Abstract
While some previous studies disaggregated the Asian, Native Hawaiian, and Pacific Islander (ANHPI) population to investigate survival for childhood leukemia, further studies are needed to understand the differences between subpopulations. The aim of our study was to estimate 5‐year relative survival for patients with childhood leukemia and to investigate disparities in prognostic factors with disaggregation of the ANHPI population. We used the Surveillance, Epidemiology, and End Results Program 17 database and included 1881 ANHPI patients with childhood leukemia and 8772 non‐Hispanic White (NHW) patients with childhood leukemia. The Cox proportional hazards model was used to estimate hazard ratios for the risk of death. We observed lower 5‐year relative survival rates for Southeast Asian and East Asian compared to NHW patients with childhood leukemia for acute lymphoid leukemia (ALL). The survival rates were higher for patients diagnosed at 1–9 years of age, more recent years of diagnosis, and patients residing in urban areas. The risk of death was 42% higher for East Asian patients and 50% higher for Southeast Asian patients compared to NHW patients for childhood ALL. For prognostic factors among East Asian patients with childhood leukemia, higher risks of death were observed for patients diagnosed at <12 months old and for acute myeloid leukemia compared to ALL. Further studies are needed to elucidate the reasons behind the disparities in survival rates for Southeast Asian and East Asian patients with childhood leukemia, including socioeconomic and genetic contributions to leukemia risk and clinical responses to different therapeutic modalities.
Keywords: Asian, childhood leukemia, Native Hawaiian and Pacific Islander, survival
What's new?
Survival of childhood acute lymphoid leukemia (ALL) may vary significantly among Asians, Native Hawaiians, and Pacific Islanders (ANHPI). However, these populations often are investigated as a single group, which ignores socioeconomic and racial/ethnic heterogeneity. Here, the authors disaggregated the ANHPI population to explore disparities in childhood ALL survival. Compared to Non‐Hispanic Whites, five‐year survival for Southeast Asians and East Asians was poorer. Patients in these subgroups experienced better survival when they lived in urban areas or were more recently diagnosed or diagnosed between ages 1 and 9. The results provide insight into differences in childhood ALL outcomes for ANHPI subgroups.
1. INTRODUCTION
Incidence rates of acute myeloid leukemia (AML) in the Asian, Native Hawaiian, and Pacific Islander (ANHPI) population may be increasing statistically significantly in contrast to acute lymphoid leukemia (ALL) rates, which did not show changes over the 2000–2019 period. 1 Upon disaggregation of the ANHPI population, incidence rate ratios showed no differences for East Asian, Filipino, and Asian Indian and Pakistani populations for childhood leukemia when compared to the non‐Hispanic white (NHW) population. 2 In the Southeast Asian population, however, the incidence rate ratios indicated lower incidence for childhood leukemia than in the NHW and East Asian populations.
A study focusing on childhood leukemia in Filipino patients reported much lower relative survival in the Philippines (32.9%) compared to Filipinos in the United States (80.1%) and NHW (81.9%). 3 In the Childhood Oncology Group Cohort Study, non‐Hispanic Asian patients with childhood leukemia, not including Native Hawaiian or Pacific Islanders, had the highest 5‐year event‐free survival (88.1%) compared to the other race and ethnicity groups. A Surveillance, Epidemiology, and End Results Program (SEER) database analysis of childhood ALL cases diagnosed from 1988 to 2008 reported that the 5‐year survival probability was 0.80 (95% confidence interval [CI] = 0.76, 0.84) for East Asian patients and 0.88 (95%CI = 0.86, 0.88) for NHW patients. 4 Hazard ratios (HRs) for all cause survival were higher for East Asian (1.48, 95%CI = 1.17, 1.86), Filipino (1.64, 95%CI = 1.13, 2.38), and Vietnamese (2.44, 95%CI = 1.50, 3.97) patients compared to NHW patients. In California, similar results were observed with a 40% increase in the risk of death (95%CI = 1.11, 1.76) among ANHPI vs. NHW patients with childhood ALL diagnosed from 1988 to 2012. 5 These studies did not investigate prognostic factors among ANHPI patients with childhood leukemia for possible reasons for differences in survival. Considering the heterogeneity in socioeconomic status, healthcare access, culture, and levels of acculturation among East Asian, South Asian, Southeast Asian, Native Hawaiian, and Pacific Islander populations, investigating possible reasons for childhood leukemia survival differences is needed.
Many previous descriptive studies of cancer grouped the ANHPI populations together. There have been more calls for the need to separate out the ANHPI subgroups or disaggregate the ANHPI population in cancer studies to understand the heterogeneity across the ANHPI subgroups. While some previous studies disaggregated the ANHPI population, further studies are needed to understand the detailed differences between subpopulations and potential explanations for these discrepancies. Five‐year relative survival for more recently diagnosed patients is also needed for childhood leukemia in the ANHPI population. This study will add patients diagnosed from 2012 to 2020 and investigate prognostic factors in contrast to previous studies. The aim of our study was to estimate 5‐year relative survival for patients with childhood leukemia, with disaggregation of the ANHPI population, and to investigate disparities in prognostics factors among ANHPI subgroups of patients with childhood leukemia.
