Elvira Singh

Childhood cancer incidence patterns by race, sex and age for 2000-2006: a report from the South African National Cancer Registry.

Higher childhood cancer incidence rates are generally reported for high income countries although high quality information on descriptive patterns of childhood cancer incidence for low or middle income countries is limited, particularly in Sub‐Saharan Africa. There is a need to quantify global differences by cancer types, and to investigate whether they reflect true incidence differences or can be attributed to under‐diagnosis or under‐reporting. For the first time, we describe childhood cancer data reported to the pathology report‐based National Cancer Registry of South Africa in 2000–2006 and compare our results to incidence data from Germany, a high income country. The overall age‐standardized incidence rate (ASR) for South Africa in 2000–2006 was 45.7 per million children. We observed substantial differences by cancer types within South Africa by racial group; ASRs tended to be 3–4‐fold higher in South African Whites compared to Blacks. ASRs among both Black and White South Africans were generally lower than those from Germany with the greatest differences observed between the Black population in South Africa and Germany, although there was marked variation between cancer types. Age‐specific rates were particularly low comparing South African Whites and Blacks with German infants. Overall, patterns across South African population groups and in comparison to Germans were similar for boys and girls. Genetic and environmental reasons may probably explain rather a small proportion of the observed differences. More research is needed to understand the extent to which under‐ascertainment and under‐diagnosis of childhood cancers drives differences in observed rates.

South African National Cancer Registry: Effect of withheld data from private health systems on cancer incidence estimates

BACKGROUND The National Cancer Registry (NCR) was established as a pathology-based cancer reporting system. From 2005 to 2007, private health laboratories withheld cancer reports owing to concerns regarding voluntary sharing of patient data. OBJECTIVES To estimate the impact of under-reported cancer data from private health laboratories. METHODS A linear regression analysis was conducted to project expected cancer cases for 2005-2007. Differences between actual and projected figures were calculated to estimate percentage under-reporting. RESULTS The projected NCR case total varied from 53,407 (3.8% net increase from actual cases reported) in 2005 to 54,823 (3.7% net increase) in 2007. The projected number of reported cases from private laboratories in 2005 was 26,359 (19.7% net increase from actual cases reported), 27,012 (18.8% net increase) in 2006 and 27,666 (28.4% net increase) in 2007. CONCLUSION While private healthcare reporting decreased by 28% from 2005 to 2007, this represented a minimal impact on overall cancer reporting (net decrease of <4%).

Establishment of a cancer surveillance programme : the South African experience

Cancer is projected to become a leading cause of morbidity and mortality in low-income and middle-income countries in the future. However, cancer incidence in South Africa is largely under-reported because of a lack of nationwide cancer surveillance networks. We describe present cancer surveillance activities in South Africa, and use the International Agency for Research on Cancer framework to propose the development of four population-based cancer registries in South Africa. These registries will represent the ethnic and geographical diversity of the country. We also provide an update on a cancer surveillance pilot programme in the Ekurhuleni Metropolitan District, and the successes and challenges in the implementation of the IARC framework in a local context. We examine the development of a comprehensive cancer surveillance system in a middle-income country, which might serve to assist other countries in establishing population-based cancer registries in a resource-constrained environment.

Hematologic malignancies in South Africa 2000-2006: analysis of data reported to the National Cancer Registry.

Little is known about the incidence patterns of hematologic malignancies in Sub‐Saharan Africa, including South Africa. We estimated incidence rates of pathology‐confirmed adult cases of leukemia, myeloma and related diseases (myeloma), Hodgkin lymphoma (HL), and non‐Hodgkin lymphoma (NHL) reported to the National Cancer Registry of South Africa (NCR) between 2000 and2006, by age, gender, and population group (Black, White, Coloured, Asian/Indian). Gender‐specific age‐standardized rates were calculated overall and by population group and incidence rate ratios (IRRs) were estimated using Poisson regression models. Between 2000 and 2006, there were 14662 cases of leukemia, myeloma, HL, and NHL reported to the registry. Incidence rates of reported hematologic malignancies were generally 20–50% higher among males than females. Our analyses suggested marked differences in the rates of reported hematologic malignancies by population group which were most pronounced when comparing the White versus Black population groups (IRRs ranging from 1.6 for myeloma to 3.8 for HL for males and females combined). Challenges related to diagnosis and reporting of cancers may play a role in the patterns observed by population group while the set‐up of the NCR (pathology‐based) could lead to some degree of under‐ascertainment in all groups. This is the first country‐wide report of the incidence of hematologic malignancies in South Africa. Despite challenges, it is important to analyze and report available national cancer incidence data to raise awareness of the cancer burden and to characterize patterns by demographic characteristics so as ultimately to improve the provision of cancer‐related health care.

