Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Incidence of haematological malignancy by sub-type: a report from the Haematological Malignancy Research Network

Incidence of haematological malignancy by sub-type: a report from the Haematological Malignancy... Clinical Studies British Journal of Cancer (2011) 105, 1684 – 1692 & 2011 Cancer Research UK All rights reserved 0007 – 0920/11 www.bjcancer.com Incidence of haematological malignancy by sub-type: a report from the Haematological Malignancy Research Network ,1 1 2 3,4 1,4 A Smith , D Howell , R Patmore , A Jack and E Roman 1 2 Epidemiology and Genetics Unit, Department of Health Sciences, University of York, York, UK; Queens Centre for Oncology and Haematology, Castle Hill Hospital, Hull, UK; Haematological Malignancy Diagnostic Service, St James’s Institute of Oncology, Bexley Wing, St James’s University Hospital, Leeds, UK; Hull York Medical School, Heslington, York, UK BACKGROUND: Ascertainment of cases and disease classification is an acknowledged problem for epidemiological research into haematological malignancies. METHODS: The Haematological Malignancy Research Network comprises an ongoing population-based patient cohort. All diagnoses (paediatric and adult) across two UK Cancer Networks (population 3.6 million, 42000 diagnoses annually, socio-demographically representative of the UK) are made by an integrated haematopathology laboratory. Diagnostics, prognostics, and treatment are recorded to clinical trial standards, and socio-demographic measures are routinely obtained. RESULTS: A total of 10 729 haematological malignancies (myeloid¼ 2706, lymphoid¼ 8023) were diagnosed over the 5 years, that is, from 2004 to 2009. Descriptive data (age, sex, and deprivation), sex-specific age-standardised (European population) rates, and estimated UK frequencies are presented for 24 sub-types. The age of patients ranged from 4 weeks to 99 years (median 70.6 years), and the male rate was more than double the female rate for several myeloid and lymphoid sub-types, this difference being evident in both children and adults. No relationship with deprivation was detected. CONCLUSION: Accurate population-based data on haematological malignancies can be collected to the standard required to deliver reproducible results that can be extrapolated to national populations. Our analyses emphasise the importance of gender and age as disease determinants, and suggest that aetiological investigations that focus on socio-economic factors are unlikely to be rewarding. British Journal of Cancer (2011) 105, 1684–1692. doi:10.1038/bjc.2011.450 www.bjcancer.com Published online 1 November 2011 & 2011 Cancer Research UK Keywords: descriptive epidemiology; incidence; lymphoma; leukaemia; myeloma; socio-economic status To originate and test hypotheses about pathogenesis, it is vitally common, accounting for around 9% of all cancers and being the important to accurately describe the underlying disease patterns fourth most frequently diagnosed cancer in both men (after and trends (Parkin, 2006; Boyle, 2008; Ferlay et al, 2010). This prostate, lung, and colorectum) and women (after breast, lung, requires complete ascertainment of cases within a defined and colorectum) in economically developed regions of the world. population, as well as the application of appropriate disease However, over and above basic tallies, the usefulness of classifications, and it is the attainment of these two key these descriptive data for epidemiological research is constrained components that continues to beleaguer epidemiological research by the classification system applied, which for haematological into the haematological malignancies (National Institute for malignancies is largely rooted in the gradual recognition of disease Clinical Excellence, 2003; Sant et al, 2009, 2010). entities at the beginning of the twentieth century. In the 1980s and Traditionally, the descriptive epidemiology of haematological 1990s, however, several competing classifications emerged as malignancies considers four broad categories – leukaemia, understanding about the relationship between the various Hodgkin lymphoma, non-Hodgkin lymphoma, and myeloma; with haematological malignancies, the bone marrow, and the immune national and global organisations including the USA’s Surveillance system, and the cellular and genetic basis of malignant transfor- Epidemiology and End Results Program (www.seer.cancer.gov), mation gradually increased. In the early 1980s, for example, the the UK’s National Cancer Intelligence Network (www.ncin.org.uk), Working Formulation, which was developed as a method of and the World Health Organization (WHO)’s International Agency translating between the many competing lymphoid classifications, for Research on Cancer (http://globocan.iarc.fr/), routinely pub- rapidly became the standard in North America, and many lishing data in this format (National Cancer Intelligence Network, epidemiological studies conducted there were based on this 2008; Westlake, 2009; Jemal et al, 2010, 2011). Such counts show system. At the same time, the majority of European centres used that, as a group, haematological neoplasms are comparatively the Kiel classification, making effective comparison of results between North America and Europe almost impossible (Harris et al, 1994). In 2001, the WHO produced, for the first time, a consensus *Correspondence: Dr A Smith; E-mail: alex.smith@egu.york.ac.uk classification that defined all haematological malignancies in terms Received 20 June 2011; revised 26 September 2011; accepted 27 of immunophenotype, genetic abnormalities and clinical features September 2011; published online 1 November 2011 Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al (World Health Organization, 2001), and this is incorporated into 2010). In recognition of this fact, a number of methods have been the current version of the International Classification of Diseases applied in an attempt to generate more informative descriptive for Oncology (ICD-O3); (Fritz, 2000). The strategy adopted was data, including, for example, the application of bridge-coding based on the principle devised for the Revised European– algorithms to historically coded data (Sant et al, 2010; Turner et al, American Classification of Lymphoid Neoplasms, which was 2010; Maynadie et al, 2011) and the reporting of specialist hospital- introduced in the mid-1990s (Harris et al, 1994). Unfortunately, based case-series frequencies (Yoon et al, 2010; Mozaheb et al, however, although the new WHO classification and its successor 2011). Inevitably, however, the accuracy and completeness of data (WHO, 2008) were adopted into clinical practice almost uniformly generated by such initiatives has continued to pose serious around the world, there was no immediate effect on population- interpretative problems for both researchers and health service based cancer information systems, where the practice of grouping planners. haematological malignancies into the four broad groups defined in In response to these challenges, the Haematological Malignancy the tenth revision of ICD-10 (www.who.int/classifications/icd/en/) Research Network (HMRN) was established in the UK in 2004 has tended to continue (National Cancer Intelligence Network, (Smith et al, 2010). HMRN is predicated on the framework of the 2008; Westlake, 2009; Ferlay et al, 2010; Jemal et al, 2010, 2011). UK National Health Service, where 37 cancer networks are This is largely because unlike many other cancers, haematological responsible for bringing together health service commissioners neoplasms are diagnosed using multiple parameters, including a and providers, the voluntary sector and local authorities to deliver combination of histology, cytology, immunophenotyping, cytoge- high quality cancer care. HMRN presently covers two such cancer netics, imaging, and clinical data. This range and depth of data is networks (Figure 1A), which comprise a single clinical network difficult for cancer registries and other researchers to access (population 3.6 million, with over 2000 new haematological systematically, potentially forming a barrier not only to the neoplasms diagnosed each year), and the present report examines collection of diagnostic data at the level of detail required to the socio-demographic characteristics of patients diagnosed over systematically implement the latest WHO classification (WHO, the first 5 years of the project, that is from 1 September 2004 to 31 2001; WHO, 2008), but also to complete ascertainment (Sant et al, August 2009. HMRN % England % Village, hamlet & isolated 0.8 0.7 dwellings - sparse Village, hamlet & isolated 7.8 8.4 dwellings - less sparse Town & fringe - sparse 0.6 0.5 Town & fringe - less sparse 10.9 9.0 Urban > 10K - sparse 0.4 0.2 Urban > 10K - less sparse 79.6 81.2 90+ HMRN males 85–89 UK males 80–84 HMRN females 75–79 UK females 70–74 65–69 60–64 55–59 HMRN % England % 50–54 1 (Affluent) 16.2 20.0 45–49 2 21.2 20.0 40–44 3 18.1 20.0 35–39 4 19.1 20.0 30–34 5 (Deprived) 25.4 20.0 25–29 20–24 15–19 10–14 5–9 0–4 8642 0 2468 Percent (%) Figure 1 Socio-demographic structure of Haematological Malignancy Research Network (HMRN). (A) Map of study area. (B) Population, age, and sex structure. (C) Office for National Statistics urban/rural definition. (D) Index of multiple deprivation – income domain. & 2011 Cancer Research UK British Journal of Cancer (2011) 105(11), 1684 – 1692 Clinical Studies Clinical Studies Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al MATERIALS AND METHODS analyses were conducted in the statistical package STATA 11. (Stata-Corp., 2010) Incidence rates, sex rate ratios, and 95% Haematological Malignancy Research Network (www.hmrn.org) is an confidence intervals (CIs) were estimated by Poisson regression; ongoing population-based cohort of patients (adult and paediatric) directly age-standardised rates were calculated using the Stata newly diagnosed with a haematological malignancy. It is a unique command dstdize, and indirectly standardised-incidence ratios venture, combining the expertise of a single integrated haemato- (SIR) were calculated using the Stata command istdize. pathology laboratory, a unified clinical network (comprising the Descriptive findings are presented here for 10 729 haemato- Yorkshire and Humber and Yorkshire Coast Cancer Networks), and a logical malignancies diagnosed within the HMRN region over specialist epidemiology unit, and full details of its structure, data- 5 years spanning September 2004 to August 2009. For analytical collection methods, and ethical approvals have been described in purposes, these diagnoses coded to ICD-O3 are grouped into detail elsewhere (Smith et al, 2010). Briefly, as a matter of policy, all 24 main WHO categories; the codes that comprise these groups diagnoses within the clinical network are made and coded by clinical are published on our website and in Supplementary Table 1 specialists to the latest WHO classification at a single integrated (www.hmrn.org/Info/Disease_Classification.aspx). haematopathology laboratory – the Haematological Malignancy Diagnostic Service (www.HMDS.info) – which was cited in the UK Department of Health’s Cancer Reform Strategy as ‘the model for RESULTS delivery of complex diagnostic services’. Following diagnosis, and with an emphasis on obtaining primary-source data, information is With a combined population of around 3.6 million, the socio- abstracted from medical records and laboratory reports to clinical demographic structure of HMRN is broadly representative of the trial standards, and all diagnostic, prognostic, treatment, and national population in terms of age, sex, urban/rural status, and outcome data are linked and held in a central database. deprivation (Figure 1). The 2001 age–sex distribution is compared Populations and area-based measures of urban/rural status, and with the UK (58.8 million) in Figure 1B, but in line with national data deprivation are routinely obtained from the UK census and other release policies, the urban/rural and deprivation configurations are national data sources (Office for National Statistics, 2001; ONS compared with England alone (49.1 million) in Figure 1C and 1D, Geography, 2004). For the purposes of the present report, subjects respectively. Although the age/sex distributions (Figure 1B) and were given a measure of area-based deprivation assigned on the urban/rural residence patterns (Figure 1C) closely mirror those of lower super output area, where they were resident at the time of the national population, the HMRN region contains proportionately diagnosis. In common with other reports (Shack et al, 2008; more people living in areas classified as deprived and fewer in areas National Cancer Intelligence Network, 2009; Department of Health, classified as affluent (Figure 1D). 