Smoking and Lung Cancer

Slide 1 · Smoking and Lung Cancer

Carcinogenic exposures • molecular injury • tumor evolution • clinicopathologic correlation

Pathology teaching lecture

Dr Abu Suraih Sakhri

MBBS, MD (Pathology)

www.abusuraihsakhri.com

Slide 2 · Learning objectives

Link tobacco exposure to lung carcinogenesis through metabolic activation, DNA adducts and acquired mutations.

Explain exposure–response relationships, susceptibility, co-exposures and the residual risk after cessation.

Compare tobacco-associated lung-carcinoma phenotypes and their major precursor lesions.

Apply pathological reasoning to clinical scenarios without confusing exposure with a tumor diagnosis.

Slide 3 · Two complementary Robbins perspectives

CONTEXT A

Tobacco as an environmental carcinogen

Smoke chemistry, exposure intensity and DNA injury

Exposure interactions with occupational carcinogens

CONTEXT B

Lung carcinoma as the clinical outcome

Acquired driver alterations and precursor lesions

Histologic categories, morphology and consequences

Slide 4 · A mechanistic argument for causation

Inhaled tobacco smoke

Carcinogen exposure of respiratory mucosa

DNA adducts and acquired mutations

Clonal expansion and lung carcinoma

Causal interpretation

Follow the direction of the arrows; Each transition represents a plausible biological step, not a deterministic event.

Schematic — The sequence is a conceptual path; Cancer is not inevitable after any particular exposure.

Slide 5 · Tobacco smoke contains many distinct hazards

Approximately 7,000 constituents in tobacco smoke.

More than 60 constituents are identified as carcinogens.

Different constituents produce DNA injury, mucosal irritation, ciliary toxicity or impaired oxygen transport.

These mechanisms must not be collapsed into a single “smoke toxin.”

Slide 6 · Selected smoke constituents: Distinct effects

Constituent

Principal effect described in Robbins

Polycyclic aromatic hydrocarbons

Carcinogenesis

Benzo[a]pyrene

Carcinogenesis

Nitrosamines

Carcinogenesis

Phenol

Tumor promotion; mucosal irritation

Formaldehyde / nitrogen oxides

Ciliary toxicity; mucosal irritation

Carbon monoxide

Impaired oxygen transport and utilization

Slide 7 · Airway irritation is not equivalent to cancer

CONTEXT A

Irritant effect on tracheobronchial mucosa

Increased mucus production

Bronchitis and mucociliary dysfunction

CONTEXT B

Chemical bioactivation

Electrophilic DNA-reactive intermediates

Mutations that may contribute to malignant transformation

Slide 8 · Chemical injury begins at the airway surface

Inhalation

Contact with tracheobronchial epithelium

Mucosal irritation and mucus production

Impaired airway function

Causal interpretation

Follow the direction of the arrows; Each transition represents a plausible biological step, not a deterministic event.

Schematic — Parallel to, not synonymous with, the pathway that produces oncogenic mutations.

Slide 9 · Direct and secondhand smoke both matter

Active smokers inhale a complex mixture of chemicals produced during tobacco combustion.

Environmental (secondhand) smoke exposes nonsmokers to carcinogens.

Robbins explicitly connects passive inhalation with lung cancer in nonsmokers.

No safe secondhand-smoke exposure level is asserted in the pulmonary carcinoma discussion.

Slide 10 · Nicotine: Dependence and promotion, not the whole causal story

CONTEXT A

Nicotine is strongly addictive.

Nicotinic receptors and catecholamine release help explain acute physiological effects.

Tumor promotion among nicotine-related effects.

CONTEXT B

Major carcinogenic exposures include polycyclic aromatic hydrocarbons and tobacco nitrosamines.

DNA-reactive metabolites directly explain a central mutagenic pathway.

Nicotine exposure cannot substitute for an inventory of smoke carcinogens.

Slide 11 · From inhaled compounds to cellular exposure

Mixture in smoke

Many chemicals, multiple effects

Airway deposition / contact

Respiratory epithelial target

Uptake and metabolism

Enzyme-dependent conversion

DNA-reactive species

Only some intermediates are electrophilic

Schematic — conceptual stages, not a diagnostic specimen.

Slide 12 · Procarcinogens require biological activation

Some chemicals in cigarette smoke are metabolized by cytochrome P450 (CYP) enzymes.

