Cancer incidence and type of treatment hospital among children, adolescents, and young adults in Japan, 2016–2018

Abstract

Cancer in children, adolescents, and young adults (AYAs) although rare, is the leading disease‐specific cause of death in Japan. This study aims to investigate cancer incidence and type of treatment hospital among children and AYAs in Japan. Cancer incidence data (2016–2018) for those aged 0–39 years were obtained from the Japanese population‐based National Cancer Registry. Cancer types were classified according to the 2017 update of the International Classification of Childhood Cancer (Third Edition), and AYA Site Recode 2020 Revision. Cases were also categorized into three groups: those treated at core hospitals for pediatric cancer treatment (pediatric cancer hospitals [PCHs]), those treated at designated cancer care hospitals, and those treated at nondesignated hospitals. The age‐standardized incidence rate was 166.6 (per million‐person years) for children (age 0–14 years) and 579.0 for AYAs (age 15–39 years) (including all cancers and benign or uncertain‐behavior central nervous system [CNS] tumors). The type of cancer varied with age: hematological malignancies, blastomas, and CNS tumors were common in children under 10 years, malignant bone tumors and soft tissue sarcomas were relatively common in teenagers, and in young adults over 20 years, carcinomas in thyroid, testis, gastrointestinal, female cervix, and breast were common. The proportion of cases treated at PCHs ranged from 20% to 30% for children, 10% or less for AYAs, and differed according to age group and cancer type. Based on this information, the optimal system of cancer care should be discussed.

The type of cancer in children, adolescents, and young adults in Japan varies according to age: hematological malignancies, blastomas, and central nervous system tumors were common in children under 10 years, malignant bone tumors and soft tissue sarcomas were relatively common in teenagers, and in young adults over 20 years, carcinomas were common.

Abbreviations

ASR

age‐standardized incidence rate

AYA

adolescents and young adult

CNS

central nervous system

DCCH

Designated Cancer Care Hospital

DCO

death certificate only

ICCC

International Classification of Childhood Cancer

ICD‐O‐3

International Classification of Diseases for Oncology, 3rd edition

MHLW

Ministry of Health, Labour and Welfare

NCR

National Cancer Registry

PCH

pediatric cancer hospital

P‐DCCH

prefectural Designated Cancer Care Hospital

R‐DCCH

regional Designated Cancer Care Hospital

1. INTRODUCTION

Although cancers are rare in children and AYAs, ¹ they are the leading disease‐specific cause of death in Japan ² and when they do occur, they cause a range of medical, psychological, and social problems. ¹ In contrast to cancers that occur in adults, those that affect this generation are often systemic in nature and are classified according to unique classification systems such as the ICCC ³ , ⁴ or the AYA Site Recode. ⁵ , ⁶ According to a report from the Japanese population‐based NCR that began nationwide in 2016, approximately 2000 children (age 0–14 years) and 20,000 AYAs (age 15–39 years) are newly diagnosed with cancer each year in Japan ⁷ ; however, this report only records incidence according to the International Classification of Diseases, 10th revision (ICD‐10) codes, which mainly define by anatomical site.

The National Cancer Control Act in Japan was established in 2006 and, based on the Basic Plan to Promote Cancer Control Programs, a total of nearly 400 hospitals are now designated as Designated Cancer Care Hospitals (DCCHs) by the MHLW. ⁸ , ⁹ These hospitals have to meet national criteria for the number of cancer patients, physician expertise, and availability of support programs for adult cancer patients. In 2012, the 2nd term Basic Plan to Promote Cancer Control Programs raised the issue of care for children and 15 hospitals in seven areas were designated as core hospitals for pediatric cancer treatment (PCHs) by the MHLW. In 2018, the 3rd term Basic Plan to Promote Cancer Control Programs raised the issue of care for AYA patients and the need to promote a certain degree of centralization and psychosocial support. ¹⁰ , ¹¹ However, due to the variety of cancers that occur and the diversity of patient needs in this generation, ¹⁰ , ¹¹ , ¹² the status of centralized cancer care could vary according to the type of cancer and age group. ¹³ , ¹⁴ Despite this, there is a lack of data showing type of treatment hospital, such as how many children and AYAs receive cancer treatment at the designated cancer hospitals (PCHs or DCCHs) across the country.

This study aims to describe the incidence of cancer in children and AYAs by cancer type in detail and identify the proportion of patients who are treated in designated cancer hospitals in Japan, using the national population‐based cancer registry data.

2. MATERIALS AND METHODS

Cancer incidence data (2016–2018) were obtained by applying for use of the Japanese NCR. The individual cancer records include coded information on sex, age at diagnosis, date of diagnosis, most valid basis of diagnosis, tumor sequence (i.e., the numerical order of occurrence of the neoplasm), topography, morphology, behavior, laterality, and treatment hospital code. When a new cancer patient is treated in a Japanese hospital, the Cancer Registry must be given details of any surgery, chemotherapy, radiotherapy, or other treatment the patient has undergone as well as information on the cancer itself. A “treatment hospital” was defined as a hospital that has submitted treatment information. If more than one hospital had submitted treatment information for the same cancer in the same person, we selected one hospital as the “treatment hospital,” and priority was given according to the order in which treatment was performed; surgery, chemotherapy, and radiation. The tumor topography, morphology, behavior, and basis of diagnosis were coded according to ICD‐O‐3. ¹⁵ , ¹⁶