2010), the income domain of the index of multiple deprivation The 10 729 haematological malignancies diagnosed over the 5 (IMD) was used (quintile one containing the most affluent fifth of years from September 2004 to August 2009 are distributed by England’s lower super output areas, and quintile five the least). All sub-type in Table 1. Myeloid malignancies, which comprise around Table 1 Numbers and annual crude rates per 100 000 (95% CI), median ages at diagnosis: Haematological Malignancy Research Network (HMRN), 2004–2009 Median age Annual rates (95% CI) per 100 000 Sex-rate ratio: Total at diagnosis male/female Neoplasm (common abbreviation/synonym) diagnoses (years) Total Males Females (95% CI) All diagnoses 10 729 70.6 60.1 (58.9 – 61.2) 68.5 (66.8 – 70.3) 52.1(50.7 – 53.6) 1.31 (1.26 – 1.36) Total myeloid 2706 72.0 15.1 (14.6 – 15.7) 16.9 (16.0 – 17.8) 13.5 (12.8 – 14.3) 1.25 (1.16 – 1.35) Chronic myelogenous leukaemia (CML) 165 59.1 0.9 (0.8 – 1.1) 1.1 (0.9 – 1.4) 0.7 (0.6 – 0.9) 1.48 (1.07 – 2.05) Primary myelofibrosis 73 73.0 0.4 (0.3 – 0.5) 0.6 (0.4 – 0.7) 0.3 (0.2 – 0.4) 2.04 (1.24 – 3.46) Chronic myeloproliferative neoplasms (MPN) 961 71.1 5.4 (5.0 – 5.7) 4.8 (4.3 – 5.3) 6.0 (5.5 – 6.5) 0.80 (0.70 – 0.91) Chronic myelomonocytic leukaemia (CMML) 96 76.1 0.6 (0.5 – 0.7) 0.8 (0.6 – 1.0) 0.4 (0.3 – 0.5) 2.10 (1.38 – 3.25) Myelodysplastic syndromes (MDS) 653 76.1 3.7 (3.4 – 3.9) 5.0 (4.5 – 5.5) 2.4 (2.1 – 2.7) 2.09 (1.78 – 2.48) Acute myeloid leukaemias (AML) 717 68.7 4.0 (3.7 – 4.3) 4.5 (4.0 – 4.9) 3.6 (3.2 – 4.0) 1.25 (1.07 – 1.45) Total lymphoid 8023 70.1 44.9 (44.0 – 45.9) 51.6 (50.1 – 53.2) 38.6 (37.4 – 39.9) 1.34 (1.28 – 1.40) Precursor B-lymphoblastic leukaemia (B-ALL) 167 12.7 0.9 (0.8 – 1.1) 1.0 (0.8 – 1.2) 0.9 (0.7 – 1.1) 1.13 (0.82 – 1.55) Precursor T-lymphoblastic leukaemia (T-ALL) 50 18.5 0.3 (0.2 – 0.4) 0.4 (0.4 – 0.5) 0.2 (0.1 – 0.3) 1.89 (1.03 – 3.58) Monoclonal B-cell lymphocytosis (MBL) 445 71.7 2.5 (2.3 – 2.7) 2.7 (2.4 – 3.1) 2.3 (2.0 – 2.6) 1.18 (0.98 – 1.43) Chronic lymphocytic leukaemia (CLL) 1145 71.6 6.4 (6.0 – 6.8) 8.1 (7.5 – 8.7) 4.8 (4.4 – 5.3) 1.69 (1.50 – 1.91) Marginal zone lymphomas (MZL) 530 71.5 3.0 (2.7 – 3.2) 3.4 (3.1 – 3.9) 2.5 (2.2 – 2.9) 1.37 (1.15 – 1.63) Hairy-cell leukaemia (HCL) 55 65.7 0.3 (0.2 – 0.4) 0.5 (0.3 – 0.7) 0.1 (0.1 – 0.2) 3.44 (1.81 – 6.98) Monoclonal gammopathy of undetermined significance (MGUS) 1135 72.2 6.4 (6.0 – 6.7) 7.4 (6.8 – 8.0) 5.4 (4.9 – 5.9) 1.38 (1.23 – 1.56) Plasma cell myeloma (multiple myeloma) 1127 73.0 6.3 (5.9 – 6.7) 7.5 (6.9 – 8.1) 5.2 (4.8 – 5.7) 1.43 (1.27 – 1.61) Plasmacytoma 99 69.2 0.6 (0.5 – 0.7) 0.8 (0.6 – 1.0) 0.4 (0.3 – 0.5) 2.03 (1.32 – 3.18) Follicular lymphomas (FL) 547 64.6 3.1 (2.8 – 3.3) 2.9 (2.6 – 3.3) 3.2 (2.8 – 3.6) 0.92 (0.77 – 1.09) Mantle-cell lymphoma 144 74.0 0.8 (0.7 – 0.9) 1.1 (0.9 – 1.3) 0.6 (0.4 – 0.7) 1.88 (1.33 – 2.70) Diffuse large B-cell lymphomas (DLBCL) 1417 70.4 7.9 (7.5 – 8.4) 8.4 (7.8 – 9.1) 7.5 (6.9 – 8.1) 1.13 (1.01 – 1.26) Burkitt lymphoma (BL) 66 52.2 0.4 (0.3 – 0.5) 0.6 (0.4 – 0.8) 0.2 (0.1 – 0.3) 3.33 (1.86 – 6.26) Lymphoproliferative disorders NOS (LPD) 330 76.9 1.8 (1.7 – 2.3) 2.0 (1.7 – 2.3) 1.7 (1.4 – 2.0) 1.19 (0.95 – 1.87) T-cell leukaemia 64 74.7 0.4 (0.3 – 0.5) 0.3 (0.2 – 0.5) 0.4 (0.3 – 0.5) 0.94 (0.56 – 1.58) T-cell lymphoma 188 65.7 1.1 (0.9 – 1.2) 1.2 (1.0 – 1.5) 0.9 (0.7 – 1.1) 1.44 (1.07 – 1.94) Nodular lymphocyte predominant Hodgkin lymphoma (NLPHL) 50 42.9 0.3 (0.2 – 0.4) 0.4 (0.4 – 0.5) 0.2 (0.1 – 0.3) 2.07 (1.12 – 3.95) Classical Hodgkin lymphoma (CHL) 464 41.2 2.6 (2.4 – 2.8) 2.9 (2.5 – 3.3) 2.3 (2.0 – 2.7) 1.23 (1.02 – 1.49) Abbreviations: CI¼ confidence interval; NOS¼ not otherwise specified. British Journal of Cancer (2011) 105(11), 1684 – 1692 & 2011 Cancer Research UK Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al a quarter of the total (N¼ 2706) are presented first, and lymphoid, within the traditional lymphoma and myeloma groupings, there is which account for the remaining malignancies (N¼ 8023), are less diversity in the cell type of origin; with mature B-cell second. Data on median ages at the time of diagnosis, annual rates, malignancies dominating. Indeed, with an annual rate of 7.9 per and sex-rate ratios (male rate/female rate) with 95% CIs are also 100 000 per year, diffuse large B-cell lymphoma is the most given in Table 1. The rates in Table 1 are ordered by magnitude common haematological malignancy, and chronic lymphocytic in Figure 2, and the bars are colour coded, differentiating the leukaemia (CLL), which like diffuse large B-cell lymphoma is also a traditional groupings of leukaemia, non-Hodgkin lymphoma, mature B-cell neoplasm, is the next most common. Hodgkin lymphoma, and myeloma from other haematological As detailed in the Introduction, the classification of haemato- neoplasms that are less consistently registered by national logical malignancies has changed markedly in recent decades, with schemes. The classic ICD-10 leukaemia group contains a mix of several conditions once classified as neoplasms of uncertain or myeloid and lymphoid conditions, the latter including both unknown behaviour now being categorised as malignant, and precursor and mature B-cell and T-cell subtypes. By contrast, other conditions being recognised as part of the cancer continuum for the first time; the disorders falling into this category are shaded grey in Figure 2. Chronic myeloproliferative neoplasms and 0 2468 myelodysplastic syndromes now comprise around two-thirds of Chronic myeloproliferative neoplasms the myeloid neoplasms assigned a behaviour code of 3 (malignant Acute myeloid leukaemia Myelodysplastic syndromes primary site) in the WHO ICD-O3. Within the lymphoid category, Chronic myelogenous leukaemia lymphoproliferative disorders not otherwise specified also Chronic myelomonocytic leukaemia contains a mix of malignancies, all with behaviour codes of 3. Primary myelofibrosis However, although monoclonal gammopathy of undetermined Diffuse large B-cell lymphoma significance (MGUS) and monoclonal B-cell lymphocytosis (MBL) Chronic lymphocytic leukaemia are both conditions in which neoplastic B-cells are detectable, the Monoclonal gammopathy of undetermined significance Plasma cell myeloma diagnostic criteria for MBL being where the peripheral blood Follicular lymphoma 9 B-cell count is less than 5 10 /l lymphocytes (and in which risks Marginal zone lymphoma Classical Hodgkin lymphoma of progression to myeloma in the case of the former and CLL in the Monoclonal B-cell lymphocytosis case of the latter are elevated), their behaviour remains uncertain. Lymphoproliferative disorder NOS As with most other cancers, the likelihood of being diagnosed T-cell lymphoma Precursor B-lymphoblastic leukaemia with a haematological malignancy increases markedly with age, Mantle cell lymphoma the median age at diagnosis being 70.6 years within the HMRN Plasmacytoma region (Table 1). However, unlike many other common cancers, Burkitt lymphoma T-cell leukaemia haematological malignancy can be diagnosed at any age, with Hairy-cell leukaemia different subtypes dominating at different ages. More information Precursor T-lymphoblastic leukaemia about the age distributions of the various subtypes is presented in Nodular lymphocyte predominant Hodgkin lymphoma Figure 3, which shows box-and-whisker (boxplots) summary age Hodgkin lymphoma Non-Hodgkin lymphoma Leukaemia plots ordered by the magnitude of the median for myeloid and Unknown or uncertain behaviour lymphoid malignancies separately. The interquartile range is Myeloma represented by the box, with outliers occurring outside the maximum data series of 1.5 times the interquartile range being Figure 2 Annual crude rates per 100 000: Haematological Malignancy Research Network (HMRN), 2004–2009. shown as separate points. Chronic myelogenous leukaemia Acute myeloid leukaemia Chronic myeloproliferative neoplasms Primary myelofibrosis Myelodysplastic syndromes Chronic myelomonocytic leukaemia Precursor B-lymphoblastic leukaemia Precursor T-lymphoblastic leukaemia Classical Hodgkin lymphoma Nodular lymphocyte predominant Hodgkin lymphoma Burkitt lymphoma Follicular lymphoma Hairy cell leukaemia T-cell lymphoma Plasmacytoma Diffuse large B-cell lymphoma Marginal zone lymphoma Chronic lymphocytic leukaemia Monoclonal B-cell Lymphocytosis Monoclonal gammopathy of undetermined significance Plasma cell myeloma Mantle cell lymphoma T-cell leukaemia Lymphoproliferative disorder NOS 0 10 20 30 40 50 60 70 80 90 100 Age at diagnosis (years) Figure 3 Age at diagnosis distributions: Haematological Malignancy Research Network (HMRN), 2004–2009. & 2011 Cancer Research UK British Journal of Cancer (2011) 105(11), 1684 – 1692 Clinical Studies Clinical Studies Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al The majority of myeloid conditions are diagnosed above 70 the point estimates at younger ages being similar to those at older years of age, but sporadic cases arise at younger ages (Figure 3). ages, although the CIs are wider, reflecting the comparative Likewise, lymphoid malignancies generally occur in older people, sparsity of the data. but nonetheless span the entire age range, our youngest diagnosis In the 24 main diagnostic categories listed in Table 1, 16 had 100 having been made at 4 weeks of age and oldest at 99 years. The diagnoses or more, and for these the SIRs and 95% CIs are plotted precursor B-cell and T-cell lymphoblastic leukaemias tend to occur by index of multiple deprivation income domain quintile in at the youngest ages, the medians being 12.7 years and 18.5 years, Figure 6, (group 1 being the most affluent and group 5 being the respectively (Table 1). However, as with some of the conditions most deprived). No trends with deprivation are evident, although that principally occur at older ages, such as diffuse large B-cell for some malignancies there is an indication of a deficit in the lymphoma, these too are periodically diagnosed outside their most deprived quintile; the most notable being myeloma where the normal age range. Such wide age spans are however not seen for all SIR in category 5 is significantly below 1.0 (0.82 (95% CI 0.71– lymphoid conditions, including the rarer forms like hairy-cell 0.95). leukaemia and mantle-cell lymphoma, and also comparatively The lack of a trend with deprivation (Figure 6) is particularly common conditions like CLL and myeloma – all of which seldom, pertinent to precursor B-lymphoblastic leukaemia and classical if ever, occur below the age of 30 years. A further conspicuous Hodgkin lymphoma (CHL), both of which have been suggested to feature of lymphoid neoplasms is the similarity in the age be increasingly common in more affluent families and communities. distributions of certain closely related conditions such as MBL B-lymphoblastic leukaemia is primarily paediatric (Figure 3), and it (median 71.7 years) and CLL (median 71.6 years), as well is in this age group that an association with socio-economic status as monoclonal gammopathy of underdetermined significance has been suggested. In our data, however, the results were similar (median 72.2 years) and myeloma (median 73.0 years), lying when the analysis was restricted to cases diagnosed before the adjacent to each other in Figure 3. age of 15 years (95 out of 167); SIRs (95% CIs) for deprivation In general, haematological malignancies tend to occur more categories 1 through 5, respectively, being 0.7 (0.4–1.3), 1.0 frequently in males than females, and for many conditions, the rate (0.6–1.6), 1.1 (0.6–1.7), 1.0 (0.5–1.6), and 1.2 (0.8–1.6). Likewise, among males is more than twice that of females (Table 1). The for CHL, the strongest effects have been reported at younger ages consistency of the gender difference is plainly visible in Figure 4, where the nodular sclerosis form of CHL predominates. In our data, which shows the sex-specific rate ratios (male rate/female rate) no associations with deprivation were observed, either for total CHL ordered by magnitude. Indeed, conditions with no apparent sex or for any of the CHL subtypes (data not shown). bias, such as the chronic myleoproliferative neoplasms (male rate/ The size and demographic similarity of HMRN’s population to female rate¼ 0.80, 95% CI 0.70–0.91) and follicular lymphoma the general UK population (Figure 1) means that the HMRN’s data (male rate/female rate¼ 0.92, 95% CI 0.77–1.09), stand out from can reasonably be extrapolated to the country as a whole. The the rest (Table 1). The lymphoid group exhibits some of the most estimated UK totals, calculated by applying HMRN’s age-specific striking sex differences, the rates of the comparatively rare Burkitt rates to the corresponding general population age strata are shown lymphoma and hairy-cell leukaemia being more than three times in Table 2. For the purposes of wider comparability, age- higher in males than in females. These sex differences occur across standardised rates (European population) are also given in Table 2; the full age spectrum, being seen in conditions with comparatively these rates are in general lower than the actual rates (Table 1), low, as well as high, median ages at diagnosis such as mantle-cell reflecting the fact that unlike the real population (Figure 1), the lymphoma (median age at diagnosis 74 years) and precursor hypothetical standard has a younger age structure with no excess T-lymphoblastic leukaemia (median age at diagnosis 18.5 years), of females in the older age groups. For the sake of completeness, for example, both with ratios approaching 2.0 lying adjacent to information on MBL and monoclonal gammopathy of undeter- each other in Figure 4. The consistency of the gender bias is further mined significance are included in Table 2, but their data are illustrated in Figure 5, which shows the sex-rate ratios plotted in excluded from the overall totals. 