Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Dec 16.
Published in final edited form as: Int J Cancer. 2020 Jul 13;147(12):3339–3348. doi: 10.1002/ijc.33153

Childhood cancer incidence among specific Asian and Pacific Islander populations in the U.S.

Kristin J Moore 1, Aubrey K Hubbard 2, Lindsay A Williams 2,3, Logan G Spector 2,3
PMCID: PMC7736474  NIHMSID: NIHMS1642357  PMID: 32535909

Abstract

Despite the vast genetic and environmental diversity in Asia, individuals of Asian and Pacific Islander (API) descent are often combined into a single group for epidemiologic analyses within the United States. We used the Surveillance, Epidemiology, and End Results (SEER) Detailed Asian/Pacific Islander Database to calculate incidence rates for discrete groups among children aged 0–19 years. Due to sample size constraints we pooled incidence among regional groups based on countries of origin: East Asians (Chinese, Japanese, Korean), Southeast (SE) Asians (Vietnamese, Laotian, Cambodian), Asian Indian/Pakistani, Oceanians (Guamanian, Samoan, Tongan), and Filipinos. Incidence rate ratios (IRR) and 95% confidence intervals (CI) were calculated comparing each API regional group to Non-Hispanic Whites (NHW) and East Asians. Lastly, we calculated the correlation between incidence of cancer in specific API ethnicities in SEER and in originating countries in the Cancer Incidence i n Five Continents. Incidence rates among API regional groups varied. Acute lymphoblastic leukemia (ALL) was lower in children of SE Asian descent (IRR 0.65, 95% CI 0.44, 0.96) compared to NHW. Acute myeloid leukemia (AML) was more common among children from Oceania compared to NHW (IRR 3.88, 95% CI 1.83, 8.22). East Asians had higher incidence rates than SE Asians but lower rates compared to children from Oceania. Correlation of some incidence rates between US-based API ethnicities and originating countries were similar. The variation observed in childhood cancer incidence patterns among API groups may indicate differences in underlying genetics and/or patterns of exposure.

Keywords: Disparities, Incidence, Childhood cancer, Pediatric cancer, Asian and Pacific Islander

INTRODUCTION

The incidence of childhood cancer has long been known to vary by demographic factors in the United States (US). Racial and ethnic differences in childhood cancer incidence are well established in the literature1 and are likely due to differences in genetic susceptibility and environmental exposures. Many childhood cancer studies in the US have focused on comparisons between white, black and Hispanic children while often combining Asians and Pacific Islanders into a single “API” category.2–4 However, this practice has the effect of overlooking the substantial diversity of national origins among Americans of API descent. Potential disparities in the incidence of childhood cancer have been obscured and potential clues into etiology have gone unexamined.

Over the past two decades, the Asian population has grown faster than any other group in the US.5 This rapid growth is driven by immigration as opposed to increases in US births.6 Though Census questionnaires have included various API groups since the 1950s, it was not until the 2000 Census that individuals were able to select their respective race/ethnicity from more detailed and specific ethnicity categories, as well as indicate multiple race and ethnic groups. Data from the 2000 Census was then used to create two Surveillance Epidemiology and End Results (SEER) datasets which include different race-specific population denominators among detailed API groups. This data has shown differences in both incidence and mortality in cancers in adults7, but has not yet been used to investigate possible differences in incidence within childhood cancer. Therefore, in this analysis, we sought to explore whether there was also variation in childhood cancer incidence among these important and growing populations within the US by estimating incidence rates for specific API groups using both the Detailed Asian/Pacific Islander Database7 in SEER and the Cancer Incidence in Five Continents8,9 data from the International Agency for Research on Cancer.

MATERIALS AND METHODS

Surveillance, Epidemiology, and End Results (SEER) Databases

We used two SEER databases to derive incidence rate estimates for this analysis. For estimates of individuals of API ethnicities, we used the Detailed Asian and Pacific Islander Database (2000-Centered)10 provided by SEER. Individuals aged 0 to 19 were included if they were diagnosed with cancer between January 1, 1988 to December 31, 2002. Estimates for Non-Hispanic White (NHW) children, which served as the referent group for many analyses, were derived using years 1998–2002 from the SEER 1311 database. This database was chosen to reflect the same years that were available for the detailed API ethnicities.

The Detailed Asian and Pacific Islander Database (2000-Centered) has been described by Miller et al.7 It includes individuals from one of the SEER reporting areas (Atlanta, Detroit, Seattle/Puget Sound and the states of California, Connecticut, Hawaii, Iowa, Kentucky, Louisiana, New Jersey, New Mexico, and Utah) which has previously been reported to cover 54% of the US API population7, though these proportions may vary by individual racial/ethnic group. SEER includes twelve pre-specified detailed API groups including Asian Indian/Pakistani, Chinese, Filipino, Guamanian, Native Hawaiian, Japanese, Kampuchean, Korean, Laotian, Samoan, Tongan, and Vietnamese. Individuals in the 2000 Census were able to indicate multiple race and ethnic responses, which has been used to calculate two populations for detailed API groups within SEER: individuals who selected a single API race (called the Low Population) and a combination group of those individuals who selected a single API race alone or in combination with one or more other race or ethnic groups (called the High Population). As outlined in Miller et al (2008)7, these may represent a maximum incidence rate (based on the single race/ethnicity respondents) and minimum rate (based on combination group which includes single race/ethnicity and multiple race/ethnicity respondents). We have chosen to focus on the Low Population in the main text of the manuscript because it is thought to represent the maximum incidence rate, however we have also included statistics on the High Population in the Supplementary Materials.

