Epidemiology of cancer and screening

Chapter 15 Epidemiology of cancer and screening



David Hole, Paul Symonds



Chapter contents



The cancer problem


Cancer in the USA


Cancer in Europe 

Epidemiology of cancer


Terminology 

Survival and cure in cancer


Outcome of palliative care 

Epidemiology and the prevention of cancer


Criteria for causality 

Aetiology


Lung cancer 

Tobacco 

Involuntary smoking 

Asbestos 

Radon 

Chemoprevention 

Colon cancer 

Screening 

Breast 

Screening 

Stomach 

Screening 

Prostate 

Screening 

Cervix 

Screening 

Oesophagus 

Melanoma 

Head and neck 

Lymphoma 

Leukaemia 

Reducing the risks of developing cancer


Reducing tobacco smoking 

Modifying alcohol consumption 

Ultraviolet light 

Occupational exposure 

Diet 

Ionizing radiation 

Pollution


Chemoprevention 

Conclusion


Further reading



The cancer problem


In 2000, malignant tumours were responsible for 12% of the nearly 56 million deaths worldwide from all causes. In many countries, more than a quarter of deaths are attributable to cancer. In 2000, 5.3 million men and 4.7 million women developed a malignant tumour and altogether 6.2 million died from the disease. Cancer has now emerged as a major public health problem in developing countries, matching its effect in industrialized nations.


Cancer rates could further increase by 50% from 10 million new cases globally in 2000 to 15 million new cases in the year 2020. This increase will mainly be due to steadily ageing populations in both developed and developing countries and also to current trends in smoking prevalence and the growing adoption of unhealthy lifestyles. However, there is clear evidence that healthy lifestyles and public health action by governments and health practitioners could stem this trend, and prevent as many as one third of cancers worldwide.


Lung cancer is the most common cancer worldwide, accounting for 1.2 million new cases annually; followed by cancer of the breast, just over 1 million cases; colorectal, 940   000; stomach, 870   000; liver, 560   000; cervical, 470   000; oesophageal, 410   000; head and neck, 390   000; bladder, 330   000; malignant non-Hodgkin’s lymphomas, 290   000; leukaemia, 250   000; prostate and testicular, 250   000; pancreatic, 216   000; ovarian, 190   000; kidney, 190   000; endometrial, 188   000; nervous system, 175   000; melanoma, 133   000; thyroid, 123   000; pharynx, 65   000; and Hodgkin’s disease, 62   000 cases.


In developed countries, the probability of being diagnosed with cancer is more than twice as high as in developing countries. However, in rich countries, some 50% of cancer patients die of the disease while, in developing countries, 80% of cancer victims already have late-stage incurable tumours when they are diagnosed, pointing to the need for much better detection programmes. The main reasons for the greater cancer burden of affluent societies are the earlier onset of the tobacco epidemic, the earlier exposure to occupational carcinogens, and the Western nutrition and lifestyle. However, with increasing wealth and industrialization, many countries undergo rapid lifestyle changes that will greatly increase their future disease burden.



Cancer in the USA


Cancer, after heart disease, is the second leading cause of death. One of every four deaths in the USA is due to cancer. It is estimated that, in 2003, 1.33 million Americans would receive a new diagnosis of invasive cancer (excluding basal and squamous cell skin cancers) and over 0.5 million Americans will die of this disease. The most common cancers among men were prostate, lung and colorectal and among women were breast, lung and colorectal. For both sexes, these cancers made up over 50% of the new cases diagnosed in 2000. Age-adjusted incidence rates for all cancer sites combined were essentially stable from 1995 through to 2000 with increases in breast cancer among women and prostate cancer in men being offset by long-term decreases in lung cancer among men.


The four most common cancer killers – lung, breast, prostate and colorectal – all declined in the late 1990s. The steep decline in lung cancer rates in men and the recent slowing of an increase in rates in women are steps in the right direction and further progress will require rigorous application of strategies known to be effective in reducing tobacco use. Death rates from breast cancer are falling despite a gradual, long-term increase in the rate of new diagnoses. This has been attributed to the successful implementation of more effective adjuvant treatment and the increased use of mammography screening. However, higher rates of late-stage disease exist in some population groups such as women who lacked health insurance and recent immigrants.