10-year age groups for all haematological malignancies combined; Sex rate ratio DISCUSSION (95% confidence intervals) 0 2 4 6 8 Our ability to calculate reliable incidence rates for clinically Chronic myeloproliferative neoplasms meaningful haematological malignancy subtypes is a fundamental Acute myeloid leukaemia Chronic myelogenous leukaemia Primary myelofibrosis Myelodysplastic syndromes Chronic myelomonocytic leukaemia Follicular lymphoma T-cell leukaemia Diffuse large B-cell lymphoma 1.5 Precursor B-lymphoblastic leukaemia Monoclonal B-cell lymphocytosis Lymphoproliferative disorder NOS Classical Hodgkin lymphoma Marginal zone lymphoma Monoclonal gammopathy of undetermined significance Plasma cell myeloma T-cell lymphoma Chronic lymphocytic leukaemia Mantle cell lymphoma Precursor T-lymphoblastic leukaemia 0.5 Plasmacytoma Nodular lymphocyte predominant Hodgkin lymphoma 0−9 10−19 20−29 30−39 40−49 50−59 60−69 70+ Burkitt lymphoma Age at diagnosis (years) Hairy cell leukaemia Sex-rate ratio 95% confidence interval Sex Rate Ratio 95% Confidences Interval Figure 4 Sex-rate ratios: Haematological Malignancy Research Network Figure 5 Sex-rate ratios by age: Haematological Malignancy Research (HMRN), 2004 –2009. Network (HMRN), 2004 –2009. British Journal of Cancer (2011) 105(11), 1684 – 1692 & 2011 Cancer Research UK Sex-rate ratio (95% confidence intervals) Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al Chronic myelogenous leukaemia Chronic myeloproliferative neoplasms Myelodysplastic syndromes Acute myeloid leukaemia Precursor B-lymphoblastic leukaemia Monoclonal B-cell lymphocytosis Chronic lymphocytic leukaemia Marginal zone lymphoma MGUS Plasma cell myeloma Follicular lymphoma Mantle cell lymphoma Diffuse large B-cell lymphoma Lymphoproliferative disorder NOS T-cell lymphoma Classical Hodgkin lymphoma 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Deprivation quintile SIR 95% confidence intervals Figure 6 Standardised-incidence ratios (SIR) by index of multiple deprivation (IMD) income domain. key research achievement; the analyses revealing notable associa- another aspect that challenges cancer registries. For example, the tions with both age and sex, contrasting somewhat starkly with the present report is based on 10 729 diagnoses, but these relate to comparative lack of variation with area-based measures of 10 306 people diagnosed with a haematological cancer for the first deprivation. In addition, the size and representative nature time, of whom 407 (3.9%) had a second haematological neoplasm diagnosed, either concurrently or because their disease progressed of our study population mean that our data can be extrapolated to the UK as a whole, providing for the first time, national or transformed, and 16 (o1%) had a third diagnosis. Investigating estimates for the main WHO-defined disease entities (WHO, 2008). the epidemiology of transformation and progression, as well as Indeed, HMRN rates could be applied to any well-characterised other outcomes, will be the subject of future reports. population, generating estimated or expected frequencies, depend- Comparing patterns and trends is a general feature of most ing on the assumptions made. descriptive epidemiological reports; and although frequencies for Haematological Malignancy Research Network was established most subtypes cannot be compared with national programmes, with the aim of providing robust data to inform epidemiological because data are not coded in the same way, we can nonetheless research and clinical practice, the project being predicated on a confirm that our incidence rates are in line with expectation for comprehensive population-based patient cohort. Within HMRN’s those few clinically evident conditions where comparisons can be population of 3.6 million, which comprises 6% of the UK’s made. For example, our acute leukaemia and Hodgkin lymphoma estimated total, over 2000 new haematological malignancies are rates are broadly similar to the most recent estimates published by diagnosed each year. All of these diagnoses – irrespective of the SEER (www.seer.cancer.gov) and Cancer Research UK (http:// patient’s age, treatment intent, or management within the National info.cancerresearchuk.org/cancerstats). Indeed, our annual UK Health Service/private sector – are made and coded by clinical incidence estimate of 1664 diagnoses for all Hodgkin lymphomas specialists working within a single integrated haematopathology combined is almost identical to the UK 2007 cancer registration laboratory (www.hmds.info). Critically, an HMDS diagnosis is a count of 1673 (Cancer Research UK, 2010). Such agreements are fundamental policy requirement of the clinical network, and reassuring not only for HMRN, but also for the national without it, treatment cannot occur. Furthermore, although outside registration scheme. Moreover, a recent collaboration between the remit of the current report, it is important to note the HMRN and the National Cancer Data Repository, comparing longitudinal nature of HMRN’s data collection processes, which observed registrations in England 2004–2007 with numbers include the collection of full sequential diagnostic and treatment expected on the basis of HMRN rates, showed good agreement histories (with response and outcome recorded for all episodes), for the conditions that could be compared nationally and by and linkage to death certificates (‘flagging’) in the national scheme. Cancer Network/Registry (Oliver et al, 2011). Haematological malignancies, unlike other cancers, are charac- Additional comparisons with the few specialist registries and/or terised by their ability to transform and progress, and this is yet consortia that have attempted to generate more informative data & 2011 Cancer Research UK British Journal of Cancer (2011) 105(11), 1684 – 1692 Standardised incidence ratio (95% confidence intervals) Clinical Studies Clinical Studies Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al Table 2 Estimated annual frequencies for the UK and European age-standardized rates per 100 000: based on HMRN sex- and age-specific rates data, 2004–2009 Estimated cases: UK European age-standardised rate (95% CI) Neoplasm (common abbreviation/synonym) Total Males Females Total Males Females a a a a a a a All diagnoses 29 017 16 264 12 752 40.8 (40.4 – 41.2) 51.1 (50.4 – 51.7) 32.7 (32.3 – 33.2) Total myeloid 8549 4693 3855 11.7 (11.5 – 11.9) 14.5 (14.1 – 14.8) 9.6 (9.3 – 9.9) Chronic myelogenous leukaemia (CML) 533 313 220 0.8 (0.8 – 0.9) 1.0 (0.9 – 1.2) 0.7 (0.6 – 0.7) Primary myelofibrosis 232 155 77 0.3 (0.3 – 0.4) 0.5 (0.4 – 0.6) 0.2 (0.1 – 0.2) Chronic myeloproliferative neoplasms (MPN) 3138 1382 1756 4.3 (4.2 – 4.5) 4.4 (4.2 – 4.6) 4.4 (4.2 – 4.5) Chronic myelomonocytic leukaemia (CMML) 300 204 96 0.4 (0.4 – 0.5) 0.6 (0.5 – 0.7) 0.2 (0.2 – 0.3) Myelodysplastic syndromes (MDS) 2049 1382 667 2.5 (2.4 – 2.6) 4.0 (3.8 – 4.2) 1.5 (1.4 – 1.6) Acute myeloid leukaemias (AML) 2275 1245 1029 3.2 (3.1 – 3.4) 4.0 (3.8 – 4.1) 2.7 (2.5 – 2.8) a a a a a a a Total lymphoid 20 468 11 571 8897 29.1 (28.8 – 29.5) 36.6 (36.0 – 37.1) 23.1 (22.7 – 23.5) Precursor B-lymphoblastic leukaemia (B-ALL) 540 279 261 1.0 (0.9 – 1.1) 1.1 (0.9 – 1.2) 1.0 (0.9 – 1.1) Precursor T-lymphoblastic leukaemia (T-ALL) 159 102 57 0.3 (0.2 – 0.3) 0.4 (0.3 – 0.4) 0.2 (0.1 – 0.3) a a a a a a a Monoclonal B-cell lymphocytosis (MBL) 1405 752 653 1.9 (1.8 – 2.0) 2.3 (2.2 – 2.5) 1.6 (1.5 – 1.7) Chronic lymphocytic leukaemia (CLL) 3624 2259 1365 5.0 (4.8 – 5.1) 7.0 (6.8 – 7.2) 3.3 (3.1 – 3.4) Marginal zone lymphomas (MZL) 1682 959 723 2.3 (2.2 – 2.4) 3.0 (2.8 – 3.1) 1.8 (1.7 – 1.9) Hairy-cell leukaemia (HCL) 177 136 41 0.3 (0.2 – 0.3) 0.4 (0.4 – 0.5) 0.1 (0.1 – 0.2) a a a a a a a Monoclonal gammopathy of undetermined significance (MGUS) 3601 2059 1542 4.9 (4.8 – 5.0) 6.3 (6.0 – 6.5) 3.9 (3.7 – 4.1) Plasma cell myeloma (multiple myeloma) 3553 2073 1480 4.7 (4.6 – 4.9) 6.3 (6.1 – 6.6) 3.5 (3.4 – 3.7) Plasmacytoma 318 211 107 0.5 (0.4 – 0.5) 0.7 (0.6 – 0.8) 0.3 (0.2 – 0.4) Follicular lymphomas (FL) 1754 821 933 2.7 (2.6 – 2.8) 2.7 (2.5 – 2.8) 2.7 (2.5 – 2.8) Mantle-cell lymphoma 454 295 159 0.6 (0.6 – 0.6) 0.9 (0.8 – 0.1) 0.4 (0.3 – 0.4) Diffuse large B-cell lymphomas (DLBCL) 4502 2353 2149 6.3 (6.1 – 6.4) 7.3 (7.1 – 7.6) 5.5 (5.3 – 5.7) Burkitt lymphoma (BL) 213 161 52 0.4 (0.3 – 0.4) 0.6 (0.5 – 0.7) 0.2 (0.1 – 0.2) Lymphoproliferative disorders NOS (LPD) 1026 557 469 1.3 (1.2 – 1.4) 1.7 (1.6 – 1.8) 1.0 (0.9 – 1.1) T-cell leukaemia 199 96 104 0.3 (0.2 – 0.3) 0.3 (0.2 – 0.4) 0.2 (0.2 – 0.3) T-cell lymphoma 601 351 249 0.9 (0.8 – 1.0) 1.1 (1.0 – 1.3) 0.7 (0.6 – 0.8) Nodular lymphocyte predominant Hodgkin lymphoma (NLPHL) 163 108 55 0.3 (0.2 – 0.3) 0.4 (0.3 – 0.5) 0.2 (0.1 – 0.2) Classical Hodgkin lymphoma (CHL) 1501 809 692 2.5 (2.4 – 2.6) 2.8 (2.4 – 2.6) 2.2 (2.1 – 2.4) Abbreviations: CI¼ confidence interval; NOS¼ not otherwise specified. Data for monoclonal B-cell lymphocytosis (MBL) and monoclonal gammopathy of undetermined significance (MGUS) are excluded from the totals. by applying bridge-coding algorithms are less rewarding. In Service, within which our study area (www.hmrn.org) is located addition to problems associated with defining catchment popula- (Northern and Yorkshire Cancer Registry and Information Service, tions, bridge coding is inevitably associated with unquantifiable 2004). However, in contrast to many other cancers, no such levels of misclassification, and with large numbers of neoplasms systematic trends have been observed in the UK haematological being categorised as ‘unknown’. For example, a recent attempt to malignancy data (National Cancer Intelligence Network, 2009). Hence, in this regard, our findings are broadly consistent with the bridge-code data for haematological malignancies diagnosed during 2000–2002 across 44 European registries produced national data, as using the same deprivation measure; we failed to disease-specific estimates for some, but not all, of the groups uncover evidence of any significant trends for the subtypes presented in the current report (Sant et al, 2010). Discrepancies examined. However, although no significant trends with depriva- were particularly marked for the lymphoid neoplasms, where some tion were found within the HMRN region, a statistically significant of the estimates were almost halved; for example, the UK age- reduction in the most deprived quintile for myeloma was found; standardised (European) rate estimate for diffuse large B-cell and this has similarly been reported in the national data (National lymphoma was 3.7 per 100 000, which compares poorly with the 6.3 Cancer Intelligence Network, 2009) The explanation for these (95% CI 6.1–6.6) per 100 000 estimated by HMRN. The low rate findings are unclear, but could reflect socio-economic variations in reported by EUROCARE may be explained by the relatively high the likelihood of a diagnosis being made, the symptoms of rate of ‘unknown’ lymphoid neoplasms (4.8 per 100 000), myeloma often extending back over several months, and perhaps demonstrating how challenging it can be to apply the WHO even years, before diagnosis (Friese et al, 2009). Indeed, the classification retrospectively. This differs from the present study in intermittent and non-specific nature of the symptoms associated which all diagnoses are coded to the latest WHO classification by with the onset of several haematological malignancies including clinical staff making the diagnosis. follicular and marginal zone lymphomas pose similar diagnostic Within most national and regional populations, the incidence of problems (Allgar and Neal, 2005; Howell et al, 2006, 2008). certain cancers is commonly observed to vary systematically with Interestingly, these diseases also showed similar deprivation socio-economic factors for reasons that are known to be related patterns to myeloma, although these were not statistically either to their aetiology or to the likelihood of their detection. In significant. England as a whole, for example, the most recent analysis of cancer It has long been known that most myeloid and lymphoid registration data showed that as