The South African National Cancer Registry: an update

We would like to take this opportunity to correct the misconception created by the News piece published in The Lancet Oncology regarding the “demise” of the South African National Cancer Registry (NCR). The NCR, a division of the National Health Laboratory Service, is the primary cancer surveillance system and largest repository of cancer data in South Africa. It was established in 1986 as a voluntary, pathologybased cancer surveillance system, and continues to operate; its database contains over 1·2 million cancer records with about 80 000 new cases added each year. The misconception about the NCR referred to above might have come about because the NCR has had diffi culties in the past due to changes in leadership, high staff turnover, and a lack of political and fi nancial prioritisation of cancer surveillance caused by competing national health priorities at the peak of the HIV epidemic. As a result of severe resource constraints, the NCR continued to obtain data but was unable to regularly produce reports. Additionally, between 2004 and 2010, some private laboratories withheld data because of concerns about voluntarily submitting confi dential patient information to the registry without the protection of government legislation. However, in 2011, Regulation 380 of the National Health Act (Act 61 of 2003) formally established the NCR as South Africa’s main cancer surveillance agency; the legislation makes reporting all confi rmed cancer diagnoses to the registry obligatory. Additionally, the Regulation mandated the NCR to implement population-based cancer registration in South Africa. With renewed political support, NCR has been revitalised. To expedite data processing, the registry has moved to electronic reporting of pathology data from laboratories. Cancer incidence reports for 2000–07 have been published online with further reports in progress. After great eff ort from NCR staff , we aim to soon report cancers within timeframes similar to those of leading international cancer registries. The NCR has developed a 10-year business plan for the implementation of both new population-based and existing pathology-based registration for the country. Fundraising activities have started, and a pilot populationbased registry is operational in one district of the country (Ekurhuleni District, Gauteng), with more to follow. In view of the concern among stakeholders about the decline in reporting from private laboratories in the past, our researchers did a thorough investigation to quantify the eff ect of withheld reports on national cancer surveillance. We found a marginal eff ect on overall reporting; NCR will submit these fi ndings for peer-reviewed publication soon. The NCR remains the main source of cancer data for South Africa, with laboratory-based reporting yielding high specifi city. With the imminent implementation of national health insurance, it is vital for the country to be able to quantify the burden of cancer for national resource planning. Increased investment in the NCR will assist the organisation to improve South African cancer surveillance.

Design and methodology of a study on colorectal cancer in Johannesburg, South Africa: Design and methodology of a study on colorectal cancer

Cancer is one of the foremost causes of morbidity and mortality worldwide. Globally, colorectal cancer (CRC) is the third most diagnosed and fourth most important cause of cancer death. A total of 70% of all CRC‐related deaths occur in low‐ and middle‐income countries. In Sub‐Saharan Africa (SSA), estimating the burden of CRC is difficult. Only 27 of 43 SSA countries have formalized cancer registration systems; data quality is variable and national coverage rare.

Temporal trends in the epidemiology of cervical cancer in South Africa (1994-2012): Cervical cancer epidemiology in South Africa