Metabolism can generate intermediates with greater water solubility.

Phase II enzyme pathways contribute to the handling and excretion of xenobiotics.

However, CYP processing can also yield electrophilic, DNA-reactive species.

Slide 13 · Detoxification and bioactivation are competing outcomes

CONTEXT A

Metabolic handling

Increased water solubility

Facilitated elimination

CONTEXT B

Electrophilic intermediate formation

Chemical attack on cellular DNA

Potential fixation of mutations

Slide 14 · The critical DNA-adduct pathway

CYP-mediated conversion

Electrophilic intermediate

Covalent DNA adduct

Imperfect repair → mutation

Causal interpretation

Follow the direction of the arrows; Each transition represents a plausible biological step, not a deterministic event.

Schematic — Specific chemical intermediates depend on the smoke constituent; This is a generalized pathway.

Slide 15 · What is a DNA adduct?

An adduct forms when a reactive chemical species covalently binds to DNA.

The altered template can interfere with faithful copying or repair.

Error-prone processing can convert chemical damage into a stable sequence mutation.

Not every adduct persists or produces an oncogenic driver event.

Slide 16 · How damage becomes a permanent mutation

DNA lesion

Repair or DNA replication

Incorrect sequence restoration / fixation

Mutated descendant cell

Causal interpretation

Follow the direction of the arrows; Each transition represents a plausible biological step, not a deterministic event.

Schematic — The diagram shows a possible causal route; Accurate repair may prevent mutation.

Slide 17 · DNA damage is not the same as a driver mutation

CONTEXT A

DNA damage: Chemical lesion in a molecule

May be repaired correctly

May never be inherited by daughter cells

CONTEXT B

Mutation: Persistent sequence alteration

Can be propagated during cell division

Only a subset promotes malignant growth

Slide 18 · Genomic sequencing provides mechanistic corroboration

Lung cancers arising in smokers often contain thousands of somatic mutations.

The mutation patterns include changes linked to tobacco carcinogens in experimental models.

This connects the chemical exposure to molecular alterations within the tumors.

Sequencing strengthens the causal inference; It does not prove every mutation is a driver.

Slide 19 · The mutational landscape contains passengers and drivers

Many smoke-related DNA lesions

Numerous acquired mutations

Selection of advantageous driver combinations

Malignant clone

Causal interpretation

Follow the direction of the arrows; Each transition represents a plausible biological step, not a deterministic event.

Schematic — High mutation burden is not equivalent to high driver count.

Slide 20 · Carcinogenesis requires more than one genetic event

Stepwise accumulation of oncogenic driver alterations in lung carcinomas.

Respiratory epithelial cells can acquire mutations before a tumor is apparent.

Only cells with the relevant combination of functional changes acquire malignant capabilities.

A long latency is compatible with multistep acquisition and selection.

Slide 21 · A simplified timeline of tumor evolution

Exposed epithelium

Carcinogens and repair

Early altered cell

Acquired somatic alterations

Expanding precursor clone

Additional selected changes

Invasive carcinoma

Invasion follows accumulated changes

Schematic — conceptual stages, not a diagnostic specimen.

Slide 22 · A field of altered epithelium may remain after quitting

That genetic changes predating lung cancer may persist in bronchial epithelium for years.

Macroscopically normal mucosa may already harbor acquired abnormalities.

This helps explain why cancer risk decreases slowly after smoking cessation.

Persistence of field changes does not imply that every altered cell becomes malignant.

Slide 23 · Why do many smokers never develop lung cancer?

CONTEXT A

Total carcinogen dose and duration differ.

Random acquisition and survival of driver combinations differ.

Individuals differ in chemical activation.

CONTEXT B

DNA damage responses and repair capacity vary.

Interactions with additional exposures vary.

Smoking is highly causal but not deterministically sufficient.

Slide 24 · Cytochrome P450 polymorphism modifies susceptibility

Polymorphic P450 monooxygenase enzymes.

Some genetic variants convert procarcinogens into active carcinogens more efficiently.

Smokers carrying such variants may incur greater lung-cancer risk.

These observations support gene–environment interaction, not simple Mendelian inheritance.

