Fax: (604) 875-2815
A population-based study of pediatric anaplastic large cell lymphoma
Version of Record online: 15 MAR 2002
Copyright © 2002 American Cancer Society
Volume 94, Issue 6, pages 1830–1835, 15 March 2002
How to Cite
Alessandri, A. J., Pritchard, S. L., Schultz, K. R. and Massing, B. G. (2002), A population-based study of pediatric anaplastic large cell lymphoma. Cancer, 94: 1830–1835. doi: 10.1002/cncr.10396
- Issue online: 15 MAR 2002
- Version of Record online: 15 MAR 2002
- Manuscript Accepted: 1 NOV 2001
- Manuscript Revised: 7 APR 2001
- Manuscript Received: 29 NOV 2000
- non-Hodgkin lymphoma;
- anaplastic large cell lymphoma;
Anaplastic large cell lymphoma (ALCL) is a relatively rare form of non-Hodgkin lymphoma (NHL) in children constituting 10–15% of this entity. To the authors' knowledge, there are no population-based pediatric studies in the literature, and incidence estimates have not been attempted.
A population-based study of all children in British Columbia, Canada, younger than 15 years of age presenting between January 1988 and December 1999 with an immunohistologically confirmed diagnosis of CD30+ ALCL was completed. Demographic, ethnic, clinical, treatment, and outcome details were collected on all patients. Population figures were obtained from census data through the BC STATS.
Ten patients were identified and confirmed to have ALCL on the basis of morphology and immunohistochemical stains. This equates to an annual incidence of 1.2 per million children younger than 15 years of age. The mean age at diagnosis was 8.23 years with a range of 1.4–13.0 years. There was an overrepresentation of East and Southeast Asian patients (40%) in the ALCL group compared with other subtypes of NHL and the pediatric population of British Columbia. Twenty percent of the patients had evidence of central nervous system (CNS) disease at diagnosis.
The annual incidence of ALCL in children younger than 15 years of age in British Columbia is 1.2 per million. The overrepresentation of East and Southeast Asian ethnic origin remains unexplained. The rate of CNS involvement is much higher than that previously reported in large non-population-based series (0–3%). Although the low numbers do not allow firm conclusions, it would seem prudent to continue to investigate all newly diagnosed patients for CNS involvement in British Columbia. Cancer 2002;94:1830–5. © 2002 American Cancer Society.
CD30+ anaplastic large cell lymphoma (ALCL) was first described in 19851 and is now widely recognized as a distinct clincopathologic entity. It constitutes approximately 10–15% of non-Hodgkin lymphoma (NHL) in the pediatric population.2, 3 Although the neoplastic cells exhibit a wide range of morphologic features in ALCL, their preferential involvement of the paracortical and intrasinusoidal areas of the lymph node is characteristic.1 With respect to lymphocyte lineage markers, most cases are associated with a T- or null cell immunophenotype. In fact, while the existence of sporadic cases of B-cell phenotype continues to be debated, the recently proposed Revised European–American Lymphoma (REAL) classification limits the term ALCL only to cases with T and null immunophenotype.4 By definition, ALCL cells are positive for the activation antigen Ki-1 or CD30. In addition, most coexpress the epithelial membrane antigen (EMA).5 Anaplastic large cell lymphoma was found to be associated with the t(2;5)(p23:q35) translocation several years ago.6–8 This translocation results in the fusion of the nucleophosmine gene (NPM) on chromosome 5 and the anaplastic lymphoma kinase gene (ALK) on chromosome 2.9 The NPM/ALK fusion gene encodes for an 80-kilodalton NPM–ALK chimeric protein that is thought to play a key role in lymphomagenesis by aberrant phosphorylation of intracellular substrates.9–12 The fusion product can be detected by immunohistochemistry in tumor biopsies using either polyclonal13 or monoclonal14 antibodies.