Cervical cancer (CC) is the leading cause of cancer death among female South Africans (SA). Improved access to reproductive health services following multi‐ethnic democracy in 1994, HIV epidemic, and the initiation of CC population‐based screening in early 2000s have influenced the epidemiology of CC in SA. We therefore evaluated the trends in CC age‐standardised incidence (ASIR) (1994–2009) and mortality rates (ASMR) (2004–2012) using data from the South African National Cancer Registry and the Statistics South Africa, respectively. Five‐year relative survival rates and average per cent change (AAPC) stratified by ethnicity and age‐groups was determined. The average annual CC cases and mortalities were 4,694 (75,099 cases/16 years) and 2,789 (25,101 deaths/9 years), respectively. The ASIR was 22.1/100,000 in 1994 and 23.3/100,000 in 2009, with an average annual decline in incidence of 0.9% per annum (AAPC = −0.9%, p‐value < 0.001). The ASMR decreased slightly by 0.6% per annum from 13.9/100,000 in 2004 to 13.1/100,000 in 2012 (AAPC = −0.6%, p‐value < 0.001). In 2012, ASMR was 5.8‐fold higher in Blacks than in Whites. The 5‐year survival rates were higher in Whites and Indians/Asians (60–80%) than in Blacks and Coloureds (40–50%). The incidence rate increased (AAPC range: 1.1–3.1%, p‐value < 0.001) among young women (25–34 years) from 2000 to 2009. Despite interventions, there were minimal changes in overall epidemiology of CC in SA but there were increased CC rates among young women and ethnic disparities in CC burden. A review of the CC national policy and directed CC prevention and treatment are required to positively impact the burden of CC in SA.

Abstract A34: The genetic etiology of esophageal cancer in South African Black populations

Esophageal squamous cell carcinoma (ESCC) is common in Black populations of the sub-Saharan Africa, with high incidence regions in East and South Africa. Clinical presentation in Africa is late, and treatment is mainly palliative with a very poor prognosis. Various environmental risk factors have been identified, but the possible contribution of genetics to disease risk is an important question, which is unresolved. The aim of this project is to test the hypothesis that genetic variation in the South African Black population contributes significantly to the risk of ESCC. We have begun by testing genetic risk factors for ESCC which were identified by genome-wide association scans in other populations for association with ESCC in the South African Black population. Cases with a histologically confirmed diagnosis of ESCC and population controls were recruited, with informed consent and institutional ethical approval. All individuals were from the Black population of the Western Cape, with 98.2% being from the Xhosa ethnic group. A total of 513 ESCC cases and 820 controls were genotyped for 15 single nucleotide polymorphisms (SNPs) by validated TaqMan assays on a 7900HT (Applied Biosystems). SNPs were tested for association using an allelic chi-square test; odds ratios (OR) and 95% confidence intervals (CI) used the common allele as reference. A significance threshold of p=0.0033 was applied, from a Bonferroni multiple testing correction for 15 SNPs. No significant evidence of association was observed for the index SNPs at the previously reported loci ATP1B2, CASP8, c20orf54, HEATR3, PDE4D, PLCE1, PTPN2, RUNX1, SMG6, ST6GAL1, TMEM173 and UNC5L loci. However, the SNP rs2239815 at XBP1 on chromosome 22q12 was associated with ESCC (OR =1.41, P = 0.00087), as were 2 SNPs in CHEK2, rs4822983 (OR = 1.31, P = 0.0013) and rs1033667 (OR = 1.29, P = 0.0025). These 3 SNPs are within 70kb of each other and are in partial linkage disequilibrium in African populations. Fine-mapping of this locus will be required to determine the causal gene and variant/s driving this association. The lack of association with ESCC in the Black South African population at most of the loci previously associated in Asian GWAS suggests that at least some of these may not contribute to susceptibility to ESCC in the South African population. However, expansion of samples sizes and thus statistical power is required for more definitive exclusions, and the lower level of linkage disequilibrium (LD) across African genomes may produce a negative result if the SNP which is tagging the causal variant in Asian or European populations is in weak LD with the causal SNP in African populations. In order to address these issues we are now expanding the power and density of this study by genotyping 36 SNPs from loci of interest in 1000 South African ESCC cases and 940 controls from the Johannesburg Cancer Study by mass spectrometry using the MassArray genotyping system (Agena Bioscience). Citation Format: Wenlong Carl Chen, Hannah Bye, Marco Matejcic, Robyn Kerr, Elvira Singh, Natalie J. Prescott, Cathryn M. Lewis, Chantal Babb de Villers, Iqbal Parker, Christopher G. Mathew. The genetic etiology of esophageal cancer in South African Black populations [abstract]. In: Proceedings of the AACR International Conference: New Frontiers in Cancer Research; 2017 Jan 18-22; Cape Town, South Africa. Philadelphia (PA): AACR; Cancer Res 2017;77(22 Suppl):Abstract nr A34.

