# HSEF Review: Cigarette Smoking as a Major Cause of Lung Carcinoma > **Provenance:** 18 sources · 14 registered claims · 14 verified · 0 unverified · 0 contradicted · 13 Tier A sources · gate passed. Every identifier was resolved against PubMed and PubMed Central; two identifiers recalled during drafting resolved to unrelated papers and were corrected before use. Framework: HSEF v5.4; cancer-biology rubric. > **Adversarial review:** This edition was itself produced by a cross-model red-team rebuild (Claude, HSEF v5.4, 2026-09-30) rather than by Sol 6.1; see the Red-team rebuild note. Adversarially reviewed again by Claude Opus 5.5 (Anthropic) on 2026-10-14. 2 corrections applied (see ADVERSARIAL REVIEW AND APPROVAL). Status: **Approved as revised.** **Hypothesis assessed:** *Sustained cigarette smoking is a major initiating cause of human lung carcinoma.* **Gradeable strong form:** Sustained cigarette smoking causally initiates a substantial (major) share of human lung carcinomas, as shown by exposure-response across independent populations, incidence decline after cessation, and a carcinogen mechanism that leaves a dose-scaled molecular fingerprint in the tumours. **Review date:** 2026-09-30 **Mode:** Full rebuild of a concise, non-standard input review (three external sources, schema 5.3), then cross-model red team. The input was authored by a different model family; the rebuild and red team were performed by Claude. ## AT A GLANCE The claim is that smoking cigarettes for years is a major cause — not the only cause, but a large one that actually starts the disease — of human lung cancer. The evidence is about as strong as causal evidence in human biology gets, and it is strong in several independent ways at once: the more and longer people smoke the more lung cancer they get; when they quit, risk falls year by year; the carcinogens in smoke leave a specific mutational fingerprint on the exact DNA positions damaged in lung tumours; and controlled smoke exposure causes lung cancer in animals. Because these lines could not all be wrong for the same reason, the claim is graded near the top of the ladder. | HSEF Level | Confidence | Role of mechanism | Primary rubric | |---|---|---|---| | **Θ** — High predictive accuracy, minimal unresolved exceptions | High | primary driver | `02-cancer-biology.md` | | Item | Assessment | |---|---| | Headline finding | Sustained cigarette smoking causally initiates a major share of human lung carcinomas; the causal reading rests on convergent exposure-response, cessation, mechanistic and animal evidence, not on association alone. | | HSEF level | **Θ** — High predictive accuracy, minimal unresolved exceptions | | Confidence | **High** — dozens of independent populations over seven decades, multiple assay modalities, extensive replication | | Why not Ω | Smoking is neither necessary (≈10–25% of lung cancers occur in never-smokers, including a distinct EGFR-mutant subset) nor sufficient (most smokers never develop lung cancer); Ω requires both | | Why not Η | The cancer rubric names this exposure–carcinoma pair as a top-of-ladder calibration case; within the claimed "major share" scope the exceptions are a separately-attributed remainder, not unresolved contradictions | | Where the rules differ | Scope-matched (prescribed) **Θ** · primary-subclaim **Θ** · weakest-link **Ζ** (SC3 held at Ζ for partial co-dependence of its molecular lines) | | Strongest positive evidence | Dose-response across independent cohorts and case-control series [1,2,7,8]; cessation-linked incidence decline [3,4,5,6]; tumour mutational fingerprint [9,10]; animal carcinogenicity [11]. The British Doctors cohort [3,4] is counted once, on the cessation line | | Strongest scope boundary | Never-smoker lung cancer and the EGFR-mutant adenocarcinoma subset [13,14,15]; residential radon as an interacting cause [12] | | Momentum / stability / divergence | Mechanistic only / Stable / Aligned | **On the headline letter.** Θ is a verdict on the claim as written — *major* initiating cause of a substantial share — not a claim that smoking is the sole or universal cause of lung carcinoma. The never-smoker cases are not exceptions to the graded claim; they are a separately-attributed remainder that bounds necessity and is exactly why the top rung, Ω, is blocked. ## SUBCLAIM MATRIX | ID | Subclaim and tested scope | Level | What fixes the level | |---|---|---:|---| | SC1 | Sustained smoking causally increases lung-carcinoma incidence at the population level, risk rising with exposure (a major share). | **Θ** | Dose-response reproduced across independent populations and decades, plus cessation-linked decline and animal carcinogenicity; never-smoker cases bound scope rather than refute it [3,4,7,8,11,13–15]. | | SC2 | Stopping smoking reduces subsequent risk relative to continued smoking; risk falls with time since quitting, residual risk persists. | **Η** | Graded, replicated risk reduction with time since cessation across populations; incomplete return to never-smoker risk fits an initiating exposure that leaves durable damage [4,5,6,8]. | | SC3 | Tobacco-smoke carcinogens contribute mechanistically to lung carcinogenesis via DNA damage and somatic mutation. | **Ζ** | Animal in-vivo carcinogenicity crosses the threshold; the human tumour mutational signature and adduct targeting of TP53 hotspots are further, differently-failing lines; in-vitro adduct mapping alone is Γ-capped [9,10,11]. | **Aggregation rule: scope-matched, pre-registered.