feasibility of tumour genomic copy number profiling of cfDNA in neuroblastoma patients at diagnosis and reported higher concentrations of cfDNA in patients with metastatic disease.
Copy number alterations (CNAs) in CfDNA
Copy number alterations (CNAs) are somatic changes to chromosome structure that result in either a gain or loss of copies in sections of the DNA and have been reported to correlate with the development and progression of cancer.Reference Osei, Walters and Masella35,Reference Combaret, Hogarty and London83 According to Combaret et al.,Reference Combaret, Hogarty and London83 high copy numbers are strongly associated with rapid tumour progression and poor outcome, independent of tumour stage or patient age and have become an important factor in treatment stratification. Several studiesReference He, Yan and Tang41,Reference Kojima, Hiyama and Fukuba93–Reference Marrano, Irwin and Thorner101 have measured MYCN copy number in cfDNA of MYCN-amplified neuroblastoma patients using qPCR to investigate whether qPCR is capable of measuring MYCN gene amplification in serumReference Ma, Lee and Park95 and if MYCN amplification status can identify high-risk neuroblastoma patients who may benefit from early therapeutic intervention or correlates with neuroblastoma prognosis and patient survival.Reference Yagyu, Iehara and Tanaka94 These studies demonstrated that the detection of MYCN amplification using qPCR analysis of serum samples is a simple and promising method to evaluate MYCN status of neuroblastoma patients and can be used in many clinical settings, such as monitoring relapse or deciding treatment strategy.Reference Ma, Lee and Park95 Furthermore, determination of MYCN copy number in plasma DNA of patients with MYCN-amplified tumours is useful to predict prognosis, plan a therapeutic approach and to evaluate the therapeutic efficacy of chemotherapy and surgery.Reference Kojima, Hiyama and Fukuba93 In addition, serum-based MYCN gene amplification status was reported to be correlated with overall survival in patients and enabled the identification of patients with poor prognosis, and its analysis was shown to have good sensitivity (about 86%) and very high specificity (about 95%).Reference Yagyu, Iehara and Tanaka94
Lodrini et al.Reference Lodrini, Sprussel and Astrahantseff102 established a droplet digital PCR protocols for MYCN and ALK copy number status in plasma from neuroblastoma patients and reported that the technique accurately discriminated between MYCN and ALK amplification, gain and normal diploid status in a large panel of neuroblastoma cell lines. Combaret et al.Reference Combaret, Audoynaud and Iacono103 assayed serum or plasma from 102 patients with neuroblastoma and 72 controls using PCR amplification to search for the presence of MYCN DNA. They demonstrated that circulating MYCN is detectable at diagnosis in plasma in MYCN-amplified neuroblastoma patients with either localised or metastatic tumour, and it is a valuable prognostic marker that may be used as an early detection test and as a complementary non-invasive assay in the follow-up of patients at high risk of relapse. In a recent study, Combaret et al.Reference Combaret, Hogarty and London83 evaluated the clinical relevance of circulating MYCN DNA in 267 patients with locoregional or metastatic neuroblastoma in children < 18 months of age and demonstrated that MYCN-amplified neuroblastoma patients with metastatic disease show high levels of circulating neuroblastoma-derived MYCN DNA sequences. Furthermore, they reported that the detection of MYCN sequences in sera is a specific technique to assess tumour MYCN status in stage III and metastatic disease for whom a ‘wait and see’ strategy is often recommended.Reference Combaret, Hogarty and London83
Segmental chromosome aberration (SCA) in CfDNA
Segmental chromosomal aberration (SCA) is the presence of an anomaly of a particular chromosome arm, as determined by fluorescence in situ hybridisation (FISH), or a deletion or imbalance, as determined by loss of heterozygosity. A number of SCA have been reported, including deletions of chromosome arm 1p, 3p, 4p and 11q, and gains of chromosome arm 2p, 1q and 17q.Reference Yagyu, Iehara and Gotoh104,Reference Combaret, Brejon and Iacono105 Several studiesReference Chicard, Boyault and Daage87,Reference Yagyu, Iehara and Gotoh104–Reference George, Attiyeh and Li107 have demonstrated the ability to detect SCAs, such as gain of 17q, 1q and 2p and loss of 11q, 4p, 3p and 1p, in cfDNA isolated from plasma and serum of patients with neuroblastoma at