Tobacco, Cigarettes, and the Liver: The Smoking Gun

Abstract

The association between alcohol and liver disease, including alcoholic hepatitis, cirrhosis, acute-on-chronic liver failure, and hepatocellular carcinoma, has been well described, but the same cannot be said for the association between smoking, water pipe or tobacco chewing. A review of cumulative evidence suggests that smoking is independently a risk factor for liver fibrosis and contributes to carcinogenesis in HCC. Smoking-related fibrosis has been reported in patients with nonalcoholic fatty liver disease, primary biliary cholangitis, alcoholic liver disease and chronic viral hepatitis. Heavy smoking leads to systemic inflammation, oxidative stress, insulin resistance, and results in tissue hypoxia, as well as free radical damage. Other than damaging the liver, patients also suffer from the systemic effects of the 4000 chemicals associated with tobacco, which include nitrosamines, aromatic hydrocarbons, nicotine, nornicotine, and other alkaloids. These include respiratory ailments, cancer of the lungs, oral cavity, esophagus, pancreas and colon, atherosclerotic vascular disease, and stroke.

1.

About a third of males worldwide smoke tobacco, as per the estimate of the World Health Organization (WHO). Tobacco smoking was the attributable cause of death in 100 million individuals globally over the course of the 20th century.^1,2 Thus, tobacco consumption is a leading cause of preventable morbidity and mortality. The available forms of tobacco in India include cigarettes, cigars, indigenous bidi, water pipes, and smokeless tobacco (ST) mixed with additional components like betel quid and gutkha. The relationship of smoking with lung cancer is well described, but the effects of smoking on the liver are less well known. Smoking and tobacco consumption can increase the fibrogenic and carcinogenic effects of alcohol and exacerbate disease progression in nonalcoholic fatty liver disease (NAFLD), primary biliary cholangitis (PBC), and chronic liver disease (CLD) due to hepatitis B virus (HBV) and hepatitis C virus (HCV) infection.^3,4 The risk of death increases in proportion to the extent of exposure to cigarette smoke. The latter can be estimated by the age of the first initiation of the tobacco habit, number of cigarettes smoked per day, duration of smoking, and depth of inhalation.⁵ Another matter of concern is second-hand smoking and environmental cigarette smoke exposure to nonsmokers, which has shown an increased risk of cardiovascular disease and asthma. The effects of second-hand smoke on liver disease progression are still unclear, though data suggests that such individuals are at an increased risk.⁶

1.1. Smokeless Tobacco

In India, ST like gutkha is frequently consumed by men and women alike, which is under-reported in epidemiological data on liver disease. India has 275 million tobacco consumers, which amounts to about 35% of adults as per the Global Adult Tobacco Survey.^7,8 This figure includes 164 million ST consumers, 69 million smokers, and 42 million consumers who partake more than one form of tobacco. All the betel quid, paan or gutkha mixtures are classified as group 1 carcinogens by the International Agency for Research on Cancer, as they contain areca nut, betel leaf, etc., and contain a high amount of nicotine.⁹ Nearly 4200 chemical constituents are found in ST. The areca nut industry has created two forms of mixtures; the first is betel quid/areca nut aromatic mixtures called “paan masala,” and the second is called gutkha if tobacco was added. This mixture contains sundried, roasted, chopped tobacco, areca nut, slaked lime, and catechu.¹⁰ Salivary, blood, and urine levels of derivatives of ST products are very much like that of smoking and constitute an equal and underestimated risk to the progression of liver disease and cancer in India. Additional data regarding additive effects of polyphenols, tannins, alkaloids, nicotine, and musk ketones extracted from these products is required due to the wide consumption of ST in India.¹¹

1.2. Pathophysiology of Smoking and Liver Disease

Prior data has shown that consuming tobacco increases the production of interleukins (IL-1, IL-6), tumor necrosis factor (TNF α), angiogenic agents like vascular endothelial growth factor (VEGF), fibrosis mediators (transforming growth factor β), and angiotensin II. Smoking also induces the nicotinamide adenine dinucleotide phosphate (NADPH) oxidase system, impairs antioxidant pathways like glutathione, and leads to free radical lipid peroxidation. This leads further to hepatocellular inflammation, cell damage, hepatic stellate cell activation, mesangial cell activation, leading to the proliferation of fibrotic mediators, matrix metalloproteinases, and extracellular matrix proteins. Another mechanism of fibrosis is iron deposition. Heavy smokers with chronic hepatitis C have polycythemia and iron deposition.¹²