** The central claim is a population-level causal-initiation claim. Scope-matched sets the overall level to the highest subclaim whose scope matches the central claim's scope. SC1 *is* that scope and sits at Θ; SC2 (reversibility) and SC3 (mechanism) are supporting subclaims at narrower scopes, reported separately. **The alternatives, as v5 requires:** scope-matched **Θ**; primary-subclaim on SC1 **Θ**; weakest-link **Ζ**. The two-rung spread between weakest-link and scope-matched is driven by rule choice on one deliberate decision: SC3 is held at Ζ because the human molecular lines (adduct mapping and the mutational signature) are partly one line read at two resolutions, whereas the population claim SC1 does not depend on fully resolving the molecular chain. There is no dedicated hypothesis-index entry for this hypothesis; the cancer rubric's calibration note and the framework's upper-ladder rules govern. ## COMPETING MECHANISTIC MODELS | Rank | Model | Level | Fit to the shared evidence | |---:|---|---:|---| | 1 | **Smoking as a major causal initiator** — nominal | **Θ** | Explains dose-response, cessation-linked decline, the tumour mutational fingerprint, and animal carcinogenicity together, each excluding a different confounder. | | 2 | **Multifactorial / multiple-cause model** | **Θ** | Complementary, not rival: specifies the non-smoking remainder (radon, never-smoker pathways) while agreeing smoking is the largest contributor. Bounds necessity and sufficiency. | | 3 | **Constitutional / confounding-only (Fisher)** | **Β** | Fails: a fixed constitutional confounder cannot produce risk decline proportional to time since cessation, a dose-scaled carcinogen signature in tumour genomes, or cancer under randomized animal exposure. | | 4 | **Reverse causation / detection artifact** | **Α** | Fails against prospective cohorts with exposure ascertained decades before disease and against the temporal fall in incidence after cessation. | The nominal model and the multifactorial model are not in genuine conflict: the multifactorial account is what makes the nominal claim "major" rather than "universal." The two failed rivals are the historically serious ones — R.A. Fisher's constitutional-confounding argument and reverse causation — and both are decisively excluded by the withdrawal, molecular and animal evidence. The clearest single separator is controlled animal inhalation carcinogenicity [11], which removes every human behavioural confounder the confounding-only account might invoke and still yields lung cancer. The tournament is reported for what it is, a comparison of explanations: it informs the role of mechanism (primary driver) and does not set the letter, which follows from the caps, thresholds and aggregation rule alone. ## FULL REVIEW ### Claim form and falsifiability The sentence is a single causal proposition — smoking is a *major initiating cause* — not a conjunction. For scoring it is decomposed into a population causal contribution (SC1), a reversibility claim (SC2), and a mechanistic contribution (SC3). "Major" is read as a substantial population-level share, explicitly not as "necessary for every case," "the sole cause," or "sufficient in every exposed person." The claim is falsifiable in each part: SC1 would fail if exposure-response and cessation patterns dissolved under careful control of competing causes; SC2 would fail if quitting did not lower risk after reverse-causation checks; SC3 would fail if carcinogen exposure produced no DNA damage, no dose-scaled tumour signature, and no animal carcinogenicity. None of these fails; the review therefore examines whether the *scope* is fairly stated, which is where the never-smoker remainder and individual susceptibility matter. ### Population-level causal contribution (SC1) The association was first established in controlled case-control studies in 1950 — Doll and Hill in the United Kingdom and Wynder and Graham's series of 684 proven bronchogenic-carcinoma cases in the United States [1,2]. Association alone, however large, is capped at Β by the framework; what carries SC1 to the top of the ladder is the convergence of exposure-response with withdrawal and mechanism. The exposure-response gradient is steep and reproducible. In the largest pooled case-control dataset — 13,169 cases and 16,010 controls from Europe and Canada — current male smokers averaging more than 30 cigarettes per day had odds ratios of 103.5 (95% CI 74.8–143.2) for squamous-cell carcinoma, 111.3 (69.8–177.5) for small-cell carcinoma, and 21.9 (16.6–29.0) for adenocarcinoma; odds ratios rose across every metric of exposure and were consistently lower for adenocarcinoma, the subtype most common in never-smokers [8]. In contemporary US cohorts, current smokers carried a lung-cancer death rate about 25 times that of never-smokers (relative risk 25.66 in women, 24.97 in men) [7]. In the 50-year follow-up of 34,439 male British doctors, men who smoked from early adult life and continued died on average about 10 years younger than lifelong non-smokers, and prolonged smoking tripled age-specific mortality among those born around 1920 [4]. The gradient holds across sexes, birth cohorts, countries and decades — a pattern no single population's confounding structure can generate. The scope is genuinely bounded, and the review states the bound rather than hiding it. Roughly 10–25% of lung cancers occur in people who never smoked, a group large enough that never-smoker lung cancer would rank among the leading causes of cancer death if counted separately [14,15]. Never-smoker adenocarcinomas are frequently a molecularly distinct entity: EGFR tyrosine-kinase-domain mutations appeared in 7 of 15 untreated never-smoker adenocarcinomas versus 4 of 81 tumours from former or current smokers [13]. Residential radon is an independent, interacting cause, raising risk about 16% per 100 Bq/m³ of usual radon and acting far more strongly in smokers [12]. These facts limit necessity and universality — which is why Ω is blocked — but they do not touch the claim that smoking is a major cause of a substantial share. ### Cessation-associated risk change (SC2) Exposure withdrawal is a field-valid causal equivalent whose confounding structure differs from exposure accumulation, and it points the same way. Among the British doctors, cessation at ages 30, 40, 50 and 60 gained about 10, 9, 6 and 3 years of life expectancy, respectively [4]. In the UK case-control-plus-national-statistics analysis, stopping before middle age avoided more than 90% of the tobacco-attributable lung-cancer risk, and former smokers' rates fell steeply with time since quitting: cumulative risk of lung cancer by age 75 for men who stopped at 60, 50, 40 and 30 was 10%, 6%, 3% and 2%, against 16% for continuing male cigarette smokers at 1990 rates [5]. A nationally representative US cohort found cessation before age 40 removed about 90% of the smoking-associated excess mortality [6]. The trajectory change with the withdrawal of exposure is the intervention-sensitive signal that defines Η, and it excludes any fixed constitutional confounder, which could not make risk fall in proportion to time since quitting. The residual: risk declines after cessation but does not fully return to never-smoker levels even about 35 years out [8] — exactly what an initiating exposure that leaves durable genomic damage predicts. SC2 sits at Η: strong convergent, intervention-sensitive causal evidence, with the persistent residual keeping it below full predictive accuracy. An older internal control deserves its own line. In the 20-year British Doctors follow-up, as the cohort progressively reduced its cigarette consumption, lung cancer became relatively less common while other cancers did not — a within-cohort withdrawal contrast the authors offered as an unusual demonstration of the causal, rather than merely associational, nature of the relationship [3]. References [3] and [4] are two follow-up reports of that one cohort, not two studies, and the independence basis counts them once: the British Doctors Study carries the exposure-withdrawal line and is deliberately excluded from the exposure-accumulation line, which rests on the two 1950 case-control series, the pooled US cohorts and the pooled European and Canadian case-control dataset [1,2,7,8]. One cohort's systematic error would defeat all of its reports, so it cannot support two independent lines at once. ### Mechanistic contribution (SC3) The mechanism is causally coherent and confirmed beyond a single assay class. Benzo[a]pyrene diol epoxide adducts form preferentially at guanine positions in TP53 codons 157, 248 and 273 in bronchial epithelial and HeLa cells — the same positions that are the major mutational hotspots in human lung cancers — indicating targeted adduct formation rather than phenotypic selection, and providing what the authors called a direct etiological link between a defined tobacco carcinogen and human cancer [9]. Across 5,243 cancers, smoking was associated with an increased burden of a specific mutational signature attributable to misreplication of DNA damage from tobacco carcinogens, concentrated in tissues directly exposed to smoke, consistent with smoking raising cancer risk by increasing the somatic mutation load [10]. And controlled life-span inhalation of mainstream cigarette smoke induced lung cancer in mice — a 7.25-fold increase in pulmonary adenocarcinomas and a 4.6-fold increase in adenomas and papillomas versus sham-exposed controls (n = 330 vs 326) [11]. Consensus and mechanistic reviews classify tobacco smoke as an IARC Group 1 human carcinogen and map how its carcinogens and metabolites damage DNA and activate oncogenic signalling [16,17,18]. SC3 is graded Ζ rather than higher for a specific, honest reason. The adduct mapping (in