area-based affluence increased the neoplasms are more common in males than females (National incidence of cancers such as lung, stomach, and cervix fell, Cancer Intelligence Network, 2008; WHO, 2008; Smith et al, 2010); whereas the incidence of cancers such as melanoma, breast, and a favoured justification for this being that men are more likely than prostate increased (National Cancer Intelligence Network, 2009); women to be exposed to potentially carcinogenic occupational and and on a smaller scale, similar associations have been reported by environmental agents (Alexander et al, 2007a, b). However, this the Northern and Yorkshire Cancer Registry and Information seems an unlikely explanation for the patterns seen within our British Journal of Cancer (2011) 105(11), 1684 – 1692 & 2011 Cancer Research UK Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al data, as the male excess is evident in children as well as adults, and malignancies that includes acute myeloid leukaemia. The data no relationship with deprivation was detected. Interestingly, the presented in the report clearly show that such additions have a subtypes with the largest male excesses – Burkitt lymphoma and major impact on estimates of the overall disease burden, hairy-cell leukaemia – are both characterised by specific genetic particularly in the myeloid group, in which the incidence is more abnormalities (WHO, 2008); and as it seems highly unlikely that than double. For epidemiology, it is equally important to recognise gender influences rates of mutation, other explanations, including that the rapid rate of progress in understanding tumour biology, gender-specific differences in immune system regulation (Fish, and the introduction of new diagnostic technologies and 2008) may well be involved. treatments mean WHO classifications, will inevitably require In addition to differences with gender, haematological malignan- ongoing revision (World Health Organization, 2001; WHO, 2008; cies exhibit characteristic age patterns that could also provide Jaffe, 2009; Vardiman et al, 2009; Campo et al, 2011). Fortunately aetiological clues. This is particularly so for the lymphoid for the present study, the Haematological Malignancy Diagnostic malignancies, where three broad overlapping patterns are discern- Service (www.hmds.info) which is at the heart of HMRN, is at the able. Precursor T- and B-cell malignancies are primarily diseases of forefront of these developments; and such changes are incorpo- children and young adults, with sporadic cases occurring at older rated as they occur. It seems highly unlikely, however, that these ages. On the other hand, malignancies arising from mature new technologies and concepts will be adopted in a uniform and immunocompetent cells (mostly B lineage) predominate in adults, timely fashion across all centres and countries; and hence, in the with sporadic cases of some, but not all, subtypes occurring at future, extrapolating data from initiatives such as HMRN may younger ages. Finally, a few disorders – notably the Hodgkin and prove to be the best way of generating reliable information on Burkittlymphomas –havemorecomplex bimodalage distributions. haematological malignancies. All of these lymphoid neoplasms exhibit characteristic, but different, In conclusion, we have demonstrated that accurate population- genetic abnormalities, and it would seem unlikely that the probability based data collection for the whole range of haematological of any one individual mutation would be related directly to age. A malignancies is achievable, and that this can be done across a more likely explanation is that the variations with age reflect the sufficiently large and diverse area to deliver reproducible data that varying proportions of cell populations across the age range, with an can be extrapolated to national populations. Our analyses immune system rich in precursor cells in young people and a emphasise the importance of gender and age as disease predominance of germinal centre and memory B-cells in older adults. determinants, and suggest that aetiological investigations that The publication of the WHO classification of haematological focus on socio-economic factors are unlikely to be rewarding. malignancies was groundbreaking in that an international consensus was finally achieved. From an epidemiological perspec- ACKNOWLEDGEMENTS tive, it was a major advance, as it stressed the unity of the haematological malignancies as a group, emphasising the links HMRN is supported by Leukaemia and Lymphoma Research. We between them, and in doing so, highlighting some of the arbitrary thank John Blase and William Curson for formatting the figures distinctions that had previously been made in many epidemiolo- and Daniel Painter for his assistance with data management and gical studies – small lymphocytic lymphoma and CLL, for example. analyses. Furthermore, the WHO classification unequivocally recognised several entities as malignant disorders that are categorised as Supplementary Information accompanies the paper on British benign/uncertain in ICD-10: myelodysplastic syndromes for example, being placed within the broad spectrum of myeloid Journal of Cancer website (http://www.nature.com/bjc) REFERENCES Alexander DD, Mink PJ, Adami H-O, Chang ET, Cole P, Mandel JS, Friese CR, Abel GA, Magazu LS, Neville BA, Richardson LC, Earle CC Trichopoulos D (2007a) The non-Hodgkin lymphomas: a review of the (2009) Diagnostic delay and complications for older adults with multiple epidemiologic literature. Int J Cancer 120(Suppl 12): 1–39 myeloma. Leuk Lymphoma 50: 392–400 Alexander DD, Mink PJ, Adami H-O, Cole P, Mandel JS, Oken MM, Fritz A (2000) International Classification of Diseases for Oncology: ICD-O, Trichopoulos D (2007b) Multiple myeloma: a review of the epidemiologic 3rd edn, World Health Organization: Geneva literature. Int J Cancer 120(Suppl 12): 40–61 Harris N, Jaffe E, Stein H, Banks P, Chan J, Cleary M, Delsol G, Allgar VL, Neal RD (2005) Delays in the diagnosis of six cancers: analysis of De Wolf- Peeters C, Falini B, Gatter K (1994) A revised European- data from the National Survey of NHS Patients: Cancer. Br J Cancer 92: American classification of lymphoid neoplasms: a proposal from 1959– 1970 the International Lymphoma Study Group (see comments). Blood 84: Boyle P (2008) World Cancer Report 2008. IARC Press: Lyon 1361– 1392 Campo E, Swerdlow SH, Harris NL, Pileri S, Stein H, Jaffe ES (2011) The Howell DA, Smith AG, Roman E (2006) Lymphoma: variations in time to 2008 WHO classification of lymphoid neoplasms and beyond: evolving diagnosis and treatment. Eur J Cancer Care 15: 272–278 concepts and practical applications. Blood 117: 5019 –5032 Howell DA, Smith AG, Roman E (2008) Help-seeking behaviour in patients Cancer Research UK (2010) Hodgkin’s lymphoma - UK incidence statistics. with lymphoma. Eur J Cancer Care (Engl) 17(4): 394–403 Available at: http://info.cancerresearchuk.org/cancerstats/types/hodgkin- Jaffe ES (2009) The 2008 WHO classification of lymphomas: implications slymphoma/incidence/ (accessed 14 June 2011) for clinical practice and translational research. Hematology/the Educa- Department of Health (2010) Reducing cancer inequality: evidence, tion Program of the American Society of Hematology. American Society progress and making it happen: a report by the National Cancer of Hematology. Education Program: 523–531 Equality Initiative. Available at: http://www.dh.gov.uk/en/Publication- Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D (2011) Global sandstatistics/Publications/PublicationsPolicyAndGuidance/DH_114353 cancer statistics. CA Cancer J Clin 61: 69–90 (accessed 21 May 2011) Jemal A, Siegel R, Xu J, Ward E (2010) Cancer statistics, 2010. CA Cancer J Ferlay J, Shin H-R, Bray F, Forman D, Mathers C, Parkin DM (2010) Clin 60: 277–300 Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Maynadie M, Girodon F, Manivet-Janoray I, Mounier M, Mugneret F, Bailly Int J Cancer Available at: http://www.ncbi.nlm.nih.gov/pubmed/20560135 F, Favre B, Caillot D, Petrella T, Flesch M, Carli P-M (2011) Twenty-five (accessed 26 August 2010) years of epidemiological recording on myeloid malignancies: data from Fish EN (2008) The X-files in immunity: sex-based differences predispose the specialized registry of hematologic malignancies of Cote d’Or immune responses. Nature Rev Immunol 8: 737–744 (Burgundy, France). Haematologica 96: 55–61 & 2011 Cancer Research UK British Journal of Cancer (2011) 105(11), 1684 – 1692 Clinical Studies Clinical Studies Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al Mozaheb Z, Aledavood A, Farzad F (2011) Distributions of major sub-types Shack L, Jordan C, Thomson CS, Mak V, Møller H (2008) Variation of lymphoid malignancies among adults in Mashhad, Iran. Cancer in incidence of breast, lung and cervical cancer and malignant Epidemiol 35: 26–29 melanoma of skin by socioeconomic group in England. BMC Cancer 8: National Cancer Intelligence Network (2008) Cancer Incidence and 271 Mortality By Cancer Network, UK, 2005, Available at: www.ncin.org.uk Smith A, Roman E, Howell D, Jones R, Patmore R, Jack A (2010) The National Cancer Intelligence Network (2009) Cancer Incidence by Haematological Malignancy Research Network (HMRN): a new informa- Deprivation England 1995–2004. Available at: http://www.ncin.org.uk/ tion strategy for population based epidemiology and health service publications/reports/default.aspx (accessed 16 February 2011) research. Br J Haematol 148: 739 –753 National Institute for Clinical Excellence (2003) Improving Outcomes in Stata-Corp (2010) STATA 11. Stata-Corp: College Station, TX, USA Haematological Cancer: The Manual. National Institute for Clinical Turner JJ, Morton LM, Linet MS, Clarke CA, Kadin ME, Vajdic CM, Excellence: London Monnereau A, Maynadie M, Chiu BC-H, Marcos-Gragera R, Costantini Office for National Statistics (2001) Census: Standard Area Statistics AS, Cerhan JR, Weisenburger DD (2010) InterLymph hierarchical (England). ESRC/JISC Census Programme Census Dissemination Unit, classification of lymphoid neoplasms for epidemiologic research based University of Manchester: Manchester on the WHO classification (2008): update and future directions. Blood Oliver S, Smith A, Roman E, Bolton E, Ross J, Patmore R, Jack A (2011) 116: e90–e98 Evaluation of the completeness of national haematological malignancy Vardiman JW, Thiele J, Arber DA, Brunning RD, Borowitz MJ, Porwit A, registration: comparison of national data with a specialist population- Harris NL, Le Beau MM, Hellstrom-Lindberg E, Tefferi A, Bloomfield CD based register. In BSH 51st Annual Scientific Meeting. British Society for (2009) The 2008 revision of the World Health Organization (WHO) Haematology: Brighton classification of myeloid neoplasms and acute leukemia: rationale and ONS Geography (2004) Rural/Urban Definition and LA Classification. important changes. Blood 114: 937–951 Available at: http://www.ons.gov.uk/about-statistics/geography/products/ Westlake S (2009) Cancer incidence and mortality in the United Kingdom area-classifications/rural-urban-definition-and-la-classification/index.html and constituent countries, 2004–06. Health Statistics Quarterly/Office for (accessed 13 May 2011) National Statistics: 56–62 Parkin D (2006) The evolution of the population-based cancer registry. WHO (2001) Tumours of Haematopoietic and Lymphoid Tissues. Nature Reviews. Cancer 6: 603–612 World Health Organisation: Lyon Sant M, Allemani C, Santaquilani M, Knijn A, Marchesi F, Capocaccia R (2009) WHO (2008) WHO Classification of Tumours of Haematopoietic EUROCARE-4. Survival of cancer patients diagnosed in 1995–1999. Results and Lymphoid Tissues. International Agency for Research on and commentary. Eur J Cancer (Oxford, England: 1990) 45: 931–991 Cancer: Lyon Sant M, Allemani C, Tereanu C, De Angelis R, Capocaccia R, Visser O, Yoon SO, Suh C, Lee DH, Chi H-S, Park CJ, Jang S-S, Shin H-R, Park B-H, Marcos-Gragera R, Maynadie´ M, Simonetti A, Lutz J-M, Berrino F (2010) Huh J (2010) Distribution of lymphoid neoplasms in the Republic Incidence of hematologic malignancies in Europe by morphologic of Korea: analysis of 5318 cases according to the World Health subtype: results of the HAEMACARE project. Blood 116: 3724 –3734 Organization classification. Am J Hematol 85: 760–764 This work is published under the standard license to publish agreement. After 12 months the work will become freely available and the license terms will switch to a Creative Commons Attribution-NonCommercial-Share Alike 3.0 Unported License. British Journal of Cancer (2011) 105(11), 1684 – 1692 & 2011 Cancer Research UK http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png British Journal of Cancer Springer Journals

Incidence of haematological malignancy by sub-type: a report from the Haematological Malignancy Research Network

British Journal of Cancer , Volume 105 (11) – Nov 1, 2011

Loading next page...