Prostate cancer death rates have been declining since 1994, while incidence rates have increased dramatically since this time. Much of the increase is due to the wider use of prostate specific antigen (PSA) screening rather than a major increase in symptomatic disease. Colorectal cancer death rates have been declining since the 1970s with steeper declines beginning in the mid-1980s. This is probably due to improving surgical techniques and, more recently, the introduction of adjuvant treatment regimens.



Cancer in Europe


The International Agency for Research on Cancer has recently produced estimates of the number of new cancers developing and the number of deaths due to cancer which are likely to have occurred in 2004. Of an estimated total of 2   886   800 incident cases of cancer (excluding non-melanoma skin), the most common forms were lung (381   500   cases), colorectal (376   400   cases) and breast (370   100   cases). Among men, the top five cancers were lung, prostate, colorectal, bladder and stomach. Among women, the most common cancers were breast, colorectal, lung, endometrium and ovary. In terms of mortality, it is estimated there were 1   711   000 cancer deaths, with the main causes being lung (341   800   deaths), colorectal (203   700   deaths), stomach (137   900   deaths) and breast (129   900   deaths). With nearly 3 million new cancers and 1.75 million deaths each year, cancer remains one of the most important public health problems in Europe. As the European population ages, these numbers will increase even if the age-specific rates remain constant. Lung, colorectal, stomach and breast cancers account for nearly half of all cancer deaths in Europe and point very clearly to the priorities for cancer control action.


While many countries, particularly in the European Union, have seen falls in the age-standardized mortality rates for cancer over the past 20 years, there are a few notable exceptions (Greece, Spain, Portugal). Many of the countries from Central and Eastern Europe exhibit increasing trends with lung cancer a particular problem.



Epidemiology of cancer



Terminology


Epidemiology is the study of the occurrence, distribution and causes of disease. It can be divided into three main components – descriptive, analytical and experimental.


Descriptive epidemiology is concerned with the variation in frequency (incidence or mortality) of a disease over space, time and in relation to age, sex and socioeconomic status. Analytical epidemiology is the study of the relationship between potentially causal risk factors (e.g. cigarette smoking, asbestos, ionizing radiation) or their proxies and the development of disease. Experimental epidemiology involves observing the effects of controlling relevant adverse risk factors (e.g. stopping cigarette smoking) or promoting possible preventative factors (e.g. beta-carotene).


The incidence rate of any disease is the total number of new cases occurring over a given period of time (a year) among a given number of people (usually 100   000 for cancer). This is different from the prevalence rate, which is the number of cases of a particular disease alive at a given point in time. This should be qualified by a statement as to how far back in time cancer patients are considered to be at risk from their disease – this is often taken as referring to cancers diagnosed within the previous 5 years. The mortality rate is the number of deaths attributed to the cancer in a given period of time (a year) among a given number of people (usually 100   000 for cancer).


Crude incidence or death rates are defined as the total number of cancers or deaths divided by the total population. Crude rates are of limited value since they do not take account of the differing proportions of people in particular age groups between populations. The crude rate in a particular population may exceed that of another country simply because it contains a higher proportion of elderly people. The age-standardized incidence or death rate takes account of the effect of age distribution on rates by attributing a weighting system to each of the age-specific rates. A choice of weighting systems exists – the two most common are the world standardized population and the European standardized population. When comparing cancer incidence rates in countries with very different proportions of elderly people (e.g. those in the developing world versus those in the developed world), the world standard is recommended. This is known as the world standardized rate (WSR). For comparison of countries with more similar proportions of elderly people, the European Standard is preferable. This is known as the European standardized rate (ESR).


An alternative method of expressing relative incidence or mortality is known as the standardized incidence ratio (SIR) or the standardized mortality ratio (SMR). This is an overall measure of incidence (mortality) which compares the observed number of cases (deaths) in a particular population with the expected number of cases (deaths) that would have been anticipated in a standard population (e.g. national) if the age-specific incidence (death) rates were the same. This is usually expressed as a ratio of 100 if the observed and expected rates are the same, and greater than 100 if the rate in the population of interest is higher than that for the standard (national) rate. SIRs are useful in comparing cancer incidence between socioeconomic or particular occupational groups.



Survival and cure in cancer


What do we mean by ‘cure’ for a cancer patient? Technically, it is when the survival for that patient is the same as that for the general population from which they originate – their ‘excess’ risk of death becomes zero. Thus, it makes sense to talk of ‘cure’ of, for example, a small basal cell carcinoma treated by surgery or radiotherapy where recurrence or persistent disease is uncommon. However, for other cancers, particularly breast cancer, the possibility of local or distant recurrence remains for many years after the tumour has been ‘successfully’ removed surgically. This likelihood increases substantially for larger tumours with nodal involvement and adverse histological features. Thus, survival rates are quoted as a measure of success but should be qualified by the length of follow up – usually 5 years.


