Indian Journal of Pathology and Microbiology

: 2011  |  Volume : 54  |  Issue : 3  |  Page : 514--519

Pediatric solid malignant neoplasms: A comparative analysis

Bryan E Harmon1, Kenneth Friedman1, Barbara Nemesure2, Meenakshi Singh1,  
1 Department of Pathology, State University of New York at Stony Brook, Stony Brook, New York, USA
2 Department of Preventive Medicine, State University of New York at Stony Brook, Stony Brook, New York, USA

Correspondence Address:
Meenakshi Singh
Department of Pathology, Stony Brook University Medical Center, UHL2 762, Stony Brook, NY 11794-7025


Background: Pediatric solid malignant neoplasms (PSMNs) are a significant cause of death among children. Our aim was to evaluate the pattern and frequency of PSMNs at our hospital in the United States and compare the results to data from other regions of the world. Materials and Methods: This is a retrospective review of 127 PSMNs in the Pathology database at Stony Brook University Medical Center (SBUMC) from 2000 to 2008. We compared our cases to a cohort of 101 cases from an academic hospital in India (1975-1982) (Christian Medical College and Hospital) and to reports from other parts of the world. Results: We report a male to female ratio of 1.16 : 1 and a mean age of 4.8 years for cases at SBUMC. Lymphomas and central nervous system (CNS) neoplasms were more common in the 5-12-year-old group while other major diagnostic groups were more common in the 0-4-year-old group. The top five most frequent tumor categories included CNS, sympathetic nervous system (SNS), soft tissue, lymphoid and renal tumors. Lymphomas were more common than soft tissue and SNS tumors in the United States«SQ» registries but all three occurred with equal frequency in our study. Tumors of the soft tissue and SNS were more frequent at SBUMC compared to registries around the world. At the academic hospital in India, the male to female ratio was 4 : 1 and the five most frequent tumor categories included lymphoid, SNS, CNS, renal and bone tumors. Lymphoid tumors made up a greater percentage and CNS tumors made up a lesser percentage of tumors at the hospital in India compared with SBUMC. The differences between CNS tumors, lymphomas and retinoblastomas between the two hospitals were statistically significant (P value <0.05 by Fisher«SQ»s Exact test). Conclusions: Geographic differences in the incidence and histologic types of PSMNs exist. Despite advancements in diagnosis and treatment, PSMNs continue to be tragically lethal.

How to cite this article:
Harmon BE, Friedman K, Nemesure B, Singh M. Pediatric solid malignant neoplasms: A comparative analysis.Indian J Pathol Microbiol 2011;54:514-519

How to cite this URL:
Harmon BE, Friedman K, Nemesure B, Singh M. Pediatric solid malignant neoplasms: A comparative analysis. Indian J Pathol Microbiol [serial online] 2011 [cited 2020 Sep 22 ];54:514-519
Available from:

Full Text


Pediatric solid malignant neoplasms (PSMNs) are a global problem. Cancer is a major killer of children in developed countries. [1] In developing countries, as significant progress is made in treating infectious diseases and nutritional deficiencies, cancer is emerging as a major childhood killer. [2]

The incidence, site of origin, and histologic subtypes of PSMNs vary with geographic location. For instance, comparing studies from Chennai, India (1990-2001) and Kyadondo County, Uganda (1993-1997), the incidence of lymphomas in the pediatric population was 19.8 cases per million children in Chennai but over three times higher in Uganda where the incidence was 65.3 per million children. [3],[4] In Chennai the incidence of central nervous system (CNS)/intracranial/intraspinal neoplasms was approximately 11 cases per million children. However, in the United States from 2000 to 2007, the incidence of CNS neoplasms was nearly three times more frequent at 29.7 cases per million children. [5] Even within the United States there appears to be a significant variation in pediatric cancer incidence. [6]

Understanding the geographic variations in pediatric cancer may contribute to advances in knowledge about etiologic factors. Our aim was to evaluate the pattern and frequency of PSMNs at our academic hospital in the United States and compare these to a cohort of 101 cases from an academic hospital in India. We also reviewed data from tumor registries in multiple regions from around the world.

