The United Kingdom Children's Cancer Study Group--the first 20 years of growth and development

Subsequent cancers within 5 years from initial diagnosis of childhood cancer. Patterns and risks in the population of Great Britain

BACKGROUND

Patterns and risks of subsequent primary tumours (SPTs) among long-term survivors of childhood cancer have been extensively described, but much less is known about early SPTs (ESPTs) occurring within 5 years after initial diagnosis.

PROCEDURE

We carried out a population-based study of ESPTs following childhood cancer throughout Britain, using the National Registry of Childhood Tumours. The full study series comprised all ESPTs occurring among 56,620 children whose initial cancer diagnosis was in the period 1971-2010. Frequencies of ESPT were calculated for the entire cohort. For analyses of risk, follow-up began 92 days after initial diagnosis.

RESULTS

ESPT developed in 0.4% of children overall, 0.52% of those initially diagnosed at age less than 1 year and 0.38% of those diagnosed at age 1-14 years. Standardised incidence ratio (SIR) was 7.7 (95% confidence interval [CI]: 6.7-8.9), overall 9.5 (95% CI: 7.1-12.5) for children initially diagnosed in 1981-1990 and 6.5-7.5 for those from earlier and later decades. SIR by type of first cancer ranged from 4.4 (95% CI: 1.8-9.1) for Wilms tumour to 13.1 (95% CI: 7.7-21.0) for non-Hodgkin lymphoma. SIR by type of ESPT ranged from 2.0 (95% CI: 1.0-3.4) for acute lymphoblastic leukaemia to 66.6 (95% CI: 52.3-83.6) for acute myeloid leukaemia. Predisposition syndromes were known to be implicated in 21% of children with ESPT and suspected in another 5%.

CONCLUSIONS

This study provides an overview of the patterns and risks of ESPTs in a large population where many children received therapy that is still in widespread use. Further research will be needed to monitor and understand changes in risk as childhood cancer treatment continues to evolve.

Incidence of childhood CNS tumours in Britain and variation in rates by definition of malignant behaviour: population-based study

Background

Intracranial and intraspinal tumours are the most numerous solid tumours in children. Some recently defined subtypes are relatively frequent in childhood. Many cancer registries routinely ascertain CNS tumours of all behaviours, while others only cover malignant neoplasms. Some behaviour codes have changed between revisions of the International Classification of Diseases for Oncology, including pilocytic astrocytoma, downgraded to uncertain behaviour in ICD-O-3.

Methods

We used data from the population-based National Registry of Childhood Tumours, which routinely included non-malignant CNS tumours, to document the occurrence of CNS tumours among children aged < 15 years in Great Britain during 2001–2010 and to document the descriptive epidemiology of childhood CNS tumours over the 40-year period 1971–2010, during which several new entities were accommodated in successive editions of the WHO Classification and revisions of ICD-O. Eligible cases were all those with a diagnosis included in Groups III (CNS tumours) and Xa (CNS germ-cell tumours) of the International Classification of Childhood Cancer, Third Edition. The population at risk was derived from annual mid-year estimates by sex and single year of age compiled by the Office for National Statistics and its predecessors. Incidence rates were calculated for age groups 0, 1–4, 5–9 and 10–14 years, and age-standardised rates were calculated using the weights of the world standard population.

Results

Age-standardised incidence in 2001–10 was 40.1 per million. Astrocytomas accounted for 41%, embryonal tumours for 17%, other gliomas for 10%, ependymomas for 7%, rarer subtypes for 20% and unspecified tumours for 5%. Incidence of tumours classified as malignant and non-malignant by ICD-O-3 increased by 30 and 137% respectively between 1971-75 and 2006–10.

Conclusions

Total incidence was similar to that in other large western countries. Deficits of some, predominantly low-grade, tumours or differences in their age distribution compared with the United States and Nordic countries are compatible with delayed diagnosis. Complete registration regardless of tumour behaviour is essential for assessing burden of disease and changes over time. This is particularly important for pilocytic astrocytoma, because of its recent downgrading to non-malignant and time trends in the proportion of astrocytomas with specified subtype.

Use of patient registries and administrative datasets for the study of pediatric cancer

Analysis of data from large administrative databases and patient registries is increasingly being used to study childhood cancer care, although the value of these data sources remains unclear to many clinicians. Interpretation of large databases requires a thorough understanding of how the dataset was designed, how data were collected, and how to assess data quality. This review will detail the role of administrative databases and registry databases for the study of childhood cancer, tools to maximize information from these datasets, and recommendations to improve the use of these databases for the study of pediatric oncology.

