Contents
8.1 Introduction . . . 87
8.2 Historical Overview . . . 88
8.3 Basic Morphology . . . 88
8.3.1 Neuroblastoma (Schwannian Stroma-poor) . . . 89
8.3.2 Ganglioneuroblastoma, Intermixed (Schwannian Stroma-rich) . . . 91
8.3.3 Ganglioneuroma (Schwannian Stroma-dominant) . . . 91
8.3.4 Ganglioneuroblastoma, Nodular (Composite, Schwannian Stroma-rich/ Stroma-dominant and Stroma-poor) . . . . 91
8.4 Prognostic Classification . . . 92
8.5 Biological Relevance . . . 92
8.5.1 Schwannian Development in Neuroblastic Tumors . . . 92
8.5.2 Correlation of Histopathology with MYCN Amplification and trkA Expression . . . 93
8.5.3 Composite Tumor . . . 93
8.6 Conclusion . . . 94
References . . . 94
8.1 Introduction
Peripheral neuroblastic tumors (pNTs), which in- clude neuroblastoma, ganglioneuroblastoma, and ganglioneuroma, are common pediatric tumors (Ross et al. 1996). These tumors are derived from im- mature sympathetic neuroblasts during embryonic, fetal, or early postnatal development, and their mor- phological features appear to recapitulate develop- mental stages of sympathetic ganglia. Their primary sites are anatomically related to the embryological distribution of neural crest cells, and include adrenal gland and structures of the sympathetic nervous sys- tem.
For many years pNTs were characterized as “enig- matic” because of their unexpected clinical behav- iors, such as involution/spontaneous regression, mat- uration, or aggressive progression. Because of recent advances in clinical and basic research, pNTs now are considered to be biologically heterogeneous, and their individual molecular properties like account for their unique clinical behaviors (Brodeur and Maris 2002). Based primarily on their clinical biology, the International Neuroblastoma Pathology Classifica- tion (Shimada et al. 1999a,b) was established by adopting the concept of the original Shimada system (age-linked evaluation of the morphological features in this disease). In this chapter, histopathology of the pNTs is illustrated according to the Classification along with its biological relevance.
Pathology of Peripheral Neuroblastic Tumors
Hiroyuki Shimada, Inge M. Ambros
8.2 Historical Overview
Since the beginning of the twentieth century, at- tempts have been made to deduce prognostic infor- mation from the histological appearance of the indi- vidual tumors (Beckwith and Martin 1968; Hughes et al. 1974; Landau 1911; Mäkinen 1972; Wahl 1914). In 1914 Wahl suggested a sequence of maturation of the pNTs (Wahl 1914), and in 1968, Beckwith and Martin proposed a grading system based on the semi-quan- titative assessment of neuroblastic cytodifferentia- tion (Beckwith and Martin 1968). In 1974 Hughes and co-workers proposed their grading system, but at this time based on non-quantitative assessment of neuroblastic/ganglionic cytodifferentiation (Hughes et al. 1974). As summarized in a review article by Dehner in 1988 (Dehner 1988), however, those at- tempts could not successfully satisfy the oncologists dealing with this “enigmatic” disease.
Interestingly, in the first half of the twentieth cen- tury, it was believed that older patients had a better prognosis (Landau 1911; Wahl 1914), which was probably due to the inability to distinguish local-re- gional from stage-4 metastatic disease. In the second half of the twentieth century, however, it was clearly recognized that younger patients (especially diag- nosed before 1 year of age) had a significantly better prognosis than older patients (Gross et al. 1959). Fur- thermore, the majority of tumors in infants with clin- ically favorable outcome showed no or very limited morphological evidence of cytodifferentiation. Beck- with and Martin concluded that “differences in de- gree of maturation probably did not account for the more favorable outcome of the neuroblastomas in in- fancy” (Beckwith and Martin 1968).
In 1984 Shimada and colleagues proposed a classi- fication system based on a unique concept of age- linked evaluation of morphological indicators (Shi- mada et al. 1984). First they made an age-appropriate framework of the maturational sequence of the pNTs.
