The NCRP reports incidence rates and mortality at population level, methods of data collection and quality control measures [5]. Usually population-based registries source information from (i) treatment facilities, such as cancer centres and major hospitals, (ii) diagnostic services, especially pathological, hematological, bioche-mical and immunological laboratories; (iii) death registration system. Active collection involves registry personnel actually visiting the different sources and abstracting the data on special forms. Passive reporting involves health-care workers completing the notification forms developed and distributed by the registry, or sending copies of discharge abstracts to the registry. The PBCRs use the PBCRDM 2.1 software applications installed at their institute to submit the data to NCRP. Relevant data from the hospital-based software application and the oncology modules transmitted by the hospitals through web-based portals can be routed to the PBCR data. PBCRDM 2.1 application can subject the data captured by them to different levels of checking, quality control, and duplicate checks and matching. Latest NCRP report (2016) provides data from 30 PBCRs covering <10% of the country’s population [4]. The 30 PBCRs were categorized into 6 regions (Table I).

TABLE I Childhood Cancer Incidenceas per Location of the Registry

Childhood cancer incidence (CCI) is generally expressed per million (pm) children [4]. For both boys and girls, CCI as age-adjusted rates for 11 selected broad types of childhood cancers (leukemias, lymphomas, central nervous system (CNS) tumors, sympathetic nervous system (SNS) tumors, retinoblastoma, renal tumors, hepatic tumors, bone tumors, soft tissue sarcomas, germ cell tumors and others) have been described in this article.

A rising trend in CCI in India has been observed ranging from 38-124 pm in 2001-2004, 11.3-159.6 pm in 2009-2011, and 6-860 pm in 2011-2014. However, childhood cancer recorded a marginal fall in proportion relative to cancers in all age groups; from 3.2% in 2001-2004, to 3.1% in 2009-2014 and 2.5% in 2011-2014. The minimum and maximum age-adjusted rates per million CCI from the 30 PBCRs and of the 11 broad types of childhood cancers along with their most common subtypes from 8 major registries have been presented in Tables I and Web Table I, respectively. Tables II and III presents CCI data divided in various age categories.

TABLE II	Age-wise Distribution* of Childhood Cancer Across Different Regions in India 

TABLE III  Incidence*of Common Types of Cancers in Children

In boys, the relative proportion of childhood cancer was lowest in Nagaland PBCR (North-east region) (0.7%) and highest in Delhi PBCR (North region) (5.4%). In girls, it varied from 0.5% in East Khasi Hill district PBCR (North-east region) to 3.5% in Naharlagun excluding Papumpara PBCR (North-east region). For both sexes, it varied from 0.7% in East Khasi Hill district PBCR to 4.4% in Chennai PBCR (South region).Male pediatric cancer had the highest incidence among 0-4 years age group in North, South and North-East regions and among 10-14 years age group in East, West and Central regions. Female pediatric cancers had the highest incidence among 0-4 years age group in North, West and East regions and among 10-14 years age group in South, Central and North-East regions. The highest incidence of cancer occurred in 0-4 years age group for males and both sexes combined and 10-14 years age group for females separately. Leukemia was the most predominant childhood cancer with highest incidence among 0-4 years for both sexes, 70% being lymphoid leukemia. Both Hodgkin’s and Non-Hodgkin’s disease had the highest incidence among 10-14 years age group for both sexes. 54% of all lymphomas were Non-Hodgkin’s disease. CNS tumours had the highest incidence among 5-9 years of age group for both sexes. Genitourinary, eye and liver tumours had highest incidence among 0-4 years age group while bone and gastrointestinal tumours had highest incidence among 10-14 years age group for both sexes.

The overall incidence of pediatric cancer has risen compared to the previous reviews, but the proportion of childhood cancer relative to cancers in all age groups have marginally reduced. Rapid industrialization and urbanization, acquired genetic mutations due to pollution, industrial disasters, rampant insecticides use in the agricultural sector, tobacco and gutka addiction among child labourers from lower socio-economic class and ozone layer depletion have been postulated as causes for rising pediatric cancer incidence [5]. The reported incidence of childhood cancer in males (4-556 pm) is higher than in females (2-309 pm); the ratio being 1.6:1, which is higher than in the developed world (1.2:1), possibly reflecting gender bias in seeking healthcare [3,6]. Leukemia remains the most common pediatric cancer for both sexes, followed by lymphoma and CNS tumours in males. However, CNS tumours exceed lymphomas in females. In the developed world, CNS tumours are the second most common cancer (22-25%) and lymphomas are third (10%) [7]. Non-Hodgkin’s disease exceeds Hodgkin’s disease in this review, a pattern similar to the developed world, although previous reports [3] from India had noted higher incidence of Hodgkin’s disease.

Limitations of this analysis is that the PBCR data includes <10% of the population with inadequate rural representation (except Barshi rural registry), since majority of the PBCRs are based in bigger cities. Also, the distribution of PBCRs is highly non-uniform across the country with fewer registries in the North and East regions. Hospitals in major cities like Delhi cater to larger population referred from nearby and distant states and hence, report higher CCI. Information of incidence of different subtypes of CNS, hepatic, renal and bone tumors is available in data from hospital-based registries of 8 important cities only and not from other PBCRs.

The considerable inter-regional variation in incidence and mortality rates across India suggests a possibility of deficiency in ascertainment of cases and death notification, and variability of compliance to cancer registration particularly in the rural areas [3].Studies based on data of different PBCRs in India are few [8-13]. This will serve as a reference source for researchers and would also act as stimulus for further research on the epidemiology of childhood cancer.

Contributors: DKP: conceptualized the study and provided intellectual inputs in to manuscript writing, SD: interpretation and analysis of data and writing the manuscript; KA and SB: planned presentation of the manuscript.