Population survival from childhood cancer in Britain during 1978-2005 by eras of entry to clinical trials

BACKGROUND

Inclusion in clinical trials is generally viewed as best practice for most newly diagnosed childhood cancers, but the impact on population-based survival has rarely been examined.

PATIENTS AND METHODS

The population-based data were analysed for 25 853 children (66% of all registered childhood cancers) diagnosed in Britain during 1978-2005 with acute lymphoblastic leukaemia (ALL), acute myeloid leukaemia (AML), Hodgkin lymphoma, non-Hodgkin lymphoma, medulloblastoma, neuroblastoma, Wilms tumour, hepatoblastoma, osteosarcoma, Ewing sarcoma, rhabdomyosarcoma and germ-cell tumours. The Kaplan-Meier survival curves were compared by log-rank tests. Time trends were analysed by Cox regression. Separate analyses were done for children with ALL, medulloblastoma and neuroblastoma according to clinically relevant age thresholds.

RESULTS

Survival increased significantly during 1978-2005 for every diagnostic category; the annual reduction in risk of death ranged from 2.7% (rhabdomyosarcoma) to 12.0% (gonadal germ-cell tumours). Survival increased steadily between trial eras for ALL (age 1-14 years) and neuroblastoma (age 1-14 years), but changed little since the mid-1980s for medulloblastoma (age 0-2 years), osteosarcoma or Ewing sarcoma.

CONCLUSIONS

Changes in survival between trial eras parallel those reported by the relevant clinical trials. The increasing level of participation in trials, facilitated by the organisation of specialist care, has underpinned the substantial improvements in survival seen at the population level.

Geographical comparison of cancer survival in European children (1988–1997): Report from the Automated Childhood Cancer Information System project

The aim of this study was to assess regional survival differences among childhood cancer patients in Europe. For this exercise, the Automated Childhood Cancer Information System (ACCIS) database was utilised. Survival data from 54 population-based cancer registries on 49,651 childhood cancer patients aged 0-14 years and diagnosed in 1988-1997 were analysed using life-table method. Overall, the 5-year survival was 72% among all patients, varying from 62% to 77% between the five geographical regions. The East region generally had lower survival rates than the rest of Europe. The geographical differences indicate the need for more co-ordination, systematisation and standardisation in diagnosis, referral and the treatment of childhood cancers in Europe. Increase of resources is necessary to improve the lower survival in the East region. Continuing data collection on a European level will facilitate monitoring of population-based survival of childhood cancer patients.

Caring and curing: paediatric cancer services since 1960

This paper traces the history of the specialist meanings of 'cure' in paediatric oncology in the UK, how they have changed with increasing organization of the discipline, ever-rising survival rates for all childhood cancers, and with feedback from patients and families. It examines the differing ways in which those involved in researching, treating, and raising funds for work on childhood cancers have understood and used the language of cure, and speculates as to why talking about the 'cure' of survivors of childhood cancers is so problematic. The paper discusses the particular importance of holistic care in the development of paediatric oncology. Psychosocial support is delivered alongside surgery, radiotherapy and chemotherapy. The focus for support is the patient's whole family, building a tenet of palliative care into curative treatment. The concept of the 'truly cured child' is argued to have been crucial in the discipline's decision in the 1970s and 1980s to make the psychosocial needs of patients and their families central in the programme of curing children with cancer.

Reporting of prognostic markers: current problems and development of guidelines for evidence-based practice in the future

Prognostic markers help to stratify patients for treatment by identifying patients with different risks of outcome (e.g. recurrence of disease), and are important tools in the management of cancer and many other diseases. Systematic review and meta-analytical approaches to identifying the most valuable prognostic markers are needed because (sometimes conflicting) evidence relating to markers is often published across a number of studies. To investigate the practicality of this approach, an empirical investigation of a systematic review of tumour markers for neuroblastoma was performed; 260 studies of prognostic markers were identified, which considered 130 different markers. The reporting of these studies was often inadequate, in terms of both statistical analysis and presentation, and there was considerable heterogeneity for many important clinical/statistical factors. These problems restricted both the extraction of data and the meta-analysis of results from the primary studies, limiting feasibility of the evidence-based approach.Guidelines for reporting the results of primary prognostic marker studies in cancer, and other diseases, are given in order to facilitate both the interpretation of individual studies and the undertaking of systematic reviews, meta-analysis and, ultimately, evidence-based practice. General availability of full individual patient data is a necessary step forward and would overcome the majority of problems encountered, including poorly reported summary statistics and variability in cutoff level, outcome assessed and adjustment factors used. It would also limit the problem of reporting bias, although publication bias will remain a concern until studies are prospectively registered. Such changes in practice would help important evidence-based reviews to be conducted in order to establish the most appropriate prognostic markers for clinical use, which should ultimately improve patient care.