Cases were restricted to children and AYAs aged 0–39 years, resident in Japan. Using both ICD‐O‐3 topography codes and morphology codes, cases were classified according to the 2017 update of the ICCC (third edition) (ICCC‐3), ⁴ , ¹⁷ and AYA Site Recode 2020 Revision. ⁵ , ⁶ Cancer incidence was calculated for all malignant cancers (ICD‐O‐3, behavior code = 3) and benign (behavior code = 0) or behavior uncertain (behavior uncertain whether benign or malignant, behavior code = 1) CNS tumors, classified by ICCC‐3. Carcinomas in situ (behavior code = 2) were included for the incidence of cancer in AYAs, which was classified according to AYA Site Recode. The main tables and figures display values for the main diagnostic groups of the ICCC‐3. For cancer in AYAs, in addition to the main diagnostic groups of the AYA Site Recode, the subgroups of “7. Gonadal and related tumors” and “9. Carcinomas” are also displayed as selected diagnostic groups.

Based on the designation of cancer care hospitals by the MLHW as of March 2019, we categorized cases into three treatment hospital types: those treated at PCHs (15 hospitals), those treated at DCCHs (391 hospitals excluding PCHs), and nondesignated hospitals, using information on the treatment hospital code. The DCCHs were further divided into prefectural DCCH including national cancer center hospitals (P‐DCCH, 46 hospitals) and regional DCCH (R‐DCCH, 345 hospitals).

The population data by 5‐year age group for person‐years at risk were downloaded from the portal site of official statistics of Japan. ¹⁸

All incidence rates are expressed per million person‐years of the relevant population at risk. Age‐specific incidence rates were calculated for 5‐year age groups. Age‐standardized incidence rates for the age range 0–14 years or 15–39 years were calculated using the direct standardization method with Segi's world standard population weights, ¹⁹ as in previous publications. ¹ , ²⁰ , ²¹ The 95% confidence intervals were calculated by standard methods. ²² , ²³

The average annual number of cases is simply the number of cases divided by 3 years.

The proportion of patients treated at PCHs or DCCHs was calculated by dividing cases in each category of treatment hospital by the total number of cases with treatment information. If the number of cases was between one and nine, they were labeled 1–9, according to the instructions of the review committee of the Japanese NCR.

All analyses were carried out using Stata 16 MP (StataCorp).

3. RESULTS

A total of 7531 children (age 0–14 years) and 69,362 AYAs (age 15–39 years) were diagnosed with cancer, benign, or uncertain‐behavior CNS tumors between 2016 and 2018 in Japan (Table 1). The average annual number of cases was 2510 in children and 23,121 in AYAs. The quality indicators of the data were as follows: proportion of DCO and morphologically verified cases were 0.2% and 87.5% for children, and 0.3% and 92.8% for AYAs, respectively; the proportion of DCO reduced from 2016 (0.5%) to 2018 (0.1%) in both children and AYAs.

TABLE 1.

Overview of data included in the analysis of cancer incidence among children, adolescents, and young adults in Japan, 2016–2018.

	Age group (years)
	0–14			15–39
	Cases	DCO (%)	MV (%)	Cases	DCO (%)	MV (%)
Total (2016–2018)
All cancers (malignant only, excluding in situ)	6565	0.2	92.8	63,285	0.3	97.6
All cancers (including benign or behavior‐uncertain CNS tumors, classified by ICCC‐3) a	7531	0.2	87.5	69,362	0.3	92.8
All cancers (including benign or behavior‐uncertain CNS tumors and in situ)	8100	0.2	87.9	111,149	0.2	95.5
All cancers (including benign or behavior‐uncertain tumors and in situ)	8133	0.2	87.9	113,141	0.2	95.6
Average annual number of cases a	2510	–	–	23,121	–	–
Year of diagnosis a
2016	2470	0.5	87.9	23,639	0.5	93.0
2017	2595	0.2	87.8	22,975	0.2	93.0
2018	2466	0.1	86.7	22,748	0.1	92.3
Main diagnostic group of ICCC‐3 a
I. Leukemias, myeloproliferative and myelodysplastic diseases	2433	0.2	99.2	4571	0.4	99.3
II. Lymphomas and reticuloendothelial neoplasms	737	0.1	98.4	3751	0.2	98.5
III. CNS and miscellaneous intracranial and intraspinal neoplasms	1890	0.2	61.5	8201	0.6	52.4
IV. Neuroblastoma and other peripheral nervous cell tumors	458	0.0	92.4	159	0.0	98.7
V. Retinoblastoma	223	0.0	69.1	18	0.0	94.4
VI. Renal tumors	170	1.2	97.1	1269	0.1	95.4
VII. Hepatic tumors	187	0.0	93.6	473	0.4	62.6
VIII. Malignant bone tumors	271	0.4	94.1	640	1.1	95.9
IX. Soft tissue and other extraosseous sarcomas	307	0.0	99.0	1985	0.0	99.1
X. Germ cell tumors, trophoblastic tumors and neoplasms of gonads	536	0.0	99.6	6167	0.1	98.0
XI. Other malignant epithelial neoplasms and malignant melanomas	254	0.0	99.6	40,741	0.0	100.0
XII. Other and unspecified malignant neoplasms	65	7.7	46.2	1387	8.2	56.2

Abbreviations: CNS, central nervous system; DCO, death certificate only; ICCC, International Classification of Childhood Cancer; ICCC‐3, ICCC 3rd edition, update 2017; MV, microscopically verified.