Five-year survival is often interpreted as ‘cure’. While this is true of many tumours, it is not true of breast cancer, as suggested above. For breast cancer, 10-, 20- or even 30-year survival figures are more appropriate end-points because of the long-term pattern of relapse and death from the disease. Survival may be with or without evidence of cancer. For this reason, the terms disease-free survival or recurrence-free survival are commonly used to define the outcome of treatment. These results require an elaborate system of clinic follow up. Yet, it is only from painstaking analysis of this kind that we can derive a sound knowledge of both the natural history of cancer and the effects of treatment. Survival data are commonly presented plotted graphically as curves which allow the comparison of different treatments for different stages of disease over time.


In studies of patients treated for cancer, a variety of ways is used for describing their survival. Crude survival rate refers to the percentage of patients alive a given number of years (n) after treatment. This is not valid unless all the patients included have been followed up for at least n years. More often, a sizeable proportion of patients have been followed up for a shorter time than n years, and the life table or Kaplan-Meier method is more appropriate. This method uses information from all the patients for the time intervals for which they have been followed. Thus, someone who has been in the study for 1 year will contribute to the first year survival estimate, those in for 2 years will contribute to the first and second year survival estimates, and so on. There is an underlying assumption that all patients are subject to the same time-specific probability of dying from a particular cancer whether or not they have been followed for all or part of the follow-up period of study. This method can be used for calculating survival rates for all causes of death (overall survival rate) or for a specific type of cancer (cancer-specific survival rate). Patients who die from causes other than cancer of interest (i.e. from intercurrent disease) are considered to have been withdrawn from the study at the point in time that their death occurred.


An alternative method for dealing with the problem of intercurrent deaths or mortality from natural causes of the patients studied is to introduce an age correction (age-corrected survival rate), which adjusts for the fact that some deaths would be expected in the cohort normally, and that this will vary with age and sex. Thus, the age-corrected survival rate enables direct comparisons between cohorts of patients with different age and sex distributions.



Outcome of palliative care


Survival figures are essential for estimating the success of treatment aimed at cure but are of limited value in assessing the effects of palliative treatment. The main aim of palliative treatment is the relief of distressing symptoms. Palliative radiotherapy or chemotherapy may sometimes restrain tumour growth sufficiently to prolong the patient’s life for a few months or sometimes years. In this case, a better measure of survival would be the median survival time. This is calculated by ranking the survival times in ascending order and choosing the middle value. It is subject to the proviso that all survival times for patients still alive, which are less than the middle value, are excluded and the median recalculated.



Epidemiology and the prevention of cancer


The principal role of epidemiological studies in cancer has been the identification of risk factors, the assessment of their likelihood of being causative agents and, ultimately, whether their avoidance offers a practical solution for reducing the cancer burden. Most of the research effort has been directed at identifying factors associated with an increased risk of cancer but it is equally legitimate to try to identify factors which are associated with a decrease in risk. One, often misunderstood concept, particularly by the general media, is that the findings of epidemiological studies are indicative of causal effects. In fact, epidemiological studies will only ever identify associations.



Criteria for causality


After individual epidemiological studies of cancer have been summarized and the quality assessed, a judgment is made concerning the strength of evidence that the agent or exposure in question is carcinogenic for humans. Several criteria are considered:



• a strong association (a large relative risk) is more likely to indicate causality than a weak one


• associations that are replicated in several studies of the same design or using different epidemiological approaches or under different circumstances of exposure are more likely to represent a causal relationship than isolated observations from single studies


• if the risk of the disease increases with the amount of exposure, this is considered to be a strong indication of causality


• demonstration of a decline in risk after cessation of or reduction in exposure in individuals or in whole populations also supports a causal interpretation


• although a carcinogen may act upon more than one target organ, the specificity of an association (an increased occurrence of cancer at one anatomical site or of one morphological type) adds plausibility to a causal relationship


• where an increase in exposure in a population and an increase in the incidence of a cancer are apparent, there must be a sensible latency (20+ years for solid tumours, 3+ years for blood-borne cancers) between the timings of the two increases


• there is evidence from animal models of a link between the exposure and the occurrence of the cancer


• there is a biological mechanism in humans for a link between the exposure and the cancer


• bias and confounding can be ruled out as possible explanations for the association.