Smoking leads to vascular vasoconstriction, endothelial dysfunction, smooth muscle proliferation, reduced nitric oxide and tissue hypoxia, and hepatocellular injury. Figure 1 shows the mechanism of liver injury and inflammatory pathways linking smoking and organ damage. Smoking has pleiotropic effects through its constituent chemicals that act on the lung, the cardiovascular system through systemic effects of inflammatory pathways like mitogen-activated kinases (MAPK). Nuclear factor-kappa B and signal transducer and activator of transcription (STAT), gene induction, and epigenetic modifiers. The systemic inflammation, increased lipid peroxidation, free radical injury, and endothelial dysfunction contribute to tissue hypoxia, iron accumulation, and liver cell injury that progresses to liver fibrosis over time. Impaired wound healing leads to impaired hepatic regeneration exacerbating fibrosis and predisposing to replication errors and de novo mutations, possibly a link to carcinogenesis. The association of smoking, alcohol, and risk of cirrhosis suggests that environmental influences need to be estimated in studies on disease pathogenesis.¹³

Figure 1.

Mechanism of liver injury and inflammatory pathways in smoking. Smoking has pleiotropic effects through its constituent chemicals that act on the lung, the cardiovascular system through systemic effects of inflammatory pathways. Abbreviations: AP-1, activatory protein-1; MAPK, Mitogen-activated protein kinases; NO, nitric oxide; NADP, nicotinamide adenine dinucleotide phosphate; NF-κB, nuclear factor kappa-B; STAT, signal transducer activator of transcription; VEGF, vascular endothelial growth factor.

A common abnormality observed in persons who consume tobacco is liver injury, manifested as abnormal liver function tests. It is associated with an elevation of the liver enzyme levels like alkaline phosphatase (ALP) and gamma glutamyl transpeptidase (GGT). Jang et al showed that in a multivariate hazards model adjusted for age, gender, BMI, alcohol, and coffee drinking, current smokers had significantly lower serum levels of total protein and albumin and higher GGT levels compared with never or past smokers. Daily and lifetime smoking affected the serum levels of total protein, albumin, and GGT but did not affect aspartate transaminase (AST)

or alanine transaminase (ALT). The effect of smoking on ALP is confounded by the association of osteoporosis and bone turnover origin of this enzyme, especially in older individuals. The consumption of coffee has protective effects on transaminases, while alcohol worsens the GGT level.¹⁴ Thus, alcohol, smoking, lifestyle, and coffee and tea intake are all factors, which modify a genetic predisposition to liver disease. Smoking and ST consumption is undoubtedly an underrecognized pathophysiological link in CLD progression.^15,16

1.3. Immunological Effects of Smoking

Smoking adversely impacts innate and adaptive immunity and can aggravate pathological immune responses while reducing defensive immune mechanisms. Smoking affects cells involved in the innate immune response, including antigen-presenting cells (APCs) like dendritic cells, macrophages, and natural killer cells.^12,13 The maturation of dendritic cells and NK cells of the innate response are also impaired.¹⁷ Smoking also affects the adaptive immune response mediators, including T helper cells (Th1/Th2/Th17), CD8+ T cells, CD4+ T regulatory (Treg) cells, and memory T/B lymphocytes. The molecular pathways affected include nuclear factor kappa B (NF κB), MAP kinases, and epigenetic responses like DNA methylation, histone modification, etc. (Figure 1). Data suggests that smoking can contribute to autoimmune diseases like rheumatoid arthritis, lupus, Crohn’s disease, etc. Broadly, smoking increases the CD8+/CD4+ ratio and enhances Th17 and CD4⁺FoxP3⁺ Treg cells effects. Smoking reduces immunosuppressive Tregs in autoimmune disease and increases Treg effects in lung disease, leading to conflicting regulatory effects and predisposing to infections. Smoking impairs the function of memory T and B cells.¹⁷

