3.5 Other cancers

Last updated: June 2021  Suggested citation: Winnall, WR, Hurley, S, Greenhalgh, EM & Winstanley, MH. 3.5 Other cancers. In Greenhalgh, EM, Scollo, MM and Winstanley, MH [editors].  Tobacco in Australia: Facts and issues. Melbourne: Cancer Council Victoria; 2021. Available from  http://www.tobaccoinaustralia.org.au/chapter-3-health-effects/3-5-other-cancers

 

Smoking is a risk factor and recognised cause of many different types of cancer. This section summarises research on the association between smoking and:

• Head and neck cancers (including upper aerodigestive tract cancers)
• Pancreatic cancer
• Stomach cancer
• Kidney and bladder cancers
• Cervical cancer
• Acute myeloid leukaemia
• Liver cancer
• Colorectal (bowel) cancer
• Breast cancer
• Other cancers

3.5.1 Head and neck cancers

‘Head and neck cancers’ are a group of diseases comprising cancers of the oral cavity (mouth), sinuses, nose, salivary glands, tonsils, pharynx (throat) and larynx (voice box). Although other cancers occur in the head and neck region, such as in the brain and eye, these are not classified as head and neck cancers. In some studies, cancers in these areas are also grouped with those in the oesophagus (the tube connecting the throat to the stomach) under the term ‘upper aerodigestive tract cancers’ (UADTCs). This section discusses the association of smoking with cancers of the head and neck as well as oesophageal cancer.

Head and neck cancers are a considerable health issue in Australia. In 2019 there were an estimated 3,807 men diagnosed with a head and neck cancer, the 5th highest cancer diagnosis for men. There were an estimated 1,405 women diagnosed with a head and neck cancer in 2019, being the 11th highest cancer diagnosis in women. The age standardised incidence rate was almost three times higher for men as for women. Mortality is also higher for men in Australia, with 6.1 per 100,000 men dying from these cancers each year compared to 1.9 per 100,000 women.¹

The incidence of head and neck cancers has steadily declined in Australia for both sexes from 1976 to 2010.² However, this is not the case worldwide. There is a trend for rising incidence in some populations, particularly in young people.³

Australians diagnosed with a head and neck cancer have a 69.5% chance of surviving for at least 5 years.¹

3.5.1.1 Risk associated with smoking

Tobacco exposure is a risk factor and a cause for the majority of head and neck cancers and oesophageal cancer.^{2, 4-6} A meta-analysis published in 2009 reviewed 85 studies that included data on over 50,000 people with UADTCs. This study found an almost four-fold higher risk of these cancers for current smokers compared with never smokers.⁷ Concern about the increasing incidence of these cancers in young people prompted a case–control study in 10 European countries that included people aged under 50 years. Smoking was associated with a 5.5-fold increase in the risk of UADTC in this population.⁵ An Australian study has found that current smokers had over 2.5-fold increased risk of a head and neck cancer compared to non-smokers. This cohort of people from NSW were aged 45 and over, with smoking behaviour reported at baseline between 2006 and 2009 (from the 45 and Up Study).⁸

The strength of the association between smoking and head and neck cancers differs between the sub-types of cancers.

Laryngeal cancer (voice box cancer)

Smoking is a cause of laryngeal cancer.⁶ The larynx is directly exposed to tobacco smoke when it is inhaled through the space between the vocal chords, and smoking is particularly strongly associated with cancer of the larynx. Risks for smokers that are 20-fold higher than non-smokers have been reported in some studies.⁶ In Australia, the risk of laryngeal cancer for smokers is 11.3-fold higher than for non-smokers, for people 45 years and above.⁸ The aforementioned meta-analysis of 85 studies found that the risk of laryngeal cancer was nine times greater for current smokers than never smokers, whereas the risks of oropharyngeal and oesophageal cancers were about three-fold higher in smokers.⁷ The 2004 US Surgeon General’s report concluded that there was sufficient evidence to infer a causal relationship between smoking and cancer of the larynx, and that smoking and alcohol cause most cases of this cancer in the US.⁶

Oesophageal cancer

Oesophageal cancer has two main histological types: adenocarcinoma (which has further sub-types) and squamous cell carcinoma. The 2004 US Surgeon General’s report concluded that there was sufficient evidence to infer a causal relationship between smoking and both main forms of oesophageal cancer.⁶ Numerous studies have shown that smoking strongly increases the risk of squamous cell carcinoma and moderately increases the risk of adenocarcinoma of the oesophagus.^9-11 A meta-analysis of 52 studies found that the risk for squamous cell carcinoma was lower among former smokers (2-fold higher than non-smokers) than current smokers (over 4-fold higher than non-smokers).⁷ A pooled analysis of data from 10 case–control studies of adenocarcinoma also found that smoking doubled the risk.¹² A similar analysis of 13 case–control and two cohort studies found an 85% increase in the risk of oesophageal adenocarcinoma associated with smoking.¹³ An Australian case–control study that included more than 1,000 people found the risks of both squamous cell carcinoma and the rare gastro-oesophageal junction adenocarcinoma were about four-fold higher in smokers compared with never smokers. The risk of oesophageal adenocarcinoma was about two-and-a-half-fold higher in smokers.¹⁴ For Australian people aged 45 and over who smoke, the risk of oesophageal cancer was estimated to be 3.8-fold higher than for non-smokers.⁸ A meta-analysis of African studies showed that the odds of developing oesophageal cancer were over 3-fold higher for smokers and tobacco chewers compared to unexposed people.¹⁵ These and numerous other studies demonstrate the consistent and extensive epidemiological evidence supporting smoking as a cause of oesophageal cancer around the world.

Oral cancer (mouth cancer)

The 2004 US Surgeon General’s report concluded that there was sufficient evidence to infer a causal relationship between smoking and cancers of the oral cavity.⁶ An example of this evidence is a meta-analysis of studies from 1961 to 2003, finding that ever smokers had a 3.4 fold higher risk of oral cancer compared to non-smokers.¹⁶ There is also evidence that smokeless tobacco causes oral cancer. A meta-analysis of 37 studies showed that smokeless tobacco users in Southeast Asia had a 4.4-fold higher odds of oral cancer compared to those who refrained.¹⁷ The odds for those in the Eastern Mediterranean were 1.28-fold higher.¹⁷ Smokeless tobacco users in India had 5.5-fold higher odds of oral cancers compared to non-users in a similar meta-analysis.¹⁸ For more information, see Section 18A.3.