diagnosis and have established its prognostic significance. Combaret et al.Reference Combaret, Brejon and Iacono105 evaluated the presence of 17q gain in circulating DNA of 142 patients with neuroblastoma for whom the genomic classification of tumours was known and reported specificity of 94·4% and sensitivity of 58·8% for 17q gain determination in circulating DNA for patients < 18 months of age. Furthermore, they demonstrated that 17q gain determination in circulating DNA is feasible, could help define patient prognosis, optimise therapy stratification and provide a more appropriate treatment for young children under 18 months of age, when no reliable information on genomic alterations can be obtained by other methods of tumour samples or for patients with neuroblastoma (stage IV-S or neonatal neuroblastoma) for whom a ‘wait and see’ strategy is recommended.Reference Combaret, Brejon and Iacono105 Yagyu et al.Reference Yagyu, Iehara and Gotoh104 examined chromosome 11q loss using cfDNA fragments in the sera of 24 children diagnosed with neuroblastoma and established a serum-based, surgery-free, quick and sensitive assay capable of detecting 11q loss and which could potentially lead to better outcome of neuroblastoma in the future. George et al.Reference George, Attiyeh and Li107 analysed paired blood and primary tumour samples from 22 children with high-risk neuroblastoma for loss of heterozygosity and DNA copy number change using the Affymetrix 10 K single-nucleotide polymorphism array and observed multiple areas of loss of heterozygosity and copy number gain. They observed that 68% (15 of 22) of samples had loss of heterozygosity on chromosome arm 11q and 95% (21 of 22) samples had gain of copy number on chromosome arm 17q which was associated with allelic imbalance. They reported that chromosome 11q loss of heterozygosity was highly associated with occurrence of chromosome 3p loss of heterozygosity, and loss of heterozygosity events on chromosomes 11q and 1p were generally accompanied by copy number loss which suggest hemizygous deletion within these regions. They concluded that single-nucleotide polymorphism array is a powerful tool for high-resolution determination of loss of heterozygosity and DNA copy number change in neuroblastoma.
Anaplastic lymphoma kinase (ALK) mutations and amplification in CfDNA
The ALK gene is located on the short arm of chromosome 2 at position 23 (2p23) and encodes a ALK tyrosine kinase receptor protein which is a member of the insulin receptor superfamily and helps control cell growth.Reference Osei, Lumini and Gunasekara34,Reference Wang, Zhou and Sun108,Reference Janoueix-Lerosey, Lequin and Brugieres109 Activating mutations in ALK have been reported to be observed in about 8–12% of patients with neuroblastoma at diagnosis and are associated with poor prognosis.Reference Kojima, Hiyama and Fukuba93,Reference Wang, Zhou and Sun108,Reference Combaret, Iacono and Bellini110,Reference Trigg, Shaw and Turner111 According to Trigg et al.Reference Trigg, Shaw and Turner111, genomic alterations of ALK may serve as important biomarkers in cfDNA for therapeutic stratification and monitoring of treatment response in the near future. Wang et al.Reference Wang, Zhou and Sun108 reported that ALK gene copy number increase is a recurrent genetic aberration of neuroblastic tumours, and MYCN-ALK amplification and ALK positivity predict inferior prognosis in neuroblastoma. Furthermore, patients with MYCN-ALK gain have better prognosis compared with those with normal MYCN-ALK level. Several investigationsReference Mosse, Laundenslager and Longo23,Reference Wang, Zhou and Sun108–Reference Combaret, Iacono and Bellini110,Reference De Brouwer, De Preter and Kumps112,Reference Trigg and Turner113 have been done to determine the correlation between ALK protein expression and clinical outcome in neuroblastoma. Combaret et al.Reference Combaret, Iacono and Bellini110 evaluated the ALK mutational status of neuroblastoma from serum or plasma samples collected from 114 patients with neuroblastoma by analysis of circulating DNA using the droplet digital PCR system for the detection of ALK mutations at F1174 and R1275 hotspots found in neuroblastoma tumours. They reported that the assay offers a reliable, non-invasive blood test to assess ALK mutational status at F1174 and R1275 hotspots and would help clinicians to identify patients that harbour an ALK mutation, especially when no tumour tissue is available. Furthermore, the analysis of circulating DNA by droplet digital PCR would help clinicians to identify patients who might benefit from ALK-targeted therapy and to define which targeted treatment should be used.Reference Combaret, Iacono and Bellini110