1.4. Smoking and NAFLD

Smoking results in systemic inflammation and tissue hypoxemia, oxidative stress, and insulin resistance. This is further exacerbated in the presence of metabolic syndrome and increases the risk of diabetes, cardiovascular, and cerebrovascular events. The mechanisms and cellular effects of smoking and NAFLD are shown in Figure 2. An interesting study demonstrating the effects of smoking in an obese rat model of NAFLD was performed by Azzalini et al, who exposed obese Zucker rats to cigarette smoke and observed activation of pathways, inflammatory mediators, immunohistochemistry, and changes in liver histology. Obese rats exhibited features of the metabolic syndrome, including dyslipidemia, insulin resistance, and histological features of NAFLD. Smoking increased the expression of tissue inhibitor of metalloproteinase-1 and procollagen-alpha2 (I) peptide in obese Zucker rats who were exposed to 2 cigarettes/day, 5 days per week, for 4 weeks, but not in the control rats who were not exposed to cigarette smoke. This was mediated by increased gene expression of AKT and extracellular signal-regulated kinase (ERK). This study showed that 4 weeks of exposure to cigarette smoke exacerbated NAFLD, but the duration was not long enough to establish a link with fibrosis or change in body weight or insulin resistance.¹⁸

Figure 2.

Mechanisms and cellular effects of smoking and NAFLD. Abbreviations: BMI, body mass index; NAFLD, nonalcoholic steatohepatitis.

Prior systematic reviews have shown that NAFLD patients have a dose-response relationship with smoking, as those with advanced fibrosis reported higher pack-years of cigarette consumption as compared with early or no fibrosis.¹⁹ In a large study, 45,409 of 199,468 Korean adults developed NAFLD during 1,070,991 person-years of follow-up. In men, the multivariate-adjusted Hazard Ratio (HR) for incident NAFLD was between 1.25 [95% confidence interval (CI) 1.21–1.29] and 1.36 (955 CI 1.30–1.42), respectively for low dose and high dose exposures as compared to nonsmokers. The risk in women was higher with adjusted HR of 1.25 (95% CI 1.04–1.50), and 1.46 (95% CI 1.17–1.81), respectively for low and high smoking exposures as compared with nonsmokers.²⁰

The largest data is from the NASH Clinical Research Network database, which enrolled 1091 patients between 2004 and 2008 in patients with biopsy-proven NASH. On bivariate analysis, associations between advanced fibrosis, age, gender, ethnicities, smoking exposure, and diabetes were noted. Smoking ≥10 pack-years was associated with advanced fibrosis (OR 1.63, P < 0.0001) on multivariate analysis. In patients with diabetes, advanced fibrosis was likely regardless of smoking, while in nondiabetics, smoking ≥10 pack-years was associated with high frequency of advanced fibrosis. Smoking is associated with increased insulin resistance and the risk of type 2 diabetes.²¹

The Thai National Health Examination assessed the follow-up of 7529 persons with NAFLD and reported 928 deaths during 64,116 person-years of follow-up. After adjusting for confounders, current smoking (adjusted HR 13.8, 95% CI –1.66 to 1.45) and smoking ≥ 10 pack years (adjusted HR 310, 95% CI 78-1296) were independently associated with death.²² Munsterman et al reported the association of biopsy-proven NAFLD in 56 patients. The NAFLD activity score (NAS) ≥ 5 was not dependent on any prior or current smoking history or dose of smoking, but a significant correlation was noted with liver fibrosis in a dose-dependent manner.²³ Several studies have corroborated an association between smoking and advanced liver fibrosis (Table 1).^{24, 25, 26, 27, 28, 29, 30, 31, 32, 33} However, Yilmaz et al refuted the association, but this study had smaller numbers and comorbidities.³⁴

Table 1.

Studies Showing the Smoking Dose Response Risk of Advanced Fibrosis in Nonalcoholic Fatty Liver Disease [24, 25, 26, 27, 28, 29, 30, 31, 32, 33].