cultured cells) is Γ-capped in isolation, and the human tumour mutational signature is arguably the same molecular line read at a different resolution — the framework itself flags adduct mapping and mutational signatures as potentially one line. The animal carcinogenicity study is a genuinely independent line (a controlled in-vivo perturbation that fails differently from any human correlation), so SC3 rests on at least two independent lines and crosses into convergence — Ζ — without over-claiming that the complete human initiating chain is quantitatively mapped for every case. ### Caps, boundary check and upper-ladder reachability No cap is triggered as a ceiling on any subclaim, and the review records why rather than resolving the caps silently. The universal association cap (Β) would bind lung-cancer epidemiology taken in isolation, but SC1 is a convergent set — dose-response plus withdrawal plus animal carcinogenicity — and the framework is explicit that a cap binds inference from one evidence class and is not re-applied to a set assembled so each member covers another's weakness. The universal in-vitro cap (Γ) binds the benzo[a]pyrene adduct mapping in isolation, but SC3 also carries animal in-vivo carcinogenicity and human tumour molecular epidemiology. Each cap governs the inference from its single evidence class; the subclaims are scored on the convergent set. The Β/Γ/Δ boundary check is positive for all three subclaims: exposure cessation lowers incidence and controlled animal exposure raises it, so the field-valid perturbations are present, disease-relevant, and positive, and the ceiling from the boundary check is "Δ or higher." The upper ladder is reachable here without a randomized trial of the mechanism, which would be unethical. Each field-valid equivalent was checked: dose-response across independent populations (present, carrying the claim), withdrawal-and-rechallenge at population scale (present), a somatic genetic natural experiment in the form of the tumour mutational signature and the never-smoker EGFR-mutant subset (present), and therapeutic/controlled perturbation with a hard endpoint in the form of animal carcinogenicity (present). Each supports Θ; none reaches Ω, because none establishes necessity or sufficiency. The framework's own text names smoking and lung cancer as a case that "has no RCT and sits near the top of the ladder; a framework that penalizes it for that is measuring study design rather than causal knowledge." This review takes that instruction at face value. ### Falsification and the boundary that blocks Ω Two prespecified falsification patterns are met — and both bound the claim rather than defeat it. Necessity fails: lung cancer occurs without smoking, in a distinct never-smoker population with its own molecular biology [13,14,15]. Sufficiency fails: most smokers never develop lung cancer, and interacting causes such as radon modulate who does [5,12]. A third pattern is tested and not met: the confounding-only account would have to explain dose-response, cessation reversal, the dose-scaled tumour signature and animal carcinogenicity together, and it cannot [3,9,10,11]. Because necessity and sufficiency both fail, Ω is unavailable; because the exceptions are a separately-attributed remainder rather than unresolved contradictions within the "major share" scope, Θ stands rather than Η. ### Steelman and final adjudication **The case one rung higher (Ω), written first.** Smoking is the most firmly established environmental cause of any common human cancer, with a larger and more convergent evidence base than most Ω-eligible mechanisms. One could argue that for the smoking-attributable subset the exposure is both necessary (that subset would not exist without smoking) and sufficient at the population level (raising exposure reliably raises incidence), and that restricting the grade to "major" understates how completely the causal question is settled. **The case one rung lower (Η).** Human exposure cannot be randomized, all the population evidence is observational, and the complete molecular initiating chain is not quantitatively mapped for every case; a cautious reviewer could stop at strong convergent causal evidence. **Both are rejected.** Ω is blocked by the necessity-and-sufficiency threshold: no dataset shows smoking is necessary (never-smoker lung cancer, the EGFR-mutant subset) or sufficient (most smokers never develop the disease), and no missing experiment could restore Ω without redefining the claim to a narrower subtype. Η is rejected as too low because the rubric names this exposure–carcinoma pair as a top-of-ladder calibration case, and within the claimed "major share" scope the exceptions are a separately-attributed remainder, so predictive accuracy is high with minimal unresolved exceptions. **Θ — High predictive accuracy, minimal unresolved exceptions — is the correct letter, with High confidence.