^a All cancers, benign or behavior‐uncertain CNS tumors classified by ICCC‐3 were included.

Table 2 shows the average annual number of cases, and ASRs by ICCC‐3 main diagnostic group for both sexes combined. The ASR was 166.6 for children and 579.0 for AYAs (per million person‐years). Within the ICCC‐3 main diagnostic group, ASRs of “I. Leukemias, myeloproliferative and myelodysplastic diseases” and “III. CNS and miscellaneous intracranial and intraspinal neoplasms” were high in children (ASR = 54.9 and 40.3, respectively), while ASRs of “XI. Other malignant epithelial neoplasms and malignant melanomas” were high in AYAs (ASR = 318.7).

TABLE 2.

Average annual number of cancer cases and age‐standardized incidence rates among children, adolescents, and young adults by International Classification of Childhood Cancer, 3rd edition, update 2017 (ICCC‐3) in Japan, 2016–2018, both sexes.

	Age group (years)
	0–14				15–39
	Average annual number of cases	ASR b (per million person‐years)	95% CI		ASR b (per million person‐years)	95% CI
All cancers a	2510	166.6	162.8	170.4	579.0	574.6	583.5
Main diagnostic group of ICCC‐3 a
I. Leukemias, myeloproliferative and myelodysplastic diseases	811	54.9	52.7	57.1	43.5	42.2	44.8
II. Lymphomas and reticuloendothelial neoplasms	246	15.8	14.6	16.9	34.3	33.1	35.4
III. CNS and miscellaneous intracranial and intraspinal neoplasms	630	40.3	38.5	42.1	75.4	73.7	77.1
IV. Neuroblastoma and other peripheral nervous cell tumors	153	11.5	10.5	12.6	1.5	1.2	1.7
V. Retinoblastoma	74	5.8	5.0	6.5	0.2	0.1	0.3
VI. Renal tumors	57	4.2	3.6	4.9	9.7	9.1	10.2
VII. Hepatic tumors	62	4.6	3.9	5.3	3.8	3.4	4.1
VIII. Malignant bone tumors	90	5.2	4.6	5.8	6.7	6.2	7.3
IX. Soft tissue and other extraosseous sarcomas	102	6.5	5.8	7.3	18.1	17.3	19.0
X. Germ cell tumors, trophoblastic tumors, and neoplasms of gonads	179	11.2	10.2	12.2	56.0	54.5	57.4
XI. Other malignant epithelial neoplasms and malignant melanomas	85	5.0	4.4	5.7	318.7	315.6	321.9
XII. Other and unspecified malignant neoplasms	22	1.5	1.2	1.9	11.1	10.5	11.8

Abbreviations: ASR, age‐standardized incidence rates; CI, confidence interval; CNS, central nervous system.

^a All cancers, benign, or behavior‐uncertain CNS tumors classified by ICCC‐3 were included.

^b Segi's standard population was used for calculation of ASRs.

Figure 1 shows the distribution of ICCC‐3 main diagnostic group by 5‐year age group in both sexes. For age 0–4 and 5–9 years, the most frequent cancer was “I. Leukemias, myeloproliferative and myelodysplastic diseases”, followed by “III. CNS and miscellaneous intracranial and intraspinal neoplasms”, and “IV. Neuroblastoma and other peripheral nervous cell tumors” (age 0–4 years) or “II. Lymphomas and reticuloendothelial neoplasms” (age 5–9 years). For age 10–14 and 15–19 years, the most frequent cancer was “III. CNS and miscellaneous intracranial and intraspinal neoplasms” followed by “I. Leukemias, myeloproliferative and myelodysplastic diseases”, and “X. Germ cell tumors, trophoblastic tumors, and neoplasms of gonads” (age 10–14 years) or “XI. Other malignant epithelial neoplasms and malignant melanomas” (age 15–19 years). “VIII. Malignant bone tumors” and “IX. Soft tissue and other extraosseous sarcomas” accounted for a higher proportion of all cancers in the 10–19 years age group than in the other age groups. For age 20–39 years, the most frequent cancer was “XI. Other malignant epithelial neoplasms and malignant melanomas”. Age‐specific incidence rates by 5‐year age group, total number of cases, and ASRs by ICCC‐3 in detail, for both sexes and each sex, are shown in Tables S1–S3. The ASR was higher for male patients than female patients among children (ASR = 179.1 vs. 153.5) (Tables S2 and S3) and “I. Leukemias, myeloproliferative and myelodysplastic diseases” (ASR = 60.7 vs. 48.9), “II. Lymphomas and reticuloendothelial neoplasms” (ASR = 19.6 vs. 11.8), and “IV. Neuroblastoma and other peripheral nervous cell tumors” (ASR = 13.1 vs. 9.9) contributed to difference by sex.

FIGURE 1.