Pharyngeal cancer (throat cancer)

That the evidence is sufficient to conclude smoking is a cause of pharyngeal cancer is a conclusion of the 2004 US Surgeon General’s report.⁶ This finding is supported by a meta-analysis of studies from 1961 to 2003, which concluded people with a history of smoking had a 6.8-fold higher risk of pharyngeal cancer compared to non-smokers.¹⁶ A 2017 meta-analysis of 17 case-control and 4 cohort studies determined that smokers had significantly higher odds of nasopharyngeal cancer (occurring in the upper throat) than non-smokers. Current smokers and had a 59% greater risk and ever smokers a 56% greater risk than non-smokers. This study found a dose-dependent relationship where the risk estimate rose by 15% with every additional 10 pack-years of smoking.¹⁹

Based on its exhaustive review of available studies, the 2004 IARC report found sufficient evidence to conclude that tobacco smoking causes cancer of the oral cavity, naso-, oro- and hypopharynx, nasal cavity, paranasal sinuses, larynx and oesophagus.⁴ 

3.5.1.2 How tobacco smoke causes head and neck cancers and oesophageal cancer

The mechanisms by which smoking and tobacco exposure cause cancer are described in Section 3.3 and briefly summarised here. Carcinogens from tobacco and smoke are absorbed into cells where they are activated and bind to DNA (the part of a cell that contains the blueprint to create new cells). Sections of DNA bonded to a carcinogenic chemical are called DNA adducts. These DNA adducts lead to mutations in the sequence of genes that change the function of cellular components such as proteins and microRNAs. An accumulation of these changes culminates in dysregulation of cellular process and the uncontrolled cell proliferation that underlies cancer. Alternative mechanisms exist whereby carcinogens cause oxidative stress, epigenetic changes or hijack hormone and other signal transduction pathways, as described in Section 3.3.

The 2010 US Surgeon General’s report lists carcinogens for which there is evidence of their involvement in specific head and neck cancers.²⁰ Oral cancers are likely to be driven by PAHs (polycyclic aromatic hydrocarbons), NNK (4-(methylnitrosamino)-1(3-pyridyl)-1-butanone), and NNN (N'-nitrosonornicotine).²⁰ This finding was supported by a study from 2017 demonstrating that high levels of tobacco carcinogen-derived DNA damage in oral cells is an independent predictor of oral and head and neck cancer risk in smokers.²¹ Numerous studies have demonstrated that smokers have higher levels of DNA adducts in larynx, nasal and oral tissue samples than non-smokers.²⁰ Such carcinogens lead to DNA mutations and chromosomal changes, which are more numerous in the oral mucosa of smokers.²² For example, PAHs are predicted to cause mutations in the influential TP53 gene, found in laryngeal and oral tumours.²⁰

PAH exposure is associated with the development of cancer of the larynx. Damaging reactive oxygen species (such as free radicals) present in tobacco smoke are implicated as a cause of squamous cell carcinoma of larynx.²³ A study from 2019 analysed the patterns of mutations found in squamous cell carcinomas from the larynx, oral cavity and oropharynx. This study determined that larynx carcinomas were characterised by slow growth and increased susceptibility to mutations from tobacco carcinogen DNA adducts.²⁴

NNN is the most prevalent and most potent oesophageal carcinogen detected in cigarette smoke.²⁰ Other nitrosamines are also expected to play a role in the development of oesophageal cancer.²⁰ Laboratory experiments and clinical studies indicate that a disruption in DNA repair processes by smoking is one way that it leads to oesophageal cancer.^{25, 26}

Nasal cancers caused by smoking are likely to involve NNK, NNN, other nitrosamines and aldehydes.²⁰ Smoking also changes the microbiome of the saliva, but whether this has a causative link with head and neck cancers will need further research.²⁷

Although the links between smoking and head and neck cancers/UADTCs are well established, only a small proportion of people exposed to tobacco develop these cancers. Some possible mechanisms behind the differing individual susceptibility to cancer after exposure to tobacco are discussed in Section 3.3. In particular, differing variants of genes responsible for activating carcinogens after exposure are believed to modify the risk from that exposure. Some of these gene variants result in lower amounts of activated carcinogen available to cause DNA damage. There is also variability in the efficiency of the DNA repair process that removes DNA mutations before they can cause damage.

Aided by information from the human genome, scientists are studying the impact of gene variations on differential susceptibility to cancers, such as head and neck cancers, among tobacco users. A 2011 meta-analysis, for example, found that DNA changes that remove the expression of the glutathione S-transferase M1 (GSTM1) gene were associated with an increased risk of oral cancer in Asian people but not Caucasians.²⁸ However, for people with this change, smokers were at lower risk of oral cancer than non-smokers or light smokers. This very large field of research continues to bring insights into the molecular pathways by which carcinogens cause cancer and how these differ between people. Insights from these studies may help to identify people at greater risk of cancer from smoking.

3.5.1.3 Factors affecting risk

Intensity and duration of smoking

A 2009 meta-analysis of 85 studies found a dose–response relationship for smoking and the risk of UADTCs (collectively). Compared to people who never smoked, those who smoked fewer than 20 cigarettes per day had a 2.49-fold increased risk of any UADTC, and those who smoked 20 or more had a 4.42-fold increased risk. While the increased in risks for higher intensity smoking were subtle for cancers of the oesophagus and oropharynx, there was a 19-fold increased risk of laryngeal cancer for heavy smokers and 7.66-fold risk for lighter smokers compared to never smokers.⁷ A pooled analysis of 15 case–control studies found that the risk of laryngeal cancer was more strongly associated with greater intensity of smoking (cigarettes per day) than the duration of smoking.²⁹ In 2020, two studies were published that performed pooled analyses of 33 case–control studies.^{30, 31} These studies also predicted that duration of smoking had a stronger effect on the risk of head and neck cancers than the number of cigarettes smoked per day.

Alcohol consumption

Alcohol and tobacco are considered to be the two most important risk factors for head and neck cancers. Since people who smoke often drink, alcohol consumption levels affect the extent of risk associated with smoking.

The effects of smoking and alcohol are synergistic in relation to head and neck cancers/UADTC (meaning that the combined effect of tobacco and alcohol exceeds the sum of their individual effects).^{4, 6} In the meta-analysis of 85 studies referred to above, 21 studies investigated the combined effects of alcohol and smoking on the risk of developing any UADTC.⁷ The risk of any UADTC in people who had the highest consumption of alcohol and tobacco was almost five times higher than the risk among heavy smokers who did not drink alcohol, or who drank moderately. The synergistic effect of tobacco and alcohol on cancer risk is strongest for cancer of the larynx and weakest for cancer of the oesophagus. A pooled analysis of 17 European and American case–control studies found that 4% of head and neck cancers were attributable to alcohol alone, 33% to tobacco alone and a further 35% were due to tobacco and alcohol combined.³² For cancer of the larynx, these figures were: 3% due to alcohol alone, 52% due to tobacco alone and 33% due to tobacco and alcohol combined.