Cancer Incidence in Five Continents (CI5) Database

The CI512 includes data from cancer registries worldwide that adhere to strict quality criteria for inclusion.13 While the CI5 database is primarily classified by anatomical site, starting in 1988, histological type of cancer was added for some sites due to epidemiologic and clinical relevance.14 Incident cases and population data for children 0 to 19 years old within available cancer registries with predominant Asian and Pacific Islander ancestry in the CI5 data were used. The included registries were from China, Japan, Korea, India, Pakistan, Philippines, Thailand, Vietnam, and the Maori and Pacific Islanders of New Zealand. Data from 1998 to 2002 were used for all registries except New Zealand (2003–2007) and Vietnam (1993–1997), which did not have registries available during that time period.

Case Classification

The following International Classification of Childhood Cancer (ICCC), 3rd Edition15, categories were included based on sample size for stratification by race and ethnic group within the SEER datasets: (I) leukemias ([Ia, acute lymphoid leukemia (ALL); Ib, acute myeloid leukemia[AML]); II, lymphomas (IIa, Hodgkin lymphoma; IIb, Non Hodgkin lymphoma); III, central nervous system (CNS) overall and IIIb, astrocytomas individually; IVa, neuroblastoma; V, retinoblastoma; VII, hepatic tumors; VIII, bone (VIIIa, osteosarcoma); IX, soft‐tissue sarcomas; and X, germ cell tumors (GCT, including gonadal, extracranial and intracranial as a single entity).

In the CI5, cancer cases are classified using the International Classification of Disease (ICD) 1016 codes and International Classification of Diseases for Oncology17 (ICD-O-3) codes. We only included cancers from the CI5 that approximated ICCC-3 categories based on ICD-O-3 histological groups. Cancers included for analysis were leukemia (overall and ALL and AML individually), Hodgkin lymphoma, non-Hodgkin lymphoma (excluding Burkitt lymphoma), CNS tumors, astrocytic tumors, liver cancer, and osteosarcoma. ICD-10 and ICD-O-3 morphology codes for CI5 histological groups are listed in Supplementary Table 1.

This analysis used published data with no personal identifiers; therefore, the study was exempt from review by the University of Minnesota Institutional Review Board.

Population pooling

Within the SEER data, the rarity of childhood cancer in some of these ethnic groups required combining them into broader groups to stabilize incidence rate estimates. For instance, the smallest individual API ethnicity was Guamanian with just over 42,000 children between 0–19 years of age and only one cancer case diagnosed between 1998–2002. Hence, we formed the following groups based on geographic proximity of nation of origin, which also corresponds to genetic similarity: East Asia (China, Japan, Korea; n=377 total cases), Southeast Asia (Vietnam, Laos, Cambodia; n= 136 total cases), and Oceania (Guam, Samoa, Tonga; n=29 total cases). The Asian Indian/Pakistani (AIP) (n= 163 total cases) group was pre-defined by SEER and therefore was retained in these analyses. In addition, we chose to present incidence rates among Filipino children separately due to their large numbers (population in SEER of 1.65 million; n=199 total cases) and distinct population history. Throughout the remainder of this manuscript, we refer to these groups as “API regional groups” as they were created based on geographic proximity of countries of origin.

Statistical Analysis

Age-adjusted incidence rates were calculated for the 5-year study period (1998–2002) within the SEER datasets as cases per million and standardized to the 2000 US standard population using 5 year age categories. All rates and 95% CIs generated with SEER data used SEER*Stat software version 8.3.618. Incidence rate ratios (IRR) and 95% confidence intervals (CI), estimated using a normal distribution, were calculated with each API regional group compared to non-Hispanic Whites (NHW) as the referent group. Additionally, IRRs and 95% were calculated with East Asians serving as the referent population compared to other API regional groups.

Incidence rates using the CI5 data were calculated using SAS software version 9.4 (Cary, North Carolina, USA) as cases per million and were also standardized to the 2000 US population. In countries with multiple registries, the incidence rates were calculated by dividing pooled case numbers from registries in each country by pooled population denominators for each 5 year age category. Age standardization was performed by weighting each 5 year age category by its proportion of the 2000 US standard population and summing the weighted incidence rates for an age-standardized incidence rate.

To compare similarities in the age-adjusted incidence rates between SEER and CI5 Pearson correlation coefficients (r) were calculated using GraphPad Prism 8, which compared each US-based individual API ethnic group (e.g. Chinese) to the country of origin (e.g. China), by tumor type when available. However, Guam, Samoa, or Tonga were not included in CI5. In addition, the individual ethnicities were very small. Instead, the combined incidence rates of the Oceania regional group calculated using SEER data was compared to the incidence rates for the most closely related population available in CI5, the Pacific Islander population of New Zealand. As we were conducting observation epidemiology, we did not correct for multiple comparisons as recommended by Rothman.19

Additional Analyses

An additional analysis comparing sex differences in incidence rates of API regional groups was also performed using SEER data described above. Age-adjusted incidence rates were calculated in SEER*Stat software version 8.3.618. IRRs and 95% CIs within each API regional group and among NHWs were calculated with females serving as the referent. Finally, all analyses conducted for the Detailed Asian and Pacific Islander Low Population were also calculated in the High Population using the same methods described above. Both additional analyses are reported in the Supplementary material.