Proportional distribution of cancer type by age group, age 0–39 years, 2016–2018, Japan. Cancers were classified by the International Classification of Childhood Cancer, volume 3, updated 2017. Source: Japanese population‐based National Cancer Registry. CNS, central nervous system.

Figure 2 shows the age‐specific incidence rates for ICCC‐3 main diagnostic group excluding “XI. Other malignant epithelial neoplasms and malignant melanomas” and “XII. Other and unspecified malignant neoplasms” by 5‐year age group. The age‐specific incidence rate of “I. Leukemias, myeloproliferative and myelodysplastic diseases” was highest in children aged 0–4 years, decreased in the 5–19 years age group, and then became slightly higher with age. The age‐specific incidence rates of “IV. Neuroblastoma and other peripheral nervous cell tumors” and “V. Retinoblastoma” were highest in children aged 0–4 years and lower thereafter. The age‐specific incidence rates of “VI. Renal tumors” and “VII. Hepatic tumors” were high in children aged 0–4 and AYAs aged over 30 years. The age‐specific incidence rate of “VIII. Malignant bone tumors” was higher in the 10–19 years age group. The age‐specific incidence rates of other cancer types increased with age after 5–9 years.

FIGURE 2.

Age‐specific incidence rates by cancer type, age 0–39 years, 2016–2018, Japan. Cancers were classified by the International Classification of Childhood Cancer, volume 3, updated 2017. Classifications XI. Other malignant epithelial neoplasms and malignant melanomas and XII. Other and unspecified malignant neoplasms were excluded. Source: Japanese population‐based National Cancer Registry. CNS, central nervous system.

Focusing on AYAs aged 15–39 years, Table 3 shows the average annual number of cases, and ASRs by main and selected diagnostic groups of AYA Site Recode for each sex and both sexes. Information on all diagnostic groups is shown in Tables S4–S6. Among this generation, females have a greater ASR than males (ASRs of all cancers including benign or behavior‐uncertain CNS tumors, and in situ = 1423.6 in female patients vs. 435.7 in male patients). For male patients aged 15–39 years, the ASR of “9. Carcinomas” (ASR = 149.7), especially “9.3 Carcinoma of gastrointestinal tract” (ASR = 70.3) was the highest, followed by “7. Gonadal and related tumors” (ASR = 64.4) including “7.1 Testis” (ASR = 52.3). For female patients aged 15–39 years, the ASR of cancer “A. In situ” was the highest (ASR = 643.9), followed by “9. Carcinomas” (ASR = 514.2) such as “9.6 Carcinoma of breast” and “9.1 Thyroid carcinoma” (ASR = 176.2 and 107.7, respectively). Figure 3 shows the distribution of cancer types in AYAs by 5‐year age group and sex, excluding cancer “A. In situ”. The most common cancers in male patients were “3. CNS and other intracranial and intraspinal neoplasms” for those aged 15–19 years and “9. Carcinomas” for those aged 20–39 years. Among selected diagnostic groups, cancers in “7.1 Testis” (age 20–29 years) and “9.3 Carcinoma of gastrointestinal tract” (age 30–39 years) were the most common in male patients. The most common cancers in female patients were “3. CNS and other intracranial and intraspinal neoplasms” for those aged 15–19 years and “9. Carcinomas” for those aged 20–39 years. Among the selected diagnostic groups, “9.1 Thyroid carcinoma” (age 20–29 years), “9.6 Carcinoma of breast”, and “9.7.1 Carcinoma of uterine cervix” (age 30–39 years) were the most common in female patients.

TABLE 3.

Average annual number of cancer cases and age‐standardized incidence rates among adolescents and young adults (age 15–39 years) by adolescent and young adult (AYA) Site Recode in Japan, 2016–2018, each sex and both sexes.