Circulating cell-free Tumour DNA (ctDNA)
The ctDNA is the tumour-derived fraction in the cfDNA component of the peripheral blood in circulation and contain cancer-specific mutations.Reference Tolaney, Toi and Neven6,Reference Sausen, Parpart and Diaz49,Reference Abbou, Shulman and DuBois50,Reference Gorgannezhad, Umer and Islam114 Jiang et al.Reference Jiang and Chan Cw115 reported that the analysis of ctDNA has opened up new possibilities for performing liquid biopsies for the assessment of solid tumours, and though its clinical potential has been acknowledged, many aspects of it biological characteristics remain uncertain. According to Tolaney et al.,Reference Tolaney, Toi and Neven6 elevated levels of ctDNA have been observed in neuroblastoma patients compared to healthy control individuals and are affected by tumour burden, vascularity and treatment response; moreover, they are generally lower in localised tumours but higher in metastatic tumours. StudiesReference Tolaney, Toi and Neven6,Reference Abbou, Shulman and DuBois50,Reference Bettegowda, Sausen and Leary81,Reference Chicard, Boyault and Daage87,Reference Klega, Imamovic-Tuco and Ha116 have reported that plasma ctDNA is present in 60 to 70% of neuroblastoma cases at diagnosis, can be used to evaluate tumour burden, is correlated with treatment response and capable of predicting relapse in neuroblastoma patients. Perakis and SpeicherReference Perakis and Speicher51 reported that ctDNA is detected in over 75% of patients with advanced cancers and in 48–73% of patients with localised tumours when digital polymerase chain reaction (PCR) technique is used. Phallen et al.,Reference Phallen, Sausen and Adleff117 Siravegna et al.Reference Siravegna, Marsoni and Siena46 and Abbouhave et al.Reference Abbou, Shulman and DuBois50 reported that ctDNA carries the same genetic alterations and mutations as those existing in the primary tumour and could provide opportunities for easily accessible and high sensitivity technique for patient-specific early cancer screening and diagnosis, non-invasive tumour profiling and identification of targetable somatic variants, pre-treatment risk stratification, monitoring of treatment response, detection of minimal residual disease, surveillance for relapse and guidance for treatment. According to Uemura et al.,Reference Tolaney, Toi and Neven6 tumour-associated genetic and epigenetic changes, such as point mutations, genomic rearrangements, copy number variation, microsatellite instability, loss of heterozygosity and DNA methylation, have been detected in ctDNA. Furthermore, since neuroblastoma relapses frequently occur at multiple anatomic sites and distinct DNA aberrations can be unique to each metastatic site, the assessment of ctDNA would potentially provide unprecedented insights into the metastatic and recurrent processes of neuroblastoma.Reference Tolaney, Toi and Neven6 Klega et al.Reference Klega, Imamovic-Tuco and Ha116 investigated whether patients with childhood solid tumour malignancies express detectable levels of ctDNA from plasma and demonstrated that children with solid tumour malignancies have detectable levels of ctDNA which may be used to track disease response. Furthermore, they reported that ctDNA can identify genomic subclassifiers of disease and changes in ctDNA levels correlated with disease burden and response to therapy. A number of investigationsReference Combaret, Hogarty and London83,Reference Chicard, Colmet-Daage and Clement84,Reference Chicard, Boyault and Daage87,Reference Lodrini, Sprussel and Astrahantseff102–Reference Combaret, Brejon and Iacono105,Reference Combaret, Iacono and Bellini110,Reference Gorgannezhad, Umer and Islam114,Reference Klega, Imamovic-Tuco and Ha116,Reference Gotoh, Hosoi and Iehara118–Reference Bosse, Buongervino and Lane120 have examined ctDNA levels in peripheral blood of patients with neuroblastoma for the purpose of identifying diagnostic and prognostic genomic features, assessment of response to therapy, post-treatment disease surveillance and provision of a non-invasive method to profile tumour-associated genomic aberrations, heterogeneity and evolution. They reported that ctDNA assay has significant potential to provide a sensitive biomarker for diagnosis, treatment response and monitoring for solid tumours, and its clinical use will decrease reliance on serial imaging and invasive biopsies in paediatric patients.