RESULTS

Comparison among API regional groups to NHW

Incidence rates among API regional groups were varied (Table 1). Figure 1 shows IRR and 95% CIs of each API regional group compared to NHW children (Panel A) as well as each API regional group compared to East Asian children (Panel B). The incidence rates for many of the cancers were lower than those in NHWs, though confidence intervals often crossed the null. The incidence of ALL was lower among all API regional groups than observed among NHW, with those among SE Asian (IRR 0.65, 95% CI 0.47, 0.92) being significantly lower. Acute myeloid leukemia (AML) was significantly higher among those from Oceania (IRR 3.88, 95% CI 1.83, 8.22) compared to NHW, albeit the rate for Oceanians was based on only 7 cases. Hodgkin lymphoma was less common among East Asians (IRR 0.28, 95% CI 0.15, 0.54), SE Asians (IRR 0.50, 95% CI 0.26, 0.97) and Filipinos (0.29, 95% CI 0.14, 0.58). CNS cancers were also less common among the same groups compared to NHW children (East Asian: IRR 0.66, 95% CI 0.51, 0.85; SE Asian: IRR 0.40, 95% CI 0.25, 0.65; Filipino: IRR 0.43, 95% CI 0.29, 0.64).

Table 1.

Incidence rates of childhood cancers per million among US-based Asian and Pacific Islander (API) groups and Non-Hispanic Whites, SEER Detailed Asian/Pacific Islander Database10 (1998– 2002; Low Population) and SEER 1311 (1998–2002)

East Asian SE Asian Oceania Filipino Asian Indian, Pakistani Non-Hispanic White
Population 2,609,190 1,256,085 232,820 1,654,275 947,400 27,686,359
Leukemias
Cases 121 35 15 58 52 1,216
Incidence Rate 47.2 (39.2, 56.4) 29.4 (20.5, 40.9) 64.8 (36.2, 107.0) 36.4 (27.6, 47.1) 50.1 (37.7, 66.0) 44.3 (41.8, 46.8)
Acute Lymphoblastic Leukemias
Cases 89 26 8 43 35 934
Incidence Rate 34.8 (27.9, 42.9) 22.2 (14.5, 32.4) 33.9 (14.6, 67.0) 27.6 (19.9, 37.1) 33.2 (23.0, 46.5) 34.0 (31.9, 36.3)
Acute Myeloid Leukemias
Cases 21 8 7 12 9 227
Incidence Rate 8.2 (5.1, 12.6) 6.6 (2.8, 13.0) 30.9 (12.4, 63.7) 7.2 (3.7, 12.7) 8.5 (3.8, 16.4) 8.2 (7.2, 9.4)
Lymphomas
Cases 43 25 <5 27 30 744
Incidence Rate 16.3 (11.8, 22.0) 19.0 (12.3, 28.2) 17.3 (4.7, 44.5) 15.6 (10.3, 22.8) 34.1 (23.0, 48.7) 26.7 (24.8, 28.7)
Hodgkin Lymphomas
Cases 10 9 <5 8 19 396
Incidence Rate 3.7 (1.8, 6.8) 6.8 (3.1, 13.0) 4.5 (0.1, 24.7) 4.4 (1.9, 8.8) 22.2 (13.3, 34.5) 14.2 (12.8, 15.6)
Non-Hodgkin Lymphomas
Cases 22 14 <5 10 <5 241
Incidence Rate 8.3 (5.2, 12.7) 10.6 (5.8, 17.9) 12.9 (2.7, 37.8) 5.8 (2.8, 10.8) 4.9 (1.3, 12.1) 8.7 (7.6, 9.8)
CNS
Cases 60 17 <5 24 25 960
Incidence Rate 23.1 (17.6, 29.8) 13.8 (8.0, 22.1) 17.3 (4.7, 44.4) 15.1 (9.7, 22.5) 25.7 (16.5, 38.1) 34.8 (32.6, 37.0)
Astrocytomas
Cases 26 7 <5 9 12 474
Incidence Rate 10.0 (6.5, 14.7) 5.7 (2.3, 11.8) 13.1 (2.7, 38.3) 5.9 (2.7, 11.1) 13.0 (6.6, 22.7) 17.2 (15.6, 18.8)
Neuroblastoma and ganglioneuroblastoma
Cases 12 8 0 11 11 253
Incidence Rate 4.9 (2.5, 8.6) 7.0 (3.0, 13.8) 0.0 7.2 (3.6, 12.9) 9.2 (4.6, 16.9) 9.3 (8.2, 10.5)
Retinoblastoma
Cases 6 <5 <5 <5 6 92
Incidence Rate 2.5 (0.9, 5.4) 3.5 (1.0, 8.9) 0.0 2.0 (0.4, 5.8) 5.0 (1.8, 11.2) 3.4 (2.7, 4.2)
Hepatic tumors
Cases 17 5 <5 <5 <5 60
Incidence Rate 6.8 (4.0, 10.9) 3.8 (1.2, 9.0) 4.4 (0.1, 24.5) 1.3 (0.2, 4.6) 2.7 (0.5, 8.2) 2.2 (1.7, 2.8)
Malignant bone tumor
Cases 18 15 <5 13 5 257
Incidence Rate 6.6 (3.9, 10.5) 11.5 (6.4, 19.0) 13.1 (2.7, 38.3) 7.6 (4.0, 13.0) 5.9 (1.9, 13.5) 9.2 (8.1, 10.4)
Osteosarcomas
Cases 17 11 <5 11 <5 128
Incidence Rate 6.2 (3.6, 10.1) 8.5 (4.2, 15.2) 8.7 (1.0, 31.4) 6.4 (3.2, 11.5) 4.7 (1.3, 11.7) 4.6 (3.8, 5.4)
Soft tissue and other extraosseous sarcomas
Cases 17 9 <5 15 8 360
Incidence Rate 6.5 (3.8, 10.4) 6.7 (3.0, 12.8) 0.0 8.6 (4.8, 14.4) 8.8 (3.8, 17.3) 13.0 (11.7, 14.4)
Germ cell & trophoblastic tumors & neoplasms of gonads
Cases 34 8 <5 22 12 329
Incidence Rate 12.9 (8.9, 18.1) 6.3 (2.7, 12.5) 4.1 (0.1, 23.6) 13.1 (8.2, 19.9) 13.3 (6.8, 23.1) 11.8 (10.6, 13.2)
Malignant gonadal germ cell tumors
Cases 18 <5 0 9 10 226
Incidence Rate 6.7 (4.0, 10.6) 2.4 (0.5, 7.0) 0.0 5.4 (2.5, 10.4) 11.3 (5.4, 20.6) 8.1 (7.1, 9.2)
Open in a new tab
*