	Male patients				Female patients				Both sexes
	Average annual number of cases	ASR b (per million person‐years)	95% CI		Average annual number of cases	ASR b (per million person‐years)	95% CI		ASR b (per million person‐years)	95% CI
All cancers (malignant only, excluding in situ)	7045	359.3	354.2	364.3	14,049	692.5	685.7	699.3	523.2	519.0	527.4
All cancers (including benign or behavior‐uncertain CNS tumors and in situ) a	8544	435.7	430.2	441.2	28,504	1423.6	1413.8	1433.4	920.8	915.2	926.4
Main and selected diagnostic group of AYA Site Recode a
1. Leukemias and related disorders	867	48.8	46.8	50.7	657	37.9	36.2	39.7	43.5	42.2	44.8
2. Lymphomas	678	36.7	35.1	38.4	572	31.7	30.2	33.3	34.3	33.1	35.4
3. CNS and other intracranial and intraspinal neoplasms	1136	61.3	59.1	63.4	1619	91.5	88.8	94.1	76.1	74.4	77.7
4. Sarcomas	438	24.6	23.2	26.0	384	22.0	20.7	23.3	23.3	22.4	24.3
5. Blood and lymphatic vessel tumors	136	7.3	6.6	8.1	115	6.5	5.8	7.3	6.9	6.4	7.4
6. Nerve sheath tumors	181	9.5	8.7	10.4	170	9.2	8.4	10.1	9.4	8.8	10.0
7. Gonadal and related tumors	1194	64.4	62.2	66.5	824	45.3	43.4	47.1	55	53.6	56.5
7.1 Testis	1006	52.3	50.4	54.2	1–9	0.1	0.0	0.1	26.7	25.7	27.7
7.2 Ovary	0	0.0	0.0	0.0	762	41.6	39.8	43.3	20.4	19.5	21.2
7.3 Germ cell and trophoblastic – CNS	99	6.8	6.0	7.5	24	1.7	1.3	2.0	4.3	3.8	4.7
8. Melanoma – malignant	56	2.8	2.4	3.2	67	3.4	2.9	3.9	3.1	2.8	3.4
9. Carcinomas	3200	149.7	146.6	152.7	10,828	514.2	508.5	520.0	328.8	325.6	332.0
9.1 Thyroid carcinoma	504	25.5	24.2	26.8	2021	107.7	104.9	110.5	65.7	64.2	67.3
9.2 Other carcinoma of head and neck	371	18.3	17.2	19.4	295	15.4	14.3	16.4	16.9	16.1	17.6
9.3 Carcinoma of gastrointestinal tract	1543	70.3	68.2	72.4	1393	65.7	63.7	67.7	68.1	66.6	69.5
9.4 Carcinoma of lung, bronchus, and trachea	235	10.6	9.8	11.4	240	11.2	10.4	12.0	10.9	10.3	11.5
9.6 Carcinoma of breast	1–9	0.2	0.1	0.4	3926	176.2	173.0	179.5	86.8	85.3	88.4
9.7.1 Carcinoma of uterine cervix	0	0.0	0.0	0.0	1892	87.8	85.4	90.1	43.1	42.0	44.3
9.7.2 Corpus uteri	0	0.0	0.0	0.0	682	31.7	30.3	33.1	15.6	14.9	16.3
9.8 Carcinoma of urinary tract	329	14.6	13.7	15.5	147	7.0	6.3	7.6	10.8	10.3	11.4
10. Miscellaneous specified neoplasms	43	2.5	2.0	2.9	62	3.3	2.8	3.8	2.9	2.6	3.2
11. Unspecified malignant neoplasms except CNS	136	6.7	6.0	7.3	300	14.4	13.5	15.4	10.5	9.9	11.1
A. In situ	479	21.5	20.4	22.7	12,905	643.9	637.4	650.5	327	323.7	330.2

Abbreviations: ASR, age‐standardized incidence rate; CI, confidence interval; CNS, central nervous system.

^a All cancers, benign or behavior‐uncertain CNS tumors, and carcinomas in situ were included.

^b Segi's standard population was used for calculation of ASRs. Numbers between 1 and 9 are displayed as “1–9”.

FIGURE 3.

Proportional distribution of cancer type by age group and sex, age 15–39 years, 2016–2018, Japan. Cancers were classified by adolescent and young adult (AYA) Site Recode 2020 Revision. Source: Japanese population‐based National Cancer Registry. CNS, central nervous system.

Out of a total of 76,893 cases, 63,660 cases (83.0%) had treatment information. Table 4 shows the number and proportion of cases treated at PCHs, DCCHs (P‐DCCHs or R‐DCCHs), and nondesignated hospitals by 5‐year age group, year of diagnosis, and diagnostic group for children and AYAs. The proportion of cases treated at PCHs decreased with age, ranging from 25.1% to 33.8% for children, and 4.1% to 10.3% for AYAs. For both children and AYAs, there was no change in the proportion of patients treated at PCHs or DCCHs by year of diagnosis. The proportion of children treated at PCHs was 40% or more in children with “VII. Hepatic tumors” (47.2%), “V. Retinoblastoma” (42.2%), and “IV. Neuroblastoma and other peripheral nervous cell tumors” (41.9%). The proportion of cases treated at P‐DCCHs was 16.5%–20.1% in each age group. By cancer type, it was higher for “VIII. Malignant bone tumors” (41.6%) in children and for “4. Sarcomas” (33.1%) in AYAs. The proportion of AYAs treated at DCCHs (P‐DCCHs or R‐DCCHs) was 63.7% in total, decreased with age and varied by diagnostic group, from 50.5% in “9.1 Thyroid carcinoma” to 74.9% in “2. Lymphomas”.

TABLE 4.

Number and proportion of cases treated at the pediatric cancer hospital (PCH) or designated cancer care hospital (DCCH) among children, adolescents and young adults, 2016–2018, Japan.

