DEVELOPMENT OF AN ATLAS OF CANCER IN INDIA

First All India Report: 2001-2002

Discussion

The Indian Council of Medical Research (ICMR) initiated a network of cancer registries under the National Cancer Registry Programme (NCRP) in 1981 and data collection commenced in these registries from 1st January 1982. Since then, the registries have provided information on incidence and patterns of cancer that in terms of quality and validity meet international standards. This is evidenced by the fact that the data from the population based cancer registries under the NCRP has been continuously published in successive volumes of the World Health Organization (WHO) publication - Cancer Incidence in Five Continents (Muir et al, 1987; Parkin et al, 1992, 1997, 2002). This volume is published every five years by the International Agency for Research on Cancer - the cancer research arm of the WHO. Data on childhood cancer and cancer occurrence in developing countries have also appeared in the Agency's publications (Parkin et al, 1988; Sankarnaryanan et al, 1998). The NCRP, itself has been bringing out its own annual and consolidated reports since 1982 (NCRP Reports, 1985… 2002). The preparation for the 1999-2000 report is underway. Besides, the registries under NCRP there are eight population based cancer registries with funds met from other sources. The data of some of these registries has also been published in Cancer Incidence in Five Continents.

Thus, in India, for cancer, and perhaps for only this disease, we have a systematic programme of data collation so as to have reliable incidence and mortality rates, thereby laying a foundation for scientific research - whether that research be epidemiological, basic, clinical or in cancer control. However, India being a vast country, extensive areas need to be covered. The NCRP cancer registries cover selected urban centres and just one rural pocket. The other population cancer registries also cover essentially urban centres except for parts of two districts in Kerala State and part of one district in Tamil Nadu State. Consequently, the patterns of cancer in several urban centres and rural regions remain largely unknown. Setting up of new registries throughout the country as in some Western countries would involve enormous cost in establishing and maintaining the same.

Therefore, under this project, a cost-effective design and plan using advances in modern electronic information technology, was conceived, to collate and process relevant data on cancer so as to fulfill the objectives of:

The data of the NCRP shows that 80-85% of registered cases of cancer has microscopy as the basis of diagnosis. Modern electronic information technology needed to be harnessed in a cost-effective way. Thus the department of pathology and pathologist became the focal point of data capture and the internet was identified as the primary communication medium for data acquisition and transmission. Internet as a tool for data collection on patient information was a unique concept being tried for the first time (in India and to the best of our knowledge anywhere else in the world) under the project.

A user friendly 'core proforma' for collecting the patient information was hosted on the web-site www.canceratlasindia.org. Internet Browser based data entry eliminated the need for software to be installed on every system and the hassles of administration and maintenance. Collaborating centres were given an individual login-ID and password with detailed instructions on entering the core patient information. Care was taken to code/encrypt the data entered so that the identity or the nature of the data could not be deciphered by any one except those concerned with the project. This ensured confidentiality of patient information and security of data transmitted. The collaborating centres transmitted the required information (mainly patient identification details including area of living, and site and morphology of tumour) on all malignant cases reported by the department of pathology.

The successful working of this concept was reflected, first, in the data that was and is downloaded on a regular basis for the past two and a half years. The core data of approximately 1200-1500 cancer cases is received every week. Secondly, in the feedback received from the participating centres during the All India Workshop - 95% of the respondents, felt, that the web-site was easy to use and 80% of them had a fairly stable Internet connection. Thirdly, because most of the collaborating centres were able to transmit the required information as soon as a diagnosis was made, this report of 2001-2002 could be brought out fairly early (comparable to international standards). In due course the experience should enable us to provide the main tables of incidence rates and report soon after the end of the calendar year and then on-line. The fact that through this project, one could receive, analyse and provide the 2001-02 report in early 2004 posed problems for comparison, since international rates on a global basis were available only for 1993-97, and the rates of the PBCRs under NCRP were for the years 1997-99. However, from the epidemiological stand-point and knowing that incidence rates of cancer take several years before showing significant variations in time trends, these differences are unlikely to have notable impact in the interpretation of geographic patterns of cancer.

The data received through the web-site was downloaded periodically at the Coordinating Unit of the NCRP and the details of checks carried out have been described earlier (Chapter 2). Based on the above, a total of 2,17,174 microscopically diagnosed cancers for the two year period (1 January 2001 to 31 December 2002) from 105 centres across the country was taken up for analysis. The centres included the cancer registries under the NCRP and other functioning cancer registries. A condensed profile with tabular and graphic presentation of cancer patterns in these 105 centres is given in Chapter 7. This report is the culmination of sustained interest and efforts made by these participating centres.