Abstract A11: A deep learning approach for extracting clinically relevant information from pathology reports

Data-Driven Healthcare, IBM Research Africa, Johannesburg, South Africa The National Cancer Registry (NCR) in South Africa plays a significant role in reporting nationwide cancer statistics and raising the global awareness of the massive impact of cancer. The government requires confirmed cancer cases to be reported to the NCR. Due to manual processes and the increasing magnitude of reports received annually, a considerable lag time exists in cancer statistics, which means the extent of the cancer cases is currently not understood. In addition, the unstructured free-text also needs to be processed in order to identify clinical information that could be important for public health planning. We present initial results from a deep learning approach to address this time lag. Deep learning is a powerful machine-learning algorithm that has made strides in the area of medical image recognition and speech processing. The deep learning system takes as input 2000 de-identified breast cancer pathology reports provided by the NCR in collaboration with the University of Witwatersrand Medical School. The pathology reports are first preprocessed using the Tf-idf (term frequency-inverse document frequency) method, which suggests how important a word is to a document in a corpus by assigning a numerical statistic to each word and hence obtain a term frequency document matrix. The high dimensional data matrix is, input into an unsupervised learning autoencoder, a data compression algorithm used to attain rich features that best represents the specific breast cancer topography and morphology. Unlike other approaches, our approach relies on non-dictionary sources such as clinical empirical knowledge extracted from the reports and dictionary sources such as the 12,000 medical diagnoses available in the International Statistical Classification of Diseases and Related Health Problems (ICD-10). The output from the deep learning system can be used to automate the classification of reports into their corresponding topography and morphology. The system could also be used to create a visual analytics system to aid data exploration and trend analysis of the current state of cancer in South Africa. Citation Format: Waheeda Banu Saib, Pavan Kumar, Geoffrey Siwo, Gciniwe Dlamini, Elvira Singh, Sue Candy, Michael Klipin. A deep learning approach for extracting clinically relevant information from pathology reports [abstract]. In: Proceedings of the AACR International Conference: New Frontiers in Cancer Research; 2017 Jan 18-22; Cape Town, South Africa. Philadelphia (PA): AACR; Cancer Res 2017;77(22 Suppl):Abstract nr A11.

Circulating microRNA’s as a diagnostic tool for hepatocellular carcinoma in a hyper endemic HIV setting, KwaZulu-Natal, South Africa: a case control study protocol focusing on viral etiology

BackgroundA wide range of studies has investigated the diagnostic proficiency of extracellular microRNAs (miRNAs) in hepatocellular cancer (HCC). HCC is expected to increase in Sub-Saharan Africa (SSA), due to endemic levels of viral infection (HBV/HIV), ageing and changing lifestyles. This unique aetiological background provides an opportunity for investigating potentially novel circulating miRNAs as biomarkers for HCC in a prospective study in South Africa.MethodsThis study will recruit HCC patients from two South African cancer hospitals, situated in Durban and Pietermaritzburg in the province of KwaZulu-Natal. These cases will include both HBV mono-infected and HBV/HIV co-infected HCC cases. The control group will consist of two (2) age and sex-matched healthy population controls per HCC case randomly selected from a Durban based laboratory. The controls will exclude patients if they have any evidence of chronic liver disease. A standardised reporting approach will be adopted to detect, quantify and normalize the level of circulating miRNAs in the blood sera of HCC cases and their controls. Reverse transcription quantitative polymerase chain reaction (RT-qPCR) will be employed to quantity extracellular miRNAs. Differences in concentration of relevant miRNA by case/control status will be assessed using the Wilcoxon rank-sum (Mann-Whitney U) test. Adjustment for multiple testing (Bonferroni correction), receiver operating curves (ROC) and optimal breakpoint analyses will be employed to identify potential thresholds for the differentiation of miRNA levels of HCC cases and their controls.DiscussionAlthough there is a growing base of literature regarding the role of circulating miRNAs as biomarkers, this promising field remains a ‘work in progress’. The aetiology of HBV infection in HCC is well understood, as well as it’s role in miRNA deregulation, however, the mediating role of HIV infection is unknown. HCC incidence in SSA, including South Africa, is expected to increase significantly in the next decade. A combination of factors, therefore, offers a unique opportunity to identify candidate circulating miRNAs as potential biomarkers for HBV/HIV infected HCC.