Author, Year	Country	Smoking Dose	N	Study Design	NAFLD N (%)	Non-NAFLD N (%)	OR (95% Confidence interval)
Hamabe et al, 201124	Japan		1553	Retrospective cohort
		All smokers			93 (5.99)	216 (13.90)	2.9 (2.0–3.6)
		Light			24 (1.55)	57 (3.67)	0.9 (0.6–1.5)
		Heavy			69 (4.44)	159 (10.24)	2.7 (2–3.7)
		Nonsmoker			172 (11.1)	1072 (69)
Liu et al, 201325	China		2426	Cross sectional
		Total smokers			420 (17.3)	962 (39.6)	1.05 (0.8–1.2)
		Light			162 (6.7)	421 (17.4)	0.9 (1.3–1.4)
		Heavy			258 (10.6)	541 (22.3)	1.1
		Current			420 (17.3)	962 (39.6)	1.1
		Former			106 (4.4)	204 (8.4)	1.3
		Passive			5	70	1.4
		Nonpassive smokers			4	225
		Nonsmoker			294 (12.1)	705 (29.06)
Chavez-Tapia et al, 200626	Mexico		885	Cross sectional
		Total smokers			87 (9.8)	232 (26.2)	0.9 (0.6–1.2)
		Nonsmoker			168 (18.9)	398 (44.9)
Caballeria et al, 201027	Spain		766	Cross sectional
		Total smokers			92 (12.0)	250 (32.6)	0.1 (0.8–1.5)
		Current smokers			39 (5.0)	150 (19.5)	0.7 (0.4–0.9)
		Former smokers			53 (6.9)	100 (13.0)	1.6 (0.8–1.8)
		Nonsmokers			106 (13.8)	318 (41.5)
Oniki et al, 201328	Japan		696	Cross sectional
		Total smoker			61 (8.7)	221 (31.7)	1.4 (0.9–2.0)
		Current smokers			21 (3.0)	57 (8.1)	1.8 (0.8–1.2)
		Former smokers			40 (5.7)	164 (23.5)	1.2 (0.7–1.9)
		Nonsmokers			69 (9.9)	345 (49.5)
Zhang et al, 201429	China		1800 of 2183 with Metabolic Syndrome	Prospective cohort 2005–2011
		Total smokers			129 (52.4)	124 (36.4)	1.14 (1.04, 1.26)
		Nonsmokers			177 (49.5)	217 (63.5)
Koehler et al, 201230	Netherlands		2811
		Total smokers			660 (23.48)	1123 (40.00)	1.3 (1.08–1.5)
		Current smokers			75 (2.67)	165 (2.31)
		Former smokers			585 (20.81)	958 (34.08)
		Nonsmokers			326 (11.60)	702 (24.97)
Chang et al, 201331	Korea		43,166
		Smokers			5133 (44)	8068 (25.6)	2.29 (2.2–2.4)
		Nonsmokers			6519 (56)	23,446 (74.4)
Singh et al, 201532	India		645	Case Control
		Smokers			49 (10.6)	12 (6.6)	1.67 (0.86–3.2)
		Nonsmokers			415 (89.4)	169 (93.4)
Otgonsuren et al, 201333	United States		10,565
		Total smokers			2241 (89.3)	7137 (88.6)	1.07 (–.93 to 1.24)
		Heavy/moderate smokers			733 (29.2)	2658 (33)
		Light smokers			1508 (60.1)	4479 (55.6)
		Never smokers			269 (10.7)	918 (11.4)

Abbreviations: OR, odds ratio; RR, relative risk.

1.5. Smoking and Alcoholic Liver Disease

A large database from the general Danish population enrolled in the Copenhagen City Heart Centre, including 9889 women and 8590 men, were assessed for smoking and covariables at four time points over 2 decades. The computed adjusted HR for developing alcoholic cirrhosis were (HR 1.6 CI 1.6–9.4) for women and 1.6 (95% CI 0.9–3) for men smoking> 10 g of tobacco daily. Similar rates were also noted for all-cause cirrhosis with HR 2.2 95% CI 1.4–3.4 for women and HR 1.4 (95% CI –0.9 to 2.2) for men. Adjustments were made for age, alcohol use, body mass index, and educational status.³⁵

Liu B et al also demonstrated the association of smoking, cirrhosis, and gall bladder disease in 1,290,413 women, with RR of 3.76 (95% CI 3.25–4.34) for current smokers as compared with persons who never smoked.²⁵ Klatsky and Armstrong showed a tripling risk of developing alcoholic cirrhosis when smoking a pack or more of cigarettes per day as compared with nonsmokers in 128,934 individuals.³⁶ In addition to alcohol consumption, smoking enhances the systemic inflammation in chronic HCV infection and increases fibrosis progression.