To our knowledge, the incidence of ALCL in the pediatric population has not been accurately determined in a population-based study. Although the SEER Pediatric Monograph recognizes ALCL as one of the major histopathologic categories of NHL in children, it is not analyzed as a separate category in the incidence data for NHL.15 Similarly, a review of the literature (Medline, 1960 to June 2000) identified several recent articles reporting incidence rates for NHL in children, but histologic subtype incidence rates were not included.16–20 Although a pediatric population-based study from Western Sweden provides a breakdown by histologic subtype in terms of raw numbers, it does not provide sufficient details to calculate an incidence for ALCL.21 There are little data on the ethnic origin of children with ALCL in the literature, because most studies have been undertaken in populations with homogeneous racial backgrounds.
Typically, ALCL is a high-grade malignancy with a predilection for skin, other extranodal sites, and a higher frequency of B symptoms and hepatosplenomegaly when compared with other types of pediatric NHL.22–24 Involvement of the central nervous system at diagnosis, however, is reportedly rare (0–3%).24–28 An optimal therapy strategy has yet to be established for this entity and event free survivals range from 60% to 80%.25–27
The objectives of this study were to calculate an incidence for ALCL in children in the British Columbian population and examine the ethnic and clinical characteristics of patients in this setting.
MATERIALS AND METHODS
Patients and Incidence Calculations
Ten children younger than 15 years of age presented with a histologically confirmed diagnosis of CD30+ ALCL to the British Columbia's Children's Hospital (BCCH) between January 1988 and December 1999. The BCCH is the only tertiary referral center for pediatrics in British Columbia with a catchment population of approximately 3,700,000. All patients were living in the catchment area for at least 6 months before diagnosis. The average annual number of patients in the younger than 15 years age group in the referral area was 709,173. Provincial population figures and ethnic origin data were obtained from Canadian census data through BC STATS, British Columbia Ministry of Finance and Corporate Relations, and the Health Data Warehouse, British Columbia Ministry of Health. East and Southeast Asian ethnicity included the following ethnic groups: Chinese, Japanese, Korean, Filipino, Indonesian, Malayan, and Indo-Chinese (Burmese, Cambodian, Laotian, Thai, and Vietnamese). The term Asian is used to denote patients of East and Southeast Asian origin on the ethnicity figures. Incidence figures for the other histologic subtypes of NHL were calculated using the same population figures and patient numbers from the BCCH Oncology Division database.
Demographic, clinical, treatment, and outcome details were collected on all children with ALCL in a retrospective manner from clinical charts. Ethnic origin data were confirmed by review of the BCCH Oncology Division database and personal inquiry with treating physicians. Information on the ethnic origin of patients with the other subtypes of NHL was obtained from the same sources.
Diagnosis and Histopathologic Examination
The diagnosis of ALCL was based on histopathology and immunohistochemistry according to the working formulation criteria for NHL.29 All diagnostic material was reviewed retrospectively by the same pathologist (B.G.M.). Immunohistochemical stains were performed using fixed, paraffin embedded sections and an avidin-biotin peroxidase complex technique using the following antibodies: LCA (leucocyte common antigen, CD45), UCHL (CD45 Ro), L26 (CD20), CD30, and EMA. Archived material (paraffin section slides) was stained for the presence of the NPM–ALK fusion product using the ALK-1 monoclonal antibody by Dr. Richard Cartun at the Hartford General Hospital, Hartford, Connecticut in 9 of the 10 cases as described previously.14 The t(2;5) rearrangement was not confirmed by reverse transcription–polymerase chain reaction, and no specific ALK breakpoints were determined. Cytogenetic analysis of direct and short-term culture chromosome preparations was undertaken when suitable biopsy material was available.
Ten patients were identified and confirmed to have ALCL on pathologic review. The mean age at diagnosis was 8.2 years with a range of 1.4–13 years. The male:female ratio was 1:1.5. Forty percent of the patients were of East or Southeast Asian ethnic origin. Further clinical characteristics are presented in Table 1. The morphologic, immunohistochemica1, and cytogenetic features of the 10 cases are presented in Table 2.