Human papilloma virus infection

Human papilloma virus (HPV) infection can be a cause of cancer; it is suspected to cause nearly all cases of cervical cancer. HPV also infects the mouth and throat, putting these regions at risk of cancer.³³ It is estimated that 12% of pharyngeal cancers and 30–60% of oropharyngeal carcinomas are caused by HPV infection.³³ A 2016 meta-analysis concluded that there is a potentially significant casual relation between HPV and oral and oropharyngeal cancers.³⁴ HPV-driven oropharangeal cancer patients have a better prognosis than those who do not have the infection. However, smoking increases the risk of poor outcomes and mortality among these people.^{35, 36} Heavy smokers (≥20 per day) with HPV had over twice the risk of cancer recurrence and of mortality from oropharyngeal cancer.³⁷ The effects of smoking on the prognosis of head and neck cancers was stronger if the smoking started earlier in life than sexual activity started.³⁸

Smoking cessation

The 2020 US Surgeon General’s report into Smoking Cessation determined that there was sufficient evidence to infer that smoking cessation reduces the risks of cancers of the larynx, oral cavity, pharynx and oesophagus.³⁹ The risks of laryngeal, oral and pharyngeal cancer are approximately halved within 10 years of cessation. In 2007, the International Agency for Research on Cancer reported that, even after a long period of abstinence, the risks for laryngeal and oesophageal cancer do not return to that of never smokers.⁴⁰ Conversely, a meta-analysis in 2009 found that risk of developing any UADTC declines when people quit smoking relative to continued smoking, and can reach the level of non-smokers by about 10 years after smoking cessation.⁷ A meta-analysis from 2010 showed that quitting smoking for 1 to 4 years resulted in a 30% reduction in the odds of developing a head and neck cancer compared to that for current smokers. Quitting for over 20 years was necessary for the risks to return to that of never-smokers.⁴¹ A large meta-analysis of 52 studies showed that smoking cessation decreased the risk of oesophageal squamous cell carcinoma, while it had limited influence on the risk of oesophageal adenocarcinoma.¹¹ Collectively, these and other studies suggest that smoking cessation can reduce the risk of head and neck and oesophageal cancers.

3.5.1.4 Impact of smoking on prognosis

Smoking is also associated with poor outcomes from head and neck cancers. A 2019 meta-analysis showed that for people with head and neck cancers receiving radiotherapy, continued smoking was associated with lower overall survival, lower loco-regional control and higher risk of late toxicities.⁴²

A meta-analysis involving 4,286 patients showed that current and former smokers with oesophageal squamous-cell carcinoma had a poorer prognosis and lower survival rates than never smokers. This was also the case for oesophageal adenocarcinoma.⁴³

Continued smoking after diagnosis of a UADTC increases the risk of a second primary tumour in the region. This is consistent with the concept of ‘field cancerisation’, which is the simultaneous occurrence of carcinogenic alterations in multiple areas of the upper aerodigestive tract.⁴

Studies of patients with head and neck cancer have shown poorer local-regional control,^{44, 45} disease-free survival⁴⁵ and overall survival^44-47 for patients who were smoking at diagnosis or had smoked in the past, compared with never smokers. Patients who continue to smoke while receiving treatment have poorer loca-regional control,⁴⁸ disease-free survival^{44, 48} and overall survival^{44, 47-49} compared with those who quit.

3.5.2 Pancreatic cancer

The pancreas is an abdominal organ that secretes enzymes and other compounds to aid digestion, as well as hormones for controlling blood sugar levels. Pancreatic cancer often goes without symptoms in the early stages, therefore detection usually occurs once it has spread, leading to a poor prognosis. Risk factors for pancreatic cancer include tobacco, heavy alcohol consumption, diabetes, obesity and genetic factors.⁵⁰

There were an estimated 3,933 people diagnosed with pancreatic cancer in Australia in 2020 and 3,300 deaths from this disease.⁵¹ Pancreatic cancer has a relatively low survival rate, with only 11% of people surviving for five years or more after diagnosis.^{1, 51} It is estimated to be the third most common cause of death from cancer in Australia in 2020.¹

3.5.2.1 Risk associated with smoking

Smoking is a well-recognised cause of pancreatic cancer.^{4, 6, 50} A pooled analysis of 12 international prospective cohort studies involving almost 1,500 cases found that current smokers had a 77% higher odds of pancreatic cancer than never smokers.⁵² The US Surgeon General’s report of 2004 concluded that the evidence is sufficient to infer a causal relationship between smoking and pancreatic cancer.⁶ Since this report, numerous meta-analyses have strengthened this conclusion.⁵³ An Australian study showed that, for people aged 45 and above, the risk of pancreatic cancer for smokers was 2.7-fold higher than for non-smokers.⁸ Another study from 2018 has determined that the risk of pancreatic cancer diagnosis sharply increases with a low number of cigarettes, or after a few years of smoking, and that it rapidly decreases a few years after cessation, though taking almost 20 years to reach the same risk as that of never smokers.⁵⁴  

3.5.2.2 How tobacco smoke causes pancreatic cancer

NNN and its break-down product NNK, mentioned in Chapter 3, Section 3.5.1.2, are the two known pancreatic carcinogens in tobacco products.²⁰ Cigarette smoke is known to increase the presence of DNA adducts, causing cancer in human pancreatic cells.⁵⁵

3.5.2.3 Factors affecting risk

The risk of pancreatic cancer increases with smoking intensity (cigarettes per day), smoking duration (years smoked) and cumulative smoking dose (pack-years). This risk decreases after quitting, reaching the same level as that of non-smokers after about 15 years.⁵² In Australia, male pancreatic cancer mortality has declined in line with reductions in tobacco consumption approximately 15 years previously.⁵⁶ However, pancreatic cancer mortality has continued to rise in Australian women. This may be due to the later peak in female tobacco consumption (compared with male tobacco consumption) or due to other factors, such as obesity, that are affecting female pancreatic cancer mortality.⁵⁶

Heavy drinking of alcohol is also associated with an increased risk of pancreatic cancer. A comprehensive review investigating a potential interaction between smoking and alcohol as risk factors determined that the combined effect of smoking and total alcohol risk of pancreatic cancer is likely to be less than additive.⁵⁷

3.5.2.4 Impact of smoking on prognosis

A meta-analysis from 2019 has shown that current and formers smokers had an elevated risk of dying from pancreatic cancer compared to never smokers.⁵⁸

3.5.3 Stomach cancer

The stomach is part of the digestive system, located between the oesophagus and the small intestine. The stomach secretes enzymes and acid, breaking down food using these secretions and through churning via muscular contractions.

Gastric (stomach) cancer is one of the most common and deadly cancers in the world, particularly for men. It is the third leading cause of cancer death on a global scale. Incidence rates are highest in Eastern and Central Asia.⁵⁹ In Australia the incidence of this cancer is relatively low compared to these regions. There were an estimated 2,246 cases and 1,140 deaths in Australia from gastric cancer in 2020. People in Australia diagnosed with gastric cancer have a 31% chance of surviving for at least five years, based on data from 2012 to 2016.⁶⁰ The age-standardised incidence rate of gastric cancers has approximately halved in Australia since 1982.