1.6. Smoking and Chronic Viral Hepatitis

In a French study of 310 persons living with chronic HCV, a history of >15 pack-years had an odds ratio (OR) of 1.9 (95% CI 1.1–3.6) as a higher Knodell fibrosis scores on adjusted analysis as compared with nonsmokers.³⁷ Yu M et al reported relative risk of cirrhosis in hepatitis B surface antigen-positive persons were 1.62 (95% CI 0.94–2.77) and 2.13 (95% CI 1.21–3.74), respectively, in those who smoked <20 or ≥ 20 cigarettes per day, confirming a dose-response risk.³⁸ A North American Hepatitis B research network assessed the risk of smoking in 1330 participants (71% Asian). An association was determined between smoking, alanine transaminase (ALT) levels, and FIB-4 levels, but not on multivariate analysis.³⁹ Tsochatzis et evaluated the risk of heavy smokers (≥20 pack-years), current and past smokers and found an OR for severe fibrosis (staging score ≥ 4) of 3.9 (95% CI 1.36–11.35) and OR of steatosis of 2.86 (95% CI 1.07–7.68) for heavy smokers on multivariate analysis with chronic HCV infection.⁴⁰ Thus, smoking and viral hepatitis appear to act in an additive if not synergistic fashion. Some studies failed to find an association between smoking and viral hepatitis. Hezode et al did not find any significant relationship between fibrosis score and smoking in 244 HCV-mono-infected patients.⁴¹ Again, Dev et al did not find any association between the METAVIR score and smoking history in persons with chronic HCV infection.⁴²

1.7. Smoking and Primary Biliary Cholangitis

Corpechot et al reported that 58 (26%) persons were smokers prior to the diagnosis of PBC, 11% of whom were active smokers. Males (OR 4.5), alcohol intake (OR 4.2), and F3/F4 fibrosis (OR 2.7) were associated with smoking. Grade of inflammation, ductular changes, and biochemical tests were not linked to smoking. The dose-dependent association was tested, which showed that those with F3/F4 fibrosis had had 8.1 ± 14.2 pack-years vs. 3.0 ± 7.0 pack-years in those with F1/F2 fibrosis. On logistic regression, smoking history and increased exposure were predictive of advanced fibrosis in PBC. One of the confounders in this study was the association with heavy drinking in 14 (6%) of individuals, but the association of fibrosis with dose-dependent smoking remained after adjusted analysis.⁴³ Smoking can accelerate the disease progression in primary biliary cholangitis by several mechanisms. First, smoking reinforces autoimmunity by sensitizing cholangiocytes, hepatocytes, and stellate cells to apoptosis by free radical injury. In PBC, the release of metalloproteinases, activation of the Fas pathway, and activation of liver sinusoidal endothelial cells, and production of inflammatory cytokines and neo angiogenic factors like VEGF also contribute to fibrogenesis. Smoking also causes the production of mediators like 2-hydroxy estrogen that inhibit the regeneration of cholangiocytes.^44,45 One more reason to investigate the role of smoking in liver disease is to estimate the therapeutic response to specific therapies like ursodeoxycholic acid. For example, smoking interferes with the therapeutic response to hydroxychloroquine in lupus and antithyroid drugs in Graves’s disease, other autoimmune conditions. The association between PBC, smoking, and advanced fibrosis was reported by epidemiological studies in Britain and the United States. [3].