Epidemiology of cancer in adolescents

In western populations, the annual incidence rate of cancer among adolescents aged 15-19 years is around 150-200 per million, intermediate between the rates for older children and young adults. The most frequent diagnostic groups are acute leukemia, lymphomas, central nervous system tumors, bone and soft tissue sarcomas, germ cell tumors, thyroid carcinoma, and malignant melanoma. While the causes of most cancers in teenagers are still unknown, health education and promotion and public health programs offer some scope for prevention among people of this age group. Reduction in sun exposure should lead to a reduction in incidence of melanoma, and elimination of hepatitis B in regions where it is endemic should result in a decrease in hepatic carcinoma. Five-year survival of patients diagnosed around 1990 exceeded 70% in the USA and UK. Entry to clinical trials appears to be much less frequent for adolescents with cancer than for children. There is some evidence that higher survival is associated with entry to trials or centralized treatment for certain cancers in this age group.

Variation in survival of children with central nervous system (CNS) malignancies diagnosed in Europe between 1978 and 1992: the EUROCARE study

EUROCARE is a population-based survival study including data from European Cancer Registries. The present paper analyses survival after a malignant neoplasm of the central nervous system (CNS) in childhood (aged 0--14 years at diagnosis). The database includes 6130 cases from 34 population-based registries in 17 countries: 1558 were primitive neuroectodermal tumours (PNET) and 4087 astrocytoma, ependymoma or other gliomas: these morphologies were grouped in the analyses in order to reduce the diagnostic variability among the registries. 87% of cases were microscopically diagnosed (range among registries 71--100%) and losses to follow-up were limited to 2% (range 0--14%). Actuarial analyses indicate that the European (weighted) average of 5 years cumulative survival for cases diagnosed in 1978--1989 was 53% (95% confidence interval (CI) 49--57) for CNS neoplasms, 44% (95% CI 37--50) for PNET and 60% (95% CI 55--65) for the glioma-related types. Analysis of the sub-set of cases diagnosed in 1985--1989 revealed better results: cumulative survival at 5 years was 61% (95% CI: 55--65) for all CNS neoplasms; 48% (95% CI 41--56) for PNET and 68% (95% CI 62--73) for glioma-related types. Compared with older children, infants showed poorer prognosis: in 1978--1989 the 5-year survival rate was 33% (95% CI 23--45) and in 1985--1989 it was 46% (95% CI 34--59). Variability among countries was very large, with 5-year survival for CNS tumours diagnosed in 1985--1989 ranging from 28% in Estonia (95% CI 17--43) to 73% Sweden (95% CI 59--83) and 75% in Iceland (95% CI 35--95) and 73% in Finland (95% CI 66--79). Time trends were studied in a multivariate analysis observing a reduction in the risk of death in periods of diagnosis 1982--1985 (hazard ratio (HR)=0.85; 95% CI 0.78--0.93) and 1986--1989 (HR=0.70; 95% CI 0.64--0.77) compared with 1978--1981. The analysis were extended to 1990--1992 for the countries whose registries provided data for that period did not indicate any further progress. Results of this study confirm the large variability in European countries and indicate a positive trend in the survival probability for cases diagnosed in the 1980s.

Childhood cancer survival in Europe: an overview

Other articles in this issue of the European Journal of Cancer have described population-based survival analyses of specific types of childhood cancer included in the EUROCARE database, diagnosed since 1979. The present paper summarises the relevant estimates and comments on intercountry differences, focusing on possible distortions in the intercountry comparisons based on data produced by the cancer registries. Potential biases include a lack of exhaustiveness of both case ascertainment and follow-up for living status and also a lack of consistency in the use of classification of the childhood cancer types. Nevertheless, despite such biases, consistent differences are observed between European countries in the probability of survival following the diagnosis of a paediatric cancer. In most cases, poor population-based survival rates are probably explained by inadequacies in the adoption and implementation of therapeutic protocols that have been proved to be effective. In some instances, the cause of unsatisfactory estimates was the inclusion of a sizeable proportion of children with cancer in clinical trials which were found to be ineffective. A regression analysis of incidence, mortality and survival rates during 1978-1989 over the whole EUROCARE database strongly indicates that the prognostic improvements over time are real and cannot be attributed to changes in diagnostic procedures.