Gastric cancer is estimated to be the 11^th most common cause of cancer death in Australia in 2019. The mortality rate from gastric cancers in Australia has also fallen in the past 50 years. In 1968 the mortality rate was over 20 per 100,000 people, compared to 3.8 per 100,000 in 2018. Australian men have had a higher mortality rate from gastric cancer than women over these years, consistent with higher rates of gastric cancer and of smoking for men.⁶⁰

There are numerous different classification systems for gastric cancers.⁶¹ Pertinent to the research discussed here is the cardia versus non-cardia classification system, based on the region of origin in the stomach. Cancers occurring at the gastric cardia, which is near the junction of the oesophagus and stomach, are referred to as gastric cardia cancers. Cancers of the gastric fundus (dome shaped region near and above the cardia), corpus (the main, large part) and the antrum (final region before the lower outlet), and are termed non-cardia cancers.

3.5.3.1 Risk associated with smoking

Smoking is a cause of gastric cancer.^{4, 6} In 2004, the US Surgeon General’s report concluded that the evidence is sufficient to infer a causal relationship between smoking and gastric cancer.⁶ Numerous epidemiological studies since this report have supported this finding. A meta-analysis of 46 case–control studies found an estimated 50% increased odds of gastric cancer for people who have smoked.⁶² For current smokers, the odds of gastric cancer were about 70% higher than for never smokers. A meta-analysis of 21 case–control and three cohort studies found that smoking increased the risk of gastric cardia adenocarcinoma by 76%.¹³ A 2020 meta-analysis of 232 studies (with over 33 million people) found that current smoking increased the odds of gastric cancers by 61%.⁶³

Smoking causes the deaths of about 10% of people who die from stomach cancer and almost half the deaths of people who die from oesophageal cancer. This was about 770 out of 2700 deaths from these types of cancers in Australia in 2015.⁶⁴

3.5.3.2 How tobacco smoke causes stomach cancer

Unlike lung and head and neck cancers, the carcinogens from tobacco exposure that lead to gastric cancer are not well established. Laboratory studies indicate that NNK may play a role in the formation of gastric cancer.^{65, 66} DNA adducts (DNA bonded to a carcinogenic chemical) have been identified in tissue samples from the stomach of smokers.²⁰ Nicotine itself also affects gastric physiology, although it is currently unclear whether this affects carcinogenesis.⁶⁷

3.5.3.3 Factors affecting risk

Intensity and duration of smoking

The 2004 International Agency for Research on Cancer (IARC) report stated that eight of 16 cohort studies reported a significant dose–dependent relationship between intensity of smoking and risk of gastric cancer. Five studies found a relationship between duration of smoking and risk. Most of the case–control studies examining dose–responses discussed in this report have found relationships between both intensity and duration of smoking and the risk of gastric cancer.⁴ The IARC concluded that gastric cancer risk increases with duration of smoking and number of cigarettes smoked.

Helicobacter pylori infection

Helicobacter pylori is a major risk factor and cause of non-cardia gastric cancer. A prospective study of 1,071 Japanese men who were followed up for 14 years found that smoking and H. pylori infection had a synergistic effect on the risk of gastric cancer.⁶⁸ The risk of gastric cancer for men who smoked was almost 2-fold higher than non-smokers, and the risk of gastric cancer from H. pylori infection in non-smokers was 6.9-fold higher than those uninfected. But for people who both smoked and were infected, the risk of gastric cancer was 11.4-fold higher than those with neither risk factor. The effects of both risk factors together were therefore greater than the sum of their individual effects, termed synergism. In this population, 7.3% of gastric cancer was estimated as due to smoking alone, 30.1% was due to H. pylori infection alone and 49.6% was due to cigarette smoking with simultaneous H. pylori infection.

Cancer type

The pathogenesis of gastric cancer is thought to differ between cardia and non-cardia types. H. pylori, a strong risk factor for gastric cancer in general, increases the risk of non-cardia cancers.⁶⁹

Studies that distinguish between cardia and non-cardia gastric cancers show that smoking is significantly associated with both types.⁴ A 2009 meta-analysis identified five studies of non-cardia gastric cancer and four studies of cardia gastric cancer.⁶² Smoking significantly increased the risk of both types of cancer; the risk of cardia cancer was increased by 47% and the risk of non-cardia cancer by 32%.

Whilst concluding that smoking is a cause of gastric cancer in general, the 2004 US Surgeon General’s report stated that: ‘The evidence is suggestive but not sufficient to infer a causal relationship between smoking and non-cardia gastric cancers’.⁶ This issue has been difficult to resolve because many studies of the risk of gastric cancer associated with smoking have not distinguished between cancer subtypes and have not accounted for infection with Helicobacter pylori.

Gender differences

A meta-analysis of one nested case–control study and nine prospective cohort studies from 2019, included over three million people, examined risk factors for gastric cancer. This study showed that the risk for male current smokers was 1.3 fold higher than for female smokers, indicating a potential sex difference in the risk of gastric cancers.⁷⁰

Smoking cessation

The 2020 US Surgeon General’s report stated that the evidence is sufficient to infer that smoking cessation reduces the risk of stomach cancer.³⁹ Numerous studies and meta-analyses have demonstrated that the risk of gastric cancer is lower in former smokers compared to current smokers and decreases with increasing duration of quitting.^{4, 39}

3.5.4 Kidney and bladder cancers

The urinary system includes the bladder, ureters, urethra and kidneys, which filter the blood and excrete waste as urine. Urine is transferred to the bladder via the ureters, and exits the body through the urethra. The kidneys also have homeostatic roles in regulating blood pressure, acid-base balance and electrolytes.

Kidney and bladder cancers are a significant concern in Australia. Kidney cancer is estimated to be the seventh most common cancer in Australia, with 4,193 people diagnosed and 917 deaths in 2020. The proportion of people surviving for at least five years after diagnosis with kidney cancer is 79%.⁷¹ There were an estimated 3,098 people diagnosed with bladder cancer and 1,016 deaths from this disease in 2020. People with bladder cancer have a 54% chance of survival for at least five years.⁷² There were 503 people diagnosed with cancer of other urinary organs, including the urethra and ureter in 2015, causing 355 deaths in 2016.¹  

3.5.4.1 Risk associated with smoking

Smoking causes cancer of the kidney,⁶ bladder and renal pelvis (region where the ureter joins the kidney).^{4, 6} The 2004 US Surgeon General’s report stated that the evidence is sufficient to infer a causal relationship between smoking and renal cell (kidney), renal pelvis and bladder cancers.⁶ More recent meta-analyses have confirmed a greater risk of kidney cancer⁷³ and bladder cancer⁷⁴ for those with a history of smoking. The IARC has estimated that 66% of bladder cancer in men and 30% in women is due to smoking.⁴ In Australia, an estimated 35% of bladder cancers in men and 29% in women are caused by smoking.⁷⁵ Australian data also show that, for people aged 45 and over, the risk of bladder cancer is over 3-fold higher in current smokers compared to non-smokers.⁸