|Stage (Ann Arbor)|
|I and II||1||10|
|III and IV||9||90|
|Extranodal disease (any)||9||90|
|Central nervous system||2||20|
|1||Asian||I||Monomorphous large cells, inflammatory infiltrate||ND||Neg||T||Pos||Neg||Failed|
|2||White||IV||Pleomorphic, small cell component||Pos||ND||T||Pos||Pos||t(2;5)|
|5||White||IV||Predominant small cell variant||Pos||Pos||T||Pos||Pos||t(2;5)|
|6||Asian||IV||Pleomorphic large cell, sinusoidal pattern||ND||Pos||T||Pos||Pos||NA|
|7||Asian||IV||Pleomorphic large cell||Pos||Pos||T||Pos||Pos||Failed|
|8||White||III||Monomorphous large cells, inflammatory infiltrate||ND||Pos||T||Pos||Neg||Failed|
|10||White||IV||Pleomorphic large cell||Pos||Pos||T||Pos||Pos||t(2;5)|
The average annual incidences of the common forms of NHL per million children younger than 15 years of age in British Columbia are as follows: Burkitt, 4.0; lymphoblastic, 1.6; anaplastic large cell, 1.2; others, 0.8 (includes other large cell and unclassifiable subtypes). This incidence data are presented graphically in Figure 1.
The ethnic origins of children in British Columbia with the three common types of NHL are depicted in Figure 2. The percentage of children of East and Southeast Asian ethnic origin for ALCL and Burkitt and lymphoblastic lymphomas were 40%, 6%, and 0% respectively. There were a few patients of mixed or Native North American origin in the Burkitt and lymphoblastic subtypes (9%and 7%, respectively). Overall, patients of East and Southeast Asian origin accounted for 12.3% of NHL patients. According to the ethnic origin data from the 1996 Canadian census, 12.1% of British Columbian children younger than 15 years of age were of East or Southeast Asian origin. Accurate ethnic origin data for children are not available for the 1991 and 1986 census. The percentage of the total BC population that were of East and South East Asian origin increased from 6% to 14% from 1986 to 1996. Because the current study spanned this period, an accurate estimation of the incidence of ALCL in children of East and Southeast Asian origin is not possible.
The mean length of follow-up for the 10 study patients was 4.4 years (range, 0.33–10.25 years). Although treatment was not uniform over the study period, most patients (80%) received treatment according to the appropriate Children's Cancer Group or Berlin–Frankfurt–Munster protocols. The progression free survival at 2 and 5 years was 70%, and the overall survival was 90% at 2 years and 79% at 5 years. Both patients with central nervous system (CNS) involvement at diagnosis experienced recurrence in the CNS. One of these patients died from extensive CNS disease within 4 months of diagnosis. The other received second-line chemotherapy and CNS irradiation and is alive and disease free greater than 7 years from diagnosis. Another patient with extensive disease at diagnosis experienced recurrence 2 months after completion of chemotherapy in the skin, lymph nodes, liver, spleen, and kidneys. She was treated with megadose chemotherapy and peripheral stem cell rescue after achieving a clinical remission. Unfortunately, she suffered an additional recurrence 3 months after this therapy and died of widespread disease 2 years from diagnosis.
Subset analysis of the data to determine factors that correlate with prognosis was not undertaken because of the low study numbers.
The average annual incidence of ALCL in children younger than 15 years of age in British Columbia is 1.2 per million. In this population, ALCL accounts for 15.4% of all NHL cases. This latter figure is in keeping with previous, non-population-based estimates.2, 3 The calculated incidence figures for Burkitt and lymphoblastic lymphoma are in keeping with the recently published SEER data figures (3.0 and 1.6, respectively),15 suggesting our ALCL calculations are accurate. As discussed previously, to our knowledge, this is the first study to report an incidence figure for pediatric ALCL.
With the exception of the ethnic origin breakdown and the presence of CNS disease at diagnosis, the clinicopathologic characteristics of the patients in this study are in keeping with previous reports.25, 26, 30 Studies investigating the presence of the t(2;5) translocation in adult ALCL using monoclonal or polyclonal antibodies to the fusion protein (NPM–ALK) report an incidence of 20–60% positivity.31–33 The incidence of NPM–ALK positivity in pediatric series is higher at 84–95%.25, 32–34 The lower incidence of NPM–ALK positivity in the current study (70%) may be because it is population-based, whereas the others consisted of cases referred to the tissue banks of large cooperative groups.
