Cancer incidence

Chapter on cancer incidence as part of the children, teenagers and young adults  UK cancer statistics report 2021.

Numbers of newly diagnosed cancer cases: Cancer incidence measures the number of new cases of cancer in an at-risk population diagnosed over a given period. It is a useful measure of the need for diagnostic and treatment services and, in some cases, of the effectiveness of preventive interventions.

Statistics can be provided as the total number of cases or as rates. Rates are often presented as the number of cases per 100,000 population per year. In this report incidence rates are presented as the number of cases per million population as this makes smaller numbers more meaningful. Incidence rates can be ‘crude’ or age standardised.

The crude rate is the actual observed rate calculated by dividing the number of patients who live in a given area by the population of that area. Age standardised rates (ASRs) adjust the proportions of different age-groups so that they are always the same, using the World Standard Population. This minimises the effects of differences in the age structure of populations over time or between countries.

Our analysis relates to children and young people who were diagnosed with cancer under the age of 25 and who were residents of the UK during the 20-year period of 1997-2016. Incidence rates were calculated per million person years for the age groups 0-4, 5-9, 10-14, 15-19 and 20-24 years based on annual UK population estimates. ⁴

For children aged 0-14, teenagers and young adults aged 15-24, and the full 0-24 age range, age standardised rates were calculated using the World Standard Population, which assigns weights of 12, 10, 9, 9 and 8 to the age groups 0-4, 5-9, 10-14, 15-19 and 20-24 years respectively. ⁵

Cumulative risk was calculated as the sum of the age-specific incidence rates for the five-year age groups, each multiplied by 5, the number of years contained in each age group.

Table 2 in Appendix B shows numbers of cancer registrations in the 12 main ICCC-3 cancer groups, together with the percentage recorded as having microscopic verification (%MV) and the percentage registered from a death certificate only (%DCO) for the UK and the constituent countries.

Table 3 in Appendix B shows the same data as Table 2 for subgroups and divisions of the diagnostic classification and for certain other subsets of particular interest, for the UK only.

There were 75,103 young persons diagnosed with cancer over the 20 years of this study, approximately 3,755 per year.

The mean numbers of cancer registrations per year were 1,645 in children and 2,110 in TYA. The mean annual numbers in the four UK nations were 3,112 (1,378 children and 1,734 TYA) in England, 123 (53 children and 70 TYA) in Northern Ireland, 318 (129 children and 189 TYA) in Scotland and 202 (85 children and 117 TYA) in Wales.

Figures 1 and 2 show the annual numbers of registrations for the ICCC-3 main groups and subgroups among children and TYA respectively. Similar data for all subgroups are shown in table 3 in Appendix B.

Among children, leukaemia accounted for 31% of registrations, CNS and miscellaneous intracranial and intraspinal neoplasms (hereafter referred to as CNS tumours) for 25%, lymphomas for 10%, soft-tissue sarcomas for 6%, neuroblastoma and other peripheral nervous cell tumours for 6%, and renal tumours for 6%.

Among TYA, lymphomas accounted for 20% of registrations, germ cell, trophoblastic and gonadal tumours for 16%, CNS tumours for 12%, malignant melanomas for 10%, leukaemia for 9%, and soft-tissue sarcomas, skin carcinomas and thyroid carcinomas each for 5%. Carcinomas (other than renal, hepatic and gonadal) and malignant melanomas accounted for 30% of cancers among TYA. No other diagnostic groups accounted for more than 5% of registrations in the respective age ranges.

There were several distinctive patterns of occurrence by age (Figure 3 and Table 3). For many types of cancer, numbers were highest before five years of age. These included lymphoid leukaemia, ependymoma & choroid plexus tumours, intracranial & intraspinal embryonal tumours (mainly medulloblastoma), neuroblastoma, retinoblastoma, nephroblastoma & other nonepithelial renal tumours, hepatoblastoma, and rhabdomyosarcoma.

For other types, numbers were low in the first few years of life and increased throughout childhood and the TYA age range. Examples include Hodgkin lymphoma, non-Hodgkin lymphoma, malignant melanoma, and nearly all carcinomas.

Numbers were lowest at 5-9 years of age for acute myeloid leukaemia, and for malignant extracranial and extragonadal germ cell tumours. Leukaemia formed the most frequent group before five years of age, when they accounted for 37% of all cancers.

At age 5-9 years, leukaemia and CNS tumours each accounted for 32-33% of all cancers for this age group.