The reference manual - Cancer Incidence in Five Continents (Parkin et al, 2002) was used to group neoplasms by site (WHO, ICD-10, 1994), calculate incidence rates and determine leading sites of cancer. The Census of India publications gives the population according to five-year age group and gender by district. As per the 2001 census results, there were 593 districts in the country. Information on cancer cases also gives the name of the district for each case, apart from age, gender, site and type of cancer.

Cancer incidence is generally expressed as age-adjusted or age standardized (according to world population) incidence rate per 100,000 persons. Therefore, the district was taken as a unit for calculation of incidence rates, with one difference. Unlike the regular age adjusted incidence rates (AAR) used in PBCRs throughout the world, minimum age adjusted incidence rates (MAAR) based on microscopically diagnosed cancers of the districts is used here.

The most recent data (1997-99) from the established population based cancer registries (PBCR) is included for description and comparison. The MAAR of cancer (for microscopically diagnosed cases) in these PBCRs was also calculated so as to have a benchmark for analysis, and, as a baseline for comparison with the MAAR of the districts. Under the NCRP, the population based cancer registry at Barshi is the only registry as of now, that has given incidence rates representative of the rural population in the country. Majority of the districts in the country has predominantly semi-urban or rural population. The MAAR for Barshi (all sites) for the period 1997-99, is 36.21 and 45.02 per 100,000 for males and females respectively. Thus the MAAR of 36.2 has been used as the cut off level to select districts for observing and compare cancer incidence and patterns. In all there were 82 districts that had a MAAR higher than 36.2/100,000 for at least one of the two years (2001 or 2002) and in either sex.

The information on cancer cases obtainable from a specified district and the calculation of incidence rates (in this instance the MAAR of at least 36.2/100,000) to be usable or workable for measuring and depicting patterns of cancer in that district depends on a number of factors. The availability and accessibility of nearby facilities for cancer diagnosis and treatment is the most important. The participation in this project of centre or centres is crucial. If there is just one comprehensive centre in a region and that centre collaborates in the project then the chances of receiving information on almost all cancer cases in those districts is high. This becomes reflected in the MAAR. All the eleven cancer registries under the NCRP and the other Hospital and Population Based Cancer Registries (not as yet under the NCRP network) have contributed to the project. Since nearly all of these registries are located at established regional cancer centres the districts served by them are suitably represented. In addition, centres like the Tata Memorial Centre, Mumbai and Cancer Institute, Chennai draw patients from several other parts of the country. On the other hand, if there are several cancer diagnosis and treatment facilities in a particular area and only some of the centres are collaborating in the project, then the chances of the information on cancers giving practicable MAAR are slim. The literacy and the general health awareness of the population are among other factors that could contribute to the extent of coverage of data on cancers.

In all there were 82 districts that had incidence rates (MAAR) higher than 36.2/100,000 for at least one of the two years (2001 or 2002) and in either sex. Of the six PBCRs under NCRP, in males, Delhi PBCR had the highest MAAR of 103.0/100,000 for all sites (ICD-10: C00-C96) of cancer. There were ten districts under the project that had a MAAR higher than that of Delhi. These included six districts in Mizoram State, one in the state of Kerala, North and South Goa, and Chandigarh. Among the urban PBCRs, Bangalore had the lowest MAAR of 75.1 per 100,000. There were eight districts that were above this MAAR but below that of Delhi, PBCR. The remaining fifty-one districts listed had MAAR lower than the urban PBCRs but above that of the rural PBCR at Barshi. Among females, Delhi PBCR had the highest MAAR of 113.9/100,000. There were four districts that had MAAR higher than this. These were in Mizoram State (three districts) and Chandigarh. Among the urban PBCRs, Bhopal PBCR had the lowest MAAR of 94.0/100,000. There were three districts that had a higher MAAR than that of Bhopal. There were forty-four districts that had a MAAR above that of Barshi PBCR, which in females was 45.0/100,000.

Chapter 6 furnishes a summary of important specific sites of cancer. It is the key chapter that presents the essential results and outcome of this project. It acts as a ready reckoner, for comparison of the MAARs of the districts with that of the PBCRs and of international and national AARs as well. Highlights of some sites are given below.

Tongue- Males: Bhopal, PBCR has an AAR of 10.9 and Ahmedabad urban registry has also a high AAR of 9.3/100,00. The district of Aizawl in Mizoram State, has a slightly higher MAAR compared to Bhopal, PBCR. There are several districts throughout the country that have a higher MAAR compared to the urban PBCRs. Of particular importance seems to be the State of Gujarat. Several districts (Mahesana, Gandhinagar, Kheda, Ahmedabad, Anand, Bhavnagar, Sabarkantha and Banaskantha) show high incidence rates of tongue cancer.