2. METHODS
We used the SEER 17 database, 6 which includes data from 17 population‐based cancer registries funded by the National Cancer Institute (NCI). We included cancer patients with the following eligibility criteria: (1) diagnosed between 2000 and 2020; (2) first primary cancer diagnosis of lymphoid leukemia, myeloid leukemia, and unspecified and other leukemias; (3) age at diagnosis <18 years old; and (4) ANHPI or NHW. The leukemia coding was based on the International Classification of Childhood Cancer, third edition, based on ICD‐O‐3/IARC 2017. There were a total of 1881 ANHPI patients with childhood leukemia and 8772 NHW patients with childhood leukemia. We assessed age at diagnosis, sex, county level income, and first‐course cancer treatment (radiation and chemotherapy) among patients by ANHPI subgroups. We used the county characteristics to group cancer patients in metropolitan and rural residence at cancer diagnosis. From the SEER data, metropolitan counties urban and rural status was defined by the 2013 Rural–Urban Continuum Codes, using all metro categories for urban and all nonmetro as rural. 7 The University of Utah Institutional Review Board (IRB) has designated that the analysis of SEER data does not require IRB review.
We grouped the ANHPI race and ethnicity groups as: (1) East Asian: Chinese, Japanese, and Korean, (2) Southeast Asian: Filipino, Vietnamese, Laotian, Hmong, Kampuchean, and Thai, (3) Native Hawaiian and Pacific Islander (NHPI): Hawaiian, Micronesian, Chamorro, Guamanian, Polynesian, Tahitian, Samoan, Tongan, Melanesian, Fiji Islander, New Guinean, and Pacific Islander NOS, and (4) South Asian: Asian Indian or Pakistani not specified (NOS), Asian Indian (from 2010), and Pakistani (from 2010). The sum of the four ANHPI subgroups does not add to the ANHPI overall due to the “other Asian” patients who were categorized in this group by SEER and cannot be categorized in the four subgroups (n = 376).
2.1. Statistical analysis
We used SAS version 9.4 to analyze the data for survival. Five‐year relative survival rates were estimated using SEER*Stat version 8.4.2. The relative survival is “a net survival measure representing cancer survival in the absence of other causes of death. Relative survival is defined as the ratio of the proportion of observed survivors in a cohort of cancer patients to the proportion of expected survivors in a comparable set of cancer‐free individuals.” 8 Relative survival is considered to be appropriate because the measure “is not influenced by changes in mortality from other causes and, therefore, provides a useful measure for cancer control over time.” 9 The Kaplan–Meier (KM) curve was used to compare the probability of survival between the groups, and the log rank test was used to compare overall survivor functions in the KM curve. The Cox proportional hazards model was used to estimate HRs for the risk of death. We tested the Cox proportional hazards assumption by including interactions between the predictors and time in the model. Time to event was calculated as the time from cancer diagnosis to cancer death or end of follow‐up. In terms of survival months, we have confirmation of vital status for the majority of patients (97.0%) either with their date of death or if they were alive past the study cutoff date. For a small proportion of patients (3.0%), they have been lost to follow‐up, where vital status or date of death could not be ascertained. For these patients with a missing date, the case is counted as lost to follow‐up during the interval where they were last confirmed to be alive.
We used the NHW patients with childhood leukemia as the reference group to investigate differences with a broader but defined US population. We also used the Southeast Asian patients with childhood leukemia as the reference group since they were the largest ANHPI subgroup to investigate differences among the ANHPI subgroups. For the prognostic factors, we estimated HRs for each race and ethnicity subgroup (ANHPI, East Asian, South Asian, Southeast Asian, and NHPI).
We adjusted for age in 5‐year age categories (00, 1–4, 5–9, 10–14, and 15–17), sex, county‐level income, rural residence, type of leukemia (lymphoid, acute myeloid, unspecified, and other), radiation, and chemotherapy when assessing the potential association between race and ethnicity and the risk of death since these are prognostic factors for death and are associated with race and ethnicity.
3. RESULTS
The largest proportion of patients who were diagnosed at 15–17 years of age with childhood leukemia was in the East Asian patient group (Table 1). The largest proportion of patients who were 0 years of age at diagnosis were in the NHPI patient group. There were higher proportions diagnosed among males, with the highest in the South Asian patient group. In terms of income, the highest proportion of all patients were in the highest income group. Very few ANHPI patients with childhood leukemia were from rural areas. Most patients received chemotherapy and did not receive radiation therapy for their first course of treatment. For the patients who received radiation therapy, 99% also received chemotherapy.
TABLE 1.
Characteristics of ANHPI and NHW childhood leukemia patients in SEER, diagnosed 2000–2020.