	PCH (15)		DCCH (391)				Nondesignated
			P‐DCCH (46)		R‐DCCH (345)
	Cases	%	Cases	%	Cases	%	Cases	%
Age group (years)
0–4	930	33.8	477	17.3	970	35.2	376	13.7
5–9	468	30.0	259	16.6	600	38.5	231	14.8
10–14	465	25.1	352	19.0	731	39.5	303	16.4
15–19	266	10.3	520	20.1	1249	48.2	555	21.4
20–24	236	5.8	702	17.3	1956	48.1	1175	28.9
25–29	377	5.1	1235	16.8	3499	47.7	2232	30.4
30–34	723	4.8	2545	16.9	7049	46.9	4705	31.3
35–39	1167	4.1	4699	16.5	13,174	46.3	9434	33.1
Children (age 0–14 years)	1863	30.2	1088	17.7	2301	37.3	910	14.8
Year of diagnosis
2016	577	28.7	354	17.6	741	36.9	335	16.7
2017	669	31.0	383	17.8	818	37.9	287	13.3
2018	617	30.9	351	17.6	742	37.1	288	14.4
Main diagnostic group of ICCC‐3 a
I. Leukemias, myeloproliferative and myelodysplastic diseases	554	25.7	333	15.4	892	41.3	380	17.6
II. Lymphomas and reticuloendothelial neoplasms	208	35.4	114	19.4	203	34.5	63	10.7
III. CNS and miscellaneous intracranial and intraspinal neoplasms	400	32.8	188	15.4	448	36.8	183	15.0
IV. Neuroblastoma and other peripheral nervous cell tumors	161	41.9	60	15.6	122	31.8	41	10.7
V. Retinoblastoma	87	42.2	78	37.9	30–39	‐	1–9	‐
VI. Renal tumors	49	30.2	20	12.3	65	40.1	28	17.3
VII. Hepatic tumors	75	47.2	23	14.5	49	30.8	12	7.5
VIII. Malignant bone tumors	51	20.8	102	41.6	82	33.5	10	4.1
IX. Soft tissue and other extraosseous sarcomas	77	28.9	59	22.2	95	35.7	35	13.2
X. Germ cell tumors, trophoblastic tumors, and neoplasms of gonads	162	31.7	67	13.1	203	39.7	79	15.5
XI. Other malignant epithelial neoplasms and malignant melanomas	32	14.0	39	17.0	92	40.2	66	28.8
AYAs (age 15–39 years)	2769	4.8	9701	16.9	26,927	46.8	18,101	31.5
Year of diagnosis
2016	912	4.7	3333	17.1	9131	46.9	6104	31.3
2017	930	4.9	3202	16.7	8950	46.7	6073	31.7
2018	927	4.9	3166	16.8	8846	46.9	5924	31.4
Main and selected diagnostic group of AYA Site Recode b
1. Leukemias and related disorders	156	4.9	425	13.2	1798	56.0	832	25.9
2. Lymphomas	156	5.2	548	18.3	1691	56.6	592	19.8
3. CNS and other intracranial and intraspinal neoplasms	323	8.1	577	14.4	2003	50.0	1102	27.5
4. Sarcomas	193	9.5	672	33.1	786	38.7	381	18.8
7.1 Testis	72	2.5	312	10.9	1349	47.0	1135	39.6
7.2 Ovary	90	4.2	317	14.9	1163	54.7	558	26.2
9.1 Thyroid carcinoma	168	2.5	691	10.5	2636	40.0	3100	47.0
9.2 Other carcinoma of head and neck	133	7.3	445	24.4	844	46.3	399	21.9
9.3 Carcinoma of gastrointestinal tract	306	3.8	1385	17.1	3509	43.4	2882	35.7
9.4 Carcinoma of lung, bronchus, and trachea	64	4.8	330	25.0	644	48.8	283	21.4
9.6 Carcinoma of breast	377	3.4	1742	15.5	4913	43.7	4203	37.4
9.7.1 Carcinoma of uterine cervix	342	6.5	1154	22.0	2546	48.6	1195	22.8

Note: All cancers, benign, or behavior‐uncertain central nervous system (CNS) tumors classified by the International Classification of Childhood Cancer, 3rd edition, update 2017 (ICCC‐3) were included. Cases registered by death certificate only were excluded. Numbers between 1 and 9 are displayed as “1–9”. For other cells where the numbers 1–9 can be identified by addition or subtraction, a range of numbers is displayed.

Abbreviations: AYA, adolescents and young adults; DCCH, designated cancer care hospital; PCH, pediatric cancer hospital; P‐DCCH, prefectural designated cancer care hospital, R‐DCCH, regional designated cancer care hospital.

^a Main diagnostic groups of ICCC‐3 with more than 100 cases are displayed.

^b Diagnostic groups of AYA Site Recode with more than 2000 cases are displayed.

4. DISCUSSION

In this study, we assessed the incidence of cancer among children and AYAs in Japan by cancer type and proportion of patients who were treated at the designated cancer hospitals (PCHs or DCCHs), using data from the Japanese nationwide population‐based cancer registry, which was started in 2016.