The maturational sequence was defined by two mor- phological indicators, grade of neuroblastic differen- tiation, and degree of Schwannian stromal develop- ment. Prior to their study, Schwannian stromal com- ponent, which is one of the major elements in the nor-
mal ganglionic structure of the sympathetic nervous system, had never been a subject of serious investiga- tion in pNTs. According to this classification system, clinically favorable tumors can be less differentiated when diagnosed in younger patients, and should have morphological features of more advanced maturation in older children (for detailed explanation see Chap. 4). They also found increased numbers of kary- orrhectic cells in highly aggressive tumors, and intro- duced a concept of mitosis–karyorrhexis index.
In 1992 Joshi and co-workers proposed histologi- cal grading by using mitotic rate (MR: low £10/10 high-power fields, high >10/10 high-power fields) and calcification (presence or absence; Joshi et al.
1992). In their report they also proposed a risk group- ing by combining the histological grade and age of the patient at diagnosis (low risk: patients in all age groups with tumor having low MR and calcification, and patients £1 year of age with either low MR or cal- cification; high risk: patients >1 year of age with ei- ther low MR or calcification, and patients in all age group with high MR and no calcification; Joshi et al.
1992). They later published a modified histological grading by replacement of mitotic rate with MKI (Joshi et al. 1996).
In 1994 the International Neuroblastoma Patholo- gy Committee was formed to establish a prognosti- cally significant and biologically relevant classifica- tion for international use. The Committee first de- fined terminology and morphological criteria of pNTs, and then analyzed and tested mainly those two classifications proposed by Shimada et al. (1984) and Joshi et al. (1992, 1996). In 1999, after 5 years of col- laborative work, the Committee developed the Inter- national Classification based on the original Shimada classification with minor modifications (Shimada et al. 1999a,b).
8.3 Basic Morphology
As proposed by Shimada et al. in their original classi- fication (Shimada et al. 1984), pNTs are classified into four basic morphological categories (Shimada et al.
1999a): neuroblastoma (Schwannian stroma-poor);
ganglioneuroblastoma, intermixed (Schwannian stro-
ma-rich); ganglioneuroma (Schwannian stroma- dominant); and ganglioneuroblastoma, nodular (composite, Schwannian stroma-rich/stroma-domi- nant and stroma-poor). Within each category one or more subtypes are recognized (see below). The first three categories and their subtypes are based on mor- phological changes according to the maturational se- quence. The Shimada system distinguishes biologi- cally favorable pNTs with a potential of age-appropri- ate maturation and biologically unfavorable pNTs without such a potential (see Chap. 4). In the last cat- egory, the tumor is composed of clearly distinct mul- tiple clones, representing different states of matura- tion or maturational arrest. Among these categories, ganglioneuroma is not considered a separate entity, but rather as a fully mature form of tumor constitut- ing the end of the biological continuum for all the pNTs, a model which postulates that ganglioneuro- mas are neuroblastomas at a later time in their devel- opment.
To date, there is no clear distinction in molecular characteristics between pNTs with a potential of re- gression and pNTs with a potential of maturation. In fact, during the maturational sequence of pNTs, the vast majority of neuroblastic cells probably die be- fore or after reaching a certain degree of maturation (cellular death during the process of tumor matura- tion, comparable to cellular death seen in the process of normal organogenesis). By contrast, Schwannian stroma, once developed and established, are believed by many to constitute a persistent and dominant component in pNTs.
8.3.1 Neuroblastoma (Schwannian Stroma-poor)
Tumors in this category are composed of neuroblas- tic cells forming lobules separated by thin fibrovas- cular septa where Schwann cells (or their precursors) can (or may) be detected as slender S-100 positive cells (Shimada et al. 1985). Three subtypes, i.e., un- differentiated, poorly differentiated, and differentiat- ing, are distinguished based on different grades of neuroblastic differentiation. It is noteworthy that in the original Shimada Classification, there were two subtypes, undifferentiated (including undifferentiat-
ed and poorly differentiated subtype of the Interna- tional Classification) and differentiating (same as differentiating subtype), in this category. On gross examination, tumors are generally soft in consisten- cy. Cut surfaces of those in the undifferentiated and poorly differentiated subtype are often hemorrhagic, while tumors in the differentiating subtype are usual- ly tan-yellow, without hemorrhage.