| ANHPI (n = 1811) | East Asian (n = 394) | Southeast Asian (n = 490) | South Asian (n = 328) | NHPI (n = 223) | NHW (n = 8772) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n | % | n | % | n | % | n | % | n | % | n | % | |
| Age | ||||||||||||
| 00 years | 106 | 5.9 | 24 | 6.1 | 27 | 5.5 | 16 | 4.9 | 18 | 8.1 | 400 | 4.6 |
| 01–04 years | 742 | 41.0 | 153 | 38.8 | 189 | 38.6 | 160 | 48.8 | 91 | 40.8 | 3769 | 43.0 |
| 05–09 years | 426 | 23.5 | 83 | 21.1 | 111 | 22.7 | 79 | 24.1 | 51 | 22.9 | 2024 | 23.1 |
| 10–14 years | 311 | 17.2 | 69 | 17.5 | 95 | 19.4 | 48 | 14.6 | 41 | 18.4 | 1400 | 16.0 |
| 15–17 years | 226 | 12.5 | 65 | 16.5 | 68 | 13.9 | 25 | 7.6 | 22 | 9.9 | 1179 | 13.4 |
| Sex | ||||||||||||
| Female | 806 | 44.5 | 172 | 43.7 | 233 | 47.6 | 138 | 42.1 | 106 | 47.5 | 3885 | 44.3 |
| Male | 1005 | 55.5 | 222 | 56.3 | 257 | 52.4 | 190 | 57.9 | 117 | 52.5 | 4887 | 55.7 |
| County‐level income | ||||||||||||
| <$59,999 | 96 | 5.3 | a | a | 47 | 9.6 | 13 | 4.0 | a | a | 2138 | 24.4 |
| $60,000–$64,999 | 126 | 7.0 | a | a | 35 | 7.1 | 24 | 7.3 | a | a | 715 | 8.2 |
| $65,000–$69,999 | 257 | 14.2 | 86 | 21.8 | 80 | 16.3 | 31 | 9.5 | 14 | 6.3 | 1185 | 13.5 |
| $70,000–$74,999 | 145 | 8.0 | 24 | 6.1 | 52 | 10.6 | 22 | 6.7 | 13 | 5.8 | 1035 | 11.8 |
| $75,000+ | 1186 | 65.5 | 258 | 65.5 | 273 | 55.7 | 238 | 72.6 | 162 | 72.6 | 3686 | 42.0 |
| Rural | ||||||||||||
| Urban | 1746 | 96.4 | a | a | 476 | 97.1 | a | a | 190 | 85.2 | 7496 | 85.5 |
| Rural | 64 | 3.5 | a | a | 14 | 2.9 | a | a | 33 | 14.8 | 12.63 | 14.4 |
| Type | ||||||||||||
| Ia Lymphoid leukemias | 1392 | 76.9 | 297 | 75.4 | 356 | 72.7 | 279 | 85.1 | 165 | 74.0 | 7104 | 81.0 |
| Ib Acute myeloid leukemias | 387 | 21.4 | a | a | a | a | a | a | a | a | 1476 | 16.8 |
| Ie Unspecified and other specified leukemias | 32 | 1.8 | a | a | a | a | a | a | a | a | 192 | 2.2 |
| Radiation | ||||||||||||
| None | 1653 | 91.3 | 353 | 89.6 | 441 | 90.0 | 306 | 93.3 | 205 | 91.9 | 8006 | 91.3 |
| Beam radiation | 153 | 8.4 | 40 | 10.2 | 47 | 9.6 | 21 | 6.4 | 17 | 7.6 | 740 | 8.4 |
| Chemotherapy | ||||||||||||
| No/unknown | 55 | 3.0 | 11 | 2.8 | 17 | 3.5 | a | a | a | a | 346 | 3.9 |
| Yes | 1756 | 97.0 | 383 | 97.2 | 473 | 96.5 | a | a | a | a | 8426 | 96.1 |
Abbreviations: ANHPI, Asian, Native Hawaiian, and Pacific Islander; NHPI, Native Hawaiian and Pacific Islander; NHW, non‐Hispanic white.
Cell sizes <11 are suppressed to protect patient confidentiality, as are cells that could be used to calculate another small cell.
The survival curves for ALL and AML are shown in Figure 1. The p‐values for the log‐rank test for the survival curves for patients with childhood leukemia were statistically significant comparing Southeast Asian to NHW patients (p < .0001 for childhood leukemia and p = .0005 for ALL) and comparing East Asian to NHW patients (p = .0018 for childhood leukemia and p = .0202 for ALL). The corresponding p‐values for the log‐rank test were not statistically significant for AML comparing Southeast Asian and East Asian to NHW patients.
FIGURE 1.
Survival curves for childhood ALL (Ia) and AML (Ib).
The 5‐year relative survival rates were lowest for Southeast Asian patients with childhood leukemia for both ALL (85.8%) and AML (58.1%; Figure 2). The second lowest 5‐year relative survival rates were observed for East Asian patients with childhood leukemia (88.6% for ALL; 61.8% for AML). For NHPI, NHW, and South Asian patients with childhood leukemia, the 5‐year relative survival rates were similar. For AML, similar trends were observed with lower 5‐year relative survival rates for Southeast and East Asian patients with childhood leukemia and similar rates for NHPI, NHW, and South Asian patients.
FIGURE 2.
Five‐year relative survival rates and 95% confidence intervals for childhood leukemia.
The 5‐year relative survival rates for ALL by age, year of diagnosis, sex, rural residence, radiation, and chemotherapy are shown in Table 2. The survival rates were higher for patients diagnosed at 1–9 years of age for ANHPI and NHW patients than the other age groups. By year of diagnosis, survival rates were improving over the years for most subgroups except for NHPI and East Asian patients. Female patients generally had higher survival rates across the groups except among East Asian patients. Patients residing in rural areas had lower 5‐year relative survival rates for ALL. In terms of cancer treatment, patients receiving radiation and patients not receiving chemotherapy had lower survival rates for ALL.