The adolescent with cancer--at the edge of no-man's land

The cancer journey of the adolescent patient can be difficult and lonely. Provision of services to these patients should occur at the interface between paediatric and adult oncology, although for the individual, that interface can seem like a gap into which he or she may too readily fall. In the UK, the needs of adolescent patients have become more widely recognised in the past decade, but they continue to have a low priority on the national agenda. Recent guidelines on cancer referral made specific reference to children's cancers, but none to cancer in adolescents. The need for dedicated services for these patients has been accepted, but the resources to meet this need have rarely been identified. We explore current and future patterns of care and service provision for adolescents with cancer in the UK.

The United Kingdom Childhood Cancer Study: objectives, materials and methods. UK Childhood Cancer Study Investigators

An investigation into the possible causes of childhood cancer has been carried out throughout England, Scotland and Wales over the period 1991-1998. All children known to be suffering from one or other type of the disease over periods of 4-5 years have been included, and control children matched for sex, age and area of residence have been selected at random from population registers. Information about both groups of children (with and without cancer) has been obtained from parental questionnaires, general practitioners' and hospital records, and from measurement of the extent of exposure to radon gas, terrestrial gamma radiation, and electric and magnetic fields. Samples of blood have also been obtained from the affected children and their parents and stored. Altogether 3,838 children with cancer, including 1,736 with leukaemia, and 7,629 unaffected children have been studied. Detailed accounts are given of the nature of the information obtained in sections describing the general methodology of the study, the measurement of exposure to ionizing and non-ionizing radiation, the classification of solid tumours and leukaemias, and the biological material available for genetic analysis.

Patterns of care and survival for children with acute lymphoblastic leukaemia diagnosed between 1980 and 1994

AIMS

To document survival rates after acute lymphoblastic leukaemia (ALL) during the era of modern chemotherapy, to assess effects of prognostic factors at presentation, and to investigate the relation of survival to patterns of organisation of care.

PATIENTS

From a population based series of 5078 children diagnosed in the UK during 1980-94, 4988 remained for analysis after exclusion of nine children ascertained from death certificates alone and 81 who received no antileukaemia treatment.

MAIN OUTCOME MEASURES

Actuarial survival rates.

RESULTS

Between 1980-84 and 1990-94, the proportion of children treated at paediatric oncology centres rose from 77% to 89%, and the proportion entered into national trials rose from 59% to 82%. Each of age, sex, white blood count, immunophenotype, and Down's syndrome status had a highly significant effect on survival. Five year survival improved from 67% in 1980-84 to 81% in 1990-94, a 42% reduction in the risk of death within five years of diagnosis. Survival did not differ significantly between hospitals with different numbers of new patients per year or between paediatric oncology centres and other hospitals. Children who were entered into national trials had higher survival and this difference became greater in recent years; five year survival rates for children diagnosed during 1980-84 were 70% and 64% for trial and non-trial patients, respectively; in 1990-94 the rates were 84% and 68% for trial and non-trial patients, respectively.

CONCLUSIONS

Survival after ALL continues to improve. Nearly 50 children/year diagnosed during 1990-94 survived who would have died a decade before. Survival does not vary systematically with place of treatment but is higher for children entered into national trials.

Survival and distribution pattern of childhood liver cancer in Taiwan

Studies of survival and distribution of liver cancer in children are scarce. In this study, using data from the cancer registry of Taiwan, from 1979 to 1992, we identified 377 young patients (0-15 years of age) suffering from liver cancer, coded 155 according to the International Classification of Diseases. Among these patients, 122 were histopathologically proven hepatocellular carcinoma (HCC) and 43 hepatoblastoma (HB). For survival analysis, we also searched for cases of liver cancer in 0-16 year old children in the Taiwan cancer registry for the period between 1988 and 1992. We found 109 cases with identification numbers and birth dates which allowed our cases to be linked with the death registry of the National Health Department of Taiwan enabling the calculation of 5-year survival rates using actuarial life tables. Between 1979 and 1992, for 122 HCC cases, there was a peak incidence at the age of 1 year, then a decline to a trough at the age of 4 years, after which the number of cases increased to the age of 15 years. After the age of 4 years boys outnumbered the girls by 2:1. 36 (84%) of 43 HB cases were under the age of 5 years and boys tended to outnumber girls by 2.9:1. Between 1988 and 1992, of the 109 patients, 49 were diagnosed histopathologically and 60 patients clinically. Their overall 5-year survival rate was 19%. The 5-year survival rate of the 28 HCC patients was 17%, whereas that of the 17 HB patients was 47%. In conclusion, our epidemiological findings indicate that the HCC distribution among children is different according to age and to some extent sex. The overall 5-year survival rate of children suffering from liver cancer was still unfavourable.