Slide 25 · Gene–environment interaction in chemical activation

Comparable tobacco exposure

Different CYP activation capacity

Different burden of reactive metabolites

Different probability of DNA injury

Causal interpretation

Follow the direction of the arrows; Each transition represents a plausible biological step, not a deterministic event.

Schematic — This is a directional explanatory schematic; No numerical risks are implied.

Slide 26 · Variation in DNA repair also modifies risk

A mutagen-sensitivity phenotype measured by chromosomal breakage after carcinogen exposure.

Higher sensitivity was associated with substantially increased lung-cancer risk in the cited comparisons.

This suggests inherited variation in processes limiting the consequences of DNA injury.

Such research phenotypes are not presented as routine clinical diagnostic tests.

Slide 27 · Dose and duration both influence lung-cancer risk

Lung-cancer occurrence to cumulative cigarette exposure.

In general, higher exposure is associated with higher risk.

Comparisons must distinguish cigarettes per day, years smoked and calculated pack-years.

An exposure–response pattern is an important epidemiologic pillar of causal inference.

Slide 28 · Pack-years summarize cumulative smoking exposure

Pack-years = packs smoked per day × years smoked

1 pack/day × 20 years

20 pack-years

2 packs/day × 20 years

40 pack-years

½ pack/day × 40 years

20 pack-years

Same pack-years can represent different smoking histories; Cumulative dose does not encode every exposure feature.

Slide 29 · Exposure metrics answer different questions

CONTEXT A

Cigarettes per day: Intensity

Duration in years: Temporal exposure

Pack-years: Rough cumulative index

CONTEXT B

Never versus ever smoked: Category

Current versus former: Cessation status

Secondhand exposure: Environmental exposure history

Slide 30 · Heavy exposure carries a particularly high relative risk

Robbins gives a 60-fold elevation in lung-cancer risk for habitual heavy smokers compared with nonsmokers.

The example exposure is two packs daily for 20 years.

This is a comparison reported by the textbook, not an individual probability.

Absolute risk, relative risk and proportion of cases attributable to smoking are different measures.

Slide 31 · Three common epidemiologic denominators

CONTEXT A

Risk among exposed people: “How likely is disease?”

Relative risk: “Compared with whom?”

Attributable fraction: “How much disease is caused by the exposure?”

CONTEXT B

Proportion of cases with smoking history: “Who has the disease?”

Proportion of smokers who develop cancer: “Who becomes ill?”

These quantities must not be interchanged.

Slide 32 · Smoking is causal, but lung cancer is not exclusive to smokers

Smoking is the most important environmental risk factor in the supplied Robbins sections.

Not all exposed persons develop a malignancy.

Some lung cancers arise in never-smokers.

The phenotype and molecular profile of lung cancer are partly associated with exposure history.

Slide 33 · Cessation lowers risk but does not erase prior exposure

During smoking

Ongoing contact with multiple carcinogens

After cessation

Further tobacco-smoke exposure stops

Subsequent years

Lung-cancer risk declines relative to continued smoking

Long term

Excess lung-cancer risk may persist for decades

Slide 34 · A decline in risk is not equivalent to immediate normalization

The tobacco section reports a fall in lung-cancer mortality after smoking cessation.

The lung-carcinoma section emphasizes that risk may remain above never-smoker baseline.

Persisting acquired mutations offer one biological explanation.

Prevention and surveillance have different roles after quitting.

Slide 35 · The synergistic context of occupational exposures

CONTEXT A

Asbestos or uranium exposure independently increases lung-cancer risk.

Tobacco adds another source of carcinogenic injury.

Multiple processes can act upon the same airway or lung epithelium.

CONTEXT B

The text reports markedly higher rates among exposed workers who smoke.

Relative effects require a defined comparison group.

Do not assume risks simply add or multiply without data.

Slide 36 · Smoking and co-exposure: Comparisons

Exposure setting

Comparison described in source

Asbestos workers who never smoked

Higher lung-cancer risk than unexposed nonsmokers

Asbestos workers who smoked

Much greater elevation than asbestos exposure alone

Uranium miners who never smoked

Elevated risk relative to general population

Uranium miners who smoked

Further elevation over nonsmoking miners

Slide 37 · The tobacco-specific lung carcinogens include several classes

Robbins Table 9.4 identifies polycyclic aromatic hydrocarbons for lung and larynx.

It also lists tobacco-specific nitrosamine NNK.

Polonium-210 appears among the suspected lung carcinogens.