TABLE 2.
Childhood ALL: Five‐year relative survival for ANHPI and NHW patients, stratified by demographic and clinical characteristics.
| ANHPI | East Asian | Southeast Asian | South Asian | NHPI | NHW | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (n = 1392) | (n = 297) | (n = 356) | (n = 279) | (n = 165) | (n = 7104) | |||||||
| Survival | SE | Survival | SE | Survival | SE | Survival | SE | Survival | SE | Survival | SE | |
| Overall | 89.3 | 0.9 | 88.6 | 2.0 | 85.8 | 2.0 | 91.8 | 1.8 | 89.4 | 2.5 | 91.4 | 0.4 |
| Age | ||||||||||||
| 00 years | 60.6 | 7.3 | 46.9 | 12.9 | 73.0 | 13.5 | 68.8 | 15.2 | 100.0 | 0.0 | 60.3 | 3.8 |
| 01–04 years | 93.8 | 1.1 | 90.9 | 2.7 | 91.7 | 2.4 | 96.2 | 1.7 | 91.1 | 3.5 | 94.9 | 0.4 |
| 05–09 years | 90.3 | 1.7 | 96.8 | 2.3 | 84.9 | 4.0 | 91.4 | 3.7 | 86.5 | 5.2 | 94.0 | 0.6 |
| 10–14 years | 87.5 | 2.4 | 90.6 | 4.5 | 84.6 | 4.8 | 84.5 | 6.5 | 89.1 | 6.0 | 86.7 | 1.1 |
| 15–17 years | 77.1 | 4.3 | 79.8 | 8.2 | 69.1 | 7.8 | 84.6 | 10.0 | 87.5 | 11.7 | 81.7 | 1.7 |
| Year of diagnosis | ||||||||||||
| 2000–2004 | 85.3 | 2.1 | 90.1 | 3.4 | 75.7 | 4.9 | 87.6 | 5.2 | 87.0 | 5.5 | 89.5 | 0.7 |
| 2005–2009 | 88.8 | 1.8 | 86.0 | 4.1 | 84.4 | 3.6 | 94.7 | 3.0 | 92.6 | 5.0 | 91.0 | 0.7 |
| 2010–2014 | 90.6 | 1.6 | 87.8 | 3.9 | 94.9 | 2.5 | 93.6 | 3.1 | 81.3 | 5.6 | 92.4 | 0.7 |
| Sex | ||||||||||||
| Male | 88.2 | 1.3 | 89.6 | 2.5 | 83.3 | 2.9 | 89.9 | 2.7 | 87.3 | 4.0 | 90.8 | 0.5 |
| Female | 90.7 | 1.3 | 87.1 | 3.3 | 88.5 | 2.7 | 94.1 | 2.2 | 91.4 | 3.1 | 92.0 | 0.5 |
| Rural | ||||||||||||
| Urban | 89.6 | 0.9 | 88.7 | 2.0 | 86.0 | 2.0 | 91.7 | 1.8 | 90.9 | 2.6 | 91.7 | 0.4 |
| Rural | 81.3 | 6.0 | 80.0 | 17.9 | 80.0 | 12.7 | 100.0 b | 0.0 | 79.0 | 9.4 | 89.5 | 1.1 |
| Radiation | ||||||||||||
| None | 90.1 | 0.9 | 88.4 | 2.1 | 85.8 | 2.1 | 94.0 | 1.6 | 91.3 | 2.5 | 92.2 | 0.4 |
| Beam radiation | 80.1 | 3.9 | 89.5 | 5.8 | 84.5 | 6.4 | 64.8 | 11.6 | 66.7 | 13.6 | 82.6 | 1.7 |
| Chemotherapy | ||||||||||||
| No/unknown | 55.8 | 10.6 | 57.2 | 18.7 | 20.0 | 17.9 | 100.0 b | 0.0 | a | a | 81.0 | 3.0 |
| Yes | 89.9 | 0.9 | 89.4 | 2.0 | 86.7 | 1.9 | 91.7 | 1.8 | 89.4 | 2.5 | 91.7 | 0.4 |
Abbreviations: ANHPI, Asian, Native Hawaiian, and Pacific Islander; NHPI, Native Hawaiian and Pacific Islander; NHW, non‐Hispanic white.
Too few patients to calculate.
The relative cumulative survival is over 100% and has been adjusted.
For AML, similar patterns of better survival for patients diagnosed between 1 and 9 years of age were observed (Table 3). Over the years, the survival rates have improved, except for the rates in Southeast Asian patients. Female patients had higher AML survival for NHW and NHPI patients and lower survival for all other groups. The survival did not appear to differ for rural patients. AML patients receiving radiation and patients not receiving chemotherapy had lower 5‐year relative survival rates, similar to the patterns observed for ALL patients.
TABLE 3.