** The direction-of-causation procedure classifies the mechanism as a **primary driver** by the framework's five-step derivation. It is not a downstream consequence (lung cancer does not produce smoking) and not an epiphenomenon (causal dependency is demonstrated). Step 3 matches on all three conditions within the claimed scope: removing the exposure substantially prevents or delays the phenotype (cessation before middle age avoids more than 90% of tobacco-attributable risk), engaging it initiates the phenotype (controlled animal inhalation causes lung cancer; smoking initiates a major share in humans), and it is rate-limiting (monotonic dose-response with daily dose and pack-years). Primary-driver describes smoking's role for the share it causes; it is not a claim of necessity or sufficiency across all lung carcinoma, which is the scope bound that blocks Ω. ## PROPOSED DECISIVE EXPERIMENTS ### 1. Prospective cross-population dose-response with molecular confirmation Combine harmonized lifetime-exposure cohorts across populations with tumour genome sequencing, linking pack-year dose both to incidence and to the tobacco mutational-signature burden, with never-smoker and radon-exposed comparison strata and adjustment for competing causes. The claim is strengthened if incidence and signature burden both track cumulative exposure across independent populations; it is pressured if either fails to track exposure once competing causes are controlled. ### 2. Cessation natural-experiment with residual-risk genomics Follow matched continuing and quitting smokers, measuring how residual risk and, where tumours arise, carcinogen-signature burden decay with time since quitting, with explicit reverse-causation checks that exclude quitting driven by preclinical disease. A graded decline in both risk and signature burden with a persistent residual supports the initiating-exposure reading; a flat trajectory after excluding reverse causation would pressure it. ## KNOWLEDGE GAPS - **Attributable-fraction precision.** The exact smoking-attributable fraction varies by histology and population and is not fully resolved for every subtype; this refines, not challenges, the "major share" claim. - **Never-smoker etiology.** The causes of most never-smoker lung cancers beyond radon and secondhand smoke are incompletely defined, which determines the size and biology of the non-smoking remainder. - **Individual susceptibility.** Why most smokers never develop lung cancer — host genetic and stochastic modifiers — is only partly understood, which is why smoking is not sufficient and Ω is bounded out. ## REVIEWER NOTES ### Red-team rebuild note This edition is the published product of a cross-model pass on 2026-09-30. The input was a thin, non-standard concise review: three external sources (a CDC web page, one pooled case-control study, one mechanistic review), a `schema_version` of "5.3" with an `assessment` block instead of `final_score`, and no subclaim matrix, cap analysis, independence basis, or provenance gate. It asserted the grade Θ. Rather than lightly edit it, this edition is a full **rebuild**: the claim was independently re-decomposed and re-derived, and the source base was expanded from three to eighteen, every one resolved against PubMed. The letter is unchanged — Θ, High confidence — but it is now earned by the full HSEF v5.4 machinery rather than asserted. Two points of process are worth recording for the corpus. First, the input's two usable citations both resolved correctly (Pesch 2012, PMID 22052329; Wen 2011, PMID 21801603), so no confabulated identifier existed in the input itself; the CDC web page was removed and replaced by verifiable primary and consensus sources. Second, during drafting, two identifiers recalled from memory for classic papers resolved to unrelated records — one to an Arabidopsis cell-death paper, one to a Russian surgical-infections paper — and were corrected against PubMed before any use. This is the confabulation failure the corpus is known for, caught by resolving every identifier by hand. ### Validation meaning The machine validator checks schema presence, level/descriptor consistency, aggregation arithmetic, cap ceilings, evidence identifiers, source status, claim-to-source links, the independence basis, and required prose blocks. A pass means those structural and internal-consistency checks succeeded. It is not an independent replication of the underlying studies. In particular, the validator cannot detect a citation whose identifier resolves to a different paper than the one described, which is why every identifier in this edition was resolved by hand against PubMed. Full detail is in `redteam/REDTEAM_REPORT.md`; the itemised rebuild map and open issues are in `redteam/SYNTHESIS_CHANGE_LOG.md`. The unaltered input is archived under `sources/`. ## ADVERSARIAL REVIEW AND APPROVAL **Authoring model:** not Sol 6.1. This edition is a cross-model red-team **rebuild** performed by Claude (Anthropic) on 2026-09-30 under HSEF v5.4, replacing a thin three-source concise review authored by a different model family; the score card's `provenance_header.generated_by` records the same. The standard round-2 banner wording was adapted accordingly rather than asserting an authorship the document contradicts. **Adversarial reviewer:** Claude Opus 5.5 (Anthropic). **Date:** 2026-10-14. **Scope of checks.