1.8. Oncogenic Effects of Smoking and Liver Cancer

Smoking is associated with de novo malignancies, including oropharyngeal, esophageal laryngeal, and lung cancer.⁴⁶ It also increases the risk of all biliary tract cancers, i.e., intrahepatic and extrahepatic biliary duct cancers, periampullary cancers, except gall bladder cancer, where the risk is modestly increased. In the latter, additional confounders such as geographic site (Asians and North Americans have increased risk as compared with Europeans), association of anomalous pancreatobiliary malunion, and higher prevalence of gallstone disease, choledochal cyst, and gall bladder cancer in Asians.⁴⁷

The association of smoking and hepatic and extrahepatic cancers remains relevant even in the post-transplant period as de novo malignancies are reported in 5–15% of liver transplant recipients, a higher rate than the general population. [4] The European Prospective Investigation into Cancer and Nutrition (EPIC) cohort, with data of more than 4.4 million person-years of follow-up, found that smoking conferred a RR of 4.55 for HCC, more than HBV and HCV infections. A meta-analysis showed a RR of 1.5 for HCC as compared with nonsmokers and an RR of 1.12 for previous smokers. The high variation of RR reported in various studies (1.49–1.96) can be explained by differences in study design and smoking exposure.⁴⁶ In a large meta-analysis from Japan, 9 of 12 cohorts and 5 of 11 case-control studies described an association between smoking and HCC, and a dose-response relationship was demonstrated in four of these studies.⁴⁸ The mechanism of carcinogenesis has been well described as many of the constituents of tobacco are carcinogens. Smoking can cause the formation of aflatoxin B1eDNA adducts and injury-mediated changes in gene expression in a genetically predisposed individual.⁴⁹ Specifically, tobacco-specific nitrosamines in various products have carcinogenic elements like N′-nitrosonornicotine, N′-nitrosoanatabine, N′-nitrosoanabasine, and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone. These nitrosamines act by inducing mutations in genes by forming DNA adducts, and tumor progression is caused by the binding of nicotinic acetylcholine receptors, expressions of growth factors like VEGF, and systemic inflammation favoring cell proliferation, and metastases.⁵⁰

Table 2 summarizes the dose and response characteristics of smoking and HCC in various epidemiological studies and population case-control studies.^{36,51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61}

Table 2.

Association of Smoking and Liver Cancer.