Childhood AML: Five‐year relative survival for ANHPI and NHW patients, stratified by demographic and clinical characteristics.
| ANHPI | East Asian | Southeast Asian | South Asian | NHPI | NHW | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (n = 387) | (n = 89) | (n = 126) | (n = 40) | (n = 55) | (n = 1476) | |||||||
| Survival | SE | Survival | SE | Survival | SE | Survival | SE | Survival | SE | Survival | SE | |
| Overall | 65.1 | 2.6 | 61.8 | 5.6 | 58.1 | 4.7 | 65.4 | 8.4 | 67.5 | 6.8 | 67.3 | 1.3 |
| Age | ||||||||||||
| 00 years | 63.0 | 7.3 | a | a | 51.6 | 14.5 | 50.2 | 25.1 | 69.9 | 12.9 | 68.8 | 3.5 |
| 01–04 years | 70.5 | 4.8 | 63.6 | 11.0 | 65.7 | 8.5 | 91.7 | 8.0 | 57.8 | 13.3 | 70.2 | 2.5 |
| 05–09 years | 67.8 | 5.8 | 57.8 | 12.2 | 68.2 | 9.9 | 62.5 | 17.1 | 80.0 | 17.9 | 68.0 | 3.5 |
| 10–14 years | 61.9 | 5.4 | 58.7 | 12.3 | 51.7 | 8.8 | 64.8 | 16.5 | 58.4 | 14.2 | 65.9 | 2.8 |
| 15–19 years | 61.4 | 6.8 | 70.1 | 10.3 | 50.6 | 13.4 | 25.0 | 21.7 | 100.0 b | 0.0 | 63.8 | 2.9 |
| Year of diagnosis | ||||||||||||
| 2000–2004 | 54.2 | 5.9 | 44.5 | 11.7 | 56.1 | 9.9 | 54.6 | 15.0 | 54.6 | 15.0 | 62.6 | 2.5 |
| 2005–2009 | 62.2 | 5.7 | 61.2 | 11.5 | 61.6 | 9.5 | 60.1 | 21.9 | 63.8 | 14.5 | 70.5 | 2.5 |
| 2010–2014 | 68.0 | 4.8 | 63.1 | 10.4 | 55.2 | 8.4 | 77.8 | 13.9 | 81.9 | 11.6 | 68.3 | 2.6 |
| Sex | ||||||||||||
| Male | 69.1 | 3.5 | 69.5 | 7.4 | 58.9 | 6.9 | 80.3 | 8.0 | 56.5 | 9.2 | 65.4 | 1.8 |
| Female | 60.4 | 4.0 | 54.4 | 8.1 | 57.1 | 6.5 | 34.1 | 15.0 | 84.2 | 8.6 | 69.5 | 1.9 |
| Rural | ||||||||||||
| Urban | 65.0 | 2.7 | 61.3 | 5.6 | 57.8 | 4.7 | 65.4 | 8.4 | 67.6 | 7.5 | 67.8 | 1.4 |
| Rural | 66.6 | 13.9 | 100 b | 0.0 | a | a | a | a | 66.8 | 15.8 | 64.0 | 3.6 |
| Radiation | ||||||||||||
| None | 65.6 | 2.8 | 64.8 | 5.7 | 57.0 | 5.0 | 66.4 | 8.6 | 66.7 | 7.2 | 68.5 | 1.4 |
| Beam radiation | 57.7 | 9.7 | 20.0 | 17.9 | 68.6 | 15.2 | 50.0 | 35.4 | 75.0 | 21.7 | 55.8 | 4.7 |
| Chemotherapy | ||||||||||||
| No/unknown | 62.6 | 12.1 | 0.0 | 0.0 | 41.2 | 17.6 | a | a | 100.0 b | 0.0 | 47.1 | 6.0 |
| Yes | 65.1 | 2.7 | 62.5 | 5.6 | 59.4 | 4.9 | 65.7 | 8.6 | 65.5 | 7.1 | 68.5 | 1.3 |
Too few patients to calculate.
The relative cumulative survival is over 100% and has been adjusted.
For cause of death, 84.8% of NHW patients and 84.6% of ANHPI patients with childhood leukemia who died had cancer as the cause of death. The risk of all‐cause death for childhood leukemia was 29% higher for ANHPI patients than for NHW patients (Table 4). This increase in the risk of death was particularly high for East Asian (42%) and Southeast Asian patients (50%). When the Southeast Asian patients with childhood leukemia were taken as the reference, the NHW patients had a lower risk of death, but differences with the other ANHPI subgroups were not identified. When separated into ALL and AML patients, the risk of death was elevated only for childhood ALL patients.
TABLE 4.
Risk of all cause death by ANHPI race and ethnicity.