Low populations include tabulations of respondents who selected a single race on the 2000 census form (e.g. Chinese alone). Incidence rates are per 1,000,000 and age-adjusted to the 2000 US Standard Population. Majority of cases were microscopically confirmed (East Asian 98%, SE Asian 96%, Oceanian 100%, Filipino 96%, Asian Indian/Pakistani 95%, Non-Hispanic White 96%) and only 6 cases were reported on a death certificate only.

Figure 1.

Figure 1.

Open in a new tab

Incidence rate ratios (IRR) and 95% confidence intervals (CI) for US-based Asian and Pacific Islander (API) regional groups compared to a) Non-Hispanic White children, and b) East Asian children, SEER Detailed Asian/Pacific Islander Database10 (1998– 2002; Low Population) and SEER 1311 (1998–2002)

Comparison of incidence among API regional groups to East Asians

When comparing the incidence of cancers among API regional groups to that of East Asians, a few clear patterns emerged. The incidence of AML in children from Oceania was nearly four times that in East Asians (IRR 3.88, 95% CI 1.64, 9.11), though roughly the same among all other regional groups. The incidence rates of ALL (IRR 0.63, 95% CI 0.41, 0.97) and malignant gonadal GCTs (IRR 0.29 95% CI 0.08, 0.97) were lower in SE Asians compared to East Asians. Hodgkin lymphoma was over five times as frequent in Asian Indian/Pakistani children compared to East Asians (IRR 5.50, 95% CI 2.56, 11.83).

Comparison between incidence rates for US-based API groups and countries of origin

When broken out by cancer type, overall correlation coefficients were variable (Figure 2) ranging from 0.12 (CNS) to 0.82 (Hodgkin lymphoma). However, even within cancer types with overall low correlations, we still observed that incidence rates between some individual US-based API ethnic groups and their countries of origin were similar. For example, despite a low correlation overall (r =0.12), the incidence rate of CNS tumors in US-based Chinese individuals was 18.6 cases/million compared to an incidence rate of 18.0 cases/million in China. Further, in osteosarcoma (r=−0.24) the incidence rates were similar among South Asians (US-based =4.7 cases/million, Pakistan=4.6 cases/million) and those from Oceania (US-based = 8.7 cases/million, Pacific Islander population from New Zealand= 9.73).

Figure 2.

Figure 2.

Open in a new tab

Scatterplots showing comparisons between childhood cancer incidence rates of US-based Asian and Pacific Islander (API) groups and incidence rates in countries of origin, SEER Detailed Asian/Pacific Islander Database10 (1998– 2002; Low Population) and Cancer in Five Continents12 (CI5) database.

Additional Analyses

In general, patterns comparing incidence rates by sex among API regional groups did not differ substantially from NHW children. Similar to what is seen among NHW children (IRR 1.19, 95% CI 1.05, 1.36), ALL is significantly more common among boys than girls (IRR: 1.56, 95% CI 1.01, 2.39) among East Asians. The sex ratio in Non-Hodgkin lymphoma was similar between both East Asians and NHWs (East Asian IRR 1.83, 95% CI 0.77, 4.37; NHW IRR 1.43, 95% CI 1.10, 1.85). Forest plots depicting male-to-female incidence rate ratios and 95% CIs are shown in Supplementary Figure 1. As expected, the magnitude of the incidence rates calculated using the High Population were lower than what was observed using the Low Population, though patterns remained similar when comparing across groups (Supplementary Table 2 and Supplementary Figures 2–4). Overall correlation coefficients were higher when compared to the High Population (Supplementary Figure 4). Incidence rates among Hawaiians were some of the highest rates compared to other API regional groups and NHWs.

DISCUSSION

We identified a number of differences in the incidence of childhood cancer between specific API groups in the US. In general, incidence rates among API regional groups were lower than those seen in NHWs. Incidence was higher among East Asians compared to SE Asians, though individuals from Oceania often had the highest incidence rates among all API regional groups. Incidence rates of specific cancers among some US-based API ethnic groups were similar to rates seen in originating countries, even when overall correlation coefficients were low. As expected, incidence rates completed in the Low Population were higher in magnitude, though general patterns were recapitulated in the High Population. Additionally, patterns in male-to-female incidence rate ratios were similar to published patterns showing a male excess, though interpretation should be careful due to small sample sizes in some of the comparisons.