Mouth - Males: Again Bhopal, PBCR has a high AAR of 9.6/100,000. There were five districts that had a higher MAAR than that of Bhopal, PBCR. Wardha district in Maharashtra State has a MAAR of 14.1. Of the other four districts that had a higher MAAR than Bhopal, two were in Tamil Nadu State and two in Kerala State. At least twentysix other districts across the country had a higher MAAR than that of the other PBCRs. Other than several districts in Tamil Nadu State, many districts in Assam State (Kamrup, Goalpara, Darrang, Nalbari, Marigaon, Jorhat) showed a high MAAR.

Hypopharynx - Males: Among the PBCRs the AAR of cancer of the hypopharynx is high in Bhopal and Ahmedabad urban PBCR. Aizawl district in Mizoram State had a higher MAAR (16.1/100,000). Besides, numerous districts in Assam State (Dibrugarh, Kamrup, Darrang, Jorhat, Nalbari, Golaghat, Barpeta, Sibsagar, Goalpara to name a few) had high incidence rates.

Oesophagus - Males: The urban PBCRs have AARs varying from 6.3 to 10.3/100,000. The district-wise comparison of MAAR showed that Aizawl had a higher MAAR (26.7/100,000). Several districts especially in Assam and Karnataka State had MAAR comparable with the rates in the urban PBCRs. The districts of North and South Goa also had high MAARs.

Stomach - Males: Among males, Chennai and Bangalore PBCRs have had cancer stomach as the leading site of cancer since the commencement of the NCRP in 1982. But the AARs in these urban areas have been much lower than that seen in Japan or in other high incidence areas of the world. The district wise comparison of MAARs with that of Chennai and Bangalore showed that the district of Sercchip in Mizoram State had eight and a half times higher rate of stomach cancer than that of Chennai. Several districts in the North Eastern states of Mizoram, Nagaland, Manipur and Sikkim had MAARs equivalent to the AARs of high incidence regions of the world.

Gall Bladder - Females: Delhi females have shown a high incidence rate (AAR: 10.6/100,000) of cancer of the gall bladder. The district-wise comparison showed that Imphal East and West districts of Mizoram State and the Union Territory of Chandigarh had comparable incidence rates.

Lung - Males: The district-wise figures revealed that Aizawl in Mizoram State and Imphal West in Manipur State, had 1½ times the MAAR of the highest urban PBCR - Delhi (11.5/100,000). Further nine other districts had MAARs higher than the MAAR of Delhi.

Lung - Females: Except in Mumbai PBCR, cancer of the lung in females has not been a leading site of cancer in women, in the PBCRs under NCRP. Even the rate (AAR of 4.2/100,000) in Mumbai is lower than that seen in Indians in Singapore and in other women in areas of high incidence in the world. Observation of the MAARs in the districts showed that Aizawl women had almost ten times (26.2 compared to 2.8/100,000) the MAAR of women in Mumbai. Imphal West and East in Mizoram State and South Goa had much higher MAARs than that seen in Mumbai.

Breast - Females: Cancer of the breast has been replacing cancer of the cervix as the leading site of cancer in all urban PBCRs, except Chennai and the AARs of this site of cancer have also been on the rise. Among the Indian PBCRs, Delhi has the highest AAR of breast cancer. At least four districts led by Chandigarh (followed by North Goa, Aizawl in Mizoram State and Panchkula in Haryana State) had higher MAAR than that of Delhi. The rates were also similar in South Goa and three districts (Kollam, Thiruvananthapuram, Thrissur) in Kerala State.

Cervix Uteri: Chennai PBCR has had the highest incidence rate of cervical cancer among the Indian PBCRs. The district-wise MAARs indicate a region of high incidence rates even higher than Chennai in the North Eastern districts of Tamil Nadu State including Pondicherry which had the highest MAAR of 39.2/100,000.

Penis: In the Indian PBCRs penile cancer has been high in Chennai and Barshi. A high incidence of penile cancer was seen in the north eastern districts of Tamil Nadu and Villupuram district had a high MAAR of 3.1/100,000.

Thyroid - Females: Of the PBCRs under NCRP, Bangalore PBCR has shown the highest AAR of cancer of the thyroid. The PBCR at Thiruvananthapuram has shown a high incidence of cancer of the thyroid where it is the third leading site of cancer. Similarly, the district-wise distribution showed a higher MAAR in Thiruvananthapuram district, with a belt of high incidence right from the southern tip of the country - Kanniyakumari in Tamil Nadu State along the coast of the States of Kerala and Karnataka extending on to South Goa.