| Childhood Leukemia | Childhood ALL | Childhood AML | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Total | Deaths | HR a | 95%CI | Total | Deaths | HR a | 95%CI | Total | Deaths | HR a | 95%CI | |
| NHW | 8772 | 1260 | 1.00 | 7104 | 720 | 1.00 | 1476 | 488 | 1.00 | |||
| ANHPI | 1811 | 312 | 1.29 | (1.13, 1.46) | 1392 | 173 | 1.36 | (1.15, 1.62) | 387 | 126 | 1.05 | (0.86, 1.29) |
| East Asian | 394 | 80 | 1.42 | (1.13, 1.79) | 297 | 44 | 1.31 | (0.96. 1.79) | 89 | 31 | 1.18 | (0.81, 1.70) |
| NHPI | 223 | 41 | 1.25 | (0.91, 1.72) | 165 | 21 | 1.58 | (1.02, 2.45) | 55 | 19 | 1.10 | (0.69, 1.75) |
| South Asian | 328 | 46 | 1.33 | (0.99, 1.80) | 279 | 30 | 1.22 | (0.84, 1.78) | 40 | 12 | 1.08 | (0.61, 1.93) |
| Southeast Asian | 490 | 108 | 1.50 | (1.23, 1.83) | 356 | 56 | 1.66 | (1.26, 2.18) | 126 | 49 | 1.23 | (0.91, 1.65) |
| Southeast Asian | 490 | 108 | 1.00 | 356 | 56 | 1.00 | 126 | 49 | 1.00 | |||
| NHW | 8772 | 1260 | 0.67 | (0.55, 0.81) | 7104 | 720 | 0.60 | (0.46, 0.79) | 1476 | 488 | 0.82 | (0.61, 1.10) |
| East Asian | 394 | 80 | 0.95 | (0.71, 1.27) | 297 | 44 | 0.79 | (0.53, 1.18) | 89 | 31 | 0.96 | (0.61, 1.51) |
| NHPI | 223 | 41 | 0.83 | (0.58, 1.20) | 165 | 21 | 0.95 | (0.58. 1.58) | 55 | 19 | 0.90 | (0.53, 1.54) |
| South Asian | 328 | 46 | 0.89 | (0.63, 1.26) | 279 | 30 | 0.74 | (0.47, 1.16) | 40 | 12 | 0.88 | (0.47, 1.67) |
Note: Estimates in bold are statistically significant.
Adjusted for age group, sex, income, rural residence, type of leukemia, radiation, and chemotherapy.
For prognostic factors among ANHPI patients with childhood leukemia, lower risks of death were observed for more recent years of diagnosis and for treatment with chemotherapy (Table 5). For East Asian patients with childhood leukemia, fewer prognostic factors were identified, though high risks were observed for the youngest age at diagnosis and for AML and unspecified types compared to ALL. For Southeast Asian patients, a lower risk of death was observed for more recent years of diagnosis, and higher risks were observed for the youngest and oldest age at diagnosis. For South Asian patients with leukemia, the lower risks of death were not observed for the more recent years of diagnosis, though increased risks of death were observed for the youngest and oldest ages of diagnosis. For the NHPI patients with leukemia, the only increased risk of death observed was for AML compared to ALL.
TABLE 5.
Prognostic factors for all cause death among ANHPI and NHW childhood leukemia patients.
| ANHPI | East Asian | Southeast Asian | South Asian | NHPI | NHW | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (n = 1971) | (n = 435) | (n = 543) | (n = 351) | (n = 236) | (n = 8772) | |||||||
| HR | 95%CI | HR | 95%CI | HR | 95%CI | HR | 95%CI | HR | 95%CI | HR | 95%CI | |
| Years of diagnosis | ||||||||||||
| 2000–2004 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | ||||||
| 2005–2009 | 0.61 | (0.45, 0.83) | 0.57 | (0.31, 1.05) | 0.67 | (0.41, 1.10) | 0.57 | (0.23, 1.42) | 0.05 | (0.18, 1.40) | 0.80 | (0.69, 0.92) |
| 2010–2014 | 0.58 | (0.43, 0.79) | 0.78 | (0.43, 1.40) | 0.56 | (0.32, 0.98) | 0.49 | (0.19, 1.28) | 0.90 | (0.40, 2.06) | 0.72 | (0.61, 0.84) |
| Age | ||||||||||||
| 00 years | 2.91 | (1.94, 4.37) | 5.05 | (2.43, 10.5) | 2.30 | (1.08, 4.94) | 5.76 | (2.04, 16.3) | 1.61 | (0.52, 5.00) | 3.08 | (2.49, 3.81) |
| 01–04 years | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | ||||||
| 05–09 years | 1.34 | (0.96, 1.88) | 0.91 | (0.46, 1.78) | 1.28 | (0.71, 2.31) | 2.05 | (0.87, 4.84) | 1.28 | (0.49, 3.35) | 1.16 | (0.97, 1.40) |
| 10–14 years | 1.86 | (1.33, 2.71) | 1.31 | (0.63, 2.74) | 2.08 | (1.19, 3.65) | 1.78 | (0.71, 4.47) | 1.62 | (0.71, 3.68) | 2.16 | (1.83, 2.56) |
| 15–18 years | 2.21 | (1.57, 3.13) | 1.05 | (0.53, 2.11) | 2.51 | (1.40, 4.48) | 4.22 | (1.59, 11.23) | 0.71 | (0.20, 2.59) | 2.67 | (2.26, 3.15) |
| Sex | ||||||||||||
| Male | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | ||||||
| Female | 1.04 | (0.83, 1.30) | 1.33 | (0.83, 2.15) | 0.79 | (0.54. 1.17) | 1.24 | (0.67, 2.32) | 0.56 | (0.28, 1.14) | 0.89 | (0.80, 1.00) |
| Income | ||||||||||||
| <$59,999 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | ||||||
| $60,000–$64,999 | 1.86 | (1.22, 2.83) | 2.24 | (0.65, 7.81) | 1.93 | (1.09, 3.43) | 1.65 | (0.37, 7.50) | 1.80 | (0.38, 8.46) | 1.24 | (1.05, 1.47) |
| $65,000–$69,999 | 1.34 | (0.87, 2.06) | 1.65 | (0.56, 4.90) | 0.92 | (0.40, 2.14) | 1.02 | (0.31, 3.39) | 1.72 | (0.51, 5.77) | 1.10 | (0.89, 1.36) |
| $70,000–$74,999 | 0.98 | (0.72, 1.34) | 0.60 | (0.32, 1.10) | 1.18 | (0.69, 2.00) | 2.08 | (0.92, 4.71) | 0.67 | (0.17, 2.65) | 1.22 | (1.03, 1.45) |