Sr. No.	Author, year	Country	Study design	Number of Cases	Underlying Etiology of liver disease	Odds Ratio (OR)/Relative risk (RR)/Hazard ratio (HR) with 95% Confidence interval (Cl)
1	Stewart SL et al., 201951	Asian Americans (4 countries)	Multifactor estimation cohort.	1004 persons of Chinese, Hmong, Korean and Vietnamese descent.	Liver disease classed as •Viral (chronic HBV or HCV), •Lifestyle (current smoker or alcohol user, no viral) •Metabolic (≥2 metabolic factors, no lifestyle or viral), •Low risk (≤1 metabolic, no lifestyle or viral)	RR for HCC not available. Ethnic risk factors identified, including smoking. Vietnamese men (16.3%, 95% CI 7.4%–25.3%) and Hmong women (15.1%, 95% CI 7.8%–22.5%) had the highest viral pattern prevalence. Current smokers at increased risk of progression and HCC.
2	Wang YH et al., 201952	China	HBV disease	209 HCC cases in 4841 male HBV carriers	HBV	High HBV DNA viral load (>4.39 log copies/mL), increased ALT ≥80 U/L raised the risk of HCC incidence. Risk of HCC vs. nonsmokers •10-19 pack years = OR 3.14; (95% CI 1.03–9.64) •≥20 pack-years = OR 6.06; (95% CI 1.10–33.25), respectively, versus nonsmokers
3	Kai K et al., 201853	Japan	Retrospective cohort	341 cases of HCC •HCV-HCC (n = 273) •HBV-HCC (n = 68)	Chronic viral hepatitis (HCV, HBV)	Current smokers had lower Overall Survival (P = 0.0039) and Disease-free survival (P = 0.0416)
4	Kolly P et al., 201754	Switzerland	Prospective cohort	238 eligible patients with HCC; 64 were smokers and 174 were nonsmokers	Chronic viral hepatitis (HCV, HBV)	HR 2.99 (95% CI 1.7–5.23, P < 0.001) Smokers had a worse overall survival than nonsmokers (HR 1.77, 95% CI 1.22–2.58, P = 0.003)
5	Abdel Rahman O et al., 201755	Multicentric	Cohort, Case-control	Variable through 81 sites	Various	Overall OR 1.90 (95% CI 1.68 to 2.14) •Pooled OR for HCC in current smokers was 1.55 (95% CI: 1.46 to 1.65; P < 0.00001). •Pooled OR for HCC with former smokers was 1.39 (95% CI: 1.26 to 1.52; P < 0.00001) •Pooled OR for HCC with heavy smokers was 1.90 (95% CI: 1.68 to 2.14; P < 0.00001).
6	Chiang CH et al.,201656	China	Prospective cohort	51,164 (253 deaths from HCC) without chronic hepatitis B or C, i.e., MAFLD	Diabetes/Metabolic syndrome	•Never smokers (HR, 0.46; 95% CI, 0.32–0.65) had lowest risk of HCC. •Quitters (HR, 0.62; 95% CI, 0.39–0.97) had a lower adjusted risk of HCC deaths compared with current smokers
7	Raffetti E et al., 201557	Italy	Retrospective cohort	552 cases of HCC	Etiology, diabetes, hypertension, heavy alcohol intake	•Alcohol increased risk by HR 1.32 (0.99–1.75) •Diabetes increased risk by HR 1.25 (1.02–1.54). •Alcohol and the presence of diabetes were positively associated with patient mortality and with HCC development. •Tobacco smoking and hypertension were not associated with higher risk.
8	Hassan MM et al., 200858	United States of America	Case-control	319	HCV	•Regular cigarette smoking increased risk of HCC by adjusted OR, 1.9 (95% CI, 1.1–3.1) in males. •Heavy alcohol intake increased HCC risk in women by adjusted OR, 7.7 (95% CI, 2.3–25.1)
9	Hara M et al., 200859	Japan	Case- control	209 HCC vs. 381 patients with CLD without HCC	Various causes	Risk of HCC noted by: •For Former smokers OR 1.0 (95% CI 0.6–1.7) •For Current Smokers OR 2.5 (1.4–4.6) against CLD patients who never smoked.
10	Zhu K et al., 200760	United States of America	Case-control	168	HCV	Relative to nonsmokers, the risks of HCC were: •Former smoker: OR 1.85 (95% CI1.05-3.25) •Current smoker: OR1.49 (95% CI, 0.83–2.68)
11	Chen Z et al., 200361	China	Case-control	200 HCC cases detected in 24 Chinese cities and 74 rural counties. Smoking habits compared between 36,000 adults who succumbed to HCC (Cases) and 17,000 who died of cirrhosis (Controls).	HBV, ALD. tobacco is currently responsible for about 50,000 liver cancer deaths each year in China, chiefly among men with chronic HBV infection	•Among adult men who smoked, there was a 36% excess risk of death from liver cancer. •Smoker vs. nonsmoker RR = 1.36, with 95% CI 1.29–1.43, P < 0.00001

Abbreviations: ALD, alcoholic liver disease; CI, confidence interval; CLD, chronic liver disease; HR, hazard ratio; HBV, hepatitis B virus; HCV, hepatitis C virus; HCC, hepatocellular carcinoma; OR, odds ratio; RR, relative risk.

1.9. Tobacco Smoke Exposure- dose and Characteristics

Table 2 shows the association between exposure rates of smoking and oncogenic potential. The tobacco habit data and smoking behaviour are reported using patient self-reports in most epidemiological studies on smoking and liver disease. Addiction is due to the nicotine component. Age of initiation of smoking, ever smoking, age of discontinuation, number of packs/cigarettes smoked per day, current smoking status, attempts at quitting, recurrence of habit if they ever quit, and use of ST must all be considered. Cumulative lifetime exposure is measured as pack years computed by the number of packs of cigarettes per day multiplied by the number of years smoked. In ex-smokers, the duration of smoking cessation must also be noted. The confounding effect of concomitant alcohol and tobacco use has been cited by early data. However, most epidemiological studies done in the last 5–10 years and reported from Asian centers in Japan, China, and South Korea, suggest that a dose exposure risk is noted with smoking and liver disease even when adjusted for age, gender, alcohol use, and diabetes.^{12,13,16,21,23,61}