A case–control study and accompanying editorial published in 2009 suggest that the association between bladder cancer and smoking has increased substantially between 1994 and 2004.^{76, 77} The risk of bladder cancer for current smokers relative to never smokers was 2.9-fold higher in 1994–1998, 4.2-fold in 1998–2001 and 5.5-fold in 2001–2004. The editorial points out that the concentrations of specific carcinogens in tobacco smoke have increased, and speculates that this may be the cause of the observed increase in smoking-attributable bladder cancer risk.⁷⁷

3.5.4.2 How tobacco smoke causes kidney and bladder cancers

Aromatic amines, including 2-naphthylamine and 4-aminobiphenyl (4-ABP), are combustion products of cigarette smoke and are known bladder carcinogens. These compounds are believed to contribute to bladder cancer caused by smoking.^{4, 20}

3.5.4.3 Factors affecting risk

A 2005 meta-analysis of 24 studies concluded that the risk of renal cell carcinoma increases with smoking intensity and declines after quitting.⁷⁸ A later meta-analysis showed that the risk of kidney cancer rose non-linearly with smoking intensity; even smoking only a few cigarettes per day significantly increased the risk.⁷³

The risks of bladder cancer and cancers of the renal pelvis and the ureter increase with the amount of tobacco consumed and the duration of smoking. For bladder cancer there was a levelling off of risk at high daily consumption levels, possibly due to under-reporting of consumption by heavy smokers.⁴ A 2019 meta-analysis of 15 case–control studies showed that for a fixed number of pack-years, smoking for a longer duration at lower intensity was more deleterious for bladder cancer risk than smoking more cigarettes per day for a shorter duration.⁷⁹

Smoking appears to be more strongly associated with risk of bladder cancer in women than in men. The reasons for this may be differences in metabolism, smoking behaviours, exposure patterns and DNA repair mechanisms between the genders.⁸⁰

The 2020 US Surgeon General’s report stated that the evidence is sufficient to infer that smoking cessation reduces the risks of bladder and kidney cancers.³⁹ The risk of bladder cancer declines after smoking cessation, rapidly in the first one to four years. However, even after 25 years the risk is not as low as for non-smokers.⁴

3.5.4.4 Impact of smoking on prognosis

A 2002 systematic review found that smoking cessation might favourably alter the course of bladder cancer, but the evidence was insufficient to conclusively recommend to patients that quitting will improve their prognosis.⁸¹ A recent study found that smoking status and a higher cumulative smoking exposure are associated with worse prognoses in patients with primary non–muscle-invasive bladder cancer, who had higher rates of disease recurrence and progression, and lower overall survival.⁸² A Japanese study of 963 people with renal cell carcinoma showed that smoking 20 or more cigarettes daily at diagnosis was associated with poorer overall and cancer-specific survival.⁸³ Smokers with bladder cancer have a lower chance, compared to non-smokers, of their chemotherapy treatment working well, which is a strong predictor of survival.^{84, 85} A meta-analysis of 14 studies also showed that active smoking was associated with poorer outcomes for people with renal cell carcinoma.⁸⁶ Smokers with bladder cancer had a 1.87-fold higher risk than non-smokers of major complications after surgery to remove the bladder.⁸⁷   

3.5.5 Cervical cancer

The cervix is the lower portion of the uterus, where it joins the vagina. Cervical cancer has a relatively low survival rate, but its incidence and mortality have dropped in Australia since the introduction of screening programs.⁸⁸

In 2020 there were an estimated 933 people diagnosed with cervical cancer and 238 deaths from this disease. The chances of a person surviving for at least five years after diagnosis are 74%, based on data from 2012 to 2016. In 2015, cervical cancer was the 14^th most common cancer in females and the 19^th most common cause of cancer death in 2016.⁸⁹

The most common risk factor for cervical cancer is infection with the human papilloma virus (HPV), see Section 3.9.7.

3.5.5.1 Risk associated with smoking

Women who smoke are more likely to develop cervical cancer. The IARC and US Surgeon General reports in 2004 both concluded that smoking is a cause of cervical cancer.^{4, 6} Subsequent case–control studies^{90, 91} provided additional evidence that smoking is an independent risk factor for cervical cancer. A 2019 meta-analysis of two cohort and three case–control studies among Japanese women found convincing evidence that cigarette smoking increases the risk of cervical cancer.⁹² A nested case–control study also supported this conclusion.⁹¹ These studies found that whether or not someone has the HPV infection, those with a history of smoking were about twice as likely to get cervical cancer compared to non-smokers.^{91, 92} Two meta-analyses have found evidence of an independent association of passive smoking with risk of cervical cancer.^{93, 94}

Two small studies have reported associations between smoking and neoplasia of other genital tract tissue.^{95, 96} Neoplasia is the abnormal proliferation of cells that may progress to cancer.

3.5.5.2 How tobacco smoke causes cervical cancer

Polyaromatic hydrocarbons (PAH) and NNK are tobacco carcinogens considered likely to be involved in causing cervical cancer.⁶ Cervical samples from smokers have higher levels of DNA adducts (DNA bonded to a carcinogenic chemical) than non-smokers.⁴ It is predicted that in combination with HPV these compounds may contribute to the development of cervical cancer in smokers.^{4, 20}

Tobacco-related carcinogens are suspected to increase the infection of HPV, a cause of cervical cancer.^{97, 98}

3.5.5.3 Factors affecting risk

HPV infection is a strong risk factor for cervical cancer. HPV is more common in women who have sex at a younger age, who have many partners or are in a lower socioeconomic tier. These risk factors are shared with women who smoke, meaning that HPV infection status will affect the risk of cervical cancer in smokers.⁶ In studies where statistical methods were used to adjust for the effects of HPV infection, smoking was shown to be an independent risk factor for cervical cancer.^{91, 92, 97} After adjusting for HPV infection, a Nordic study determined that smokers had a 2.7 increased in odds of developing cervical cancer.⁹¹

There is a dose–response relationship between smoking and cervical cancer; the risk of cervical cancer increases with the duration of smoking.^{6, 91, 92}

The risk from smoking differs for different types of cervical cancer There is evidence that smoking increases the risk of squamous cell carcinoma, but not adenocarcinoma of the cervix.⁹⁷

The 2020 US Surgeon General’s report concluded that the evidence is sufficient to infer that smoking cessation reduces the risk of cervical cancer.³⁹ A 2003 meta-analysis of 28 studies compared the risks of current and former smokers to that of never smokers. The relative risk of cervical cancer for former smokers was lower than that of current smokers (1.26 fold compared to 1.83 fold).¹⁶

3.5.5.4 Impact of smoking on prognosis

Several studies suggest that smoking affects prognosis for women with cervical cancer. A study in the US involving approximately 2,500 women with cervical cancer, who were followed-up for five years, found that smokers were about 20% more likely to die from cervical cancer.⁹⁹ Former and current smoking were also associated with decreased survival and reduced disease control for cervical cancer patients undergoing radiotherapy.¹⁰⁰ 

3.5.6 Acute myeloid leukaemia

Acute myeloid leukaemia (AML) is a cancer that affects the blood and bone marrow (the part of the bone that produces blood cells). Myeloid leukaemias involve overproduction of immature white blood cells called myeloblasts, preventing the bone marrow from making normal blood cells. AML develops quickly, with anaemia, bleeding and bruising occurring because of inadequate numbers of red cells and platelets. If untreated, AML is rapidly fatal. In contrast, chronic myeloid leukaemia develops more slowly and urgent treatment is usually unnecessary. There are at least eight different sub-types of AML.