| $75,000+ | 0.79 | (0.49, 1.25) | 1.44 | (0.58, 3.58) | 0.60 | (0.27, 1.32) | 1.26 | (0.40, 3.96) | 0.60 | (0.08, 4.70) | 1.22 | (1.01, 1.47) |
| Type | ||||||||||||
| Ia Lymphoid leukemias | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | ||||||
| Ib Acute myeloid leukemias | 2.51 | (1.96, 3.20) | 4.39 | (2.61, 7.39) | 2.37 | (1.57, 3.56) | 2.92 | (1.38, 6.19) | 3.12 | (1.51, 6.44) | 2.76 | (2.44, 3.13) |
| Ie Unspecified and other specified leukemias | 3.15 | (1.77, 5.59) | 3.94 | (1.40, 11.1) | 1.26 | (0.37, 4.35) | 6.08 | (2.03, 18.18) | 6.93 | (0.80, 60.2) | 2.00 | (1.48, 2.70) |
| Rural | ||||||||||||
| Urban | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | ||||||
| Rural | 0.74 | (0.41, 1.34) | 0.27 | (0.03, 2.11) | 0.69 | (0.16, 2.90) | ‐‐ | 0.83 | (0.24, 2.89) | 1.21 | (1.02, 1.45) | |
| Radiation | ||||||||||||
| None | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | ||||||
| Beam radiation | 1.45 | (1.04, 2.02) | 1.49 | (0.75, 2.95) | 0.80 | (0.40, 1.58) | 3.6 | (1.61, 8.06) | 1.83 | (0.73, 4.58) | 1.55 | (1.32, 1.82) |
| Chemotherapy | ||||||||||||
| No/unknown | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | ||||||
| Yes | 0.48 | (0.30, 0.76) | 0.10 | (0.04, 0.27) | 0.42 | (0.20, 0.85) | 0.73 | (0.17, 3.55) | ‐‐ | 0.48 | (0.39, 0.60) | |
Note: All variables are adjusted for in the table. Estimates in bold are statistically significant.
The youngest and oldest age groups of diagnosis had higher risks of death for most subgroups. The lower risk of death for female patients was observed only for NHW patients. Increased risk of death for the highest income group and for rural residence was only observed for NHW patients. Compared to lymphoid subtypes, the acute myeloid and unspecified subtypes were associated with higher risks of death for NHW patients. NHW patients who had radiation had higher risks of death, whereas NHW patients who had chemotherapy had lower risks of death.
4. DISCUSSION
We observed lower 5‐year relative survival rates for Southeast Asian and East Asian compared to NHW patients with childhood leukemia for ALL. The survival rates were higher for patients diagnosed at 1–9 years of age, for more recent years of diagnosis, and for patients residing in urban areas. The risk of death from childhood leukemia was 42% higher for East Asian patients and 50% higher for Southeast Asian patients compared to NHW patients. For ALL, the risk of death was 66% higher for Southeast Asian patients and 58% higher for NHPI patients compared to NHW patients. In the Cox proportional hazards model adjusting for various factors, lower risks of death were observed for more recent years of diagnosis for NHW and Southeast Asian patients with childhood leukemia.
Our results are similar to previous studies that report lower 5‐year survival for East Asian, Filipino, and Vietnamese patients compared to NHW patients with SEER data for childhood ALL patients diagnosed from 1988 to 2008. Our study included 12 more years of more recently diagnosed childhood ALL cases, and the disparity still exists. The reasons for the higher risks of death for East Asian and Southeast Asian patients with childhood ALL are not clear. Our prognostic factor analysis for childhood leukemia showed a very high risk of death for East Asian patients diagnosed at <12 months of age and for AML and unspecified leukemia. For Southeast Asian patients with childhood leukemia, the prognostic factor analysis showed higher risks of death for the youngest and oldest age groups of diagnosis and for AML. For NHPI patients, the AML subtype was the only prognostic factor identified, possibly due to lower statistical power to detect prognostic factors with the small patient numbers.
The higher risks of death for the youngest and oldest age groups for childhood leukemia were expected. These age groups are considered high risk by the Children's Oncology Group. 10 For patients receiving radiation therapy, we observed higher risks of death as expected since these patients are likely to have had involvement of the central nervous system. 10 The lower risk of death for childhood leukemia patients diagnosed in more recent years suggests that 5‐year relative survival is improving for Southeast Asian and NHW patients. However, it is a concern that the lower risk for more recent years was not observed for East Asian, South Asian, and NHPI patients. In terms of socioeconomic status, higher risks of death for the highest income category group were observed only in NHW patients. Since these are census‐level socioeconomic variables, further investigation into the role of individual level socioeconomic status is needed for ANHPI patients with childhood leukemia.