The average annual number of cases was 2510 in children and 23,121 in AYAs and the ASR was 166.6 (per million‐person years) in children and 579.0 in AYAs (including all malignant cancer, benign, or uncertain‐behavior CNS tumors). The average annual number of child‐specific solid tumors such as neuroblastoma, retinoblastoma, nephroblastoma, and hepatoblastoma in children (age 0–14 years) was 151, 74, 45, and 51, respectively. These numbers were similar to, or slightly higher than, those reported by the Japanese Society of Pediatric Hematology/Oncology in 2018 ²⁴ (neuroblastoma, 124; retinoblastoma, 43; nephroblastoma, 30; and hepatoblastoma, 37 in patients aged 0–19 years). Before the Japanese NCR was established, childhood cancer incidence was estimated using data from several prefectural cancer registries. ²¹ , ²⁵ A report from Marugame et al. in 2007, which collected data on newly diagnosed cancers between 1993 and 2001 from 13 population‐based cancer registries in Japan, estimated the ASRs of children aged 0–14 years as 103.7 in male patients and 80.1 in female patients. ²⁵ In 2017, these statistics were updated to 134.6 in male patients, 118.6 in female patients, and 126.8 in both sexes, as reported by Katanoda et al. using cases diagnosed between 2009 and 2011 from the 27 population‐based cancer registries. ²¹ Although the different time periods and regions covered could have affected the results, the reason for the increase in incidence compared to the previous study might be explained by several factors such as improved data quality and completion of missed notifications (proportion of DCO was 0.2% or 0.3% in the present study vs. 2% in a previous study ²¹ ), as the previous data collection was not mandated by law and was only carried out at selected prefectural cancer registries. Compared to childhood cancer ASRs in Asian countries that were reported in the International Incidence of Childhood Cancer, volume 3 (ASR from 75.6 in Pakistan to 157.2 in Turkey), ²⁶ , ²⁷ the ASR of Japan in this study (ASR = 166.6) was slightly higher and nearer to that of the United States (ASR = 166.9) (Figure S1). As found in a previous report, ²¹ , ²⁸ by cancer type, the incidence of Hodgkin's lymphoma (ASR = 1.0), astrocytoma (ASR = 8.4), nephroblastoma (ASR = 3.4), Ewing tumor and related sarcomas of bone (ASR = 1.2) was lower in Japan than in North America and European regions, while the incidence of intracranial and intraspinal germ cell tumors (ASR = 4.5) was higher in Japan. Although some of the etiology of these cancer types has been reported in recent years, ²⁹ , ³⁰ , ³¹ , ³² much remains unknown and further investigation, including consideration of environmental factors and genetic predisposition, is required.

Within the 0–39 years age group, the type of cancer varies with age: hematological malignancies, blastomas, and CNS tumors were common in children under 10 years of age, malignant bone tumors and soft tissue sarcomas were relatively common in teenagers, and in those over 20 years, carcinomas in thyroid, testis, gastrointestinal, female cervix, and breast were common (Figures 1, 2, 3). These characteristics were consistent with previous reports. ²¹ Given the different types of cancer according to age and gender, it is likely that different specialists will be involved in the treatment of this generation of cancer patients.

Focusing on AYAs, the ASRs in Japanese AYAs in this study were also higher than in the previous report (ASRs of all cancers, malignant only increased from 429.7 to 523.2 in both sexes), ²¹ and ranked 40th out of 185 countries reported in GLOBOCAN2020 ³³ (Figure S2). The AYA Site Recode was adapted in this study for cancers in AYAs as cancers occurring in this generation are unique and different from cancers in both children and adults. ⁶ Compared to reports from the United States and the Netherlands using AYA Site Recode, ³⁴ , ³⁵ the proportion of melanoma was low among the Japanese AYAs in this study, and also among Korean AYAs, ³⁶ which indicates genetic background could be contributing to the etiology of melanoma. ³⁷ Also, the incidence rate of thyroid cancer in Japanese AYAs was lower than that in Korea. ³⁶ This could be explained by differences in prevention systems for thyroid cancer, ³⁸ , ³⁹ for example, in the Korea thyroid screening is included in government‐paid health check‐ups, whereas this is not the case in Japan.

We have identified, for the first time, the proportion of child and AYA patients in Japan who were treated in designated cancer hospitals. For children, the proportion of cases treated at the PCHs was only 20%–30%. In Japan, children with cancer were previously treated at more than 200 hospitals, ⁴⁰ that had less treatment experience per hospital than in other countries. ⁴¹ A European report showed that centralized treatment for childhood cancer was associated with improved survival. ¹³ In 2012, the 2nd term Basic Plan to Promote Cancer Control Programs raised the issue of care for children and 15 hospitals that fulfill criteria such as “experience in treating more than 30 children with cancer per year” were designated as PCHs, similar to designation under the European Standards of Care for Children with Cancer. ⁴² , ⁴³ The study found that the proportion of children receiving treatment at PCHs was low overall, and varied by cancer type. It was higher in child‐specific solid tumors such as hepatic tumors (47.2%), retinoblastoma (42.2%), and neuroblastoma (41.9%), while lower for those with malignant bone tumors (20.8%) who are often treated at P‐DCCHs (41.6%). As some DCCHs focus only on adult cancers, it is necessary to check whether the care given to children with cancer, who are treated at DCCHs, is developmentally appropriate. Based on these results, the government and relevant stakeholders need to reconsider the cancer care system for children, including the number of PCHs, to ensure that all children receive equally high‐quality pediatric cancer care. In 2019, in addition to the PCHs, shared pediatric cancer care hospitals were identified in seven areas and a system was established for collaborative treatment of children with cancer between them. It will be necessary to continue to add this information and to monitor the pattern of care for children with cancer in the future.