 Materials and Methods

A retrospective review was conducted to reveal 127 patients with PSMNs in the Pathology database from 2000 to 2008 at Stony Brook University Medical Center (SBUMC). The pathology reports for all cases were reviewed. We evaluated the pattern and frequency of these tumors and the histologic type. The histologic slides of all lymphoma cases were reviewed with a hematopathologist to ensure that all diagnoses fit the current WHO lymphoma classification system as this was introduced during the course of our study period. We compared the cases from SBUMC to a group of 101 PSMNs from Christian Medical College and Hospital (CMCH) in Ludhiana, Punjab, India from 1975 to 1982 to compare regional as well as time-related differences. A comparison of PSMN incidence from reported series from the United States, [5] France, [7] Europe, [8] South India, [3] China, [9] Australia, [10] Uruguay, [11] and Uganda, [4] was performed.


The data from a study performed by the senior author at an academic hospital in India (CMCH) [Table 1] and [Figure 1]a (from 1975 to 1982) revealed that tumors were most frequent during the first three years of life with neuroblastoma being the commonest neoplasm in this age group. In the 10-12 year age group, Hodgkin disease and CNS neoplasms were the most common, while neuroblastomas on the other hand were virtually absent. The overall male to female ratio was 4 : 1. Lymphoid tumors were the most common followed in frequency by neuroblastomas and CNS tumors, and the remainder of the cases consisted of bone tumors, Wilms tumor, soft tissue sarcomas, and retinoblastoma.{Table 1}{Figure 1}

Lymphoid tumors comprised one-third of the total cases. Hodgkin disease was more common than non-Hodgkin lymphomas. The commonest form of Hodgkin disease was mixed cellularity type (per the classification in use at that time). Infants as young as 8 months of age were affected, but the majority of the cases were 7-9 years of age. Cervical lymph nodes were the most common site of involvement. Most cases were at an advanced stage at the time of diagnosis, often with histologic evidence of splenic involvement. Non-Hodgkin lymphomas were less frequent and were generally nodal in origin. The commonest type was poorly differentiated lymphocytic lymphoma (per the classification in use at that time).

Neuroblastoma and CNS tumors together accounted for one-third of the total cases. Neuroblastoma was the commonest tumor of early childhood and some of these tumors were present at birth. Fibrillary astrocytoma was the most frequent CNS neoplasm and histologically often showed microcyst formation.

Bone tumors, Wilms tumors, soft tissue sarcomas, and retinoblastomas combined accounted for one-third of the total cases. Osteogenic sarcoma and Ewing's sarcoma occurred with equal frequency. Wilms tumor was more common in the left kidney. Histologically, tubule formation was more common than pseudo-glomeruloid structures and it was present in all of the Wilms tumors. Embryonal rhabdomyosarcoma was the commonest soft tissue sarcoma in children, and it was prone to arise from mucosal surfaces. Histologically it could be mistaken for a benign polyp. The presence of a cellular cambial layer was necessary for a definite diagnosis (in the pre-immunohistochemistry era of the 1970s). The only intraocular tumor seen in children was retinoblastoma. Pseudorosettes were seen more frequently than true rosettes. Yolk sac tumor was the most frequent gonadal tumor in both male and female children and it was also the commonest type of malignant tissue observed in teratomas. Multiple blocks were submitted for the identification of this component and Schiller-Duval bodies were pathognomonic of this neoplasm. Granulosa cell tumor of the ovary occurred in children and many of these had the adult-type histology with classic grooved nuclei and Call-Exner bodies.

For the 127 PSMNs in the Pathology database at SBUMC [Table 1], [Table 2] and [Figure 1]b, we report a male to female ratio of 1.16 : 1 and a mean age of 4.8 years, with the 0-4-year-old group comprising 52.8% of the cases. The top five sites in our series are CNS tumors (35.4%), neuroblastomas and peripheral nervous system tumors (12.6%), soft tissue tumors (12.6%), lymphomas (12.6%), and renal tumors (9.4%). In children less than 5 years of age, neuroblastomas were the most common malignancy followed by Wilms tumor. All hepatoblastomas occurred in this age group as did the majority of neuroblastoma, Wilms tumor, soft tissue sarcoma, and germ cell malignancies. In children 5-12 years of age, CNS and lymphoid tumors were most common. Amongst the 5-9 year age group, CNS neoplasms were more common with pilocytic astrocytoma and medulloblastoma accounting for the majority of cases. However, amongst the 10-12 year age group, lymphoid neoplasms outnumbered CNS neoplasms. Non-Hodgkin lymphomas slightly outnumbered Hodgkin lymphomas.{Table 2}