In Australia, there were 1,042 people diagnosed with AML in 2015 and 971 deaths from this disease in 2016. People diagnosed with AML have a 28% chance of surviving for at least five years, based on data from 2011 to 2015.¹ AML is the second most common myeloid cancer,¹⁰¹ and myeloid cancers were the sixth most common cancer reported in 2007.¹⁰²

3.5.6.1 Risk associated with smoking

Smoking is a cause of AML.^{4, 6} The 2004 US Surgeon General’s report concluded that the evidence is sufficient to infer a causal relationship between smoking and AML, and that the risk increases with the number of cigarettes smoked and the duration of smoking.⁶ In the studies assessed by the 2004 Surgeon General’s report, the risk of AML for people with a history of smoking was 1.3 to 1.5-fold higher than the risk for those who had never smoked.⁶ A 2016 meta-analysis of 27 studies of adults showed that current smokers had a 1.36-fold higher odds of developing AML than never smokers, and former smokers had a 1.21-fold higher odds of AML.¹⁰³ A 2019 meta-analysis of 20 case–control studies, involving 7,538 people with AML and 137,924 healthy controls, showed a 1.42-fold higher odds of AML for current smokers compared to never smokers.¹⁰⁴

Smoking by parents is associated with childhood AML. There was a 1.34-fold increased odds of childhood AML with perinatal parental smoking found by a meta-analysis from 2016.¹⁰⁵    

3.5.6.2 How tobacco smoke causes acute myeloid leukaemia

Cigarette smoke contains benzene, a known carcinogen that causes leukaemia.¹⁰⁶ A study that combined epidemiological data on the health effects of smoking with risk assessment techniques for low-dose extrapolation estimated that benzene from cigarette smoke was responsible for 12–58% of deaths from AML induced by smoking.¹⁰⁷ A similar study from 2018 calculated that between 9% and 24% of smoking-associated deaths from AML were due to benzene.¹⁰⁸ Polonium-210 and lead-210, both of which emit ionising radiation, are also found in cigarette smoke. Ionising radiation is a recognised cause of leukaemia, however more evidence is required to determine whether these elements are a cause of the leukaemias attributable to smoking.^{6, 20}

3.5.6.3 Factors affecting risk

The risk of AML increases with the amount smoked and the duration of smoking.⁶ A meta-analysis of 22 case–control and five cohort studies from 2016 concluded that the risk of AML increases with the number of smoking years, the number of cigarettes per day and the pack-years. People who smoked more than 20 years, and those who smoked between 1 to 20 cigarettes per day, and those who smoked over 20 cigarettes per day all have a significantly higher risk of developing AML.¹⁰³

Smoking cessation reduces the risk of AML. The 2020 US Surgeon General’s report concluded that the evidence is sufficient to infer that smoking cessation reduces the risk of acute myeloid leukaemia.³⁹ This report detailed a pooled analysis showing the risk of AML declined with increasing time since smoking cessation. There was no statistically significant reduction in risk for former smokers who had quit within 10 years compared with continuing smokers. However, the risk was lower for those who had quit for 10–20 years (odds ratio of 0.74) and lower still for those who had quit for more than 20 years (odds ratio of 0.59).³⁹

3.5.6.4 Impact of smoking on prognosis

Smoking status was an independent prognostic factor for survival for people with AML. People with AML who had never smoked lived for an average 60 months after diagnosis, compared to smokers who lived an average 30 months.¹⁰⁹ A small study of 148 patients undergoing stem cell transplantation for treatment of acute myeloid leukaemia found that smokers had longer hospitalisation and poorer survival.¹¹⁰ 

3.5.7 Liver cancer

The liver is an abdominal organ essential for metabolism and detoxification. It has multiple functions including production of bile for fat digestion, breakdown of toxic compounds, regulation and storage of carbohydrates and breakdown of aging red blood cells. About 80% of liver cancers are hepatocellular carcinomas, with numerous other types of liver cancer that are less common. Risk factors for liver cancer include cirrhosis from hepatitis infection, use of tobacco, alcohol, illicit drugs, unsafe sex, diabetes and obesity.¹

In Australia, liver cancer is considered a less-common cancer. In 2019 there were an estimated 2,662 people diagnosed with liver cancer and 2,297 deaths. It was the 10^th most-common cancer for men and the 19^th most common for women.¹ People with liver cancer in Australia have a 20% chance of surviving for at least five years, based on data from the period 2012–2016.¹¹¹

Mortality from liver cancer is relatively high. It was estimated to be the fifth most common cause of cancer mortality for men and the eighth most common for women in 2019.¹

3.5.7.1 Risk associated with smoking

The IARC concluded in 2004 and the US Surgeon General concurred in 2014 that tobacco smoking is a cause of liver cancer.^{4, 112} Researchers from the IARC conducted a meta-analysis of 38 cohort studies and 58 case–control studies to clarify the potential association. The results showed a statistically significant 51% increase in risk for current smokers compared with never smokers.¹¹³ Similarly, the 2014 US Surgeon General’s report found a 60-70% increased risk of liver cancer in current smokers compared with never-smokers.¹¹² Australians aged 45 and over have a 4-fold higher risk of liver cancer if they are current smokers, compared to non-smokers.⁸

A large meta-analysis from 2017 provided further evidence that smoking is associated with liver cancer. This study used data from 81 case–control and cohort studies. Compared with never smokers, current smokers had a 1.55-fold higher odds of hepatocellular carcinoma. The odds for former smokers were 1.39-fold higher and for heavy smokers were 1.9-fold higher.¹¹⁴

3.5.7.2 How tobacco smoke causes liver cancer

The liver metabolises many circulating carcinogens from tobacco smoke,¹¹² with NNK, other nitrosamines and furan identified as likely liver carcinogens.²⁰ A number of potential mechanisms have been identified for liver carcinogenesis, including long-term exposure to carcinogens in cigarette smoke increasing the risk of liver cell damage, promoting the development of cancer. Smoking also increases the risk of liver fibrosis, primary biliary cirrhosis, and chronic liver disease, which can progress to liver cancer.¹¹²

3.5.7.3 Factors affecting risk

A meta-analysis found some evidence of an increase in risk with the number of cigarettes smoked per day, but this effect differed between studies.¹¹³

The incidence of liver cancer in Australia has increased almost 4-fold since 1982, despite decreases in smoking rates in this time-period. This rise indicates that risk factors other than smoking are making a major contribution to liver cancer.