By contrast, numbers increased to reach a maximum at age 10-14 or 15-19 and decreased thereafter for osteosarcoma, Ewing sarcoma family of tumours, and intracranial & intraspinal germ cell tumours. CNS tumours were the most frequent group at age 10-14, accounting for 26% of all cancers.

At age 15-19 years, lymphomas accounted for 23% of all cancers.

At age 20-24 years, 36% of all cancers were carcinomas. Comparisons of incidence by sex are covered in the incidence rates section below.

Overall, 92.7% of registrations were recorded as microscopically verified. In general, a pathological sample (microscopic verification) is needed to confirm a cancer diagnosis and guide treatment. However, for some cancer sites (for example, those of the central nervous system) it is sometimes not safe or not needed to take a tissue sample and the diagnosis is then usually based on radiological findings. We would therefore expect the MV proportion to be high but not 100%.

There was some variation between countries in the proportion of cases overall and by disease group with microscopic verification. To some extent this may be a coding issue: practices have changed differently over time in the four national registries. In general, recording of microscopic verification is better now than at the start of the period, and is a focus of attention for the individual registries.

The only main diagnostic groups to have less than 90% MV were CNS tumours (80.8%), retinoblastoma (68.1%) and other and unspecified malignant neoplasms (38.3%). The lower %MV for CNS tumours and retinoblastoma are a consequence of the relatively low proportions of tumours in these categories that are biopsied, while the low %MV for other and unspecified malignant tumours reflects the provisional nature of the data for a high proportion of patients in this small and miscellaneous group.

Overall, only 0.2% of all registrations were recorded from only the death certificate (Death Certificate Only, DCO); the percentage of DCO was below 0.4% in all the main diagnostic groups except other and unspecified malignant neoplasms, where they accounted for 1.1% of registrations.

The relative frequencies of the %MV and %DCO overall and for the main diagnostic groups were typical of those for cancer registries with high-quality data in Europe, North America and Oceania. ⁶

Notes:

⁴ Population Estimates for UK, England and Wales, Scotland and Northern Ireland, Crown Copyright Office for National Statistics, accessed December 2019

⁵ Glossary of Terms, Cancer Incidence in Five Continents (CI5) series, International Association of Cancer Registries

⁶ International Incidence of Childhood Cancer 3 (IICC-3)

Figure 1 - Mean number of newly diagnosed cancer cases per year registered among children under 15 years of age  and resident in the UK, 1997-2016, grouped according to ‘International Classification of Childhood Cancer, Third  Edition’ (ICCC-3)

Source: National Cancer Registration and Analysis Service for England (Public Health England), the Northern Ireland Cancer Registry, the Scottish Cancer Registry, and the Welsh Cancer Intelligence and Surveillance Unit Mean number of cases for persons are shown in brackets by the cancer classifications labels. The mean number for persons does not always sum the mean number of males and females due to rounding. Only cancer groupings I-XI with 10 cases per year and over are displayed on the chart.

Figure 2 - Mean number of newly diagnosed cancer cases per year registered among teenagers and young adults  15-24 years of age and resident in the UK, 1997 to 2016, grouped according to ‘International Classification of  Childhood Cancer, Third Edition’ (ICCC-3

Source: National Cancer Registration and Analysis Service (NCRAS) for England (Public Health England), the Northern Ireland Cancer Registry, the Scottish Cancer Registry, and the Welsh Cancer Intelligence and Surveillance Unit Mean number of cases for persons are shown in brackets by the cancer classifications labels. The mean number for persons does not always sum the mean number of males and females due to rounding. Only cancer groupings I-XI with 20 cases per year and over are displayed on the chart.

Figure 3: Mean number of newly diagnosed cancer cases per year registered among those under 25 years of age  and resident in the UK, 1997-2016, grouped according to ‘International Classification of Childhood Cancer, Third  Edition’ (ICCC-3) and age.

Source: National Cancer Registration and Analysis Service (NCRAS) for England (Public Health England), the Northern Ireland Cancer Registry, the Scottish Cancer Registry, and the Welsh Cancer Intelligence and Surveillance Unit

Table 4 in Appendix B shows cancer incidence rates in children, teenagers and young adults by ICCC-3 main group for males and females separately for the UK and the constituent countries. Corresponding data for subgroups and divisions of the diagnostic classification and for certain other subsets of particular interest are shown in Table 5, for the UK only.

The total ASR for age 0-24 years was 202 per million in males and 182 per million in females. For children aged 0-14 years, the ASR was 162 per million in males and 141 per million in females. For TYA aged 15-24 years, the ASR were 273 per million in males and 258 per million in females. The sex ratio of ASRs was M/F=1.11 overall, M/F=1.15 at age 0-14, and M/F=1.06 at age 15-24.