DNA methylation in CtDNA
Changes in DNA methylation (epigenetic alterations) status is one of the most common molecular alterations in human neoplasia including neuroblastoma, and several potential tumour-suppressor genes have been reported as frequently silenced by hypermethylation in neuroblastomas.Reference Yagyu, Gotoh and Iehara119,Reference Misawa, Tanaka and Yagyu121 The Ras-association domain family 1, isoform A (RASSF1A) gene is reported to be frequently silenced and inactivated by promoter region hypermethylation in many adult and childhood cancers, including neuroblastoma, and aberrant hypermethylation of the decoy receptor 2 (DCR2) promoter has been associated with rapidly progressing neuroblastoma tumour.Reference Yagyu, Gotoh and Iehara119 Furthermore, methylation of promoter CpG islands is reported to inhibit transcriptional initiation and cause permanent silencing of downstream genes.Reference Misawa, Tanaka and Yagyu121
Several studiesReference Yagyu, Gotoh and Iehara119,Reference Misawa, Tanaka and Yagyu121,Reference Sun, Jiang and Chan122 have evaluated the potential of circulating tumour-related methylated DNA in serum for the molecular diagnosis and prognosis of various types of cancer. Misawa et al.Reference Misawa, Tanaka and Yagyu121 used a series of matched neuroblastoma tumour and serum DNA obtained from 68 patients to investigate the potential detection of the RASSF1A gene epigenetic alterations in the serum of neuroblastoma patients and also if aberrant RASSF1A methylation in patient pre-therapeutic serum is of prognostic significance in neuroblastoma. They observed serum methylation of RASSF1A in 25% (17 of 68) of patients and demonstrated the value of detecting circulating methylated RASSF1A in serum of patients as a potential predictive marker of neuroblastoma outcome. They reported that serum methylation of RASSF1A was statistically associated with age ≥12 months, stage IV and MYCN amplification at diagnosis and found the influence of serum RASSF1A methylation on prognosis to be comparable to that of MYCN amplification (which is currently the most reliable marker). They concluded that the methylation status of RASSF1A in the serum of patients with neuroblastoma has the potential to become a prognostic predictor of outcome. Yagyu et al.Reference Yagyu, Gotoh and Iehara119 investigated the potential development of a non-invasive DCR2 methylation assay for patients with neuroblastoma using serum ctDNA and DNA-based real-time PCR in 86 patients (18 patients with MYCN amplification neuroblastoma and 68 patients had non-MYCN amplification neuroblastoma) and established a non-invasive, sensitive and rapid assay for quantifying aberrant methylation of DCR2 gene promoter in patient serum. They reported that DCR2 methylation was associated with stage and DCR2-methylated patients showed significantly poorer 5-year event-free survival. They demonstrated that the detection of methylated-DCR2 in serum DNA hold promise for predicting prognosis and therapeutic efficacy in neuroblastoma, especially in non-MYCN-amplified patients. Furthermore, they indicated that DCR2 methylation status is a sensitive and useful marker of tumour recurrence in DCR2-methylated cases and can potentially distinguish neuroblastoma cases with poor outcome before initial therapy, and its serial monitoring can be a sensitive indicator of therapeutic efficacy in DCR2-methylated cases.
