Other biological prognostic factors that have been extensively investigated include tumor cell telomere length, telomerase activity, and telomerase ribonucleic acid;[49,50] urinary VMA, HVA, and their ratio; dopamine; CD44 expression; TrkA gene expression; and serum neuron-specific enolase level, serum lactic dehydrogenase level, and serum ferritin level. High-level expression of the MRP1 drug resistance gene is an independent indicator of decreased survival. The profile of GABAergic receptors expressed in neuroblastoma is predictive of prognosis regardless of age, stage, and MYCN gene amplification. Gene expression profiling may prove useful for prognosis prediction. Whole chromosome copy number changes do not predict recurrence, while segmental chromosome number changes do.[55,56] In addition, response to treatment has been associated with outcome. The persistence of neuroblastoma cells in bone marrow during or after chemotherapy, for example, is associated with a poor prognosis.[57,58]
Unique Aspects of Neuroblastoma
Biologically discrete types of neuroblastoma
Based on biologic factors and an improved understanding of the molecular development of the neural crest cells that give rise to neuroblastoma, the tumors have been categorized into three biological types. These types are not used to determine treatment at this time; however, type 1 has a very favorable prognosis, while types 2 and 3 have poor prognoses.
- Type 1: Expresses the TrkA neurotrophin receptor, is hyperdiploid, and tends to spontaneously regress.[59,60]
- Type 2: Expresses the TrkB neurotrophin receptor and its ligand, has gained an additional copy of chromosome 17q, has loss of heterozygosity of 14q or 11q, and is genomically unstable.[59,60]
- Type 3: Has a gain of chromosome 17q, loss of chromosome 1p, and the MYCN gene becomes amplified.[59,60]
Children whose tumors have lost a copy of 11q are older at diagnosis, and their tumors contain more segmental changes in gene copy number compared with children whose tumors show MYCN amplification.[61,62] Moreover, segmental chromosome changes not detected at diagnosis may be found in neuroblastomas at relapse. This suggests that clinically important tumor progression is associated with accumulation of segmental chromosomal alterations.
Current data do not support neuroblastoma screening. Screening infants for neuroblastoma by assay of urinary catecholamine metabolites was initiated in Japan. A large population-based North American study, in which most infants in Quebec were screened at the ages of 3 weeks and 6 months, has shown that screening detects many neuroblastomas with favorable characteristics [65,66] that would never have been detected clinically, apparently due to spontaneous regression of the tumors. Another study of infants screened at the age of 1 year shows similar results. Screening at the ages of 3 weeks, 6 months, or 1 year caused no reduction in the incidence of advanced-stage neuroblastoma with unfavorable biological characteristics in older children, nor did it reduce the number of deaths from neuroblastoma in infants screened at any age.[66,67] No public health benefits have been shown from screening infants for neuroblastoma at these ages. (Refer to the PDQ summary Neuroblastoma Screening for more information.)