As in the study from India, histologic analysis was the primary means for diagnosing PSMNs in the more recent SBUMC study. However, many advancements occurred in the decades between the two studies resulting in a greater array of diagnostic tools being available during the SBUMC study, for example: Imaging technologies were available which helped identify tumors, immunohistochemical stains were available to support histologic diagnoses (particularly useful for classifying lymphomas), and molecular tests were available to help determine prognosis (e.g. N-myc status in neuroblastoma).

While PSMNs were found more frequently in males than females in both studies, this inequality was more extreme at CMCH. The overall distribution of diagnoses is significantly different between SBUMC and CMCH (P <0 0.001). CNS tumors composed a larger percentage of the overall PSMNs at SBUMC as compared to CMCH (P < 0.001) [Table 1]. At CMCH there were no cases of hepatoblastoma. At SBUMC there were no cases of retinoblastoma whereas at CMCH retinoblastoma made up 4.9% of overall tumors, and this difference was statistically significant (P = 0.016). Soft tissue tumors were more frequent at SBUMC but the difference was not statistically significant. Lymphomas accounted for a significantly higher percentage of PSMNs at CMCH than at SBUMC (P < 0.001).

We analyzed the tumor registry data from regions around the world (including data from cities, countries, and continents) to try to identify important differences based on geography. The following regions were included in the comparison: USA, France, Europe, Chennai (India), Shanghai (China), Australia, Uganda, and Uruguay [Table 3] and [Figure 1]c-j.[3],[4],[5],[7],[8],[9],[10],[11]{Table 3}


Comparing the recent data from SBUMC with the study conducted at CMCH, we find that PSMNs were most common in the youngest age groups in both studies. The reasons behind the markedly higher male to female ratio observed at CMCH are unclear but may be related to access to medical care. Much of the difference in frequency of CNS tumors can likely be attributed to the sophisticated imaging technology available during the SBUMC data collection period, which was not available during the CMCH study period. At CMCH there were no cases of hepatoblastoma. Hepatoblastoma is a rare tumor and having no cases in the cohort may be secondary to small cohort size rather than a decreased risk for hepatoblastoma in the population seeking care at CMCH. It is unclear why lymphomas accounted for a larger percentage of PSMNs at CMCH as compared with SBUMC. Comparison between specific sub-types of lymphoma is limited due to significant changes that occurred in lymphoma classification in the decades between these two studies.

We compared our data at SBUMC with data from throughout the United States obtained from the SEER database from 2003 to 2007 [Figure 1]c.[5] The SBUMC data and the national data share the same top five sites of origin but with differences in tumor distribution. Studies report that cancer incidences are higher in the Northeast United States (SBUMC is within this northeast region) as a result of higher rates of CNS tumors and lymphomas, [12] though our population suggests that soft tissue tumors and neuroblastomas and PNS tumors may play a larger role.

Our analysis of tumor registry data from regions around the world revealed that the incidence of PSMNs varied from 69.75 cases per million children in Chennai, India, up to 166.0 cases per million children in Kyadondo County, Uganda. The high incidence observed in Uganda is due in large part to endemic Burkitt's lymphoma and Kaposi's sarcoma. CNS tumor incidence appears greater in developed regions compared with less developed regions. This is likely due at least in part to the greater availability of CT and MRI scanners in the developed areas which allow for easier detection of CNS malignancies. Retinoblastoma comprises a larger percentage of the PSMNs in Chennai than in any of the other regional databases studied. It is unclear whether children in Chennai are at greater risk of retinoblastoma or whether the proportion of this tumor is relatively high because of under- diagnosis of other malignancies (e.g. less access to CT/MRI scanners could lead to under-diagnosis); [13] however, the former is supported by an increased incidence for retinoblastoma compared with the other regions. While the number of neuroblastoma cases from Uganda was unavailable, neuroblastoma rates in East Africa in general are very low. [4] It is likely that at least part of this low incidence can be attributed to under-diagnosis in developing countries. Germ cell tumor incidence was markedly lower in the data from Uganda. These cancers are rare in Africa for reasons that are unclear. [4] The incidence of pediatric renal tumors is lowest in Shanghai. A markedly lower incidence of Wilms tumor amongst East Asians has been noted previously and it was concluded that the incidence varied by ethnic group rather than geography and that genetic predisposition is an important etiologic factor. [14] There is little variation in incidence rates for hepatic tumors amongst the regions. Uruguay had the highest rates of malignant bone tumors for reasons that are unclear.