Neuroblastoma, Undifferentiated Subtype (Fig. 8.1 a):
In this rare subtype, tumor tissue is composed of un- differentiated neuroblastic cells without identifiable neuropil or rosettes. In order to establish the diagno- sis, supplementary tests, such as immunohistochem- istry, electron microscopy, and/or molecular/cytoge- netic analysis, are usually required.
Neuroblastoma, Poorly Differentiated Subtype (Fig.
8.1 b): Diagnosis for tumor in this subtype is rela- tively easy because of the presence of varying amount of neuropil and/or rosettes of the Homer-Wright type. Most of the tumor cells are undifferentiated:
less than 5% of the population has morphological ev- idence of differentiation (see below).
Neuroblastoma, Differentiating Subtype (Fig. 8.1 c):
Tumor of this subtype usually has abundant neu- ropil. Five percent or more of the tumor cells are dif- ferentiating neuroblasts: they are characterized by synchronous differentiation of the nucleus (enlarged, eccentrically located with vesicular chromatin pat- tern, and a single prominent nucleolus), and of the cytoplasm (eosinophilic/amphophilic with a diame- ter two or more times larger than the nucleus).
Mitosis–karyorrhexis index (MKI): One of three MKI classes is assigned to the given neuroblastoma tumor.
Those classes are low MKI (<2% or <100 of 5000 mi- totic and karyorrhectic cells), intermediate MKI (2–4% or 100–200 of 5000 mitotic and karyorrhectic cells), and high MKI (>4% or >200 of 5000 mitotic and karyorrhectic cells). The MKI is defined by counting the number of tumor cells in mitosis and in the process of karyorrhexis (Fig. 8.1d), and should reflect an average for all tumor sections available.
Karyorrhectic cells show condensed and fragmented
Figure 8.1 a–f
Histology of peripheral neuroblastic tumors.aNeuroblastoma (Schwannian stroma-poor), undifferentiated subtype.
bNeuroblastoma (Schwannian stroma-poor), poorly differentiated subtype.cNeuroblastoma (Schwannian stroma- poor), differentiating subtype.d Neuroblastoma (Schwannian stroma-poor) with a high mitosis–karyorrhexis index.
eGanglioneuroblastoma, intermixed (Schwannian stroma-rich).fGanglioneuroma (Schwannian stroma-dominant), maturing subtype.
nuclear material, usually accompanied by condensed eosinophilic cytoplasm. Simple hyperchromatic nu- clei without chromatin fragmentation are not includ- ed in MKI counting.
8.3.2 Ganglioneuroblastoma,
Intermixed (Schwannian Stroma-rich)
The international classification has stipulated that tumors having prominent Schwannian stromal de- velopment occupying more than 50% of the tumor tissue are upgraded to this category. Tumor histology is consistent with a transition to the full differentia- tion/maturation of ganglioneuroma (see below), but the process is not complete, as evidenced by scattered
“residual” microscopic foci where neuroblastic cells in various stages of differentiation as well as varying numbers of maturing ganglion cells are found in the background of neuropil (Fig. 8.1e).
8.3.3 Ganglioneuroma
(Schwannian Stroma-dominant)
Tumors are predominantly composed of Schwannian stroma with individually distributed maturing/ma- ture ganglion cells. Two subtypes, ganglioneuroma, maturing, and mature, are included in this category.
Ganglioneuroma, Maturing Subtype (Fig. 8.1 f): Tu- mor of this subtype was previously named “gan- glioneuroblastoma, well differentiated” in the origi- nal Shimada classification. Some of the neuroblastic components appear to be on their way to fully mature ganglion cells and have appearances of differentiat- ing neuroblasts and/or maturing ganglion cells.