That incidence rates of most childhood cancers examined were lower among API regional groups compared to NHW was unsurprising, as previous studies have shown that childhood cancer rates in combined API (i.e. the usual practice in SEER analyses) are often lower than NHW children.20 In addition, reports of incidence rates from Asian countries are often lower than those in the US.21,22 Liver cancer is one of the childhood cancer types where this trend did not hold. Instead, incidence rate of hepatic tumors were higher among almost all API regional groups, especially among East Asians. It has been well-documented in the literature on cancer in adults that liver cancer is common in this region of the world23,24 and in Asians and Pacific Islanders in the US7,25. Risk factors that are prevalent in this region of the world include, among others, endemic hepatitis B infection and consumption of aflatoxin-contaminated foods produced in the humid areas of SE Asia.23,26 These same dietary and infectious issues could underlie the observed increase in incidence we observed for hepatic tumors in children from API regional groups when compared to NHWs.

When comparing incidence rates among API regional groups, our findings highlight patterns which have been reported to some extent at the country-level in global analyses. Higher incidence rates of many cancer types among East Asians compared to SE Asians has been observed in both the adult27 and childhood cancer literature.9,28 We also found a much higher incidence rate of AML among those in the US from Oceania compared to those of East Asian children. This finding is similar to a previous report28, which showed a higher incidence of leukemia in registry data from Oceania compared to registries in East Asia. Finally, we also observed a higher incidence rate of Hodgkin lymphoma among US children of Asian Indian or Pakistani descent compared to those from East Asia. This confirms findings of an earlier study in the US in which Glaser and Hsu (2002)29 found a higher rate of Hodgkin’s lymphoma in US Asian Indians compared to US Chinese and US Japanese, though this study only included Asian Indians from California. These differences in incidence rates between Asian ethnic populations could reflect differences in Epstein-Barr Virus (EBV) infection. EBV-associated Hodgkin’s lymphoma has been shown to vary across countries as well as within different ethnic groups in the US.30,31 These studies, however, only report infection among grouped “Asian” or “Asian and Pacific Islanders” categories, which hinder any determination about whether differences exist among Asian ethnic groups in the US. A study from the United Kingdom reported EBV+ infection rates among South Asian (Indian, Pakistani, or Bangladeshi individuals) Hodgkin lymphoma cases. They found that among 55 cases of childhood Hodgkin’s disease, South Asians had over 20 times higher risk of being EBV+ compared to non-South Asians.32 Future studies should examine whether these differences in EBV positivity are also present in the US.

Differences in the incidence rates observed between racial and ethnic groups may be explained by observed by differences in the frequency of single nucleotide polymorphisms (SNPs) between these groups. Most of the existing genome-wide association studies (GWAS) of pediatric cancer have been conducted in children of European ancestry, though some of the genome-wide significant SNPs have been examined in other race and ethnic groups including Non-Hispanic Blacks33,34, Hispanics33,35, and East Asians36,37. The most extensive transethnic replication in childhood cancer has been done within ALL. To explore the role that SNP frequencies may play in explaining differences observed among incidence rates our findings, we obtained allele frequencies and corresponding ORs and 95% CIs from the published literature38–45 among East Asians, SE Asians, and European populations and summarized the information in Figure 3. Several of the SNPs which have been reported to have the strongest effect in ALL, including ARID5B, IKZF1, CEBPE, CDKN1A/B, PIP4K2A, and GATA3, are included in the figure. All ORs reported in European populations show significant associations with ALL. ORs among the East Asian and SE Asian populations are relatively similar, albeit with wide confidence intervals. Allele frequencies, extracted from 1000 Genomes, also appear to be relatively similar between East Asian, SE Asian, and European populations. This suggests that if genetics explained differences in incidence rates of ALL in SE Asians, NHW, and East Asians, the particular variants have not yet been described. Understanding the role of genetic differences in cancer susceptibility is especially important across Asia and the Pacific Islands, areas which are known for rich genetic diversity46, and areas where differences in the amount of DNA present from an ancient, extinct hominid have been reported.47 A recent article from Hsieh et al (2019)48 characterized large copy number variants arising from this this ancient hominid population among Oceanians which map near genes associated with metabolism, development and cell cycle or cell signaling, and immune response. Discovering whether this structural variation is associated with childhood cancer in API populations will only be possible with the inclusion of more diverse populations in future GWAS and genetic predisposition studies.

Figure 3.

Figure 3.

Open in a new tab

a) Odds ratios and 95% confidence intervals for SNPs identified from published GWAS literature in acute lymphoblastic leukemia (ALL)38–45 and b) corresponding SNP allele frequencies from 1000 Genomes in ARID5B, IKZF1, CEBPE, CDKN2A/B, PIP4K2A, GATA3 among Asian and European populations.

Much of the growth in the Asian population within the US is driven by immigration6, so we compared how similar incidence rates were between US-based API ethnic groups and Asian or Pacific Island countries of origin. While some incidence rates were similar, suggesting shared genetic predisposition or similar environmental exposures among unassimilated recent immigrants, others were quite different. Other reasons for these differences might include the fact that cancers may be less well diagnosed in countries of origin, especially those which have more limited resources.

A male excess of childhood cancer, especially among NHWs, has been well described in the literature.49 In addition, this pattern has been seen at the global level, including, a previous report from Steliarova-Foucher et al (2010)28, which collected data from hundreds of registries worldwide and found a global excess in cancer among males. We confirmed these findings in API regional groups, though confidence intervals were large and care should be taken in making interpretations.