In summary, there exist regional and geographic differences in the incidence and histologic types of pediatric cancers. Understanding these patterns and differences may contribute to advances in knowledge about etiologic factors. We hope this will raise awareness about the global nature of pediatric neoplasms and their histologic distribution. Despite major advances over the past decades in tumor classification and treatments, PSMNs are often tragically lethal even today.


Dr. M. Singh expresses deep appreciation for her MD thesis mentor, the Late Dr. S. D. Khanna, Professor, CMCH, India, for being a teacher who will be remembered for a life time. The authors thank Vencine Kelly of the SBUMC tumor registry for assistance with the data. The authors also thank Dr. Youjun Hu for his assistance in reviewing the slides of SBUMC lymphoma cases. This work is dedicated to all the children, past and present, who have suffered from malignant tumors and to their families and caregivers. It is also an appreciation of all those who provide Pediatric Oncology care and the funds to support Pediatric Cancer research. There is still a long way to go.


1Chirdan LB, Bode-Thomas F,Chirdan OO. Childhood cancers: Challenges and strategies for management in developing countries. Afr J Paediatr Surg 2009;6:126-30.
2Magrath I, Shad A, Epelman S, De Camargo D, Petrilli AS, El Mawla NG, et al. Pediatric oncology in countries with limited resources. In: Pizzo PA, Poplack DG, editors. Principles and Practice of pediatric oncology. Philadelphia: JB Lippincott Co.; 1997. p. 1395-420.
3Swaminathan R, Rama R, Shanta V. Childhood cancers in Chennai, India, 1990-2001: Incidence and survival. Int J Cancer 2008;122:2607-11.
4Cancer in Africa: Epidemiology and prevention. International Agency for Research on Cancer. IARC Sci Publ 2003;153:1-414.
5Altekruse SF, Kosary CL, Krapcho M, Neyman N, Aminou R, Waldron W, et al., editors. SEER Cancer Statistics Review, 1975-2007. Available from:, based on November 2009 SEER data submission, posted to the SEER web site, 2010. Bethesda, MD: National Cancer Institute; 2010.
6Oregon State Cancer Registry (OSCaR), Mississippi Cancer Registry, and the CDC's National Program of Cancer Registries Cancer Surveillance System (NPCR-CSS). November 2009/January 2010 data submission. Available from:
7Lacour B, Guyot-Goubin A, Guissou S, Bellec S, Désandes E, Clavel J. Incidence of childhood cancer in France: National Children Cancer Registries, 2000-2004. Eur J Cancer Prev 2010;19:173-81.
8Kaatsch P. Epidemiology of childhood cancer. Cancer Treat Rev 2010;36:277-85.
9Bao PP, Zheng Y, Wang CF, Gu K, Jin F, Lu W. Time trends and characteristics of childhood cancer among children age 0-14 in Shanghai. Pediatr Blood Cancer 2009;53:13-6.
10Baade PD, Youlden DR, Valery PC, Hassall T, Ward L, Green AC, et al. Trends in incidence of childhood cancer in Australia, 1983-2006. Br J Cancer 2010;102:620-6.
11Castillo L, Fluchel M, Dabezies A, Pieri D, Brockhorst N, Barr R. Childhood cancer in Uruguay: 1992-1994. Incidence and mortality. Med Pediatr Oncol 2001;37:400-4.
12Li J, Thompson TD, Miller JW, Pollack LA, Stewart SL. Cancer incidence among children and adolescents in the United States, 2001-2003. Pediatrics 2008;121:1470-7.
13Arora RS, Eden TO, Kapoor G. Epidemiology of childhood cancer in India. Indian J Cancer 2009;46:264-73.
14Stiller CA, Parkin DM. International variations in the incidence of childhood renal tumours. Br J Cancer 1990;62:1026-30.