It is not necessary for all of these criteria to hold for a judgment to be made that causality exists.


One current area of controversy that currently concerns public health is the impact of chronic low levels of exposure of known carcinogens. Two examples would be the impact of ionizing radiation to the general population and leukaemia risk, and the effect of inhaling second-hand tobacco smoke among non-smokers and lung cancer risk. In both cases, the first criterion listed above would not be expected to hold – the relative risks are likely to be low because the level of exposure is (relatively) low. In these circumstances, a more appropriate interpretation would be that the level of risk should be compatible with an acceptable extrapolation from the dose–response relationship observed for higher levels of exposure.


Such studies have played a major role in establishing occupational hazards in several industries (e.g. bladder cancer in aniline dye workers) and the link between lung cancer and smoking.


Epidemiology can assist in the prevention of cancer in a number of ways. First, it can show differences in the incidence of cancer in different populations and correlate them with differences in the prevalence of a potential causal factor. Secondly, it can test a hypothesis about the relationship between the occurrence of the disease to an aspect of the affected individual’s constitution or exposure to some environmental factor. Thirdly, it can test the validity of a causal relationship by seeing whether the disease can be prevented or its incidence reduced by changing the prevalence of the suspected agent. A good example of the latter is the reduction in lung cancer observed among doctors since they gave up smoking cigarettes.



Aetiology



Lung cancer



Tobacco


Lung cancer is currently still the largest single cause of death from any cancer in the world as well as being the most common incident cancer (excluding non-melanoma skin cancer). However, it was a rare disease at the start of the 20th century, but exposures to new etiological agents, particularly tobacco, and an increasing lifespan led to enormous increases in numbers of cases. While tobacco had been used throughout the world for many centuries, the 20th century pandemic was a result of the introduction of manufactured cigarettes on a mass scale. Its addictive properties led to sustained exposure of inhaled carcinogens. Scientists in Nazi Germany conducted some of the earliest research on the links between smoking and lung cancer, but their results were generally dismissed as propaganda. In the early 1950s, case-control studies in Britain and the USA suggested a strong association. Three major cohort studies were initiated at that time – UK doctors, US veterans and the American Cancer Society volunteers – and their findings corroborated the earlier observations as well as identifying other cancers and causes of death as also being strongly linked to cigarette smoking. For a smoker, lung cancer risk is related to cigarette smoking in accordance with the basic principles of chemical carcinogenesis. Risk is determined by the number of cigarettes smoked (dose), the number of years of smoking (duration) and the intensity of exposure (inhalation, tar levels). After taking these factors into account, women are at least as susceptible as men. An increase in risk of lung cancer (relative to a non-smoker) is consistently evident at the lowest level of daily consumption, and is at least linearly related to increasing consumption.


Doll et al reported on 50 years’ worth of observations of the UK doctors study initiated in 1951. It is only now that the emergence of the full hazards for persistent smokers can be gagged as it requires a study of men whose cigarette consumption as young adults was already substantial when those who are now old were young. Men born in the 1920s in the UK may well have had the most intense early exposure as widespread military conscription of 18-year-old men, which began in 1939 and continued for decades, routinely included provision of low cost cigarettes to the conscripts. The (relative) risks of lung cancer for all levels of smoking range from 7.5 for light smokers (<15/day) to 25.4 for heavy smokers (≥25/day) as compared with lifelong non-smokers. These levels of risk are higher than previously observed, highlighting the fact that earlier analyses have underestimated the full effect of persistent cigarette smoking.


The introduction of lower tar cigarettes was one mechanism by which it was hoped to reduce the risk of lung cancer. In practice, the reduction in risk has been far less than might be have been expected on the basis of reduced exposure and, in a recent analysis of the American Cancer Society cancer prevention study II, there was no difference in risk for men or women who smoked brands rated as very low (≤7   mg tar/cigarette) or low (8–14   mg) compared with those who smoked medium tar (15–21   mg) brands. There was an increase in risk for those who smoked high tar (≥22    mg) brands, these tended to be non-filter brands. The explanation for this phenomenon is described as ‘compensatory’ smoking. Addicted smokers who switch from a higher to lower tar cigarette can maintain their nicotine intake by blocking ventilation holes, increasing the puff volume or the time during which the smoke is retained in the lungs. As a result, the actual dose of toxicants to the smoker may be higher than is predicted by machine-measured yields. Thus, reducing the use of higher tar cigarettes will provide limited public health benefits.