The cumulative risk of being diagnosed with cancer in the first 15 years of life was 1 in 422 for males and 1 in 488 for females. The cumulative risk within the TYA age range was 1 in 361 for males and 1 in 382 for females. The total cumulative risk in the first 25 years of life was 1 in 194 for males and 1 in 214 for females. Within childhood, incidence was highest in the first five years for both males and females, fell to a minimum at age 5-9 years, and was slightly higher at age 10-14 years, marking the start of the unbroken rise in incidence that continues into the teenage years and throughout adulthood. Incidence rates were within the ranges reported from other countries in Europe, North America and Oceania. ^7,8

Generally, the incidence rates across the UK nations are comparable. Where there was some variation in incidence rates between UK nations, possible explanations for the differences include chance findings (made more likely by multiple statistical testing), differences in the proportion of registrations with information on morphology, and real differences in incidence. Differences in rates for CNS tumours might also result from differences in coding or recording of cases between registries (for instance, increased recording of certain asymptomatic or low grade tumours). Higher incidence of melanoma in Scotland and Northern Ireland than in England and Wales is consistent with the pattern throughout adulthood. ⁸

The variation between UK nations in the incidence of carcinoma of the cervix below age 25 during 1997-2016 reflects differences between the nations over time in the age at first invitation to cervical cancer screening. At the start of the study period in 1997, the age at first invitation was 20 years throughout the UK. It was increased to 25 years in England in 2004, then slightly lowered to 24.5 years in 2012 in order to increase the proportion of young women screened as close as possible to their 25th birthday. ⁹ Age at first invitation was increased from 20 to 25 in Northern Ireland in January 2011, in Wales in September 2013, and in Scotland in June 2016.

The sex ratio varied between diagnostic categories. Among diagnostic categories with at least 50 registrations (Table 5), the highest sex ratio of ASRs was for malignant gonadal germ cell tumours, M/F=5.9. This is largely because testicular tumours are mainly of germ cell origin. There were also relatively marked male excesses for intracranial & intraspinal germ cell tumours, M/F=2.8, non-Hodgkin lymphoma (including Burkitt lymphoma), M/F=2.0, and medulloblastoma, M/F=1.8. For several categories of epithelial cancers, incidence was higher among females than males. Breast carcinoma, which is extremely rare in males of all ages, had a sex ratio of F/M>100. Other categories with at least 50 registrations that had marked female excesses were gonadal carcinoma, M/F=0.1, thyroid carcinoma, M/F=0.3, malignant melanoma, M/F=0.5, salivary carcinoma, M/F=0.5, and skin carcinoma, M/F=0.8. There was also a pronounced female excess for nephroblastoma (Wilms tumour), M/F=0.8.

There were also some notable variations in sex ratio by age. For lymphoid leukaemia, astrocytoma other than pilocytic, osteosarcoma, Ewing sarcoma family of tumours, and CNS germ cell tumours, the male excess was much more pronounced among TYA than among children. Contrastingly, Hodgkin and non-Hodgkin lymphomas both had substantial male excesses in childhood, whereas at age 15-24 years the male excess of non-Hodgkin lymphoma was much smaller, and males and females had approximately equal ASRs for Hodgkin lymphoma; Hodgkin lymphoma in fact had the highest incidence of any cancer among female TYA. Malignant melanoma was more frequent among females than males at all ages, but the sex ratio F/M increased with age from 1.3 in childhood to 1.9 at aged 15-19 and 2.2 at age 20-24; malignant melanoma was the most frequent cancer among females aged 20-24 years. Gonadal germ cell tumours were more frequent among males than females at age 0-4 and, especially, at age 15-24 years, whereas at age 5-14 there was a marked female excess. Testicular germ cell tumours were the most frequent cancer among male TYA and accounted for 29% of all cancers among males aged 20-24 years.

Notes:

⁷ International Incidence of Childhood Cancer 3 (IICC-3)

⁸ Bray F, Colombet M, Mery L, Piñeros M, Znaor A, Zanetti R and Ferlay J, editors (2017). Cancer Incidence in Five Continents, Vol. XI (electronic version). Lyon: International Agency for Research on Cancer. Available from: https://ci5.iarc.fr, accessed January 2021

9 Castanon A, Sasieni P. Is the recent increase in cervical cancer in women aged 20-24 years in England a cause for concern? Prev Med. 2018;107:21-28