 
/lp/springer-journals/incidence-of-haematological-malignancy-by-sub-type-a-report-from-the-0GDmwas0yZ

References (75)

Publisher
Springer Journals
Copyright
Copyright © 2011 by The Author(s)
Subject
Biomedicine; Biomedicine, general; Cancer Research; Epidemiology; Molecular Medicine; Oncology; Drug Resistance
ISSN
0007-0920
eISSN
1532-1827
DOI
10.1038/bjc.2011.450
Publisher site
See Article on Publisher Site

Abstract

Clinical Studies British Journal of Cancer (2011) 105, 1684 – 1692 & 2011 Cancer Research UK All rights reserved 0007 – 0920/11 www.bjcancer.com Incidence of haematological malignancy by sub-type: a report from the Haematological Malignancy Research Network ,1 1 2 3,4 1,4 A Smith , D Howell , R Patmore , A Jack and E Roman 1 2 Epidemiology and Genetics Unit, Department of Health Sciences, University of York, York, UK; Queens Centre for Oncology and Haematology, Castle Hill Hospital, Hull, UK; Haematological Malignancy Diagnostic Service, St James’s Institute of Oncology, Bexley Wing, St James’s University Hospital, Leeds, UK; Hull York Medical School, Heslington, York, UK BACKGROUND: Ascertainment of cases and disease classification is an acknowledged problem for epidemiological research into haematological malignancies. METHODS: The Haematological Malignancy Research Network comprises an ongoing population-based patient cohort. All diagnoses (paediatric and adult) across two UK Cancer Networks (population 3.6 million, 42000 diagnoses annually, socio-demographically representative of the UK) are made by an integrated haematopathology laboratory. Diagnostics, prognostics, and treatment are recorded to clinical trial standards, and socio-demographic measures are routinely obtained. RESULTS: A total of 10 729 haematological malignancies (myeloid¼ 2706, lymphoid¼ 8023) were diagnosed over the 5 years, that is, from 2004 to 2009. Descriptive data (age, sex, and deprivation), sex-specific age-standardised (European population) rates, and estimated UK frequencies are presented for 24 sub-types. The age of patients ranged from 4 weeks to 99 years (median 70.6 years), and the male rate was more than double the female rate for several myeloid and lymphoid sub-types, this difference being evident in both children and adults. No relationship with deprivation was detected. CONCLUSION: Accurate population-based data on haematological malignancies can be collected to the standard required to deliver reproducible results that can be extrapolated to national populations. Our analyses emphasise the importance of gender and age as disease determinants, and suggest that aetiological investigations that focus on socio-economic factors are unlikely to be rewarding. British Journal of Cancer (2011) 105, 1684–1692. doi:10.1038/bjc.2011.450 www.bjcancer.com Published online 1 November 2011 & 2011 Cancer Research UK Keywords: descriptive epidemiology; incidence; lymphoma; leukaemia; myeloma; socio-economic status To originate and test hypotheses about pathogenesis, it is vitally common, accounting for around 9% of all cancers and being the important to accurately describe the underlying disease patterns fourth most frequently diagnosed cancer in both men (after and trends (Parkin, 2006; Boyle, 2008; Ferlay et al, 2010). This prostate, lung, and colorectum) and women (after breast, lung, requires complete ascertainment of cases within a defined and colorectum) in economically developed regions of the world. population, as well as the application of appropriate disease However, over and above basic tallies, the usefulness of classifications, and it is the attainment of these two key these descriptive data for epidemiological research is constrained components that continues to beleaguer epidemiological research by the classification system applied, which for haematological into the haematological malignancies (National Institute for malignancies is largely rooted in the gradual recognition of disease Clinical Excellence, 2003; Sant et al, 2009, 2010). entities at the beginning of the twentieth century. In the 1980s and Traditionally, the descriptive epidemiology of haematological 1990s, however, several competing classifications emerged as malignancies considers four broad categories – leukaemia, understanding about the relationship between the various Hodgkin lymphoma, non-Hodgkin lymphoma, and myeloma; with haematological malignancies, the bone marrow, and the immune national and global organisations including the USA’s Surveillance system, and the cellular and genetic basis of malignant transfor- Epidemiology and End Results Program (www.seer.cancer.gov), mation gradually increased. In the early 1980s, for example, the the UK’s National Cancer Intelligence Network (www.ncin.org.uk), Working Formulation, which was developed as a method of and the World Health Organization (WHO)’s International Agency translating between the many competing lymphoid classifications, for Research on Cancer (http://globocan.iarc.fr/), routinely pub- rapidly became the standard in North America, and many lishing data in this format (National Cancer Intelligence Network, epidemiological studies conducted there were based on this 2008; Westlake, 2009; Jemal et al, 2010, 2011). Such counts show system. At the same time, the majority of European centres used that, as a group, haematological neoplasms are comparatively the Kiel classification, making effective comparison of results between North America and Europe almost impossible (Harris et al, 1994). In 2001, the WHO produced, for the first time, a consensus *Correspondence: Dr A Smith; E-mail: alex.smith@egu.york.ac.uk classification that defined all haematological malignancies in terms Received 20 June 2011; revised 26 September 2011; accepted 27 of immunophenotype, genetic abnormalities and clinical features September 2011; published online 1 November 2011 Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al (World Health Organization, 2001), and this is incorporated into 2010). In recognition of this fact, a number of methods have been the current version of the International Classification of Diseases applied in an attempt to generate more informative descriptive for Oncology (ICD-O3); (Fritz, 2000). The strategy adopted was data, including, for example, the application of bridge-coding based on the principle devised for the Revised European– algorithms to historically coded data (Sant et al, 2010; Turner et al, American Classification of Lymphoid Neoplasms, which was 2010; Maynadie et al, 2011) and the reporting of specialist hospital- introduced in the mid-1990s (Harris et al, 1994). Unfortunately, based case-series frequencies (Yoon et al, 2010; Mozaheb et al, however, although the new WHO classification and its successor 2011). Inevitably, however, the accuracy and completeness of data (WHO, 2008) were adopted into clinical practice almost uniformly generated by such initiatives has continued to pose serious around the world, there was no immediate effect on population- interpretative problems for both researchers and health service based cancer information systems, where the practice of grouping planners. haematological malignancies into the four broad groups defined in In response to these challenges, the Haematological Malignancy the tenth revision of ICD-10 (www.who.int/classifications/icd/en/) Research Network (HMRN) was established in the UK in 2004 has tended to continue (National Cancer Intelligence Network, (Smith et al, 2010). HMRN is predicated on the framework of the 2008; Westlake, 2009; Ferlay et al, 2010; Jemal et al, 2010, 2011). UK National Health Service, where 37 cancer networks are This is largely because unlike many other cancers, haematological responsible for bringing together health service commissioners neoplasms are diagnosed using multiple parameters, including a and providers, the voluntary sector and local authorities to deliver combination of histology, cytology, immunophenotyping, cytoge- high quality cancer care. HMRN presently covers two such cancer netics, imaging, and clinical data. This range and depth of data is networks (Figure 1A), which comprise a single clinical network difficult for cancer registries and other researchers to access (population 3.6 million, with over 2000 new haematological systematically, potentially forming a barrier not only to the neoplasms diagnosed each year), and the present report examines collection of diagnostic data at the level of detail required to the socio-demographic characteristics of patients diagnosed over systematically implement the latest WHO classification (WHO, the first 5 years of the project, that is from 1 September 2004 to 31 2001; WHO, 2008), but also to complete ascertainment (Sant et al, August 2009. HMRN % England % Village, hamlet & isolated 0.8 0.7 dwellings - sparse Village, hamlet & isolated 7.8 8.4 dwellings - less sparse Town & fringe - sparse 0.6 0.5 Town & fringe - less sparse 10.9 9.0 Urban > 10K - sparse 0.4 0.2 Urban > 10K - less sparse 79.6 81.2 90+ HMRN males 85–89 UK males 80–84 HMRN females 75–79 UK females 70–74 65–69 60–64 55–59 HMRN % England % 50–54 1 (Affluent) 16.2 20.0 45–49 2 21.2 20.0 40–44 3 18.1 20.0 35–39 4 19.1 20.0 30–34 5 (Deprived) 25.4 20.0 25–29 20–24 15–19 10–14 5–9 0–4 8642 0 2468 Percent (%) Figure 1 Socio-demographic structure of Haematological Malignancy Research Network (HMRN). (A) Map of study area. (B) Population, age, and sex structure. (C) Office for National Statistics urban/rural definition. (D) Index of multiple deprivation – income domain. & 2011 Cancer Research UK British Journal of Cancer (2011) 105(11), 1684 – 1692 Clinical Studies Clinical Studies Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al MATERIALS AND METHODS analyses were conducted in the statistical package STATA 11. (Stata-Corp., 2010) Incidence rates, sex rate ratios, and 95% Haematological Malignancy Research Network (www.hmrn.org) is an confidence intervals (CIs) were estimated by Poisson regression; ongoing population-based cohort of patients (adult and paediatric) directly age-standardised rates were calculated using the Stata newly diagnosed with a haematological malignancy. It is a unique command dstdize, and indirectly standardised-incidence ratios venture, combining the expertise of a single integrated haemato- (SIR) were calculated using the Stata command istdize. pathology laboratory, a unified clinical network (comprising the Descriptive findings are presented here for 10 729 haemato- Yorkshire and Humber and Yorkshire Coast Cancer Networks), and a logical malignancies diagnosed within the HMRN region over specialist epidemiology unit, and full details of its structure, data- 5 years spanning September 2004 to August 2009. For analytical collection methods, and ethical approvals have been described in purposes, these diagnoses coded to ICD-O3 are grouped into detail elsewhere (Smith et al, 2010). Briefly, as a matter of policy, all 24 main WHO categories; the codes that comprise these groups diagnoses within the clinical network are made and coded by clinical are published on our website and in Supplementary Table 1 specialists to the latest WHO classification at a single integrated (www.hmrn.org/Info/Disease_Classification.aspx). haematopathology laboratory – the Haematological Malignancy Diagnostic Service (www.HMDS.info) – which was cited in the UK Department of Health’s Cancer Reform Strategy as ‘the model for RESULTS delivery of complex diagnostic services’. Following diagnosis, and with an emphasis on obtaining primary-source data, information is With a combined population of around 3.6 million, the socio- abstracted from medical records and laboratory reports to clinical demographic structure of HMRN is broadly representative of the trial standards, and all diagnostic, prognostic, treatment, and national population in terms of age, sex, urban/rural status, and outcome data are linked and held in a central database. deprivation (Figure 1). The 2001 age–sex distribution is compared Populations and area-based measures of urban/rural status, and with the UK (58.8 million) in Figure 1B, but in line with national data deprivation are routinely obtained from the UK census and other release policies, the urban/rural and deprivation configurations are national data sources (Office for National Statistics, 2001; ONS compared with England alone (49.1 million) in Figure 1C and 1D, Geography, 2004). For the purposes of the present report, subjects respectively. Although the age/sex distributions (Figure 1B) and were given a measure of area-based deprivation assigned on the urban/rural residence patterns (Figure 1C) closely mirror those of lower super output area, where they were resident at the time of the national population, the HMRN region contains proportionately diagnosis. In common with other reports (Shack et al, 2008; more people living in areas classified as deprived and fewer in areas National Cancer Intelligence Network, 2009; Department of Health, classified as affluent (Figure 1D). 