Circulating tumour cells (CTCs)
CTCs are tumour cells circulating in the peripheral blood or lymphatic vessels of patients, shed from either the primary tumour or from its metastasis and are considered the main mechanism for tumour metastasis.Reference Siravegna, Marsoni and Siena46,Reference Hao and Zhang47,Reference Man, Wang and Kemmner123,Reference Joosse, Gorges and Pantel124 The first account of the existence of CTCs in the blood was reported by Thomas Ashworth in 1869 in which he postulated that CTC cells might potentially offer better understanding of the metastatic process in cancer patients.Reference Perakis and Speicher51,Reference Ferreira, Ramani and Jeffrey125,Reference Ashworth148 Metastasis caused by the entry of CTC into the bloodstream or lymphatic vessels as single cells or in clusters is the major factor contributing to disease progression, relapse, poor survival and death in cancer patients.Reference Man, Wang and Kemmner123,Reference Ferreira, Ramani and Jeffrey125,Reference Liu, Zhang and Zhang126 Ferreira et al.Reference Ferreira, Ramani and Jeffrey125 reported that elevated counts of CTC in serum is associated with poor prognosis. The detection and characterisation of CTCs offer opportunities to understand the metastatic process, early detection of metastasis and to provide improved means of treatment and independent prognostic information for neuroblastoma patients.Reference Hao and Zhang47,Reference Man, Wang and Kemmner123,Reference Ferreira, Ramani and Jeffrey125 StudiesReference Siravegna, Marsoni and Siena46,Reference Hao and Zhang47,Reference Man, Wang and Kemmner123,Reference Ferreira, Ramani and Jeffrey125 have demonstrated that the presence of CTC in the blood of neuroblastoma patients correlates with treatment outcomes, poor overall survival of patients with metastatic disease and patient prognosis. Furthermore, determination of CTC is reported to be capable of monitoring tumour recurrence and therapeutic responses in real time, identify new therapeutic targets, elucidate drug resistance mechanisms and improve current understanding of tumour progression and metastatic disease.Reference Ferreira, Ramani and Jeffrey125 However, the challenges facing the clinical analysis and molecular recognition of CTCs are their extreme rarity in serum (only about 1–10 CTCs per milliliter of whole blood containing 107–109 haematologic cells) and their clinical and genetic heterogeneity.Reference Hao and Zhang47,Reference Ferreira, Ramani and Jeffrey125
Liu et al.Reference Liu, Zhang and Zhang126 used epithelial cell adhesion molecule (EpCAM)-independent assay combined with immunostaining fluorescence in situ hybridisation (i-FISH) to investigate CTCs in the peripheral blood of 28 neuroblastoma patients and reported that the technique was capable of detecting CTCs at a high sensitivity of 88·89% and specificity of 78·59% and could have clinical significance for prognosis evaluation and diagnosing metastasis when imaging data are uncertain. They observed that patients with metastasis had elevated CTCs counts and were significantly correlated with prognosis and patient overall survival. Furthermore, they reported that patients with CTCs counts per 4 mL of peripheral blood ≥3 had a high likelihood of metastasis and patients with counts ≥10 had poorer overall survival. They suggested that a cut-off value of 3 CTCs counts per 4 mL of peripheral blood could potentially predict metastasis and can help physicians verify a patient’s risk for metastatic disease when imaging results are uncertain. Kuroda et al.Reference Kurodaa, Morikawaa and Matsuoka127 studied the prognostic significance of CTC and indications for aggressive surgery in 29 neuroblastoma patients (stage IV, n = 24 and stage III, n = 5) using RT–PCR method and reported an overall survival rate of 58·6% (17 of 29). They detected CTCs in 55·6% (10 of 18) of stage IV patients but none in stage III patients, and all reported deaths in the study were related to systemic metastases in the CTC-positive patients. Furthermore, the survival rate was significantly lower (33·8%) in patients with CTC and/or persistent bone marrow micrometastasis compared to those without detectable micrometastasis (87·5%). They concluded that the presence of CTC and/or persistent micrometastasis may suggest the presence of a tumour cell population that is resistant to treatment and is disseminating systemically and thus can be regarded as significant risk factors in neuroblastoma.
