Screening and Diagnosis

Given the high incidence of neuroblastoma in infancy, newborn urinary screening for catecholamines, excreted in 80% to 90% of cases, has been suggested as a cost-effective way to detect this tumor. Interestingly, the results of screening at birth and 6 months reported thus far suggest that most of the cases detected by this method are biologically favorable tumors with a high rate of spontaneous regression. Although screening has revealed an increased incidence in infancy, the survival or stage distribution of cases in children over age 1 year is not substantially affected, suggesting that this method does not detect the more malignant tumors at an early stage.[2]

Approximately half of children who present with neuroblastoma already have disseminated disease. The most common symptoms are pain due to either primary tumor or to bone and bone marrow involvement, an abdominal mass, weight loss, anorexia, and irritability. Metastases to the orbit are common with resulting proptosis, while mediastinal primaries may be accompanied by Horner’s syndrome. Rare presentations of neuroblastoma due to paraneoplastic phenomena include secretory diarrhea (due to secretion of vasoactive intestinal peptide) or opsoclonus, myoclonus, and ataxia (which is possibly due to cross-reacting antineuronal antibodies elicited by the tumor, which cause acute cerebellar encephalopathy and result in this syndrome in about 3% of patients).

Primary tumors arise in the abdomen in 70% of patients, often in the adrenal gland and next most frequently in the chest. Bone marrow and bone are the most common sites of metastases, while lung and central nervous system metastases are very rare (less than 5%).[2]

Diagnostic evaluation of neuroblastoma requires tissue for both diagnosis and prognosis, since histologic grading, as well as molecular genetic studies, are essential for management decisions. Urinary catecholamine determinations and serum ferritin and LDH levels are useful both in assessing prognosis and as a baseline for disease monitoring (Table 2).

Metastatic and staging evaluation includes, at a minimum, a computed tomographic (CT) or magnetic resonance imaging (MRI) scan of the primary, bone scan, bilateral bone marrow aspirate and biopsy, and, optimally, an iodine-123- or iodine-131-metaiodobenzylguanidine (MIBG) scan (Table 2). Meta-iodobenzylguanidine is a derivative of guanethidine, similar in structure to norepinephrine, that specifically concentrates in neuroblastoma and pheochromocytoma. It is 90% sensitive for tumor detection, and has the advantage of imaging equally well in soft-tissue, bone, and bone marrow tumors, with no uptake in normal bone, thus preventing the confusion with normal epiphyses that can occur with technetium scans.[19,20] Additional sensitivity for bone marrow metastases can be achieved with immunocytologic testing using either immunofluorescence or immunoperoxidase stain, which can detect as little as 0.001% tumor infiltration, including unsuspected bone marrow tumor in patients with otherwise low-stage disease.[21,22]

Staging Systems

Several different staging systems have been used to help plan treatment for neuroblastoma: (1) the original and widely used classification of Evans, also used by the CCG; (2) the St. Jude Children's Research Hospital system, now revised and utilized by the Pediatric Oncology Group (POG); and (3) the International Union Against Cancer (UICC) tumor-node-metastasis (TNM) system. All three systems have prognostic value. An international consensus group has arrived at a new synthesis of these systems, the International Neuroblastoma Staging System (INSS; see Table 3). The international system is being tested further prospectively in order to facilitate international comparisons of treatments.[19,20]

Risk-Based Therapy

Risk Classification

Treatment of neuroblastoma must be tailored to both the extent of disease and the biological risk characteristics. A working division into risk groups, shown in Table 4, categorizes patients into three groups requiring different treatment intensities.

Further study using multivariate analysis may result in greater refinement of a combined anatomic and biological staging system. However, the current working classification used by the CCG has succeeded in identifying those children who have an excellent prognosis when given little or no therapy and those who have a guarded prognosis even with very aggressive treatment.

Low-Risk Group

The low-risk group is comprised of patients with INSS stage 1, 2, or 4s disease and usually favorable biological characteristics. These patients require minimal therapy, usually surgery alone, unless they have local disease-related symptoms, such as cord compression from a paraspinal tumor or respiratory distress from liver infiltration by a tumor in the intraventricular system (Figure 2).

Both retrospective and now prospective studies have confirmed that surgery alone is sufficient therapy for stage 1 and 2 neuroblastoma. A retrospective CCG analysis showed that the survival rate among patients with stage 2 disease was greater than 90% regardless of measurable postoperative residual tumor.[15] This has been confirmed in prospective studies conducted by both the POG and CCG.

The overall survival rate for more than 200 children with stage 2 disease treated with initial surgical management is greater than 95% regardless of other biological features, with only 13% of patients requiring any chemotherapy for symptoms. Even the 15% of patients who have a local tumor recurrence after surgery can usually be treated effectively with moderate-intensity chemotherapy or local radiation.[unpublished CCG data]

Neither radiation nor chemotherapy has been shown to contribute to survival in the low-risk group as a whole. The only exceptions are children over 1 year old at diagnosis who have stage 2 disease with tumor MYCN amplification. In previous studies, these children showed rapid progression and dissemination without aggressive therapy; therefore, it is now recommended that they receive very aggressive therapy on the treatment protocols for high-risk patients.

Children with stage 4s neuroblastoma also had a 90% overall 3-year survival in the most recently completed CCG study, CCG-3881. However, approximately half of these children required a short course of chemotherapy with cyclophosphamide (Cytoxan, Neosar), and some received hepatic irradiation for respiratory distress. The rare patient with MYCN amplification progressed rapidly, and therefore, will be treated as a high-risk patient in the future.[23,24]

Intermediate-Risk Group

The recently completed CCG-3881 study for intermediate-risk patients treated all infants with stage 3 disease; children over 1 year old with stage 3 disease and favorable Shimada classification, normal ferritin (less than 143 ng/mL), and single-copy MYCN; and infants with stage 4 disease and nonamplified MYCN with a 9-month course of chemotherapy, surgery, and local radiation to residual disease. The 3-year event-free survival rate for the stage 3 patients was 96%. Event-free survival did not differ between the infants and the children greater than 1 year old with favorable-biology tumors, nor between the infants with unfavorable biological features and those with no unfavorable biological features.

In the past, the subset of stage 3 patients greater than1 year old at diagnosis had an event-free survival of only 30% to 40%. In CCG-3881, stage 4 infants had a 75% 3-year event-free survival overall, but if only patients with known nonamplified MYCN were considered (N = 72), the 3-year event-free survival was 95%.

The excellent outcome for these intermediate-risk patients has led to the goal of decreasing cost and late sequelae in this group by decreasing the overall cumulative dose and duration of therapy without compromising dose intensity. Thus, in the next intermediate-risk protocol (a combined effort of the CCG and POG), patients will be treated with only four cycles (12 weeks) of combination chemotherapy, all administered in the outpatient setting; the only exception will be children with either unfavorable Shimada classification or diploidy, who will receive eight cycles. The stage 3 and 4 patients with MYCN amplification will be considered high risk regardless of age, and will be treated on a different protocol.[16,23,24]