“Suspected organ-specific” is the source table’s designation; The list is not exhaustive.

Slide 38 · Mechanism-oriented exposure inventory

Source group

Biologic role

Polycyclic aromatic hydrocarbons

Chemical carcinogens; DNA injury following bioactivation

Tobacco nitrosamines (e.g., NNK)

Tobacco-associated chemical carcinogens

Polonium-210

Suspected lung carcinogen listed in Robbins

Irritants / ciliary toxicants

Airway injury; not interchangeable with DNA adduct formation

Slide 39 · Major lung-carcinoma categories differ biologically

Robbins groups tumors into small cell and non–small cell lung carcinomas.

Adenocarcinoma and squamous cell carcinoma are major non–small cell subtypes.

Small cell carcinoma has particularly strong smoking association.

Morphology and molecular alterations are not identical across histotypes.

Slide 40 · Histotypes: Relationship to tobacco exposure

Histologic category

Relation to smoking in supplied Robbins

Squamous cell carcinoma

Strongly associated with tobacco smoke

Small cell carcinoma

Virtually always smoking related

Adenocarcinoma

Associated with smoking, but also most common in never-smokers

Large cell carcinoma

Undifferentiated category; no single exposure-specific feature established here

Slide 41 · Squamous carcinogenesis has recognizable precursor stages

Basal hyperplasia

Epithelial proliferation

Squamous metaplasia

Altered epithelial lineage

Squamous dysplasia / CIS

Atypia / full-thickness abnormality

Invasive carcinoma

Basement membrane breached

Schematic — conceptual stages, not a diagnostic specimen.

Slide 42 · Metaplasia and dysplasia answer different questions

CONTEXT A

Metaplasia: A change in differentiated epithelial phenotype

May occur during chronic mucosal injury

Not itself a diagnosis of invasive malignancy

CONTEXT B

Dysplasia: Cytologic and architectural abnormalities

Loss of polarity; Nuclear atypia; Mitotic activity

More advanced abnormalities raise concern for neoplastic progression

Slide 43 · Carcinoma in situ versus invasion

Atypical epithelium

Full-thickness epithelial atypia

Intact basement membrane = in situ

Basement membrane breach = invasion

Causal interpretation

Follow the direction of the arrows; Each transition represents a plausible biological step, not a deterministic event.

Schematic — A conceptual transition; Not a diagnostic photomicrograph.

Slide 44 · Squamous cell carcinoma: Morphology and setting

A central or hilar location is characteristic, though peripheral cases occur.

Keratinization and/or intercellular bridges support squamous differentiation.

Keratin pearls may be prominent in well-differentiated lesions.

Poorly differentiated tumors may require p40 immunostaining to demonstrate squamous lineage.

Slide 45 · Why a central squamous tumor causes obstruction

CONTEXT A

A central mass can protrude into the bronchial lumen.

Complete obstruction prevents distal ventilation.

Retained secretions can predispose to infection.

CONTEXT B

Atelectasis develops downstream of complete obstruction.

Postobstructive pneumonia or abscess may reveal an occult tumor.

Local effects reflect position, not only histologic grade.

Slide 46 · Adenocarcinoma: A different precursor context

Atypical adenomatous hyperplasia

Small peripheral precursor

Adenocarcinoma in situ

Pure lepidic growth without invasion

Minimally invasive lesion

Restricted invasive focus

Invasive adenocarcinoma

Glandular / mucinous differentiation

Schematic — conceptual stages, not a diagnostic specimen.

Slide 47 · Lepidic growth is an architectural pattern

Neoplastic cells extend along pre-existing alveolar septa.

Pure lepidic growth without stromal invasion characterizes adenocarcinoma in situ.

Invasive adenocarcinomas may contain a lepidic component plus invasive patterns.

A schematic cannot substitute for authentic sections in a morphology examination.

Slide 48 · Peripheral adenocarcinoma versus central squamous cancer

CONTEXT A

Adenocarcinoma is often peripheral.

Gland formation or mucin production supports differentiation.

Many cases express TTF-1.

CONTEXT B

Squamous carcinoma has a historical central predilection.

Keratinization / intercellular bridges are defining clues.

P40 can establish squamous differentiation when morphology is limited.

Slide 49 · Adenocarcinoma spans distinct molecular backgrounds

KRAS alterations are prominent in a subset of tumors.