As was mentioned earlier, use of both the Detailed Asian and Pacific Islander Low Population and the High Population Databases represent a maximum incidence rate (based on the single race/ethnicity respondents; Low Population) and minimum rate (based on combination group which includes single race/ethnicity and multiple race/ethnicity respondents; High Population). As expected, the incidence rates calculated using the High Population were smaller in magnitude due to the fact that the underlying population is larger. Despite this, patterns comparing incidence rates among different groups remained relatively similar.

There are several strengths to this study. We used a large population-based sample to estimate incidence rates for specific API groups within the US using SEER data. The SEER database has previously been shown to cover over half of the API population within the US.7 We both compared these rates amongst API groups to highlight important differences between them and compared them to rates calculated from population-based registries arising from the well-documented CI5 data that corresponds to multiple Asian and Pacific Island countries which adhere to strict quality criteria.

Conversely, there are also some important limitations to this data. One is the lack of information about the parent and child’s country of birth and the length of time each family has been in the US. This information is especially important because of potential environmental exposures for both the mother and the child. Early studies on Asian immigrants have shown that within one or two generations, incidence of some cancer in adults eventually mirror those found in the general US population.50 It is unclear whether incidence rates of API regional groups are similar to those observed from the originating countries because of recent immigration to the US, such that their risk would still be reflective of their home country, or because of shared genetic predisposition between individuals with common countries of origin. Statistical limitations of this paper resulted in a lack of precision for some of the estimates of rate. For example, small numbers of cases, especially in the Oceania regional group, may have led to unstable and inflated estimates. However, we found incidence rates of similar magnitude among Pacific Islanders from the New Zealand registry data in CI5. This suggests that these larger incidence rate estimates may instead be due to genetic predisposition of individuals from this region of the world, though validation of this hypothesis is warranted.

In conclusion, we found differences in the incidence rates of childhood cancers among API populations in the US. Incidence rates in these groups appeared similar to the respective countries of origin, potentially underscoring differences in genetic predisposition or environmental risk factors. Future studies are necessary to better understand differences in genetic predisposition, as well as differences in environmental exposures (and how they vary with assimilation) among Asian and Pacific Islander ethnicities in the US.

Supplementary Material

Moore IJC Supplementary FINAL

FUNDING:

This work was supported by the National Institutes of Health (grant T32 CA099936; Logan Spector, PhD, MPH [LAW and AKH]) and the National Cancer Institute of the National Institutes of Health (grant 2T32CA163184; Michele Allen, MD, MS, PI [KJM]). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

ABBREVIATIONS:

ALL

Acute lymphoblastic leukemia

AML

Acute myeloid leukemia

API

Asian and Pacific Islander

CI

Confidence interval

CI5

Cancer in Five Continents

CNS

Central Nervous System

GWAS

Genome-wide association studies

ICCC

International Classification of Childhood Cancer

ICD

International Classification of Disease

IRR

Incidence rate ratio

NHW

Non-Hispanic Whites

SE

Southeast

SEER

Surveillance, Epidemiology, and End Results

US

United States

Footnotes

DATA ACCESSIBILITY: The data that support the findings of this study are openly available from the Surveillance, Epidemiology, and End Results (SEER) at https://seer.cancer.gov/seerstat/databases/api.races.2000pops/2005submission.html, reference number 10 and the Cancer Incidence in Five Continents at https://ci5.iarc.fr/CI5I-X/Pages/download.aspx, reference number 12. Datasets supporting the results in the paper will be made available upon reasonable request.

CONFLICT OF INTEREST: The authors report no conflicts of interest.

ETHICAL APPROVAL: This analysis used published data with no personal identifiers; therefore, the study was exempt from review by the University of Minnesota Institutional Review Board.