Alcohol consumption is a risk factor for liver cancer. Since people who smoke are more likely to also consume alcohol, it is difficult to separate the effects of alcohol and smoking on this risk. Some studies have concluded that alcohol and smoking are independent risk factors,¹¹⁵ but the potential for an interaction, or synergistic effect of both alcohol and smoking on the risk of liver cancer is yet to be determined.

Cirrhosis of the liver caused by hepatitis infection is a cause of liver cancer. Cigarette smoking was found to be associated with cirrhosis in people with hepatitis B.¹¹⁶ There is a dose-dependent effect of smoking on aggravation of liver lesions in people with hepatitis C, putting these people at higher risk of liver cancer.¹¹⁷ Consistent with these findings, a synergistic effect was shown between smoking and hepatitis B and C, where infected people who smoked had an increase in risk of liver disease that is greater than the addition of the individual risks of smoking and infection.¹¹⁸

That the evidence is sufficient to infer that smoking cessation reduces the risk of liver cancer was a conclusion of the 2020 US Surgeon General’s report.³⁹

3.5.7.4 Impact of smoking on prognosis

Smoking affects the prognosis for people diagnosed with liver cancer, leading to a greater chance of dying. A 2015 meta-analysis of 27 studies concluded that people with a history of smoking were more likely to die from liver cancer than never smokers.¹¹⁹ A large meta-analysis of 81 studies found that current smokers with hepatocellular carcinoma had a 1.29-fold higher odds of dying from this disease than never smokers, while former smokers had a 1.2-fold higher odds.¹¹⁴

3.5.8 Colorectal (bowel) cancer

The colon is the major part of the large intestine and its primary function is absorption of water from digested material. The rectum is the final (straight) portion of the large intestine, where faeces are stored before defecation through the anus. Colorectal (bowel) cancer is more likely to be diagnosed in people who are older, obese, smoke, have a poor diet and low rates of physical activity.

Colorectal (bowel) cancer is common in Australia. There were an estimated 15,494 people diagnosed with this cancer in 2020 and 5,322 deaths.¹²⁰ In 2020, colorectal cancer was estimated to be the fourth most common cancer diagnosis and the second most common cause of death from cancer.¹ Australians with colorectal cancer have a 70% chance of surviving for five years or longer, based on data from the years 2012–2016.¹²⁰ Over 90% of colorectal cancers are adenocarcinomas, with a number of other types being much rarer.

3.5.8.1 Risk associated with smoking

Four meta-analyses^121-124 and an analysis of a large US prospective cohort study,¹²⁵ have concluded that cigarette smoking is associated with colorectal cancer. The increase in risk is about 1.2-fold. On the basis of these and other studies, the US Surgeon General’s report from 2014 concluded that the evidence is sufficient to infer a causal relationship between smoking and colorectal cancer. This report also stated that smoking causes colorectal adenomatous polyps, a non-cancerous condition that may develop into colorectal cancer.¹¹² Subsequent genetic studies have indicated that the association of smoking and colorectal cancer is causative.¹²⁶

Australian current smokers (aged 45 and over) have a 1.31-fold increased risk of colorectal cancer compared to non-smokers.⁸

The evidence that smoking is associated with colorectal cancer has led to a suggestion that screening guidelines for bowel cancer be amended to recommend that screening for smokers start at age 45 years rather than 50 years.^{127, 128}

3.5.8.2 How tobacco smoke causes bowel cancer

The evidence strongly suggests that smoking increases the formation of polyps, the precursor of colorectal cancer, and possibly also the development of malignancy.¹¹² Many carcinogens in cigarette smoke, such as PAHs, heterocyclic aromatic amines and N-nitrosamines can reach the bowel via the bloodstream. Higher concentrations of DNA adducts to metabolites of PAHs have been found in the bowel tissue of smokers than non-smokers.¹²⁹ The identification of sub-types of colorectal cancer for which the causative link with smoking is stronger has led to suggestions that smoking’s impact on colorectal cancer is mediated through interference with normal DNA methylation pathways.¹²⁸ Animal experiments indicate that exposure to smoke increases the risk of colorectal cancer spreading to the lungs by increasing inflammation and changes to cell adhesion processes.¹³⁰

3.5.8.3 Factors affecting risk

Two meta-analyses and a large prospective cohort study have reported increases in risk with amount smoked per day, pack-years of smoking and longer duration of smoking.^{121, 123, 125} Both meta-analyses analysed risk by sub-site of cancer and found that the risk associated with smoking is higher for rectal cancer than colon cancer.^{123, 124}

Colorectal cancer is a complex collection of diseases and causation appears likely to differ between molecularly defined subsets. A study in women from Iowa in the US found only a moderately increased risk of colorectal cancer associated with smoking, but much higher risks for sub-types defined by anatomical location, phenotype and BRAF gene mutation status.^{128, 131} A study comparing the risks of colorectal cancer from smoking and from genetics predicted that smoking was an independent risk factor for this cancer.¹³²

Former smokers generally have lower risk than current smokers. A large US cohort analysis found a trend for decreased risk of colorectal cancer with longer time since cessation with no association for former smokers who had quit before age 40 years or had been non-smokers for 31 years or more.¹²⁵ A subsequent meta-analysis of 14 prospective cohort studies found that smoking cessation was associated with improved chances of survival compared to current smokers.¹³³ The 2020 US Surgeon General’s report stated that the evidence is sufficient to infer that smoking cessation reduces the risk of colorectal cancer.³⁹

3.5.8.4 Impact of smoking on prognosis

Smokers with colorectal cancer are more likely to have advanced disease than non-smokers.¹³⁴ Two meta-analyses investigated the link between smoking and death from colorectal cancer and found that mortality rates are higher in smokers.^{121, 123} A more recent meta-analysis from 2018 found that current smokers had a 1.29-fold increase in risk of death and former smokers a 1.12-fold increase compared to never smokers.¹³³ A small study of people with cancer of the anus found that recurrence rates and cancer-related deaths were higher in patients who continued to smoke.¹³⁵ 

3.5.9 Breast cancer

Breast cancer is very common in Australia and was the most common cancer diagnosis in 2020.^{1, 136} In 2020 there were an estimated 19,974 women and 167 men diagnosed with this disease. It accounts for 14% of all new cancer diagnoses.¹³⁶

Survival rates for breast cancer are relatively high, with 91% of people living for five years or longer after diagnosis. However, the high incidence means that the actual number of deaths from this disease are high. In 2020, an estimated 3,091 people died from breast cancer.¹³⁶

Known risk factors for breast cancer are age, genetic mutations, family history of breast cancer, obesity, alcohol consumption, reproductive history and some hormone treatments for menopause.¹³⁷