Ganglioneuroma, Mature Subtype (Fig. 8.1 g): Tumor in this subtype is composed of fully developed Schwannian stroma and mature ganglion cells. Those mature ganglion cells are surrounded by satellite cells.
Mature non-myelinating Schwann cells, the dominat- ing component of tumor, characteristically form mul- tiple fascicles covered with perineurial cells.
Tumors categorized as either ganglioneuroblas- toma, intermixed, or ganglioneuroma have an elastic consistency, and their cut surfaces are always tan-yel- low and homogenous with or without fibrous bands.
8.3.4 Ganglioneuroblastoma,
Nodular (Composite, Schwannian Stroma-rich/
Stroma-dominant and Stroma-poor)
Tumor in this category is characterized by the pres- ence of one or more macroscopic, usually hemor- rhagic neuroblastomatous nodule(s) (stroma-poor
Figure 8.1 g–h
g Ganglioneuroma (Schwannian stroma-dominant), mature subtype.hGanglioneuroblastoma, nodular (composite, Schwannian stroma-rich/stroma-dominant and stroma-poor)
component) coexisting with ganglioneuroblasto- ma, intermixed (stroma-rich component) or with gang-lioneuroma (stroma-dominant component;
Fig. 8.1h). On microscopic examination, there is typi- cally abrupt demarcation (pushing border or even fibrous pseudo-capsular formation) between the neu- roblastomatous nodule(s) and the stroma-rich or stro- ma-dominant tumor tissue. Some nodules, however, are not clearly demarcated but rather have a zone of neuroblastic infiltration into the adjacent Schwannian stromal tissue. In rare cases the neuroblastomatous nodule becomes so large, dominating the tumor tis- sue, that one can recognize stroma-rich/stroma-dom- inant area only by light microscopic examination.
Nodular formation is usually considered to be a feature of the primary tumor, but it may be over- looked on gross examination; thus, those cases with ganglioneuroblastoma, intermixed or ganglioneuro- ma at the primary site and neuroblastoma at the metastatic site should be classified into this category.
8.4 Prognostic Classification
The International Neuroblastoma Pathology Classifi- cation (the Shimada system) distinguishes favorable and unfavorable histology groups (Shimada et al.
1999b). Tumors in the favorable histology group fall within a conceptual framework of age-linked matu- rational sequence from poorly differentiated subtype (<1.5 years of age at diagnosis) to differentiating sub- type (<5 years of age) of neuroblastoma (Schwann- ian stroma-poor) to ganglioneuroblastoma, inter- mixed (Schwannian stroma-rich) to ganglioneuroma (Schwannian stroma-dominant). The neuroblastoma tumors in this group should have a low (for those pa- tients <5 years of age) or an intermediate (for those
<1.5 years of age) MKI. By contrast, tumors in the un- favorable histology group have immature histologies for patient’s age and include undifferentiated sub- type (at any age), poorly differentiated subtype (≥1.5 years of age), and all subtypes (≥5 years of age) of the neuroblastoma. Among the neuroblastoma tumors, those with a high MKI (at any age) or an in- termediate MKI ( ≥1.5 years of age) also qualify as unfavorable histology. Ganglioneuroblastoma, inter-
mixed and ganglioneuroma are classified into a fa- vorable histology group regardless of the patients’
age, although these tumors are usually diagnosed in older children. Ganglioneuroblastoma, nodular (composite, Schwannian stroma-rich/stroma-domi- nant and stroma-poor) can be divided into two sub- sets, favorable and unfavorable, by applying the same criteria of age-linked histopathological evaluation to the nodular (neuroblastomatous) component (Peuch- maur et al. 2003; Umehara et al. 2000).
While arriving at the proposed Classification, the International Neuroblastoma Pathology Committee tested other morphological indicators (calcification, mitotic rate) and classifications (original risk group- ing by combination of mitotic rate, calcification, and age (Joshi et al. 1992); modified risk grouping by com- bination of MKI, calcification, and age (Joshi et al.