2010), the income domain of the index of multiple deprivation The 10 729 haematological malignancies diagnosed over the 5 (IMD) was used (quintile one containing the most affluent fifth of years from September 2004 to August 2009 are distributed by England’s lower super output areas, and quintile five the least). All sub-type in Table 1. Myeloid malignancies, which comprise around Table 1 Numbers and annual crude rates per 100 000 (95% CI), median ages at diagnosis: Haematological Malignancy Research Network (HMRN), 2004–2009 Median age Annual rates (95% CI) per 100 000 Sex-rate ratio: Total at diagnosis male/female Neoplasm (common abbreviation/synonym) diagnoses (years) Total Males Females (95% CI) All diagnoses 10 729 70.6 60.1 (58.9 – 61.2) 68.5 (66.8 – 70.3) 52.1(50.7 – 53.6) 1.31 (1.26 – 1.36) Total myeloid 2706 72.0 15.1 (14.6 – 15.7) 16.9 (16.0 – 17.8) 13.5 (12.8 – 14.3) 1.25 (1.16 – 1.35) Chronic myelogenous leukaemia (CML) 165 59.1 0.9 (0.8 – 1.1) 1.1 (0.9 – 1.4) 0.7 (0.6 – 0.9) 1.48 (1.07 – 2.05) Primary myelofibrosis 73 73.0 0.4 (0.3 – 0.5) 0.6 (0.4 – 0.7) 0.3 (0.2 – 0.4) 2.04 (1.24 – 3.46) Chronic myeloproliferative neoplasms (MPN) 961 71.1 5.4 (5.0 – 5.7) 4.8 (4.3 – 5.3) 6.0 (5.5 – 6.5) 0.80 (0.70 – 0.91) Chronic myelomonocytic leukaemia (CMML) 96 76.1 0.6 (0.5 – 0.7) 0.8 (0.6 – 1.0) 0.4 (0.3 – 0.5) 2.10 (1.38 – 3.25) Myelodysplastic syndromes (MDS) 653 76.1 3.7 (3.4 – 3.9) 5.0 (4.5 – 5.5) 2.4 (2.1 – 2.7) 2.09 (1.78 – 2.48) Acute myeloid leukaemias (AML) 717 68.7 4.0 (3.7 – 4.3) 4.5 (4.0 – 4.9) 3.6 (3.2 – 4.0) 1.25 (1.07 – 1.45) Total lymphoid 8023 70.1 44.9 (44.0 – 45.9) 51.6 (50.1 – 53.2) 38.6 (37.4 – 39.9) 1.34 (1.28 – 1.40) Precursor B-lymphoblastic leukaemia (B-ALL) 167 12.7 0.9 (0.8 – 1.1) 1.0 (0.8 – 1.2) 0.9 (0.7 – 1.1) 1.13 (0.82 – 1.55) Precursor T-lymphoblastic leukaemia (T-ALL) 50 18.5 0.3 (0.2 – 0.4) 0.4 (0.4 – 0.5) 0.2 (0.1 – 0.3) 1.89 (1.03 – 3.58) Monoclonal B-cell lymphocytosis (MBL) 445 71.7 2.5 (2.3 – 2.7) 2.7 (2.4 – 3.1) 2.3 (2.0 – 2.6) 1.18 (0.98 – 1.43) Chronic lymphocytic leukaemia (CLL) 1145 71.6 6.4 (6.0 – 6.8) 8.1 (7.5 – 8.7) 4.8 (4.4 – 5.3) 1.69 (1.50 – 1.91) Marginal zone lymphomas (MZL) 530 71.5 3.0 (2.7 – 3.2) 3.4 (3.1 – 3.9) 2.5 (2.2 – 2.9) 1.37 (1.15 – 1.63) Hairy-cell leukaemia (HCL) 55 65.7 0.3 (0.2 – 0.4) 0.5 (0.3 – 0.7) 0.1 (0.1 – 0.2) 3.44 (1.81 – 6.98) Monoclonal gammopathy of undetermined significance (MGUS) 1135 72.2 6.4 (6.0 – 6.7) 7.4 (6.8 – 8.0) 5.4 (4.9 – 5.9) 1.38 (1.23 – 1.56) Plasma cell myeloma (multiple myeloma) 1127 73.0 6.3 (5.9 – 6.7) 7.5 (6.9 – 8.1) 5.2 (4.8 – 5.7) 1.43 (1.27 – 1.61) Plasmacytoma 99 69.2 0.6 (0.5 – 0.7) 0.8 (0.6 – 1.0) 0.4 (0.3 – 0.5) 2.03 (1.32 – 3.18) Follicular lymphomas (FL) 547 64.6 3.1 (2.8 – 3.3) 2.9 (2.6 – 3.3) 3.2 (2.8 – 3.6) 0.92 (0.77 – 1.09) Mantle-cell lymphoma 144 74.0 0.8 (0.7 – 0.9) 1.1 (0.9 – 1.3) 0.6 (0.4 – 0.7) 1.88 (1.33 – 2.70) Diffuse large B-cell lymphomas (DLBCL) 1417 70.4 7.9 (7.5 – 8.4) 8.4 (7.8 – 9.1) 7.5 (6.9 – 8.1) 1.13 (1.01 – 1.26) Burkitt lymphoma (BL) 66 52.2 0.4 (0.3 – 0.5) 0.6 (0.4 – 0.8) 0.2 (0.1 – 0.3) 3.33 (1.86 – 6.26) Lymphoproliferative disorders NOS (LPD) 330 76.9 1.8 (1.7 – 2.3) 2.0 (1.7 – 2.3) 1.7 (1.4 – 2.0) 1.19 (0.95 – 1.87) T-cell leukaemia 64 74.7 0.4 (0.3 – 0.5) 0.3 (0.2 – 0.5) 0.4 (0.3 – 0.5) 0.94 (0.56 – 1.58) T-cell lymphoma 188 65.7 1.1 (0.9 – 1.2) 1.2 (1.0 – 1.5) 0.9 (0.7 – 1.1) 1.44 (1.07 – 1.94) Nodular lymphocyte predominant Hodgkin lymphoma (NLPHL) 50 42.9 0.3 (0.2 – 0.4) 0.4 (0.4 – 0.5) 0.2 (0.1 – 0.3) 2.07 (1.12 – 3.95) Classical Hodgkin lymphoma (CHL) 464 41.2 2.6 (2.4 – 2.8) 2.9 (2.5 – 3.3) 2.3 (2.0 – 2.7) 1.23 (1.02 – 1.49) Abbreviations: CI¼ confidence interval; NOS¼ not otherwise specified. British Journal of Cancer (2011) 105(11), 1684 – 1692 & 2011 Cancer Research UK Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al a quarter of the total (N¼ 2706) are presented first, and lymphoid, within the traditional lymphoma and myeloma groupings, there is which account for the remaining malignancies (N¼ 8023), are less diversity in the cell type of origin; with mature B-cell second. Data on median ages at the time of diagnosis, annual rates, malignancies dominating. Indeed, with an annual rate of 7.9 per and sex-rate ratios (male rate/female rate) with 95% CIs are also 100 000 per year, diffuse large B-cell lymphoma is the most given in Table 1. The rates in Table 1 are ordered by magnitude common haematological malignancy, and chronic lymphocytic in Figure 2, and the bars are colour coded, differentiating the leukaemia (CLL), which like diffuse large B-cell lymphoma is also a traditional groupings of leukaemia, non-Hodgkin lymphoma, mature B-cell neoplasm, is the next most common. Hodgkin lymphoma, and myeloma from other haematological As detailed in the Introduction, the classification of haemato- neoplasms that are less consistently registered by national logical malignancies has changed markedly in recent decades, with schemes. The classic ICD-10 leukaemia group contains a mix of several conditions once classified as neoplasms of uncertain or myeloid and lymphoid conditions, the latter including both unknown behaviour now being categorised as malignant, and precursor and mature B-cell and T-cell subtypes. By contrast, other conditions being recognised as part of the cancer continuum for the first time; the disorders falling into this category are shaded grey in Figure 2. Chronic myeloproliferative neoplasms and 0 2468 myelodysplastic syndromes now comprise around two-thirds of Chronic myeloproliferative neoplasms the myeloid neoplasms assigned a behaviour code of 3 (malignant Acute myeloid leukaemia Myelodysplastic syndromes primary site) in the WHO ICD-O3. Within the lymphoid category, Chronic myelogenous leukaemia lymphoproliferative disorders not otherwise specified also Chronic myelomonocytic leukaemia contains a mix of malignancies, all with behaviour codes of 3. Primary myelofibrosis However, although monoclonal gammopathy of undetermined Diffuse large B-cell lymphoma significance (MGUS) and monoclonal B-cell lymphocytosis (MBL) Chronic lymphocytic leukaemia are both conditions in which neoplastic B-cells are detectable, the Monoclonal gammopathy of undetermined significance Plasma cell myeloma diagnostic criteria for MBL being where the peripheral blood Follicular lymphoma 9 B-cell count is less than 5 10 /l lymphocytes (and in which risks Marginal zone lymphoma Classical Hodgkin lymphoma of progression to myeloma in the case of the former and CLL in the Monoclonal B-cell lymphocytosis case of the latter are elevated), their behaviour remains uncertain. Lymphoproliferative disorder NOS As with most other cancers, the likelihood of being diagnosed T-cell lymphoma Precursor B-lymphoblastic leukaemia with a haematological malignancy increases markedly with age, Mantle cell lymphoma the median age at diagnosis being 70.6 years within the HMRN Plasmacytoma region (Table 1). However, unlike many other common cancers, Burkitt lymphoma T-cell leukaemia haematological malignancy can be diagnosed at any age, with Hairy-cell leukaemia different subtypes dominating at different ages. More information Precursor T-lymphoblastic leukaemia about the age distributions of the various subtypes is presented in Nodular lymphocyte predominant Hodgkin lymphoma Figure 3, which shows box-and-whisker (boxplots) summary age Hodgkin lymphoma Non-Hodgkin lymphoma Leukaemia plots ordered by the magnitude of the median for myeloid and Unknown or uncertain behaviour lymphoid malignancies separately. The interquartile range is Myeloma represented by the box, with outliers occurring outside the maximum data series of 1.5 times the interquartile range being Figure 2 Annual crude rates per 100 000: Haematological Malignancy Research Network (HMRN), 2004–2009. shown as separate points. Chronic myelogenous leukaemia Acute myeloid leukaemia Chronic myeloproliferative neoplasms Primary myelofibrosis Myelodysplastic syndromes Chronic myelomonocytic leukaemia Precursor B-lymphoblastic leukaemia Precursor T-lymphoblastic leukaemia Classical Hodgkin lymphoma Nodular lymphocyte predominant Hodgkin lymphoma Burkitt lymphoma Follicular lymphoma Hairy cell leukaemia T-cell lymphoma Plasmacytoma Diffuse large B-cell lymphoma Marginal zone lymphoma Chronic lymphocytic leukaemia Monoclonal B-cell Lymphocytosis Monoclonal gammopathy of undetermined significance Plasma cell myeloma Mantle cell lymphoma T-cell leukaemia Lymphoproliferative disorder NOS 0 10 20 30 40 50 60 70 80 90 100 Age at diagnosis (years) Figure 3 Age at diagnosis distributions: Haematological Malignancy Research Network (HMRN), 2004–2009. & 2011 Cancer Research UK British Journal of Cancer (2011) 105(11), 1684 – 1692 Clinical Studies Clinical Studies Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al The majority of myeloid conditions are diagnosed above 70 the point estimates at younger ages being similar to those at older years of age, but sporadic cases arise at younger ages (Figure 3). ages, although the CIs are wider, reflecting the comparative Likewise, lymphoid malignancies generally occur in older people, sparsity of the data. but nonetheless span the entire age range, our youngest diagnosis In the 24 main diagnostic categories listed in Table 1, 16 had 100 having been made at 4 weeks of age and oldest at 99 years. The diagnoses or more, and for these the SIRs and 95% CIs are plotted precursor B-cell and T-cell lymphoblastic leukaemias tend to occur by index of multiple deprivation income domain quintile in at the youngest ages, the medians being 12.7 years and 18.5 years, Figure 6, (group 1 being the most affluent and group 5 being the respectively (Table 1). However, as with some of the conditions most deprived). No trends with deprivation are evident, although that principally occur at older ages, such as diffuse large B-cell for some malignancies there is an indication of a deficit in the lymphoma, these too are periodically diagnosed outside their most deprived quintile; the most notable being myeloma where the normal age range. Such wide age spans are however not seen for all SIR in category 5 is significantly below 1.0 (0.82 (95% CI 0.71– lymphoid conditions, including the rarer forms like hairy-cell 0.95). leukaemia and mantle-cell lymphoma, and also comparatively The lack of a trend with deprivation (Figure 6) is particularly common conditions like CLL and myeloma – all of which seldom, pertinent to precursor B-lymphoblastic leukaemia and classical if ever, occur below the age of 30 years. A further conspicuous Hodgkin lymphoma (CHL), both of which have been suggested to feature of lymphoid neoplasms is the similarity in the age be increasingly common in more affluent families and communities. distributions of certain closely related conditions such as MBL B-lymphoblastic leukaemia is primarily paediatric (Figure 3), and it (median 71.7 years) and CLL (median 71.6 years), as well is in this age group that an association with socio-economic status as monoclonal gammopathy of underdetermined significance has been suggested. In our data, however, the results were similar (median 72.2 years) and myeloma (median 73.0 years), lying when the analysis was restricted to cases diagnosed before the adjacent to each other in Figure 3. age of 15 years (95 out of 167); SIRs (95% CIs) for deprivation In general, haematological malignancies tend to occur more categories 1 through 5, respectively, being 0.7 (0.4–1.3), 1.0 frequently in males than females, and for many conditions, the rate (0.6–1.6), 1.1 (0.6–1.7), 1.0 (0.5–1.6), and 1.2 (0.8–1.6). Likewise, among males is more than twice that of females (Table 1). The for CHL, the strongest effects have been reported at younger ages consistency of the gender difference is plainly visible in Figure 4, where the nodular sclerosis form of CHL predominates. In our data, which shows the sex-specific rate ratios (male rate/female rate) no associations with deprivation were observed, either for total CHL ordered by magnitude. Indeed, conditions with no apparent sex or for any of the CHL subtypes (data not shown). bias, such as the chronic myleoproliferative neoplasms (male rate/ The size and demographic similarity of HMRN’s population to female rate¼ 0.80, 95% CI 0.70–0.91) and follicular lymphoma the general UK population (Figure 1) means that the HMRN’s data (male rate/female rate¼ 0.92, 95% CI 0.77–1.09), stand out from can reasonably be extrapolated to the country as a whole. The the rest (Table 1). The lymphoid group exhibits some of the most estimated UK totals, calculated by applying HMRN’s age-specific striking sex differences, the rates of the comparatively rare Burkitt rates to the corresponding general population age strata are shown lymphoma and hairy-cell leukaemia being more than three times in Table 2. For the purposes of wider comparability, age- higher in males than in females. These sex differences occur across standardised rates (European population) are also given in Table 2; the full age spectrum, being seen in conditions with comparatively these rates are in general lower than the actual rates (Table 1), low, as well as high, median ages at diagnosis such as mantle-cell reflecting the fact that unlike the real population (Figure 1), the lymphoma (median age at diagnosis 74 years) and precursor hypothetical standard has a younger age structure with no excess T-lymphoblastic leukaemia (median age at diagnosis 18.5 years), of females in the older age groups. For the sake of completeness, for example, both with ratios approaching 2.0 lying adjacent to information on MBL and monoclonal gammopathy of undeter- each other in Figure 4. The consistency of the gender bias is further mined significance are included in Table 2, but their data are illustrated in Figure 5, which shows the sex-rate ratios plotted in excluded from the overall totals. 