In 2018, the 3rd term Basic Plan to Promote Cancer Control Programs raised the issue of care for AYA patients and the need to promote a certain degree of centralization, with a focus on AYA cancer care. ¹⁰ However, there are relatively large numbers of AYA patients compared to children with cancer, and this group have more diverse life stages and a greater variety of needs and psychosocial difficulties, making it harder to determine the ideal cancer care system for them. ¹⁰ In this study, 64% of AYAs were treated at DCCHs with fewer than 10% treated at PCHs; this percentage decreased with age in both types of hospitals and varied by cancer type. This could be due to generational differences in needs and treatment specificities in AYAs. For example, the relatively young age group might choose a PCH or DCCH because they require educational support, while the relatively older age group could choose a hospital closer to home or work because of financial or work‐related concerns. ¹⁰ In 2022, MHLW proposed that DCCHs should create support teams for the AYA generation and establish a network system of support, including local government in each region, with the P‐DCCH taking the lead. ⁴⁴ We need to consider how all AYAs in each region can receive appropriate cancer care and support according to their needs and cancer type.

The foremost strength of our study is that it is based on nationwide population‐based cancer registry data, not estimates, but actual measurements to determine the number and incidence of cases by cancer type, even for very rare entities of cancer in children and AYAs. Based on these statistics, the epidemiological information for children and AYAs with cancer and their families in Japan can be updated. ⁴⁵ , ⁴⁶ Our study thus provides the most complete overview of incidence patterns in children and AYAs to date and, for the first time, describes the proportion of patients treated in designated cancer hospitals.

Despite this strength, the study has several limitations. First, the Japanese NCR has only been in place for a short period of time (since 2016) and trends have not been assessed. Second, differences in the quality of the data by prefecture have been reported. ⁴⁷ For example, the incidence of thyroid cancer in Miyazaki and carcinoma of the breast in Kumamoto in 2016 increased temporarily at the start of the NCR, with the registration of cases that had not previously been registered. ⁴⁷ In the present data, there were fewer than 2% DCOs in all of the prefectures (data not shown), but caution should be exercised in interpreting the incidence and we could not analyze incidence by region or prefecture. It is expected that the accuracy of the data will stabilize in forthcoming years. Third, the category of treatment hospital was limited to only one treatment hospital per case, so that we could not observe the collaboration between hospitals. Fourth, variables required for the population‐based cancer registries were limited to essential information on a patient and a cancer, and the collection of further data items such as associated congenital malformations, stage at diagnosis, type of treatment, and relapse would be an asset as well as a further improvement of data quality. Since 2021, the International Benchmarking of Childhood Cancer Survival by Stage (BENCHISTA) project has been collecting data on childhood cancer stage, first‐line treatment, tumor biology, nonstage prognostic factors, relapse and cause of death, with partial data participation from Osaka and Tokyo. ⁴⁸ It is hoped that such studies will lead to a more detailed understanding of the risk of each childhood cancer in Japan. Finally, it should be noted that the timing of diagnosis and the target population differ with regard to comparison with other countries. For rare cancers such as childhood and AYA cancers, direct comparisons with cancer incidence in other countries are necessary in order to assess the risk of cancer incidence in one's own country and to use this information for prevention and other purposes. A balance should be struck between the protection of personal data and the need to share information so that Japan can continue to actively participate in projects that directly compare data internationally. ⁴⁹

In conclusion, this study provides an overall picture of incidence and type of treatment hospital for children and AYAs with cancer in Japan. Based on this information, the desirable system of cancer care should be discussed for this vulnerable population. Population‐based cancer registry data are important to provide accurate childhood cancer burden estimates for patients, their families, clinicians, and policy makers who make cancer control planning decisions. Further research and surveillance of comprehensive statistics, including survival and mortality, is needed.

AUTHOR CONTRIBUTIONS

KN and KK conceived the idea of the study. KK, MH, and KN applied for the Japanese population‐based NCR data and developed the statistical analysis plan. KN performed the literature search and statistical data analysis. TM, HS, KT, IM, KM, AY, JT, and CS contributed to the interpretation of the results. KN drafted the original manuscript. KK and TM supervised the conduct of this study. All authors reviewed the manuscript draft and critically revised the intellectual content. All authors approved the final version of the manuscript prior to submission.

FUNDING INFORMATION

This work was supported by Health, Labour and Welfare Sciences Research Grants from the Ministry of Health, Labour and Welfare of Japan (Grant Number JPMH20EA1026 [to KN, TM, MH, HS, IM, and KK], JPMH23EA1033 [to KN, TM, MH, HS, and KK], and JPMH23EA1017 [to KN and CS]), and a Grant‐in‐Aid for Early‐Career Scientists from the Japan Society for the Promotion of Science KAKENHI (JP20K18952 to KN).

CONFLICT OF INTEREST STATEMENT

Junko Takita is an Editorial Board Member of Cancer Science. The other authors have no conflicts of interest.

ETHICS STATEMENTS

Approval of the research protocol by an institutional review board: This study was approved by the institutional review board of the National Cancer Institute (approval number: 2019–214). We obtained the dataset anonymized by the Japanese NCR under the Cancer Registry Promotion Act and independently processed it. The data are not publicly accessible and is available only on request due to privacy and ethical restrictions.

Informed consent: N/A.

Registry and registration no. of the study/trial: N/A.

Animal studies: N/A.

Supporting information

Figures S1–S2

Tables S1–S6

Nakata K, Matsuda T, Hori M, et al. Cancer incidence and type of treatment hospital among children, adolescents, and young adults in Japan, 2016–2018. Cancer Sci. 2023;114:3770‐3782. doi: 10.1111/cas.15892