1996), and analyzed their prognostic effects (Shimada et al. 1999b). Although these indicators and classifica- tions all had prognostic effects by univariate analysis, calcification and mitotic rate did not add any signifi- cant prognostic information to the International Neu- roblastoma Pathology Classification in multivariate analysis.Furthermore,the Classification could provide significantly better prognostic information than those risk groupings. The Committee also examined the age factor of the Shimada system, and confirmed that the two cutoff points, i.e., 1.5 and 5 years of age at diagno- sis, used in the Classification distinguished prognostic groups most significantly (Shimada et al. 1999b).
8.5 Biological Relevance
In this section, the biological relevance of Interna- tional Neuroblastoma Pathology Classification (the Shimada system) is summarized.
8.5.1 Schwannian Development in Neuroblastic Tumors
Peripheral neuroblastic tumors consist of two main
cell populations: neuroblastic/ganglionic cells and
Schwann cells. As described above, the International
Neuroblastoma Pathology Classification uses mor-
phological features of both neuroblastic differentia-
tion and Schwannian stromal development for defin- ing maturational sequence of the pNTs. Based on em- bryological interactions between normal neuroblasts and Schwann cells (Reynolds and Woolf 1993), some postulate that neoplastic neuroblasts produce Schwann cell mitogens important for their prolifera- tion and development (Ambros 2001). Schwann cells, in return, can secrete anti-proliferative and differen- tiation-inducing factors crucial to neuronal differen- tiation. This mutual interaction between neuroblastic cells and Schwannian stromal cells may explain the maturational processes of biologically favorable pNTs. In biologically unfavorable pNTs, there is gen- erally less Schwannian component and limited tumor maturation. The origin of the tumor Schwann cells remains controversial. One study indicates that both cell types, i.e., neuroblastic/ganglionic cells and Schwann cells, arise from the same neoplastic neu- roblastic clone or precursor cell (Mora et al. 2001);
however, other studies present evidence to support that the Schwann cells in pNTs are reactive in nature and probably recruited from surrounding non-neo- plastic tissue by tumor neuroblastic cells (Ambros et al. 1996).
8.5.2 Correlation of Histopathology
with MYCN Amplification and trkA Expression There is a reproducible correlation between the molecular event of MYCN amplification and the mor- phological manifestations in pNTs (Shimada et al.
1995; Goto et al. 2001). Those tumors with amplified MYCN typically are of the undifferentiated or poorly differentiated subtype of neuroblastoma (Schwann- ian stroma-poor) with markedly increased mitotic (proliferating) and karyorrhectic (apoptotic) activi- ties (Shimada et al. 1995; Goto et al. 2001), an unfa- vorable histology group according to the Internation- al Neuroblastoma Pathology Classification. The pres- ence of prominent nucleoli in neuroblastic cells of undifferentiated or poorly differentiated neuroblas- toma, often associated with unfavorable prognosis (Ambros et al. 2002), can be an additional hallmark of MYCN amplification (own unpublished observa- tions).
The balance appears to favor cellular proliferation (mitosis) more than cellular death (karyorrhexis) in a MYCN amplified tumor, which is well known to have a highly aggressive and rapidly progressive clin- ical behavior. In light microscopic sections from MYCN amplified tumors, however, the number of karyorrhectic cells always exceeds that of mitotic cells. This may be explained by the fact that the histo- logically visible stage of mitosis is much shorter than that of karyorrhexis (Bursch at al. 1991). Our prelim- inary data show that neuroblastoma tumors with fa- vorable histology express significantly higher levels of trkA than those with unfavorable histology (Shi- mada et al. 2004). Favorable histology neuroblastoma tumors include both poorly differentiated and differ- entiating subtypes: although there is no difference in the level of trkA expression between these two histo- logical subtypes, tumors of differentiating subtype are diagnosed in significantly older children (usually between 1 and 5 years of age) than those of poorly differentiated subtype (newborn to 1.5 years of age).