10-year age groups for all haematological malignancies combined; Sex rate ratio DISCUSSION (95% confidence intervals) 0 2 4 6 8 Our ability to calculate reliable incidence rates for clinically Chronic myeloproliferative neoplasms meaningful haematological malignancy subtypes is a fundamental Acute myeloid leukaemia Chronic myelogenous leukaemia Primary myelofibrosis Myelodysplastic syndromes Chronic myelomonocytic leukaemia Follicular lymphoma T-cell leukaemia Diffuse large B-cell lymphoma 1.5 Precursor B-lymphoblastic leukaemia Monoclonal B-cell lymphocytosis Lymphoproliferative disorder NOS Classical Hodgkin lymphoma Marginal zone lymphoma Monoclonal gammopathy of undetermined significance Plasma cell myeloma T-cell lymphoma Chronic lymphocytic leukaemia Mantle cell lymphoma Precursor T-lymphoblastic leukaemia 0.5 Plasmacytoma Nodular lymphocyte predominant Hodgkin lymphoma 0−9 10−19 20−29 30−39 40−49 50−59 60−69 70+ Burkitt lymphoma Age at diagnosis (years) Hairy cell leukaemia Sex-rate ratio 95% confidence interval Sex Rate Ratio 95% Confidences Interval Figure 4 Sex-rate ratios: Haematological Malignancy Research Network Figure 5 Sex-rate ratios by age: Haematological Malignancy Research (HMRN), 2004 –2009. Network (HMRN), 2004 –2009. British Journal of Cancer (2011) 105(11), 1684 – 1692 & 2011 Cancer Research UK Sex-rate ratio (95% confidence intervals) Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al Chronic myelogenous leukaemia Chronic myeloproliferative neoplasms Myelodysplastic syndromes Acute myeloid leukaemia Precursor B-lymphoblastic leukaemia Monoclonal B-cell lymphocytosis Chronic lymphocytic leukaemia Marginal zone lymphoma MGUS Plasma cell myeloma Follicular lymphoma Mantle cell lymphoma Diffuse large B-cell lymphoma Lymphoproliferative disorder NOS T-cell lymphoma Classical Hodgkin lymphoma 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 Deprivation quintile SIR 95% confidence intervals Figure 6 Standardised-incidence ratios (SIR) by index of multiple deprivation (IMD) income domain. key research achievement; the analyses revealing notable associa- another aspect that challenges cancer registries. For example, the tions with both age and sex, contrasting somewhat starkly with the present report is based on 10 729 diagnoses, but these relate to comparative lack of variation with area-based measures of 10 306 people diagnosed with a haematological cancer for the first deprivation. In addition, the size and representative nature time, of whom 407 (3.9%) had a second haematological neoplasm diagnosed, either concurrently or because their disease progressed of our study population mean that our data can be extrapolated to the UK as a whole, providing for the first time, national or transformed, and 16 (o1%) had a third diagnosis. Investigating estimates for the main WHO-defined disease entities (WHO, 2008). the epidemiology of transformation and progression, as well as Indeed, HMRN rates could be applied to any well-characterised other outcomes, will be the subject of future reports. population, generating estimated or expected frequencies, depend- Comparing patterns and trends is a general feature of most ing on the assumptions made. descriptive epidemiological reports; and although frequencies for Haematological Malignancy Research Network was established most subtypes cannot be compared with national programmes, with the aim of providing robust data to inform epidemiological because data are not coded in the same way, we can nonetheless research and clinical practice, the project being predicated on a confirm that our incidence rates are in line with expectation for comprehensive population-based patient cohort. Within HMRN’s those few clinically evident conditions where comparisons can be population of 3.6 million, which comprises 6% of the UK’s made. For example, our acute leukaemia and Hodgkin lymphoma estimated total, over 2000 new haematological malignancies are rates are broadly similar to the most recent estimates published by diagnosed each year. All of these diagnoses – irrespective of the SEER (www.seer.cancer.gov) and Cancer Research UK (http:// patient’s age, treatment intent, or management within the National info.cancerresearchuk.org/cancerstats). Indeed, our annual UK Health Service/private sector – are made and coded by clinical incidence estimate of 1664 diagnoses for all Hodgkin lymphomas specialists working within a single integrated haematopathology combined is almost identical to the UK 2007 cancer registration laboratory (www.hmds.info). Critically, an HMDS diagnosis is a count of 1673 (Cancer Research UK, 2010). Such agreements are fundamental policy requirement of the clinical network, and reassuring not only for HMRN, but also for the national without it, treatment cannot occur. Furthermore, although outside registration scheme. Moreover, a recent collaboration between the remit of the current report, it is important to note the HMRN and the National Cancer Data Repository, comparing longitudinal nature of HMRN’s data collection processes, which observed registrations in England 2004–2007 with numbers include the collection of full sequential diagnostic and treatment expected on the basis of HMRN rates, showed good agreement histories (with response and outcome recorded for all episodes), for the conditions that could be compared nationally and by and linkage to death certificates (‘flagging’) in the national scheme. Cancer Network/Registry (Oliver et al, 2011). Haematological malignancies, unlike other cancers, are charac- Additional comparisons with the few specialist registries and/or terised by their ability to transform and progress, and this is yet consortia that have attempted to generate more informative data & 2011 Cancer Research UK British Journal of Cancer (2011) 105(11), 1684 – 1692 Standardised incidence ratio (95% confidence intervals) Clinical Studies Clinical Studies Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al Table 2 Estimated annual frequencies for the UK and European age-standardized rates per 100 000: based on HMRN sex- and age-specific rates data, 2004–2009 Estimated cases: UK European age-standardised rate (95% CI) Neoplasm (common abbreviation/synonym) Total Males Females Total Males Females a a a a a a a All diagnoses 29 017 16 264 12 752 40.8 (40.4 – 41.2) 51.1 (50.4 – 51.7) 32.7 (32.3 – 33.2) Total myeloid 8549 4693 3855 11.7 (11.5 – 11.9) 14.5 (14.1 – 14.8) 9.6 (9.3 – 9.9) Chronic myelogenous leukaemia (CML) 533 313 220 0.8 (0.8 – 0.9) 1.0 (0.9 – 1.2) 0.7 (0.6 – 0.7) Primary myelofibrosis 232 155 77 0.3 (0.3 – 0.4) 0.5 (0.4 – 0.6) 0.2 (0.1 – 0.2) Chronic myeloproliferative neoplasms (MPN) 3138 1382 1756 4.3 (4.2 – 4.5) 4.4 (4.2 – 4.6) 4.4 (4.2 – 4.5) Chronic myelomonocytic leukaemia (CMML) 300 204 96 0.4 (0.4 – 0.5) 0.6 (0.5 – 0.7) 0.2 (0.2 – 0.3) Myelodysplastic syndromes (MDS) 2049 1382 667 2.5 (2.4 – 2.6) 4.0 (3.8 – 4.2) 1.5 (1.4 – 1.6) Acute myeloid leukaemias (AML) 2275 1245 1029 3.2 (3.1 – 3.4) 4.0 (3.8 – 4.1) 2.7 (2.5 – 2.8) a a a a a a a Total lymphoid 20 468 11 571 8897 29.1 (28.8 – 29.5) 36.6 (36.0 – 37.1) 23.1 (22.7 – 23.5) Precursor B-lymphoblastic leukaemia (B-ALL) 540 279 261 1.0 (0.9 – 1.1) 1.1 (0.9 – 1.2) 1.0 (0.9 – 1.1) Precursor T-lymphoblastic leukaemia (T-ALL) 159 102 57 0.3 (0.2 – 0.3) 0.4 (0.3 – 0.4) 0.2 (0.1 – 0.3) a a a a a a a Monoclonal B-cell lymphocytosis (MBL) 1405 752 653 1.9 (1.8 – 2.0) 2.3 (2.2 – 2.5) 1.6 (1.5 – 1.7) Chronic lymphocytic leukaemia (CLL) 3624 2259 1365 5.0 (4.8 – 5.1) 7.0 (6.8 – 7.2) 3.3 (3.1 – 3.4) Marginal zone lymphomas (MZL) 1682 959 723 2.3 (2.2 – 2.4) 3.0 (2.8 – 3.1) 1.8 (1.7 – 1.9) Hairy-cell leukaemia (HCL) 177 136 41 0.3 (0.2 – 0.3) 0.4 (0.4 – 0.5) 0.1 (0.1 – 0.2) a a a a a a a Monoclonal gammopathy of undetermined significance (MGUS) 3601 2059 1542 4.9 (4.8 – 5.0) 6.3 (6.0 – 6.5) 3.9 (3.7 – 4.1) Plasma cell myeloma (multiple myeloma) 3553 2073 1480 4.7 (4.6 – 4.9) 6.3 (6.1 – 6.6) 3.5 (3.4 – 3.7) Plasmacytoma 318 211 107 0.5 (0.4 – 0.5) 0.7 (0.6 – 0.8) 0.3 (0.2 – 0.4) Follicular lymphomas (FL) 1754 821 933 2.7 (2.6 – 2.8) 2.7 (2.5 – 2.8) 2.7 (2.5 – 2.8) Mantle-cell lymphoma 454 295 159 0.6 (0.6 – 0.6) 0.9 (0.8 – 0.1) 0.4 (0.3 – 0.4) Diffuse large B-cell lymphomas (DLBCL) 4502 2353 2149 6.3 (6.1 – 6.4) 7.3 (7.1 – 7.6) 5.5 (5.3 – 5.7) Burkitt lymphoma (BL) 213 161 52 0.4 (0.3 – 0.4) 0.6 (0.5 – 0.7) 0.2 (0.1 – 0.2) Lymphoproliferative disorders NOS (LPD) 1026 557 469 1.3 (1.2 – 1.4) 1.7 (1.6 – 1.8) 1.0 (0.9 – 1.1) T-cell leukaemia 199 96 104 0.3 (0.2 – 0.3) 0.3 (0.2 – 0.4) 0.2 (0.2 – 0.3) T-cell lymphoma 601 351 249 0.9 (0.8 – 1.0) 1.1 (1.0 – 1.3) 0.7 (0.6 – 0.8) Nodular lymphocyte predominant Hodgkin lymphoma (NLPHL) 163 108 55 0.3 (0.2 – 0.3) 0.4 (0.3 – 0.5) 0.2 (0.1 – 0.2) Classical Hodgkin lymphoma (CHL) 1501 809 692 2.5 (2.4 – 2.6) 2.8 (2.4 – 2.6) 2.2 (2.1 – 2.4) Abbreviations: CI¼ confidence interval; NOS¼ not otherwise specified. Data for monoclonal B-cell lymphocytosis (MBL) and monoclonal gammopathy of undetermined significance (MGUS) are excluded from the totals. by applying bridge-coding algorithms are less rewarding. In Service, within which our study area (www.hmrn.org) is located addition to problems associated with defining catchment popula- (Northern and Yorkshire Cancer Registry and Information Service, tions, bridge coding is inevitably associated with unquantifiable 2004). However, in contrast to many other cancers, no such levels of misclassification, and with large numbers of neoplasms systematic trends have been observed in the UK haematological being categorised as ‘unknown’. For example, a recent attempt to malignancy data (National Cancer Intelligence Network, 2009). Hence, in this regard, our findings are broadly consistent with the bridge-code data for haematological malignancies diagnosed during 2000–2002 across 44 European registries produced national data, as using the same deprivation measure; we failed to disease-specific estimates for some, but not all, of the groups uncover evidence of any significant trends for the subtypes presented in the current report (Sant et al, 2010). Discrepancies examined. However, although no significant trends with depriva- were particularly marked for the lymphoid neoplasms, where some tion were found within the HMRN region, a statistically significant of the estimates were almost halved; for example, the UK age- reduction in the most deprived quintile for myeloma was found; standardised (European) rate estimate for diffuse large B-cell and this has similarly been reported in the national data (National lymphoma was 3.7 per 100 000, which compares poorly with the 6.3 Cancer Intelligence Network, 2009) The explanation for these (95% CI 6.1–6.6) per 100 000 estimated by HMRN. The low rate findings are unclear, but could reflect socio-economic variations in reported by EUROCARE may be explained by the relatively high the likelihood of a diagnosis being made, the symptoms of rate of ‘unknown’ lymphoid neoplasms (4.8 per 100 000), myeloma often extending back over several months, and perhaps demonstrating how challenging it can be to apply the WHO even years, before diagnosis (Friese et al, 2009). Indeed, the classification retrospectively. This differs from the present study in intermittent and non-specific nature of the symptoms associated which all diagnoses are coded to the latest WHO classification by with the onset of several haematological malignancies including clinical staff making the diagnosis. follicular and marginal zone lymphomas pose similar diagnostic Within most national and regional populations, the incidence of problems (Allgar and Neal, 2005; Howell et al, 2006, 2008). certain cancers is commonly observed to vary systematically with Interestingly, these diseases also showed similar deprivation socio-economic factors for reasons that are known to be related patterns to myeloma, although these were not statistically either to their aetiology or to the likelihood of their detection. In