1.10. Smoking and Liver Transplantation

In a study on post-transplant candidates, 60% of persons reported a lifetime history of smoking, with 155 continuing the habit in the post-transplant period. In persons who quit smoking before the surgery, 15% relapsed after the surgical period was over. Interestingly 54% of individuals reported using both tobacco and alcohol.^62,63 There is limited information on the association of smoking and OLT outcomes due to precepts regarding any linkage. Heavy smokers also tend to be heavy drinkers, and 90% of persons with ALD also smoke. ALD recipients of orthotopic liver transplantation resume smoking early, increase their consumption over time, and become dependent on tobacco again, resulting in the increased rate of de novo malignancies after transplantation, especially oropharyngeal or lung cancer.⁶⁴

1.11. COVID-19, Smoking and Liver Disease

An asymptomatic elevation of aminotransferases has been commonly reported in patients with and without liver disease who develop coronavirus disease 2019 (COVID-19) because of hypoxic injury, systemic inflammation, and cytokine storm, drug-induced liver injury, and rarely direct cytopathic injury. Largely, those affected with more severe disease and secondary organ failures are affected severely with higher aminotransferase levels, low serum albumin, and thrombocytopenia.⁶⁵ Herein lies the pathophysiological effect of smoking affecting the lung and the liver resulting in a low systemic inflammatory state. Despite the main organ affected being the lung, which is the portal of entry, acute-on-chronic liver failure, acute decompensation, and precipitation of acute liver injury have been reported. The progression of both metabolic dysfunctions-associated liver disease and alcohol-associated liver disease is accelerated, with the comorbid association with cigarette smoking.⁶⁶

1.12. Economic Cost and Public Health Interventions Related to Tobacco Consumption

The tobacco industry in India and the ancillary lobby wields significant power as it contributes to both agricultural, small- and large-scale industrial income and provides employment to Indian labor. However. the use of tobacco as cigarettes, bidis and smokeless tobacco entails significantly high cost to the individual in view of direct adverse effects on oral health, cardiovascular disease, NAFLD and risk of cancers. In turn these health hazards progress over time to require advanced care, surgery hospitalizations, and invasive investigations, which add on the cumulative economic burden on India’s healthcare. Hence the public health burden of disease in terms of direct costs and loss of quality-adjusted life-years (QALYs), social and economic productivity must be considered. The total economic cost due to all diseases and deaths in India in 2017, attributable to bidi smoking alone, for adults aged between 30 and 69 years was INR 805.5 billion, 20.9% and 79.1% direct and indirect costs, respectively.⁶⁷

Several predictive models have shown that cessation of smoking or prohibition of tobacco consumption as a means of cancer prevention, and reduced morbidity, and mortality, can be cost-effective and even cost-saving in a country like India.

Various interventions have been utilized globally, including higher taxation on tobacco products, prohibition of use in public spaces, heavy fines for littering or spitting chewing tobacco, tobacco-free zones in hospitals, public transport, educational institutes, and shared spaces, etc. The state must run effective information, education, and communication campaigns in schools.⁶⁸ An ongoing campaign restricting surrogate advertisements for tobacco and alcohol, cinema and television infomercials, packaging with medical images of cancers, etc., have also been introduced in India. Most importantly, a long-term approach will be a practical shift of tobacco cultivation to other sustainable and profitable forms of agriculture together with reduced public demand for all forms of tobacco.⁶⁹

The association between smoking and liver disease poses several clinical implications regarding disease progression, risk of hepatic and extrahepatic malignancies. In India, the risk factor of smoking and the use of smokeless tobacco is rarely discussed in epidemiological data pertaining to liver disease. Clinicians should understand that assessment of tobacco use as firsthand or passive smoking and smokeless tobacco is essential in liver disease due to the contribution to hepatic and systemic inflammation and progression of fibrosis. In a country like India, where a third of the population has NAFLD, the risk of progression, and malignancy entail significantly high cost for the individual and as a public health burden. Cessation of the use of all forms of tobacco should be emphasized in patients with CLD of any etiology and even in post-transplant individuals due to extrahepatic worsening of cardiovascular and cerebrovascular disease and malignancies.

Credit authorship contribution statement

MP and ACA were both involved in the manuscript preparation. Both the authors have read and approved the manuscript.

Conflicts of interest

The authors have none to declare.

Funding

None.

Disclosures

Neither of the authors has potential conflicts (financial, professional, or personal), which are relevant to this manuscript.