REFERENCES

  • 1.Linabery AM, Ross JA. Trends in childhood cancer incidence in the U.S. (1992–2004). Cancer 2008;112:416–32. [DOI] [PubMed] [Google Scholar]
  • 2.Zhao Y, Wang Y, Ma S. Racial Differences in Four Leukemia Subtypes: Comprehensive Descriptive Epidemiology. Sci Rep 2018;8:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Wong JR, Tucker MA, Kleinerman RA, Devesa SS. Retinoblastoma incidence patterns in the US surveillance, epidemiology, and end results program. JAMA Ophthalmol 2014;132:478–83. [DOI] [PubMed] [Google Scholar]
  • 4.Siegel DA, Li J, Jane Henley S, Wilson RJ, Buchanan Lunsford N, Tai E, Van Dyne EA. Geographic variation in pediatric cancer incidence — United States, 2003–2014. Morb Mortal Wkly Rep 2018;67:707–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Hoeffel EM, Rastogi S, Kim MO, Shahid H. The Asian Population: 2010. 2010 Census Briefs. 2012;1–24. [Google Scholar]
  • 6.Brown A U.S. Hispanic and Asian populations growing, but for different reasons 2015. http://www.pewresearch.org/fact-tank/2014/06/26/u-s-hispanic-and-asian-populations-growing-but-for-different-reasons/. Accessed September 6, 2019.
  • 7.Miller BA, Chu KC, Hankey BF, Ries LA. Cancer incidence and mortality patterns among specific Asian and Pacific Islander populations in the U.S. Cancer Causes Control 2008;19:227–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Breslow NE, Langholz B. Childhood cancer incidence: Geographical and temporal variations. Int J Cancer 1983;32:703–16. [DOI] [PubMed] [Google Scholar]
  • 9.Hubbard AK, Spector LG, Fortuna G, Marcotte EL, Poynter JN. Trends in International Incidence of Pediatric Cancers in Children Under 5 Years of Age: 1988–2012. JNCI Cancer Spectr 2019;3:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Surveillance, Epidemgiology, and End Results (SEER) Program. Detailed Asian/Pacific Islander Databases (2000-Centered) https://seer.cancer.gov/seerstat/databases/api.races.2000pops/2005submission.html. Accessed June 13, 2019.
  • 11.Surveillance, Epidemiology, and End Results (SEER) Program (www.seer.cancer.gov) SEER*Stat Database: Incidence—SEER 13 Regs Research Data, Nov 2018. Sub (1992–2016) <Katrina/Rita Population Adjustment>. http://www.seer.cancer.gov.libproxy.lib.unc.edu. Accessed February 9, 2020
  • 12.International Agency for Research on Cancer. Cancer in Five Continents Volumes I to X. 2020. https://ci5.iarc.fr/CI5I-X/Pages/download.aspx. Accessed September 1, 2019.
  • 13.Bray F, Parkin DM. Evaluation of data quality in the cancer registry: Principles and methods. Part I: Comparability, validity and timeliness. Eur J Cancer 2009;45:747–55. [DOI] [PubMed] [Google Scholar]
  • 14.Parkin DM, Ferlay J, Shanmugaratnam K, Sobin L, Teppo L, Whelan SL. Chapter 4. Histological groups. Int Classif 1992;45–55. [Google Scholar]
  • 15.Steliarova-Foucher E, Stiller C, Lacour B, Kaatsch P. International classification of childhood cancer, third edition. Cancer 2005;103:1457–67. [DOI] [PubMed] [Google Scholar]
  • 16.World Health Organization. (‎2004)‎. ICD-10 : international statistical classification of diseases and related health problems : tenth revision, 2nd ed World Health Organization. [Google Scholar]
  • 17.Fritz A, Percy C, Jack A. International classification of diseases for oncology (ICD-O) 3rd edn World Health Organization; Geneva: 2000. [Google Scholar]
  • 18.Surveillance Research Program, National Cancer Institute SEER* Stat software (seer.cancer.gov/seerstat) version 8.3.6.
  • 19.Rothman. No Adjustments Are Needed for Multiple Comparisons. EPIDEMIOLOGY 1990;1:43–6. [PubMed] [Google Scholar]
  • 20.Marcotte EL, Druley TE, Johnson KJ, Richardson M, von Behren J, Mueller BA, Carozza S, McLaughlin C, Chow EJ, Reynolds P, Spector LG. Parental Age and Risk of Infant Leukaemia: A Pooled Analysis. Paediatr Perinat Epidemiol 2017;31:563–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Lu L, Huang C, Huang H. Childhood cancer : an emerging public health issue in China. Ann Transl Med 2015;3:250. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Nakata K, Ito Y, Magadi W, Bonaventure A, Stiller CA, Katanoda K, Matsuda T, Miyashiro I, Pritchard-Jones K, Rachet B. Childhood cancer incidence and survival in Japan and England: A population-based study (1993–2010). Cancer Sci 2018;109:422–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Zhu RX, Seto W, Lai C, Yuen M. Epidemiology of Hepatocellular Carcinoma in the Asia-Pacific Region. Gut Liver 2016;10:332–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Yang JD, Hainaut P, Gores GJ, Amadou A, Plymoth A, Roberts LR. A global view of hepatocellular carcinoma: trends, risk, prevention and management. Nat Rev Gastroenterol Hepatol 2019;16:589–604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Chang ET, Keegan THM, Gomez SL, Le GM, Clarke CA, So SKS, Glaser SL. The burden of liver cancer in Asians and Pacific Islanders in the Greater San Francisco Bay Area, 1990 through 2004. Cancer 2007;109:2100–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Ko KP, Shin A, Cho S, Park SK, Yoo KY. Environmental contributions to gastrointestinal and liver cancer in the Asia–Pacific region. J Gastroenterol Hepatol 2018;33:111–20. [DOI] [PubMed] [Google Scholar]
  • 27.Ng CJ, Teo CH, Abdullah N, Tan WP, Tan HM. Relationships between cancer pattern, country income and geographical region in Asia. BMC Cancer 2015;15:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Steliarova-Foucher E, Colombet M, Ries LAG, Moreno F, Dolya A, Bray F, Hesseling P, Shin HY. International incidence of childhood cancer, 2001 – 10 : a population-based registry study. Lancet Oncol 2017;18:719–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Glaser SL, Hsu JL. Hodgkin’s disease in Asians: Incidence patterns and risk factors in population-based data. Leuk Res 2002; 26(3):261–9. [DOI] [PubMed] [Google Scholar]