significant. England as a whole, for example, the most recent analysis of cancer It has long been known that most myeloid and lymphoid registration data showed that as area-based affluence increased the neoplasms are more common in males than females (National incidence of cancers such as lung, stomach, and cervix fell, Cancer Intelligence Network, 2008; WHO, 2008; Smith et al, 2010); whereas the incidence of cancers such as melanoma, breast, and a favoured justification for this being that men are more likely than prostate increased (National Cancer Intelligence Network, 2009); women to be exposed to potentially carcinogenic occupational and and on a smaller scale, similar associations have been reported by environmental agents (Alexander et al, 2007a, b). However, this the Northern and Yorkshire Cancer Registry and Information seems an unlikely explanation for the patterns seen within our British Journal of Cancer (2011) 105(11), 1684 – 1692 & 2011 Cancer Research UK Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al data, as the male excess is evident in children as well as adults, and malignancies that includes acute myeloid leukaemia. The data no relationship with deprivation was detected. Interestingly, the presented in the report clearly show that such additions have a subtypes with the largest male excesses – Burkitt lymphoma and major impact on estimates of the overall disease burden, hairy-cell leukaemia – are both characterised by specific genetic particularly in the myeloid group, in which the incidence is more abnormalities (WHO, 2008); and as it seems highly unlikely that than double. For epidemiology, it is equally important to recognise gender influences rates of mutation, other explanations, including that the rapid rate of progress in understanding tumour biology, gender-specific differences in immune system regulation (Fish, and the introduction of new diagnostic technologies and 2008) may well be involved. treatments mean WHO classifications, will inevitably require In addition to differences with gender, haematological malignan- ongoing revision (World Health Organization, 2001; WHO, 2008; cies exhibit characteristic age patterns that could also provide Jaffe, 2009; Vardiman et al, 2009; Campo et al, 2011). Fortunately aetiological clues. This is particularly so for the lymphoid for the present study, the Haematological Malignancy Diagnostic malignancies, where three broad overlapping patterns are discern- Service (www.hmds.info) which is at the heart of HMRN, is at the able. Precursor T- and B-cell malignancies are primarily diseases of forefront of these developments; and such changes are incorpo- children and young adults, with sporadic cases occurring at older rated as they occur. It seems highly unlikely, however, that these ages. On the other hand, malignancies arising from mature new technologies and concepts will be adopted in a uniform and immunocompetent cells (mostly B lineage) predominate in adults, timely fashion across all centres and countries; and hence, in the with sporadic cases of some, but not all, subtypes occurring at future, extrapolating data from initiatives such as HMRN may younger ages. Finally, a few disorders – notably the Hodgkin and prove to be the best way of generating reliable information on Burkittlymphomas –havemorecomplex bimodalage distributions. haematological malignancies. All of these lymphoid neoplasms exhibit characteristic, but different, In conclusion, we have demonstrated that accurate population- genetic abnormalities, and it would seem unlikely that the probability based data collection for the whole range of haematological of any one individual mutation would be related directly to age. A malignancies is achievable, and that this can be done across a more likely explanation is that the variations with age reflect the sufficiently large and diverse area to deliver reproducible data that varying proportions of cell populations across the age range, with an can be extrapolated to national populations. Our analyses immune system rich in precursor cells in young people and a emphasise the importance of gender and age as disease predominance of germinal centre and memory B-cells in older adults. determinants, and suggest that aetiological investigations that The publication of the WHO classification of haematological focus on socio-economic factors are unlikely to be rewarding. malignancies was groundbreaking in that an international consensus was finally achieved. From an epidemiological perspec- ACKNOWLEDGEMENTS tive, it was a major advance, as it stressed the unity of the haematological malignancies as a group, emphasising the links HMRN is supported by Leukaemia and Lymphoma Research. We between them, and in doing so, highlighting some of the arbitrary thank John Blase and William Curson for formatting the figures distinctions that had previously been made in many epidemiolo- and Daniel Painter for his assistance with data management and gical studies – small lymphocytic lymphoma and CLL, for example. analyses. Furthermore, the WHO classification unequivocally recognised several entities as malignant disorders that are categorised as Supplementary Information accompanies the paper on British benign/uncertain in ICD-10: myelodysplastic syndromes for example, being placed within the broad spectrum of myeloid Journal of Cancer website (http://www.nature.com/bjc) REFERENCES Alexander DD, Mink PJ, Adami H-O, Chang ET, Cole P, Mandel JS, Friese CR, Abel GA, Magazu LS, Neville BA, Richardson LC, Earle CC Trichopoulos D (2007a) The non-Hodgkin lymphomas: a review of the (2009) Diagnostic delay and complications for older adults with multiple epidemiologic literature. Int J Cancer 120(Suppl 12): 1–39 myeloma. Leuk Lymphoma 50: 392–400 Alexander DD, Mink PJ, Adami H-O, Cole P, Mandel JS, Oken MM, Fritz A (2000) International Classification of Diseases for Oncology: ICD-O, Trichopoulos D (2007b) Multiple myeloma: a review of the epidemiologic 3rd edn, World Health Organization: Geneva literature. Int J Cancer 120(Suppl 12): 40–61 Harris N, Jaffe E, Stein H, Banks P, Chan J, Cleary M, Delsol G, Allgar VL, Neal RD (2005) Delays in the diagnosis of six cancers: analysis of De Wolf- Peeters C, Falini B, Gatter K (1994) A revised European- data from the National Survey of NHS Patients: Cancer. Br J Cancer 92: American classification of lymphoid neoplasms: a proposal from 1959– 1970 the International Lymphoma Study Group (see comments). Blood 84: Boyle P (2008) World Cancer Report 2008. IARC Press: Lyon 1361– 1392 Campo E, Swerdlow SH, Harris NL, Pileri S, Stein H, Jaffe ES (2011) The Howell DA, Smith AG, Roman E (2006) Lymphoma: variations in time to 2008 WHO classification of lymphoid neoplasms and beyond: evolving diagnosis and treatment. Eur J Cancer Care 15: 272–278 concepts and practical applications. Blood 117: 5019 –5032 Howell DA, Smith AG, Roman E (2008) Help-seeking behaviour in patients Cancer Research UK (2010) Hodgkin’s lymphoma - UK incidence statistics. with lymphoma. Eur J Cancer Care (Engl) 17(4): 394–403 Available at: http://info.cancerresearchuk.org/cancerstats/types/hodgkin- Jaffe ES (2009) The 2008 WHO classification of lymphomas: implications slymphoma/incidence/ (accessed 14 June 2011) for clinical practice and translational research. Hematology/the Educa- Department of Health (2010) Reducing cancer inequality: evidence, tion Program of the American Society of Hematology. American Society progress and making it happen: a report by the National Cancer of Hematology. Education Program: 523–531 Equality Initiative. Available at: http://www.dh.gov.uk/en/Publication- Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D (2011) Global sandstatistics/Publications/PublicationsPolicyAndGuidance/DH_114353 cancer statistics. CA Cancer J Clin 61: 69–90 (accessed 21 May 2011) Jemal A, Siegel R, Xu J, Ward E (2010) Cancer statistics, 2010. CA Cancer J Ferlay J, Shin H-R, Bray F, Forman D, Mathers C, Parkin DM (2010) Clin 60: 277–300 Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Maynadie M, Girodon F, Manivet-Janoray I, Mounier M, Mugneret F, Bailly Int J Cancer Available at: http://www.ncbi.nlm.nih.gov/pubmed/20560135 F, Favre B, Caillot D, Petrella T, Flesch M, Carli P-M (2011) Twenty-five (accessed 26 August 2010) years of epidemiological recording on myeloid malignancies: data from Fish EN (2008) The X-files in immunity: sex-based differences predispose the specialized registry of hematologic malignancies of Cote d’Or immune responses. Nature Rev Immunol 8: 737–744 (Burgundy, France). Haematologica 96: 55–61 & 2011 Cancer Research UK British Journal of Cancer (2011) 105(11), 1684 – 1692 Clinical Studies Clinical Studies Descriptive epidemiology: leukaemias, lymphomas, myelomas A Smith et al Mozaheb Z, Aledavood A, Farzad F (2011) Distributions of major sub-types Shack L, Jordan C, Thomson CS, Mak V, Møller H (2008) Variation of lymphoid malignancies among adults in Mashhad, Iran. Cancer in incidence of breast, lung and cervical cancer and malignant Epidemiol 35: 26–29 melanoma of skin by socioeconomic group in England. BMC Cancer 8: National Cancer Intelligence Network (2008) Cancer Incidence and 271 Mortality By Cancer Network, UK, 2005, Available at: www.ncin.org.uk Smith A, Roman E, Howell D, Jones R, Patmore R, Jack A (2010) The National Cancer Intelligence Network (2009) Cancer Incidence by Haematological Malignancy Research Network (HMRN): a new informa- Deprivation England 1995–2004. Available at: http://www.ncin.org.uk/ tion strategy for population based epidemiology and health service publications/reports/default.aspx (accessed 16 February 2011) research. Br J Haematol 148: 739 –753 National Institute for Clinical Excellence (2003) Improving Outcomes in Stata-Corp (2010) STATA 11. Stata-Corp: College Station, TX, USA Haematological Cancer: The Manual. National Institute for Clinical Turner JJ, Morton LM, Linet MS, Clarke CA, Kadin ME, Vajdic CM, Excellence: London Monnereau A, Maynadie M, Chiu BC-H, Marcos-Gragera R, Costantini Office for National Statistics (2001) Census: Standard Area Statistics AS, Cerhan JR, Weisenburger DD (2010) InterLymph hierarchical (England). ESRC/JISC Census Programme Census Dissemination Unit, classification of lymphoid neoplasms for epidemiologic research based University of Manchester: Manchester on the WHO classification (2008): update and future directions. Blood Oliver S, Smith A, Roman E, Bolton E, Ross J, Patmore R, Jack A (2011) 116: e90–e98 Evaluation of the completeness of national haematological malignancy Vardiman JW, Thiele J, Arber DA, Brunning RD, Borowitz MJ, Porwit A, registration: comparison of national data with a specialist population- Harris NL, Le Beau MM, Hellstrom-Lindberg E, Tefferi A, Bloomfield CD based register. In BSH 51st Annual Scientific Meeting. British Society for (2009) The 2008 revision of the World Health Organization (WHO) Haematology: Brighton classification of myeloid neoplasms and acute leukemia: rationale and ONS Geography (2004) Rural/Urban Definition and LA Classification. important changes. Blood 114: 937–951 Available at: http://www.ons.gov.uk/about-statistics/geography/products/ Westlake S (2009) Cancer incidence and mortality in the United Kingdom area-classifications/rural-urban-definition-and-la-classification/index.html and constituent countries, 2004–06. Health Statistics Quarterly/Office for (accessed 13 May 2011) National Statistics: 56–62 Parkin D (2006) The evolution of the population-based cancer registry. WHO (2001) Tumours of Haematopoietic and Lymphoid Tissues. Nature Reviews. Cancer 6: 603–612 World Health Organisation: Lyon Sant M, Allemani C, Santaquilani M, Knijn A, Marchesi F, Capocaccia R (2009) WHO (2008) WHO Classification of Tumours of Haematopoietic EUROCARE-4. Survival of cancer patients diagnosed in 1995–1999. Results and Lymphoid Tissues. International Agency for Research on and commentary. Eur J Cancer (Oxford, England: 1990) 45: 931–991 Cancer: Lyon Sant M, Allemani C, Tereanu C, De Angelis R, Capocaccia R, Visser O, Yoon SO, Suh C, Lee DH, Chi H-S, Park CJ, Jang S-S, Shin H-R, Park B-H, Marcos-Gragera R, Maynadie´ M, Simonetti A, Lutz J-M, Berrino F (2010) Huh J (2010) Distribution of lymphoid neoplasms in the Republic Incidence of hematologic malignancies in Europe by morphologic of Korea: analysis of 5318 cases according to the World Health subtype: results of the HAEMACARE project. Blood 116: 3724 –3734 Organization classification. Am J Hematol 85: 760–764 This work is published under the standard license to publish agreement. After 12 months the work will become freely available and the license terms will switch to a Creative Commons Attribution-NonCommercial-Share Alike 3.0 Unported License. British Journal of Cancer (2011) 105(11), 1684 – 1692 & 2011 Cancer Research UK

Journal

British Journal of CancerSpringer Journals

Published: Nov 1, 2011

There are no references for this article.