Smoking cessation is by far the most effective means of reducing the risk of lung cancer among active smokers. There is a steady upward trend in lung cancer risk between lifelong non-smokers, those stopping between the ages of 25 and 34 years, 35–44 years, 45–54 years, 55–64 years and continuing smokers. Thus, there is substantial protection even for those who stop at 55–64 years, and progressively greater protection for those who stop earlier, up to as much as a 90% reduction in those stopping between 25 and 34 years.


On a global scale, the rise in cigarette consumption in developing countries such as China and India is alarming. Going on the current worldwide smoking patterns, whereby about 30% of young adults become smokers, there will be about 7 million lung cancer deaths worldwide unless there is widespread cessation. Convincing the next generation not to start smoking remains a major public health imperative.



Involuntary smoking


The question as to whether involuntary smoking poses a significant risk of lung cancer has received considerable scientific interest since 1980. Involuntary (or passive) smoking is exposure to second-hand tobacco smoke, which is a mixture of exhaled mainstream smoke and side stream smoke released from the smouldering cigarette and diluted with ambient air. Involuntary smoking involves exposure to the same numerous carcinogens and toxic substances that are present in tobacco smoke produced by active smoking, which is the principal cause of lung cancer. This implies that there will be some risk of lung cancer from exposure to second-hand tobacco smoke.


More than 50 studies, based on over 7300 non-smoking lung cancer cases, have examined the association between involuntary smoking and the risk of lung cancer in never-smokers, especially spouses of smokers, from many countries during the last 25 years. Most showed an increased risk, especially for persons with higher exposures. To evaluate the information collectively, in particular from those studies with a limited number of cases, meta-analyses have been conducted. The excess risk of lung cancer has been identified as being of the order of 24% and this excess could not be explained by chance, potential biases or confounding. Moreover, a recent international working group of 29 experts convened by the IARC Monograph Program concluded that second-hand smoke is carcinogenic to humans.


Interestingly, an association has also been shown between exposure to environmental tobacco smoke (ETS) in infancy and risk of lung cancer in adulthood among 60,   182 lifelong non-smokers who participated in the EPIC (European Prospective Investigation into Cancer and Nutrition) study.


With passive smoking now implicated as a cause of lung cancer as well as being a contributing factor in increasing coronary heart disease, the emphasis has now moved to what action public health policy makers should advocate and what legislation government should enact. The whole of the UK and Ireland have imposed bans on smoking in enclosed workspaces to protect the health of workers and a number of other countries are actively considering or recommending legislation. One of the additional benefits of such restrictions has been a reduction in the rate of ‘active’ smoking among the population at large.



Asbestos


It has been known for many years that exposure to asbestos is a major cause of lung cancer. In 1999, a meta-analysis of 69 cohort studies found a standardized mortality ratio (SMR) of 163 for lung cancer, but a substantial heterogeneity existed largely attributable to the different cumulative exposure to asbestos in various cohorts. These cohorts comprised workers from a variety of industries – mining, textile and manufacturing, insulation, asbestos-cement factories and shipyards – and risk was closely related to level of exposure in the different industries. Type of asbestos was also observed to be relevant, with chrysotile fibres being considered less carcinogenic than amphiboles.


Asbestos dust and cigarette smoke are both causes of lung cancer. What is less clear is the effect of the combination of the two carcinogenic agents. Initially, two hypotheses existed about the way asbestos and cigarette smoking interact. In one, it is assumed that asbestos produces the same additional risk in men who smoke cigarettes as in those who do not (additive hypothesis); in the other, it is assumed that asbestos produces an effect that is proportional to the effect of the other agent (multiplicative hypothesis). The additive hypothesis explains the data less well than the second, but the fit to the latter is not particularly convincing. Until recently, no other hypotheses have been considered even though the multiplicative hypothesis is just one of infinitely many putative forms of synergism.

Related posts:

Default ThumbnailPrinciples of management of patients with cancer Radiation treatment planning: immobilization, localization and verification techniques Thyroid cancer Imaging with radionuclides Radiotherapy beam production Imaging with x-ray, MRI and ultrasound

Stay updated, free articles. Join our Telegram channel
Tags: ,
Mar 7, 2016 | Posted by in GENERAL RADIOLOGY | Comments Off on Epidemiology of cancer and screening

Full access? Get Clinical Tree
Get Clinical Tree app for offline access