This may suggest an in vivo latent period required for morphological evidence of neuroblastic differentia- tion among the neuroblastoma tumors in the favor- able-histology group, and supports the concept of an age-linked Pathology Classification.
8.5.3 Composite Tumor
The term “composite” implies that the tumor is com-
posed of histologically and, probably biologically,
different clonal populations (Schmidt et al. 1993), a
description possibly applicable to almost 10% of
pNTs. In the International Neuroblastoma Pathology
Classification, this composite form is designated as
ganglioneuroblastoma, nodular (composite, Schwann-
ian stroma-rich/stroma-dominant and stroma-
poor). In this model the neuroblastic nodule(s) rep-
resents the evolution of one or multiple clones, either
because of newly acquired genetic aberrations in late
stage of tumor progression or the persistence of ge-
netically and biologically distinct variants evolving
early in tumor formation. Clinically, two-thirds of
these composite tumors are aggressive (Peuchmaur
et al. 2003; Umehara et al. 2000).
8.6 Conclusion
Neuroblastic tumors are known to be heterogeneous and their clinical behaviors are driven by complex molecular/genetic properties. The International Neu- roblastoma Pathology Classification exploits a sys- tem of age-linked evaluation of morphological indi- cators, to distinguish among tumors with near-iden- tical histological features but vastly different clinical behaviors. This classification offers a unique forum for finding the morphological link between clinical behavior and tumor genetics of the enigmatic cancer of childhood.
References
Ambros IM, Zellner A, Roald B, Amann G, Ladenstein R, Printz D et al. (1996) Role of ploidy, chromosome 1p, and Schwann cells in the maturation of neuroblastoma. N Engl J Med 334:1505–1511
Ambros IM, Attarbaschi A, Rumpler S, Luegmayr A, Turkof E, Gadner H, Ambros PF (2001) Neuroblastoma cells provoke Schwann cell proliferation in vitro. Med Pediatr Oncol 36:163–168
Ambros IM, Hata J-I, Joshi VV, Roald B, Dehner LP, Tuchler H, Potschger U, Shimada H (2002) Morphologic features of neuroblastoma (Schwannian stroma poor tumors) in clini- cally favorable and unfavorable groups. Cancer 94:1574–
1583
Beckwith JB, Martin RF (1968) Observations on the histo- pathology of neuroblastomas. J Pediatr Surg 3:106–110 Brodeur GM, Maris JM (2002) Neuroblastoma. In: Pizzo PA,
Poplack DG (eds) Principles and practice of pediatric on- cology. Lippincott Williams and Wilkins, Philadelphia, pp 895–938
Bursch W, Liehr JG, Sirbasku DA, Putz B, Taper H, Schulte-Her- mann R (1991) Control of cell death (apoptosis) by diethyl- stilbestrol in an estrogen-dependent kidney tumor. Car- cinogenesis 12:855–860
Dehner LP (1988) Classic neuroblastoma: histopathologic grading as a prognostic indicator: the Shimada system and its progenitors. Am J Pediatr Hematol Oncol 10:143–154 Goto S, Umehara S, Gebing RB, Stram DO, Brodeur GM, Seeger
RC et al. (2001) Histopathology (International Neuroblas- toma Pathology Classification) and MYCN status in pa- tients with peripheral neuroblastic tumors. A report from the Children’s Cancer Group. Cancer 92:2699–2708 Gross RE, Farber S, Martin LW (1959) Neuroblastoma sympa-
theticum, a study and report of 217 cases. Pediatrics 23:1179–
1191
Hughes M, Marsden HB, Palmer MK (1974) Histologic patterns of neuroblastoma related to prognosis and clinical staging.