** All 18 REFERENCES identifiers verified against PubMed `get_article_metadata`: 18/18 resolve to the stated first author, journal, year and DOI; 0 failures; no retraction or expression-of-concern flag. Eleven load-bearing quantitative claims were spot-checked against the cited abstracts, and every figure matched: the pooled case-control totals (13,169 cases, 16,010 controls) with the three histology-specific odds ratios and their intervals, and the finding that risk does not return to never-smoker baseline even 35 years after cessation; the contemporary-cohort lung-cancer death relative risks of 25.66 in women and 24.97 in men; the 34,439 male British doctors, the roughly ten years of life lost, the tripling of age-specific mortality in the 1920s birth cohort, and the cessation gains of about 10, 9, 6 and 3 years at ages 30, 40, 50 and 60; the UK cumulative risks by age 75 of 10, 6, 3 and 2 per cent against 16 per cent for continuing smokers, and the more-than-90-per-cent avoidance when stopping before middle age; the US figure of about 90 per cent of excess mortality removed by cessation before 40; the mouse inhalation study's 7.25-fold adenocarcinoma and 4.6-fold adenoma-and-papilloma increases at n = 330 against 326; the 7 of 15 versus 4 of 81 EGFR-mutation split; the 16 per cent per 100 Bq/m3 usual-radon increase; and the 5,243 cancers in the mutational signature analysis. Scoring logic was re-derived: caps precede aggregation, the prescribed `scope-matched` rule sets the headline from SC1 (the only subclaim whose scope matches the central claim) and the narrower SC2 and SC3 are reported separately, which is the correct reading of that rule; weakest-link Ζ and primary-subclaim Θ are both reported; the Β/Γ/Δ boundary and the Ζ-plus independence basis were checked. Framework validator (v5.4, 2.4) re-run. **Findings** | ID | Severity | Location | Original to corrected | Evidence | |---|---|---|---|---| | K1 | Major | `independence_basis`, exposure-accumulation line; AT A GLANCE "Strongest positive evidence" row; SC2 section | Exposure-accumulation line sources [1, 2, 4, 7, 8] to [1, 2, 7, 8]; the headline evidence row re-cited as dose-response [1,2,7,8] and cessation [3,4,5,6]; an explicit shared-cohort sentence added to SC2 | References [3] and [4] are two follow-up reports of one cohort, the British Doctors Study, so [4] was being counted in the exposure-accumulation line while [3] and [4] carried the exposure-withdrawal line. The framework rule is that independence counts studies, not reports, because one cohort's systematic error would defeat all of its reports. This matters at Θ, where the letter rests on an independence basis above Ζ. The cohort is now counted once, on the withdrawal line where it uniquely supplies the cessation gradient; the accumulation line still rests on four genuinely separate investigations and is not weakened. The framework validator moved from WARN to a clean PASS on this check as a result. No level, confidence or subclaim letter changed. | | K2 | Minor | `citation_ledger` and `evidence_inventory`, all 18 rows | No `study_id` on any row to a distinct `study_id` per study, with rows 3 and 4 sharing `BRITISH-DOCTORS-STUDY` | The validator warned that 18 ledger rows carried no study identity, with the fix "tag rows so a reader does not read N records as N independent investigations". Tagging them is what surfaced K1. | **Rejected findings (checked, no change).** (i) Confidence band High: the stated rationale already rests on evidence density and reproducibility — dozens of independent populations over seven decades, multiple assay modalities, extensive replication — and says nothing about reading access. Under the lead's two-step test nothing needed striking and the band stands. This review is the model the boilerplate rationales in the Sol 6.1 batches should have followed. (ii) Momentum "Mechanistic only": the artifact states that the threshold-crossing causal work was done decades ago and that current output refines mechanism and the never-smoker remainder rather than re-testing causation. Correct, and it does not treat a completed publication as work underway. (iii) Scope-matched aggregation: SC1 is the only subclaim whose scope matches the population-level central claim and it sits at Θ; SC2 (Η) and SC3 (Ζ) are narrower and reported separately. Applied exactly as the rule is written, and the two-rung spread to weakest-link Ζ is disclosed with its reason rather than hidden. (iv) Collective authors: no defect of either known form. Reference 17's first author "Secretan" is a personal surname (Béatrice Secretan, IARC), not a truncated organisation name; no leading-comma instance and no "Group", "Network" or "Consortium" fragment standing in for a consortium. (v) No rendering defects: no glued numerals, no empty DOI field, no "PMIDnot supplied" string. (vi) No spliced "The earlier adversarial approval refers to its original edition" sentence. The consensus supplement's closing line is correctly hedged — "any earlier approval" — and