3.5.9.1 Risk associated with smoking

There is some evidence that smoking is associated with an increased risk of breast cancer, however whether smoking is a cause of breast cancer is currently unclear. The strength of the association between smoking and breast cancer risk is also unclear, due to inconsistencies between studies. A problem with confounding by alcohol consumption also makes interpretation difficult. Alcohol consumption is a known risk factor for breast cancer and many people who smoke also consume alcohol. Separating the risks of smoking from those of alcohol is difficult.⁴  

In 2004, the IARC performed a pooled analysis of case–control and cohort studies examining the risk of smoking in those who did not consume alcohol. After adjusting for many other factors such as age and family history of breast cancer, they concluded that there was no association between active smoking and risk of breast cancer in women who did not drink alcohol.⁴

The 2014 US Surgeon General’s report featured a meta-analysis of 46 case–control and cohort studies assessing the association of smoking and breast cancer incidence. The studies included were published between 2000 and 2011. For people who had ever smoked, there was a small 1.12-fold increased risk of breast cancer. For active smoking, their analyses supported a 10% increased risk of breast cancer. People who smoked with higher intensity or duration had a slightly higher risk.¹¹² The US Surgeon General in 2014 concluded that the evidence is sufficient to identify mechanisms by which cigarette smoking may cause breast cancer. The evidence is suggestive but not sufficient to infer a causal relationship between tobacco smoke and breast cancer.¹¹²

Some recent genetic studies have indicated that the association between smoking and breast cancer is causative,¹²⁶ and that lifetime smoking increases the risk of ER-positive, but not ER-negative, breast cancer.¹³⁸ More data is required to validate these findings.

In 2011, a Canadian expert panel reviewed the evidence and concluded that active smoking causes breast cancer and that the association between exposure to ETS and breast cancer in young women who have never smoked is also consistent with causality.¹³⁹

3.5.9.2 How tobacco smoke may cause breast cancer

There are biologically plausible mechanisms by which exposure to tobacco smoke could cause breast cancer. There are at least 20 known or suspected mammary carcinogens in tobacco smoke.¹³⁹ DNA adducts (carcinogens bonded to DNA) from tobacco carcinogens are found in the breast fluid and breast tissue of smokers. The 2014 US Surgeon General’s report concluded that available evidence supports biologically plausible mechanisms, particularly for DNA adduct formation and unrepaired DNA mutations, by which exposure to tobacco smoke could cause breast cancer. But this report also warned that the data are limited and a detailed mechanistic model cannot yet be assembled.¹¹²

3.5.9.3 Factors affecting risk

The intensity and duration of smoking affects then strength of association with breast cancer. Those who smoked for 20 or more years, 20 or more cigarettes per day, or 20 or more pack-years of smoking has been shown to have significantly increased risk by 13-16%, depending on the study.¹¹²

There is emerging evidence that premenopausal women may be at greater risk for breast cancer from smoking than postmenopausal women.¹¹²

3.5.9.4 Impact of smoking on prognosis

The 2014 US Surgeon General has highlighted the difficulty in inferring a causal association between smoking and breast cancer mortality; there are many confounding variables relating to treatment and other non-cancer, smoking-related comorbidities that can contribute to mortality. There is currently insufficient evidence to conclude that either active or passive smoking influences breast cancer mortality.¹¹²

Since this report, a large meta-analysis, including over 400,000 women with breast cancer, found a 28% increase in mortality from breast cancer in current smokers compared to never smokers. The mortality rates for former smokers was not significantly different to never smokers.¹⁴⁰ 

3.5.10 Other cancers

3.5.10.1 Hodgkin lymphoma

Studies of Hodgkin lymphoma and smoking reviewed by the IARC reported weak or no association.⁴ A number of studies indicate a slightly higher risk for smokers, but few have adjusted for potential confounding factors, that are likely to be alternative explanations for the apparent higher risks for smokers. One study published in 2009 found that people who smoked for 25 years or more were at increased risk of Hodgkin lymphoma, after adjusting for some potential confounding factors.¹⁴¹

3.5.10.2 Prostate cancer

Both the IARC in 2004 and the US Surgeon General in 2014 concluded that the evidence did not support a causal relationship between prostate cancer and smoking.^{4, 112} A meta-analysis of cohort studies found no increase in risk of prostate cancer overall in smokers, but did find a slightly increased risk associated with higher daily consumption of cigarettes or greater pack-years of smoking.¹⁴² A large prospective US cohort study, involving over a quarter of a million men, found that current and former smokers may be at decreased risk of being diagnosed with prostate cancer.¹⁴³ A review of the epidemiologic evidence found that smokers are not at appreciably higher risk of developing prostate cancer.¹⁴⁴

The US Surgeon General concluded that the evidence suggests that smokers have a higher mortality rate from prostate cancer than non-smokers, as well as a higher risk of advanced-stage disease, less well-differentiated cancer, and a higher risk of disease progression.¹¹² The elevated risk of death from prostate cancer noted by the US Surgeon General has also been reported in the meta-analysis¹⁴² and other studies.¹⁴³

Prostate cancer has a high survival rate as more than half of cases are localised and treatable with surgery or radiotherapy. However, some forms of the disease are more aggressive, with a high likelihood of progression to a fatal metastatic stage. There is some evidence that smoking increases the risks of a more aggressive form of prostate cancer. A 2009 review concluded that cigarette smoking is likely to be a risk factor for prostate cancer progression. Smokers had more advanced disease at diagnosis, a worse prognosis and a greater risk of fatal prostate cancer.¹⁴⁴ A later meta-analysis of smoking status in men with localised prostate cancer found that current smokers were at higher risk of their cancer returning after initial treatment, spreading by metastasis and mortality from prostate cancer.¹⁴⁵ A study from Japan indicated that current smokers were more likely than non-smokers to have their prostate cancer reoccur after surgery plus radiotherapy treatment.¹⁴⁶

3.5.10.3 Ovarian cancer

A 2012 report from the IARC reviewed 24 case—control and 6 cohort studies examining the risk of ovarian cancer in people exposed to tobacco.¹⁴⁷ Most of these studies showed no association between smoking and ovarian cancer, overall. However, 10 studies indicated a 2.4-fold increased risk for smokers of a rare subtype called mucinous ovarian cancer. This association was consistently detected over the studies that tested for it, and the risk increased with greater exposure to tobacco. The IARC has therefore concluded that smoking is likely to be a cause of mucinous ovarian cancer.¹⁴⁷

3.5.10.4 Cancer of unknown primary origin

Cancer of unknown primary (CUP) is a cancer that has spread by metastasis, but the site of tumour origin remains unidentified. Different types of cancer are including under the definition of CUP. A case—control study that adjusted for alcohol intake and other potential confounding factors showed that current smokers had a 1.6-fold increased risk of getting CUP compared to never-smokers.¹⁴⁸ Australian data show that people who smoke (aged 45 and over) have a 3.3-fold higher risk of CUP compared to non-smokers.⁸ Given the lack of information on the nature and causes of CUP, it is currently unknown whether smoking is a cause of some forms of CUP.

 

Relevant news and research

For recent news items and research on this topic, click  here. ( Last updated May 2021)