Tumour Ribonucleic Acid (tRNA)
Messenger RNA (mRNA)
mRNA is a large family of RNA molecules that carry genetic information from DNA to the ribosome, where they specify the amino acid sequence of the protein products of gene expression and hence, mRNAs can reflect both the state of the genome and the cell’s current homeostatic state.Reference Zaporozhchenko, Ponomaryova and Rykova128 According to Feng et al.,Reference Feng, Qin and Zhang129 mutations and rearrangements, such as gene fusions, as well as the presence of alternative splicing variants and consequences of epigenetic transcriptional regulation can be revealed by the analysis of circulating mRNAs. The circulating cell-free mRNA (cfmRNA) contains tumour-associated genetic alterations and was first demonstrated in the 1990s in the plasma of patients with nasopharyngeal carcinoma,Reference Siravegna, Marsoni and Siena46,Reference Zhou, Tang and Liu130 and its clinical importance has been investigated in other solid tumours and can serve as a suitable marker for cancer.Reference Tolaney, Toi and Neven6,Reference Yáñez, Grau and Oltra24,Reference Träger, Kogner and Lindskog131–Reference Burchill, Lewis and Abrams138 Tumour-derived mRNAs have been discovered in relative abundance in peripheral blood, urine, sputum and other biological fluids and have been associated with cancer aggressiveness and prognosis.Reference Zaporozhchenko, Ponomaryova and Rykova128
Several investigatorsReference Tolaney, Toi and Neven6,Reference Träger, Kogner and Lindskog131–Reference Burchill, Lewis and Abrams138 have used RT-qPCR assay to detect mRNA of neuroblastoma-associated genes [e.g., tyrosine hydroxylase (TH), paired-like homeobox 2b (PHOX2B), embryonic lethal, abnormal vision-4 (ELAV-4), dopamine decarboxylase (DDC), beta-1,4-N-acetylgalactosaminyltransferase 1 (GD2-s), doublecortin (DCX), sialyltranferase ST8SiaII (STX) and chromogranin A (CHGA)] and have found the RT-qPCR method to be specific, sensitive, reliable and easy to perform and can potentially be used for upfront staging, prognostic assessment, treatment monitoring, prediction of event-free survival and overall survival of neuroblastoma. Träger et al.Reference Träger, Kogner and Lindskog131 investigated TH mRNA expression in blood and bone marrow at diagnosis using real-time reverse transcription-PCR (RT-PCR) for the detection and monitoring of neuroblastoma cells during therapy and also evaluated the usefulness of the method for early and sensitive detection of relapsed disease during follow-up. They observed that all control blood samples were below the detection limit of 70 transcripts/mL, blood TH mRNA was elevated in children with metastatic disease (stage IV and IV-S) at diagnosis than patients with localised disease (stages I–III) and TH-mRNA concentration was correlated with clinical course. Lee et al.Reference Lee, Son and Choi132 investigated the clinical significance of TH mRNA expression in peripheral blood at diagnosis in 210 patients (median age was 22 months) with newly diagnosed neuroblastoma using RT-qPCR. They reported that TH mRNA expression was positive in 28% (60 of 210) of patients and 83% (50 of 60) of the TH mRNA positive patients had metastatic tumours and the rest (17%) had localised tumours; moreover, TH mRNA expression in peripheral blood at diagnosis was an independent poor prognostic factor for progression-free survival. Furthermore, they observed that the 5-year progression-free survival was lower in TH-mRNA positive patients and higher TH mRNA expression level was associated with high-risk features (i.e., advanced stage, older age, unfavourable pathology and MYCN amplification) at diagnosis and with worse outcome.