  • 30.Dinand V, Arya LS. Epidemiology of childhood Hodgkins disease: Is it different in developing countries? Indian Pediatr 2006; 43(2):141–7. [PubMed] [Google Scholar]
  • 31.Glaser SL, Gulley ML, Clarke CA, Keegan TH, Chang ET, Shema SJ, Craig FE, DiGiuseppe JA, Dorman RF, Mann RB, Anton-Culver H, Ambinder RF. Racial/ethnic variation in EBV-positive classical Hodgkin lymphoma in California populations. Int J Cancer 2008; 123(7):1499–1507. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Flavel KJ, Biddulph JP, Powel JE, Parkes SE, Redfern D, Weinreb M, Nelson P, Mann JR, Young LS, Murray PG. South Asian ethnicity and material deprivation increase the risk of Epstein-Barr Virus infection in Childhood Hodgkin’s Disease. B J Cancer 2001; 85(3):350–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Wiemels JL, Walsh KM, Smith AJ De, Metayer C, Gonseth S, Hansen HM, Francis SS, Ojha J, Smirnov I, Barcellos L, Xiao X, Morimoto L, et al. GWAS in childhood acute lymphoblastic leukemia reveals novel genetic associations at chromosomes 17q12 and 8q24.21. Nat Commun 2018;9:286. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Xu H, Yang W, Perez-andreu V, Devidas M, Fan Y, Cheng C, Pei D, Scheet P, Burchard G, Eng C, Huntsman S, Torgerson DG, et al. Novel Susceptibility Variants at 10p12. 31–12. 2 for Childhood Acute Lymphoblastic Leukemia in Ethnically Diverse Populations. J Natl Cancer Inst 2013;105:733–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Qian M, Xu H, Perez-Andreu V, Roberts KG, Zhang H, Yang W, Zhang S, Zhao X, Smith C, Devidas M, Gastier-Foster JM, Raetz E, et al. Novel susceptibility variants at the ERG locus for childhood acute lymphoblastic leukemia in Hispanics. Blood 2019;133:724–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Marigorta UM, Navarro A. High Trans-ethnic Replicability of GWAS Results Implies Common Causal Variants. PLoS Genet 2013;9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Zhang R, Zou Y, Zhu J, Zeng X, Yang T, Wang F, He J. The Association between GWAS-identified BARD1 Gene SNPs and Neuroblastoma Susceptibility in a Southern Chinese Population. Int J Med Sci 2016;13:133–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Yang W, Yang J, Scheet P, Pui C, William E, Relling MV. ARID5B SNP rs10821936 is associated with risk of childhood acute lymphoblastic leukemia in blacks and contributes to racial differences in leukemia incidence. Leukemia 2010;24:894–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Urayama KY, Takagi M, Kawaguchi T, Matsuo K, Tanaka Y, Ayukawa Y, Arakawa Y, Hasegawa D, Yuza Y, Kaneko T, Noguchi Y, Taneyama Y, et al. Regional evaluation of childhood acute lymphoblastic leukemia genetic susceptibility loci among Japanese. Sci Rep 2018;8:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Tao R, Liu YJ, Liu LF, Li W, Zhao Y, Li HM, Yi XL, Zhao ZY. Genetic polymorphisms of ARID5B rs7089424 and rs10994982 are associated with B-lineage ALL susceptibility in Chinese pediatric population. J Chinese Med Assoc 2019;82:562–7. [DOI] [PubMed] [Google Scholar]
  • 41.Bhandari P, Ahmad F, Mandava S, Das BR. Association of genetic variants in ARID5B, IKZF1 and CEBPE with Risk of childhood de novo B-lineage acute lymphoblastic leukemia in India. Asian Pacific J Cancer Prev 2016;17:3987–93. [PubMed] [Google Scholar]
  • 42.Wang Y, Chen J, Li J, Deng J, Rui Y, Lu Q, Wang M, Tong N, Zhang Z, Fang Y. Association of three polymorphisms in ARID5B, IKZF1and CEBPE with the risk of childhood acute lymphoblastic leukemia in a Chinese population. Gene 2013;524:203–7. [DOI] [PubMed] [Google Scholar]
  • 43.Vijayakrishnan J, Sherborne AL, Sawangpanich R, Hongeng S, Houlston RS, Pakakasama S. Variation at 7p12.2 and 10q21.2 influences childhood acute lymphoblastic leukemia risk in the Thai population and may contribute to racial differences in leukemia incidence. Leuk Lymphoma 2010;51:1870–4. [DOI] [PubMed] [Google Scholar]
  • 44.Lin CY, Li MJ, Chang JG, Liu SC, Weng T, Wu KH, Yang SF, Huang FK, Lo WY, Peng CT. High-resolution melting analyses for genetic variants in ARID5B and IKZF1 with childhood acute lymphoblastic leukemia susceptibility loci in Taiwan. Blood Cells, Mol Dis 2014;52:140–5. [DOI] [PubMed] [Google Scholar]
  • 45.Xu H, Cheng C, Devidas M, Pei D, Fan Y, Yang W, Neale G, Scheet P, Burchard EG, Torgerson DG, Eng C, Dean M, et al. ARID5B genetic polymorphisms contribute to racial disparities in the incidence and treatment outcome of childhood acute lymphoblastic leukemia. J Clin Oncol 2012;30:751–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Zhang F, Su B, Zhang YP, Jin L. Genetic studies of human diversity in East Asia. Philos Trans R Soc B Biol Sci 2007;362:987–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Browning SR, Browning BL, Zhou Y, Tucci S, Joshua M, Biology E. Analysis of human sequence data reveals two pulses of archaic Denisovan admixture. Cell 2018;173:53–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Hsieh P, Vollger MR, Dang V, Porubsky D, Baker C, Cantsilieris S, Hoekzema K, Lewis AP, Munson KM, Sorensen M, Kronenberg ZN, Murali S, et al. Adaptive archaic introgression of copy number variants and the discovery of previously unknown human genes. 2019;2083. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Williams LA, Poynter JN, Richardson M, Spector LG, Marcotte EL. Sex ratio among childhood cancers by single year of age. Pediatr bl 2019;66:p.e27620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Le GM, Gomez SL, Clarke CA, Glaser SL, West DW. Cancer incidence patterns among Vietnamese in the United States and Ha Noi, Vietnam. Int J Cancer 2002;102:412–7. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Moore IJC Supplementary FINAL
NIHMS1642357-supplement-Moore_IJC_Supplementary_FINAL.docx (446.9KB, docx)

RESOURCES