this document carried no earlier adversarial approval, so the sentence is true as written. (vii) The REVIEWER NOTES do not contradict the consensus-consistency supplement: the notes describe the 2026-09-30 rebuild and the supplement describes the separate 2026-10-09 consistency check, and `consensus-addendum.md` matches the embedded block. (viii) Dated facts: the never-smoker fraction, the IARC Group 1 classification and the EGFR-mutant subset are all correct as of the stated cutoff. (ix) Levels: SC1 Θ, SC2 Η, SC3 Ζ and the headline Θ follow from the quoted caps, thresholds and aggregation rule. The blocks on Ω (necessity and sufficiency both fail) and on Η (the rubric's calibration note) are correctly argued. No letter required a change. **Scorecard change.** `independence_basis` (one line's sources plus two explanatory notes) and `study_id` on the 18 `citation_ledger` and `evidence_inventory` rows. The level, confidence, subclaim levels and every other field are unchanged. Validator re-run: PASS, 0 fail, 0 warn. **Residual limitations.** Access depth across the ledger is mixed, with several rows at abstract or metadata depth; the review states this. The exact smoking-attributable fraction by histology and population remains unresolved, as the knowledge-gaps section records, and this review does not resolve it. **Approval:** Following adversarial review and the corrections listed above, the revised document is approved. - Claude Opus 5.5 (Anthropic), 2026-10-14. ## REFERENCES 1. Doll R, Hill AB. Smoking and carcinoma of the lung; preliminary report. *Br Med J*. 1950;2(4682):739–748. [PMID 14772469](https://pubmed.ncbi.nlm.nih.gov/14772469/) · [DOI 10.1136/bmj.2.4682.739](https://doi.org/10.1136/bmj.2.4682.739) 2. Wynder EL, Graham EA. Tobacco smoking as a possible etiologic factor in bronchiogenic carcinoma; a study of 684 proved cases. *J Am Med Assoc*. 1950;143(4):329–336. [PMID 15415260](https://pubmed.ncbi.nlm.nih.gov/15415260/) · [DOI 10.1001/jama.1950.02910390001001](https://doi.org/10.1001/jama.1950.02910390001001) 3. Doll R, Peto R. Mortality in relation to smoking: 20 years' observations on male British doctors. *Br Med J*. 1976;2(6051):1525–1536. [PMID 1009386](https://pubmed.ncbi.nlm.nih.gov/1009386/) · [DOI 10.1136/bmj.2.6051.1525](https://doi.org/10.1136/bmj.2.6051.1525) 4. Doll R, Peto R, Boreham J, Sutherland I. Mortality in relation to smoking: 50 years' observations on male British doctors. *BMJ*. 2004;328(7455):1519. [PMID 15213107](https://pubmed.ncbi.nlm.nih.gov/15213107/) · [DOI 10.1136/bmj.38142.554479.AE](https://doi.org/10.1136/bmj.38142.554479.AE) 5. Peto R, Darby S, Deo H, Silcocks P, Whitley E, Doll R. Smoking, smoking cessation, and lung cancer in the UK since 1950: combination of national statistics with two case-control studies. *BMJ*. 2000;321(7257):323–329. [PMID 10926586](https://pubmed.ncbi.nlm.nih.gov/10926586/) · [DOI 10.1136/bmj.321.7257.323](https://doi.org/10.1136/bmj.321.7257.323) 6. Jha P, Ramasundarahettige C, Landsman V, et al. 21st-century hazards of smoking and benefits of cessation in the United States. *N Engl J Med*. 2013;368(4):341–350. [PMID 23343063](https://pubmed.ncbi.nlm.nih.gov/23343063/) · [DOI 10.1056/NEJMsa1211128](https://doi.org/10.1056/NEJMsa1211128) 7. Thun MJ, Carter BD, Feskanich D, et al. 50-year trends in smoking-related mortality in the United States. *N Engl J Med*. 2013;368(4):351–364. [PMID 23343064](https://pubmed.ncbi.nlm.nih.gov/23343064/) · [DOI 10.1056/NEJMsa1211127](https://doi.org/10.1056/NEJMsa1211127) 8. Pesch B, Kendzia B, Gustavsson P, et al. Cigarette smoking and lung cancer — relative risk estimates for the major histological types from a pooled analysis of case-control studies. *Int J Cancer*. 2012;131(5):1210–1219. [PMID 22052329](https://pubmed.ncbi.nlm.nih.gov/22052329/) · [DOI 10.1002/ijc.27339](https://doi.org/10.1002/ijc.27339) 9. Denissenko MF, Pao A, Tang M, Pfeifer GP. Preferential formation of benzo[a]pyrene adducts at lung cancer mutational hotspots in P53. *Science*. 1996;274(5286):430–432. [PMID 8832894](https://pubmed.ncbi.nlm.nih.gov/8832894/) · [DOI 10.1126/science.274.5286.430](https://doi.org/10.1126/science.274.5286.430) 10. Alexandrov LB, Ju YS, Haase K, et al. Mutational signatures associated with tobacco smoking in human cancer. *Science*. 2016;354(6312):618–622. [PMID 27811275](https://pubmed.ncbi.nlm.nih.gov/27811275/) · [DOI 10.1126/science.aag0299](https://doi.org/10.1126/science.aag0299) 11. Hutt JA, Vuillemenot BR, Barr EB, et al. Life-span inhalation exposure to mainstream cigarette smoke induces lung cancer in B6C3F1 mice through genetic and epigenetic pathways. *Carcinogenesis*. 2005;26(11):1999–2009. [PMID 15944214](https://pubmed.ncbi.nlm.nih.gov/15944214/) · [DOI 10.1093/carcin/bgi150](https://doi.org/10.1093/carcin/bgi150) 12. Darby S, Hill D, Auvinen A, et al. Radon in homes and risk of lung cancer: collaborative analysis of individual data from 13 European case-control studies. *BMJ*. 