Burchill et al.Reference Burchill, Lewis and Abrams134 investigated the independent prognostic value of neuroblastoma cells detected by RT-PCR for TH mRNA detection in a prospective, multicenter study and reported the detection of TH mRNA in peripheral blood of 67% (33 of 49) of children with stage IV neuroblastoma and children aged >1year at diagnosis. In addition, they indicated that TH mRNA could potentially identify patients with an unfavourable prognosis by association with the disease stage at diagnosis and within advanced stage patients. They observed that TH mRNA is a poor prognostic indicator, a significant predictive factor for overall survival and event-free survival, and the presence of TH mRNA in peripheral blood improved the accuracy of staging for some children. They further demonstrated that when TH mRNA levels are combined with a serum LDH > 1,500 IU/L, it gives the most powerful poor prognostic model at diagnosis for children >1-year-old with stage IV disease. Yáñez et al.Reference Yáñez, Grau and Oltra24 analysed TH mRNAs and DCX mRNAs expression at diagnosis in peripheral blood and bone marrow samples from 102 patients with non-metastatic neuroblastoma to determine the association between the presence of minimal disease and outcome using RT-qPCR. They reported that the detection of minimal disease in peripheral blood of patients is useful for predicting relapse in patients with non-metastatic neuroblastoma, showed lower event-free survival, showed high survival rates for stages I and II patients, and is useful for predicting outcome and for early and sensitive detection of relapse disease in stage III patients with neuroblastoma. Viprey et al.Reference Viprey, Gregory and Corrias133 evaluated the detection of neuroblastoma mRNAs using RT-qPCR in peripheral blood and bone marrow aspirates from 290 children with stage IV neuroblastoma to determine if they are clinically useful biomarkers of risk. They reported that high level of TH, PHOX2B or DCX mRNA in peripheral blood and bone marrow at diagnosis is a strong predictor of worse event-free survival and overall survival, capable of objectively identifying children with high risk and could contribute to an algorithm to improve stratification of children for treatment.
MicroRNA (miRNA)
miRNAs are small single-stranded non-coding RNA molecules (~22 nucleotides) that are secreted into the circulatory system and are the most abundant class of circulating cell-free RNA molecules in the blood.Reference Tolaney, Toi and Neven6,Reference Chen, Ba and Ma48,Reference Zaporozhchenko, Ponomaryova and Rykova128,Reference Mitchell, Parkin and Kroh139–Reference Galardi, Colletti and Businaro142 They are important regulators of gene expression and play a key role in many biological processes. According to Zaporozhchenko et al.,Reference Zaporozhchenko, Ponomaryova and Rykova128 many non-coding RNAs are directly involved in the regulation of key signal transduction pathways and can act as cancer drivers. The expression of specific subsets of miRNAs in biological fluids has been associated with the initiation and progression of cancer and can easily be detected by molecular biology techniques.Reference Zaporozhchenko, Ponomaryova and Rykova128,Reference Peng and Croce143 Chen et al.Reference Chen, Ba and Ma48 reported that serum and plasma contain a large amount of stable miRNAs and demonstrated that the levels of miRNAs in serum are stable, reproducible and consistent among patients, and that the expression profile of miRNAs shows great promise as a novel non-invasive biomarker for diagnosis of cancer and other diseases. Furthermore, they observed that miRNAs also exist in other body fluids, including urine, tears, ascitic fluid and amniotic fluid, and the detection of miRNA has the potential to transform present clinical cancer management, including determination of cancer classification, estimation of prognosis, prediction of therapeutic efficacy, maintaining surveillance following surgery, as well as forecasting disease recurrence.Reference Chen, Ba and Ma48 According to Peng and Croce,Reference Peng and Croce143 miRNA expression is dysregulated in human cancer through various mechanisms, including amplification or deletion of miRNA genes, abnormal transcriptional control of miRNAs, dysregulated epigenetic changes and defects in the miRNA biogenesis machinery and may function as either oncogenes or tumour suppressors under certain conditions.