The strengths of the study include the availability of the SEER data that allowed the disaggregation of the childhood leukemia patients into finer subgroups of ANHPI race and ethnicity, inclusion of more recent diagnosis years, and assessment of prognostic factors for the risk of death. Using population‐based high‐quality SEER cancer registry data is also a strength. With the addition of more state cancer registries in recent years, we were able to include 17 cancer registries.
Limitations of the study include the lack of some key clinical variables such as bone marrow transplantation and high‐risk and low‐risk classification based on the combination of blood counts and age at diagnosis. Having the clinical variables would have allowed for investigation of other potential differences among the ANHPI patients with childhood leukemia. We also do not have individual‐level information for socioeconomic factors such as income or education, although we used county‐level variables as a proxy. Grouping socioeconomic status at the county level may mask some differences. Some of the analyses still had small sample sizes for the ANHPI subgroups and may have lacked statistical power to detect risks. Finally, the improvement in survival for childhood leukemia during the included study period needs to be considered when interpreting our data, although the trends appear to be similar to the years leading up to our study.
In conclusion, we observed lower 5‐year relative survival rates for Southeast Asian and East Asian patients compared to NHW patients with childhood leukemia for ALL. We investigated prognostic factors for these patient groups and observed very high risks of death for East Asian patients diagnosed at <12 months old. Furthermore, we did not observe lower risks for more recent years of diagnosis for East Asian, South Asian, and NHPI patients. It will be important to continue to monitor trends in survival among ANHPI patients with childhood leukemia. Further studies are needed to elucidate the reasons behind the disparities in survival rates for Southeast Asian and East Asian patients with childhood leukemia. This includes investigating both the socioeconomic and genetic contributions to leukemia risk and clinical responses to different therapeutic modalities.
AUTHOR CONTRIBUTIONS
Mia Hashibe: Conceptualization; data curation; formal analysis; investigation; methodology; project administration; resources; software; supervision; validation; visualization; writing – original draft; writing – review and editing. Kimberly A. Herget: Conceptualization; investigation; methodology; writing – review and editing. Joshua D. Schiffman: Conceptualization; investigation; methodology; writing – review and editing. Vivian Y. Chang: Conceptualization; investigation; methodology; writing – review and editing.
CONFLICT OF INTEREST STATEMENT
No conflicts of interest.
Hashibe M, Herget KA, Schiffman JD, Chang VY. Disparities in childhood leukemia survival for Asian, Native Hawaiian, and Pacific Islanders in the United States. Int J Cancer. 2025;156(8):1563‐1571. doi: 10.1002/ijc.35259
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available at https://seer.cancer.gov/. Further information is available from the corresponding author upon request.
REFERENCES
- 1. Monterroso PS, Li Z, Domingues AM, Sample JM, EL M. Racial/ethnic and socioeconomic disparities in childhood cancer incidence trends in the United States, 2000–2019. J Natl Cancer Inst. 2023;115(12):1576‐1585. doi: 10.1093/jnci/djad148 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Moore KJ, Hubbard AK, Williams LA, Spector LG. Childhood cancer incidence among specific Asian and Pacific islander populations in the United States. Int J Cancer. 2020;147(12):3339‐3348. doi: 10.1002/ijc.33153 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Redaniel MT, Laudico A, Mirasol‐Lumague MR, Alcasabas AP, Pulte D, Brenner H. Geographic and ethnic differences in childhood leukaemia and lymphoma survival: comparisons of Philippine residents, Asian Americans and Caucasians in the United States. Br J Cancer. 2010;103(1):149‐154. doi: 10.1038/sj.bjc.6605703 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Goggins WB, Lo FF. Racial and ethnic disparities in survival of US children with acute lymphoblastic leukemia: evidence from the SEER database 1988–2008. Cancer Causes Control. 2012;23(5):737‐743. doi: 10.1007/s10552-012-9943-8 [DOI] [PubMed] [Google Scholar]
- 5. Abrahão R, Lichtensztajn DY, Ribeiro RC, et al. Racial/ethnic and socioeconomic disparities in survival among children with acute lymphoblastic leukemia in California, 1988–2011: a population‐based observational study. Pediatr Blood Cancer. 2015;62(10):1819‐1825. Epub 20150420. doi: 10.1002/pbc.25544 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Surveillance, Epidemiology, and End Results (SEER) Program (www.seer.cancer.gov) SEER*Stat Database: Incidence—SEER Research Data, 17 Registries, Nov 2022 Sub (2000–2020)—Linked To County Attributes—Time Dependent (1990–2021) Income/Rurality, 1969–2021 Counties, National Cancer Institute, DCCPS, Surveillance Research Program, released April 2023, based on the November 2022 submission.
- 7. https://www.ers.usda.gov/data‐products/rural‐urban‐continuum‐codes.aspx
- 8. https://seer.cancer.gov/help/seerstat/survival‐session/survival‐statistic‐tab/relative‐survival
- 9. https://seer.cancer.gov/help/seerstat/survival‐session/survival‐statistic‐tab/determining‐the‐appropriate‐cancer‐survival‐measure‐for‐a‐specific‐analysis
- 10. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines®) Pediatric Acute Lymphoblastic Leukemia. Version 3.2024—October 31, 2023.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are openly available at https://seer.cancer.gov/. Further information is available from the corresponding author upon request.