Cancer 34:1706–1711
Joshi VV, Cantor AB, Altshuler A, Larkin EW, Neill JSA, Shuster JJ et al. (1992) Age-linked prognostic categorization based on a new histologic grading system of neuroblastomas: a clinical pathologic study of 211 cases from the Pediatric Oncology Group. Cancer 69:2197–2211
Joshi VV, Rao PV, Canter AB, Altshuler G, Shuster JJ, Castleber- ry RP (1996) Modified histologic grading of neuroblastoma by replacement of mitotic rate with mitosis karyprrhexis index: a clinicopathologic study of 223 cases from the Pedi- atric Oncology Group. Cancer 77:1582–1588
Landau M (1911) Verhandlungen des Internationalen Patholo- gischen Kongresses, Turin
Mäkinen J (1972) Microscopic patterns as guide to prognosis of neuroblastoma in childhood. Cancer 29:1637–1646 Mora J, Cheung NK, Juan G, Illei P, Cheung I, Akram M, Chi S,
Ladanyi M, Cordon-Cardo C, Gerald WL (2001) Neuroblastic and Schwannian stromal cells of neuroblastoma are derived from a tumoral progenitor cell. Cancer Res 61:6892–6898 Peuchmaur M, d’Amore ESG, Joshi VV, Hata J, Roald B, Dehner
LP et al. (2003) Revision of the International Neuroblas- toma Pathology Classification: Confirmation of favorable and unfavorable prognostic subsets in ganglioneuroblas- toma, nodular. Cancer 98:2274–2281
Reynolds ML, Woolf CJ (1993) Reciprocal Schwann cell-axon interactions. Curr Opin Neurobiol 3:683–693
Ross JA, Severson RK, Pollock BH, Robison LL (1996) Child- hood cancer in the United States.A geographical analysis of cases from the Pediatric Cooperative Clinical Trial groups.
Cancer 77:201–207
Schmidt ML, Salwen HR, Chagnovich D, Bauer KD, Crawford SE, Cohn SL (1993) Evidence for molecular heterogeneity in human ganglioneuroblastoma. Pediatr Pathol 13:787–796 Shimada H, Chatten J, Newton WA, Sachs N, Hamoudi AB, Chi-
ba T et al. (1984) Histopathologic prognostic factors in neu- roblastic tumors; definition of subtypes of ganglioneuro- blastoma and an age-linked classification of neuroblas- tomas. J Natl Cancer Inst 73:405–416
Shimada H, Aoyama C, Chiba T, Newton WA Jr (1985) Prognos- tic subgroups for undifferentiated neuroblastoma: im- munohistochemical study with anti-S-100 protein anti- body. Hum Pathol 16:471–476
Shimada H, Ambros IM, Dehner LP, Hata J, Joshi VV, Roald B (1999a) Terminology and morphologic criteria of neuro- blastic tumors: recommendation by the International Neu- roblastoma Pathology Committee. Cancer 86:349–363 Shimada H, Ambros IM, Dehner LP, Hata J, Joshi VV, Roald B et
al. (1999b) The International Neuroblastoma Pathology Classification (Shimada System). Cancer 86:364–372 Shimada H, Stram DO, Chatten J, Joshi VV, Hachitanda Y,
Brodeur GM et al. (1995) Identification of subsets of neu- roblastomas by combined histopathologic and N-myc analysis. J Natl Cancer Inst 87:1470–1476
Shimada H, Umehara S, Monobe Y, Hachitanda Y, Nakagawa A, Goto S et al. (2001) International Neuroblastoma Pathology Classification for prognostic evaluation of patients with pe- ripheral neuroblastic tumors: a report from the Children’s Cancer Group. Cancer 92:2451–2461
Shimada H, Nakagawa A, Peters J, Wang H, Wakamatsu PK, Lukens JN et al. (2004) TrkA expression in peripheral neu- roblastic tumors: prognostic significance and biological relevance. Cancer 101:1873–1881
Umehara S, Nakagawa A, Matthay KK, Lukens JN, Seeger RC, Stram DO et al. (2000) Histopathology defines prognostic subsets of ganglioneuroblastoma, nodular: a report from the Children’s Cancer Group. Cancer 89:1150–1161
Wahl HR (1914) Neuroblastomata; with a study of a case illus- trating the three types that arise from the sympathetic sys- tem. J Med Res 30:205–260