Some adenocarcinomas contain activating EGFR or other receptor kinase pathway alterations.

The driver spectrum varies with smoking history and other factors.

Predictive molecular testing concerns tumor biology, not proof that a particular exposure caused that cancer.

Slide 50 · Oncogenic signaling in some adenocarcinomas

Receptor kinase activation

Downstream growth signaling

Clonal proliferation / survival

Invasive adenocarcinoma phenotype

Causal interpretation

Follow the direction of the arrows; Each transition represents a plausible biological step, not a deterministic event.

Schematic — Generic signaling schematic; Molecular drivers and effects vary by tumor.

Slide 51 · Small cell carcinoma is especially smoking associated

Small cell carcinoma as virtually always related to smoking.

It has a particularly high burden of acquired mutations.

TP53 and RB loss of function is characteristic of the subtype.

Aggressive biology and early dissemination distinguish it clinically.

Slide 52 · Small cell carcinoma: Diagnostic morphology

Tumor cells are small with scant cytoplasm and finely granular chromatin.

Nuclear molding and inconspicuous nucleoli are typical.

Mitotic activity is high and necrosis often extensive.

Neuroendocrine differentiation may be supported by appropriate marker expression.

Slide 53 · Acquired alteration patterns across major histotypes

Tumor type

Illustrative genetic changes in Robbins

Squamous cell carcinoma

TP53; CDKN2A; 3p loss; sometimes FGFR1 amplification

Small cell carcinoma

Frequent TP53 and RB inactivation; MYC family amplification in some

Adenocarcinoma

KRAS, EGFR, ALK, ROS1, MET, RET, BRAF and other driver pathways

Slide 54 · Smoking-related versus never-smoker lung cancer

CONTEXT A

Smoking-related tumors often show high mutation burden.

Squamous and small cell carcinomas strongly track smoking.

Tobacco mutational patterns can be identified by sequencing.

CONTEXT B

Adenocarcinoma commonly occurs in never-smokers.

EGFR alterations are relatively enriched in some never-smoker groups.

Exposure history changes prior probability—not histologic diagnosis.

Slide 55 · Immunophenotype: A diagnostic adjunct, not exposure proof

TTF-1 expression is frequent in pulmonary adenocarcinoma.

P40 supports squamous differentiation in poorly differentiated carcinomas.

Chromogranin, synaptophysin and CD56 support neuroendocrine differentiation.

A staining pattern identifies lineage; It does not establish whether smoking caused the tumor.

Slide 56 · Pathology reasoning: Keep the levels separate

Question

Evidence to use

Was there tobacco exposure?

Patient history and exposure assessment

Does the lesion represent carcinoma?

Histology and invasion assessment

Which lung-carcinoma phenotype?

Architecture, cytology and selective immunostains

What drives this tumor?

Tumor molecular characterization

Can smoking be assigned as an individual cause?

Exposure history plus population-level and molecular context; avoid categorical attribution

Slide 57 · Local anatomy explains clinical presentations

Central airway involvement may cause cough or hemoptysis.

Bronchial blockage can lead to atelectasis and postobstructive infection.

Extension into pleura, chest wall or mediastinum produces additional local effects.

These findings signal tumor location or spread, not a specific carcinogen.

Slide 58 · How a bronchial carcinoma produces secondary disease

Tumor grows into bronchial lumen

Partial or complete obstruction

Impaired ventilation and secretion clearance

Atelectasis / infection / abscess

Causal interpretation

Follow the direction of the arrows; Each transition represents a plausible biological step, not a deterministic event.

Schematic — Anatomic complication rather than a specific etiologic marker.

Slide 59 · Dissemination makes lung cancer clinically consequential

Lung carcinomas spread through lymphatic and hematogenous pathways.

Adrenal glands, liver, brain and bone as frequent metastatic sites.

Small cell carcinoma often disseminates early.

Stage and disease distribution are distinct from the initiating tobacco exposure.

Slide 60 · Paraneoplastic associations reflect tumor phenotype

CONTEXT A

Small cell carcinoma: ADH-related hyponatremia may occur.

Small cell carcinoma: Ectopic ACTH may occur.

Other neuroendocrine/systemic manifestations are described.

CONTEXT B

Squamous carcinoma: PTHrP-related hypercalcemia is a familiar association.