The ethnic origin data reveal an apparent overrepresentation of East and Southeast Asian patients in the ALCL group. This overrepresentation is in relation to the number of East and Southeast Asian children in the BC population as well as the number of East and Southeast Asian children in the other NHL subgroups. Many of the previous studies of pediatric ALCL have been undertaken in European centers with more homogeneous populations in terms of ethnicity.25–27 These, as well as North American studies, have not reported an increased number of patients of East and Southeast Asian origin.24, 28 There are few studies addressing the incidence of ALCL in children of Asian ethnic origin. Two adult studies from Southeast Asia report that ALCL accounts for 2.3–5% of NHL cases.35, 36 The incidence of NHL in Southeast Asian pediatric populations appears to be similar to that in Europe and North America, although there is some uncertainty over the incidence in Japan.37, 38 There are no figures that estimate the incidence of ALCL in Asian children; however, the percentage of patients with large cell lymphomas are similar to white populations.38 The incidence of NHL in Asian migrants and their descendants has been studied in selected areas of the United States.39 This study excludes children younger than 15 years of age and uses the working formulation classification of NHL that does not incorporate ALCL as a category. Nevertheless, the incidence of NHL in each of the Asian groups studied (Chinese, Japanese, and Filipino) was 35–85% that of U.S.-born whites. There was no consistent trend of increasing incidence with increasing generation of residence in any group. Three of the four Asian patients in our population were first generation Canadian and the fourth was born in Japan. To date, to our knowledge, the literature provides no explanation for the overrepresentation of Asian patients in our population.
Central nervous system involvement at diagnosis is reportedly rare in pediatric ALCL,2 with only isolated cases being reported in the literature.40, 41 Large non-population-based series from a variety of centers report an incidence of 0–3%.24–28, 30, 42 In contrast, 2 of our patients (20%) presented with anaplastic large cells in their cerebrospinal fluid (CSF). Both were white and experienced recurrences (one fatal) in the CNS. Recent studies have concluded that minimal, if any, CNS prophylaxis is required in pediatric patients with ALCL who do not have CNS involvement at diagnosis.25, 26, 28 In contrast, the few patients who do present with CNS involvement at diagnosis appear to have an aggressive clinical course and often die of the disease despite combination and intrathecal chemotherapy and CNS irradiation.27, 40, 41 Although numbers are low, our data suggest that the staging of pediatric ALCL should continue to include a careful examination of the CSF and that CNS disease should be treated aggressively from the outset.
To our knowledge, this is the first population-based series to report an incidence of ALCL in children. The incidence estimate of 1.2 per million children younger than 15 years of age is useful but may have limited accuracy due to the small sample size. Similarly, the small sample size identifies but does not firmly establish an overrepresentation of patients of East and Southeast Asian origin. Our data suggest CNS disease should be excluded at diagnosis in all patients and treated aggressively when present.
The authors thank Ruth McDougall, Manager, Population Statistics, BC STATS, for her assistance in accessing the Canadian census data for British Columbia, Dr. Richard Cartun Department of Pathology, Hartford General Hospital and Richard Mah, Department of Pathology, Children's and Women's Hospital, Vancouver, for the ALK-1 antibody staining of specimens.
- 2Pediatric lymphomas: recent advances and commentary on Ki-1 positive anaplastic large-cell lymphomas of childhood. Ann Oncol. 1994; 5: 51–53..
- 11Role of nucleophosmin (NPM) portion of the non-Hodgkin's lymphoma associated NPM-anaplastic kinase fusion protein in oncogenesis. Mol Cell Biol. 1997; 176: 2312–2325., , , .
- 15RiesLAG, SmithMA, GurneyJG, et al. editors. Cancer incidence and survival among children and adolescents: United States SEER Program 1975–1995, National Cancer Institute, SEER Program. NIH Pub. No. 99-4649. Bethesda, MD: National Institutes of Health; 1999.
- 32High incidence of the t(2;5) (p25;q35) translocation in anaplastic large cell lymphoma and its lack of detection in Hodgkin's disease. Comparison of cytogenetic analysis, reverse transcriptase-polymerase chain reaction, and the P-80 immunostaining. Blood. 1996;87: 284–291., , ,et al.