Several studiesReference Murray, Raby and Saini140–Reference Galardi, Colletti and Businaro142,Reference Chen and Stallings144–Reference Ma, Jiang and Kang147 have evaluated miRNAs in plasma and serum as non-invasive biomarkers for diagnosis, therapeutic stratification and prognostication across various cancer types, including neuroblastoma. Murray et al.Reference Murray, Raby and Saini140 hypothesised that differential expression and/or release of miRNAs by solid tumours may be detected as altered serum miRNA profiles and therefore undertook a global RT-qPCR miRNA profiling (n = 741) on RNA from 53 serum samples, representing 33 diagnostic cases of common childhood cancers, including neuroblastoma. They identified a panel of miRNAs (miR-124-3p, miR-9-3p, miR-218-5p, miR-490-5p and miR-1538) of potential importance in the clinical management of neuroblastoma and found the qRT-PCR approach to be more cost-effective. They reported that sera from neuroblastoma patients with MYCN amplification showed overexpression of miR-124-3p, miR-9-3p, miR-218-5p, miR-490-5p and miR-1538 consistent with a study by Chen et al.Reference Chen and Stallings144 who demonstrated MYCN status as a determinant of global miRNA profiles in neuroblastomas. Zeka et al.Reference Zeka, Decock and Can Goethem145 investigated the abundance levels of 743 human miRNAs in serum samples collected from 185 neuroblastoma patients at diagnosis and reported that the evaluation of serum miRNA variance in a model testing for tumour stage, MYCN status, age at diagnosis, and overall survival revealed tumour stage as the most significant factor impacting miRNA abundance in neuroblastoma serum. Furthermore, they observed that miRNAs were strongly associated with metastatic stage IV disease, disease burden and treatment response among patients with metastatic and localised disease. They also reported elevated levels of miRNAs in serum from xenografted mice bearing human neuroblastoma tumours, and the levels were strongly associated with tumour volume which was validated in longitudinal serum samples from metastatic neuroblastoma patients. Ramraj et al.Reference Ramraj, Aravindan and Somasundaram146 used whole-miRnome miRNA microarray profiling to assess serum-circulating miRNAs in control mice without tumours, mice bearing non-metastatic human neuroblastoma xenografts and mice with high-risk neuroblastoma. They analysed serum levels of 852 miRNAs in clinically translatable well-characterised, reproducible mouse models of favourable and high-risk aggressive neuroblastoma. They identified clusters of serum-circulating miRNAs that showed both high and low serum levels in high-risk aggressive metastatic neuroblastoma and demonstrated the benefit of using serum-circulating miRNAs as prognostic biomarkers of neuroblastoma. They reported that utilising whole-miRnome approach to examine serum levels in favourable neuroblastoma and high-risk neuroblastoma has the potential to quantify up/down-regulated clusters, and the identification of these clusters could precisely predict neuroblastoma progression.
Conclusions
Neuroblastoma is a complex and heterogeneous disease that involves multiple gene–environment interactions and affects numerous biological pathways. Therefore, the development of reliable, sensitive and reproducible non-invasive detection method represents an important step towards the promise of precision and targeted medicine. Current investigations in liquid biopsy techniques continue to show significant potentials in the diagnosis, prognosis and stratification of cancer patients and demonstrate a potential alternative method for monitoring treatment response compared to tissue biopsy approach. The ease, non-invasiveness, frequency and opportunity of more rapid course correction of administering therapies made possible by serial liquid biopsy collection offers several advantages compared to standard surgical procedures. As our knowledge of the biology behind liquid biopsy analytes improves, in addition, as technological advancement continues and further innovations in liquid biopsy methodology evolve and the technology becomes commonplace in the clinical settings, so will the promise of precision and targeted medicine for the management of paediatric neuroblastoma patients.
Acknowledgement
The authors would like to acknowledge with much gratitude the support from the staff at the Medical Physics Department, Grand River Cancer Centre.
Conflict of Interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