An unexpected metabolic syndrome may precede tumor recognition.

These associations are not diagnostic in isolation.

Slide 61 · Prevention targets the causal exposure

Avoiding initiation of tobacco smoking prevents a major source of lung carcinogens.

Stopping smoking limits future delivery of DNA-damaging compounds.

Reducing secondhand exposure protects nonsmokers.

The persistence of residual risk does not negate the benefits of cessation.

Slide 62 · Screening and prevention serve different purposes

Smoking prevention reduces carcinogen exposure before cancer develops.

The lung-carcinoma section discusses low-dose CT detection in selected high-risk people.

Early detection may identify resectable disease but produces false-positive findings.

No specific current screening eligibility threshold is inferred from the supplied excerpt.

Slide 63 · Case: Cumulative exposure and a new mass

A 67-year-old patient smoked 1½ packs daily for 30 years and now has a centrally located lung mass with bronchial obstruction.

Interpretation questions

Compute the cumulative pack-year exposure.

Which histologic subtype is classically associated with a central mass and smoking?

Why might downstream lung become atelectatic?

Slide 64 · Case interpretation: Distinguish three inferences

Exposure: 1.5 × 30 = 45 pack-years.

Morphologic hypothesis: Squamous cell carcinoma is plausible, but tissue diagnosis is required.

Obstruction prevents distal ventilation and impairs secretion clearance; Atelectasis and infection can follow.

Slide 65 · Case: A former smoker with an interval of abstinence

A patient quit smoking many years ago. A newly detected pulmonary carcinoma is diagnosed today.

Interpretation questions

Does cessation make tobacco causation impossible?

Which cellular finding can persist after smoking stops?

Is an exposure history sufficient to identify the tumor subtype?

Slide 66 · Case interpretation: Persistent molecular effects

No. Excess lung-cancer risk can remain elevated for decades after cessation.

Previous somatic changes in bronchial epithelium may persist.

No. Histologic evaluation and, when indicated, immunophenotype and molecular assays are necessary.

Slide 67 · Case: Lung cancer in a never-smoker

A lifelong never-smoker develops a peripheral pulmonary mass composed of malignant gland-forming cells.

Interpretation questions

Does absence of tobacco history exclude lung cancer?

Which major histotype is most consistent with gland formation?

What is the role of mutation testing in this setting?

Slide 68 · Case interpretation: Morphology before attribution

Never-smokers can develop lung carcinoma.

Glandular differentiation supports pulmonary adenocarcinoma when the clinical context is appropriate.

Tumor molecular analysis can identify driver pathways; It does not retrospectively prove an environmental cause.

Slide 69 · Integrated causal model of smoking-related lung cancer

Smoke chemistry + cumulative exposure

Bioactivation / DNA adduct burden

Somatic mutations + clone selection

Precursor lesion → invasive carcinoma

Causal interpretation

Follow the direction of the arrows; Each transition represents a plausible biological step, not a deterministic event.

Schematic — Host susceptibility and non-tobacco carcinogens can modify multiple steps.

Slide 70 · Exam-oriented distinctions

Concept

Essential distinction

Nicotine vs carcinogens

Addiction / promotion is not interchangeable with DNA mutagenesis

Damage vs mutation

Repairable lesion versus persistent sequence change

Exposure vs diagnosis

Smoking history does not specify tumor histotype

CIS vs invasion

Basement membrane intact versus breached

Quitting vs risk eradication

Reduced future exposure; possible residual risk

Slide 71 · What the evidence supports—and what it does not

Smoking causes a substantial burden of lung cancer through mutagenic exposure.

Dose–response, co-exposure effects and tumor mutation patterns support causality.

The precise initiating events of a particular patient’s cancer usually cannot be reconstructed.

Clinical management still requires histologic classification and assessment of tumor biology.

Slide 72 · References and directed reading

Robbins, Cotran & Kumar: Pathologic Basis of Disease, 11th ed. (2026): Effects of Tobacco; Smoking and Lung Cancer.

Robbins, Cotran & Kumar: Pathologic Basis of Disease, 11th ed. (2026): The Lung — Carcinomas.

Review the original textbook images of squamous precursor lesions, adenocarcinoma and small cell carcinoma.

Revisit the tobacco-smoke constituent tables when interpreting different mechanisms of injury.