2005;330(7485):223. [PMID 15613366](https://pubmed.ncbi.nlm.nih.gov/15613366/) · [DOI 10.1136/bmj.38308.477650.63](https://doi.org/10.1136/bmj.38308.477650.63) 13. Pao W, Miller V, Zakowski M, et al. EGF receptor gene mutations are common in lung cancers from "never smokers" and are associated with sensitivity of tumors to gefitinib and erlotinib. *Proc Natl Acad Sci U S A*. 2004;101(36):13306–13311. [PMID 15329413](https://pubmed.ncbi.nlm.nih.gov/15329413/) · [DOI 10.1073/pnas.0405220101](https://doi.org/10.1073/pnas.0405220101) 14. Sun S, Schiller JH, Gazdar AF. Lung cancer in never smokers — a different disease. *Nat Rev Cancer*. 2007;7(10):778–790. [PMID 17882278](https://pubmed.ncbi.nlm.nih.gov/17882278/) · [DOI 10.1038/nrc2190](https://doi.org/10.1038/nrc2190) 15. Samet JM, Avila-Tang E, Boffetta P, et al. Lung cancer in never smokers: clinical epidemiology and environmental risk factors. *Clin Cancer Res*. 2009;15(18):5626–5645. [PMID 19755391](https://pubmed.ncbi.nlm.nih.gov/19755391/) · [DOI 10.1158/1078-0432.CCR-09-0376](https://doi.org/10.1158/1078-0432.CCR-09-0376) 16. Hecht SS. Tobacco carcinogens, their biomarkers and tobacco-induced cancer. *Nat Rev Cancer*. 2003;3(10):733–744. [PMID 14570033](https://pubmed.ncbi.nlm.nih.gov/14570033/) · [DOI 10.1038/nrc1190](https://doi.org/10.1038/nrc1190) 17. Secretan B, Straif K, Baan R, et al. A review of human carcinogens — Part E: tobacco, areca nut, alcohol, coal smoke, and salted fish. *Lancet Oncol*. 2009;10(11):1033–1034. [PMID 19891056](https://pubmed.ncbi.nlm.nih.gov/19891056/) · [DOI 10.1016/S1470-2045(09)70326-2](https://doi.org/10.1016/S1470-2045(09)70326-2) 18. Wen J, Fu JH, Zhang W, Guo M. Lung carcinoma signaling pathways activated by smoking. *Chin J Cancer*. 2011;30(8):551–558. [PMID 21801603](https://pubmed.ncbi.nlm.nih.gov/21801603/) · [DOI 10.5732/cjc.011.10059](https://doi.org/10.5732/cjc.011.10059) ### Consensus consistency validation — 9 October 2026 **Current grade:** Θ **Grade changed:** no **Scope:** Retrospective, transparently unblinded consistency check of load-bearing primary records already in the topic inventory. This is not a new peer review and does not establish historical Consensus use. **Consensus queries:** - `tobacco smoking causal lung cancer cohort case control cessation dose response mutational mechanism primary studies` - `primary prospective British doctors smoking lung cancer Doll Peto 50 years 2004 cessation risk cohort original study` - `Doll Peto Mortality in relation to smoking 50 years observations on male British doctors 2004 BMJ original paper` **Saved raw Consensus records:** - `D:\HypothesisCritic\deliverables\consensus_consistency_audit_2026-10-09\checks\smoking-lung-carcinoma\consensus-search-01.json` - `D:\HypothesisCritic\deliverables\consensus_consistency_audit_2026-10-09\checks\smoking-lung-carcinoma\consensus-search-02.json` - `D:\HypothesisCritic\deliverables\consensus_consistency_audit_2026-10-09\checks\smoking-lung-carcinoma\consensus-search-03.json` - `D:\HypothesisCritic\deliverables\consensus_consistency_audit_2026-10-09\checks\smoking-lung-carcinoma\consensus-fetch-23f85c7fbe72578a91a1b00491caa2e4.json` **Original-primary comparisons:** - **PMID 14772469** (claims C1) (PubMed abstract_metadata in baseline inventory): Doll-Hill hospital-based case-control study reported a strong smoking/lung carcinoma association; early case-control design, later complemented by prospective evidence. - **PMID 15213107** (claims C4,C5) (Full-text complete as documented by source inventory): Fifty-year British doctors cohort: cause-specific mortality varied with smoking; cessation at younger ages avoided much excess risk; prospective follow-up. - **PMID 10926586** (claims C5) (Full-text complete as documented by source inventory): UK lung-cancer risk and cessation trends from national statistics and case-control studies; risk declines after stopping with latency. - **PMID 8832894** (claims C8) (Full-text complete as documented by source inventory): Benzo[a]pyrene adducts preferentially form at lung-cancer p53 mutation hotspots; mechanism complements but does not alone quantify attributable fraction. **Discovery/reconciliation versus source incorporation:** Broad query returned recent mutational-signature papers and reviews; targeted exact-title query retrieved Doll et al.'s 50-year British doctors cohort (PMID 15213107). Fetch of modern mechanism and second-primary-risk records; the original source ledger supports dose-response, cessation, and mutational lines. The existing phrase 'consensus and mechanistic reviews' is not tool documentation. **Conclusion:** No grade change: Θ for sustained smoking as a major initiating cause of human lung carcinoma, with never-smoker cases bounding necessity and most smokers not developing the disease bounding sufficiency. Fresh search/fetch is retrospective, not historical service-use evidence. This dated supplement adds source reconciliation to the original review. Its original narrative, registered scope, source inventory, grades and any earlier approval remain historical records; the earlier approval does not certify this supplement. New sources described here are supplemental and are not counted in the original inventory.