"Squamous cell carcinoma" without a qualifier most often means skin cancer — this page is about the lung. A three-stage Han Chinese study found one variant that lowers risk specifically for this histotype; a much larger 2017 meta-analysis then added five more, four of which cluster in one region on chromosome 6.
Squamous cell carcinoma occurs in several organs — skin, lung, cervix, oesophagus — and each site is genetically its own disease, driven by different exposures and different biology, sharing only a cell type under the microscope. Said without a qualifier, "squamous cell carcinoma" most often refers to the skin cancer. This page is not about that — it is specifically about squamous cell carcinoma of the lung, one of the two major types of non-small-cell lung cancer alongside adenocarcinoma, and strongly linked to smoking.
Dong et al. 2013 ran a genome-wide association study in Han Chinese populations across three stages: 833 cases and 3,094 controls in discovery, 2,223 cases and 6,409 controls in replication. Then, to test whether the finding was specific to this histotype, they compared it against 4,368 lung adenocarcinoma cases and 9,486 controls.
The variant, rs12296850, sits in the SLC17A8-NR1H4 gene region at 12q23.1 and was significantly associated with lung squamous cell carcinoma risk — with a protective effect for people carrying the variant allele. Critically, the same variant showed no association with lung adenocarcinoma. That histotype-specific result is the paper's central point: squamous cell and adenocarcinoma lung cancers have at least partly distinct genetic determinants, not one shared "lung cancer" genetics.
McKay et al. 2017 combined OncoArray genotyping of 14,803 cases and 12,262 controls with existing data for an aggregated analysis of 29,266 cases and 56,450 controls overall, finding 18 genome-wide significant loci across lung cancer histologies. Five of those are specifically associated with the squamous subtype and are on this site's catalogue.
Checked one by one against their actual chromosome position rather than treated as five separate genes: four of the five sit within about 2.4 million bases of each other on chromosome 6 (roughly 30.4 to 32.8 Mb) — TRIM39-RPP21, VARS2, NOTCH4 and HLA-DQB2 — squarely inside the extended MHC/HLA region. That region turns up repeatedly across this site's catalogue (it is also schizophrenia's own strongest signal) because it carries exceptionally dense linkage disequilibrium: one real association there routinely tags several nearby genes as separate statistical "hits." Treated here as most plausibly one regional signal rather than four independent ones.
The fifth, CLPTM1L, sits at 5p15.33 — nowhere near chromosome 6 — and is a genuinely separate, well-established locus. It sits next to TERT, the gene that maintains chromosome ends, and this same regional signal has been reported for several other cancers in the wider literature, not just lung squamous cell carcinoma.
2025-11-21 · Integrating Human Proteomes with Genome-Wide Association Data Reveals Prioritized Therapeutic Candidates for Lung Squamous Cell Carcinoma. Biology. 2025. DOI:10.3390/biology14121640
Lung squamous cell carcinoma (LUSC), the second most common lung cancer type, has poorly characterized molecular drivers relative to its disease burden. This study combined plasma proteome data from 54,219 people with LUSC GWAS data (7,426 cases, 55,627 controls), running genome-wide Mendelian randomization and colocalization to find proteins causally linked to LUSC risk rather than merely correlated with it. After Bonferroni correction, sensitivity analyses and reverse-causation checks, it identified 12 potential druggable proteins, with 5 (DOK2, FKBPL, NCF2, PDIA3, TCL1A) showing strong colocalization evidence -- meaning the genetic signal for the protein and for LUSC risk are very likely driven by the same underlying variant, not two separate coincidental signals. Stratifying all 12 by colocalization strength, differential expression, protein-protein interaction networks and druggability, the study named DOK2 as its single Tier 1 target, with FKBPL, NCF2, AXL and PDIA3 as Tier 2. It also identified 14 modifiable risk factors (lifestyle/environmental) that mediate how these proteins affect LUSC risk, pointing to concrete prevention angles alongside the drug-target angle. This site's lung squamous cell carcinoma page carries 8 variants, distinct from the general lung cancer page; none of DOK2, FKBPL, NCF2, PDIA3 or AXL are currently among them.
Smoking is the dominant risk factor for lung squamous cell carcinoma by a wide margin. These variants shift risk slightly on top of that. None of this is a reason to change smoking-cessation advice or lung cancer screening eligibility, both of which are driven by smoking history and age, not genotype.
The Dong et al. specificity finding — protective for squamous cell carcinoma, neutral for adenocarcinoma — is a reminder that "lung cancer" is not one disease genetically, even though both types share the organ and often the same risk factor (smoking).
The original finding (rs12296850) comes from Han Chinese cohorts. The 2017 expansion is predominantly European-ancestry. Genetic associations found in one population do not always replicate with the same effect size in another, and this page does not claim otherwise — the six variants here come from two separately-run, separately-replicated studies rather than one combined analysis.
Squamous cell carcinoma at other sites — skin, for instance — is a genetically distinct disease and is catalogued separately where evidence supports it. Sharing a cell type under the microscope does not mean sharing genetics.
What a 23andMe/AncestryDNA export or raw VCF can and can't tell you about Lung Squamous Cell Carcinoma comes down to these specific, well-studied positions — not a diagnosis.
The studies behind these variants recruited participants from different ancestries — a result found in one population doesn't always transfer to another. Based on 8 of 8 linked studies with a resolved discovery ancestry.
Databases, guidelines and references
No. "Squamous cell carcinoma" without a qualifier usually refers to skin cancer, which is a genetically distinct disease. This page is specifically about squamous cell carcinoma of the lung.
No. Smoking remains the dominant risk factor for this cancer by a wide margin. This variant shifts risk slightly on top of that in one population, and does not change smoking-cessation advice or screening guidelines, which are driven by smoking history and age.
To test whether the finding was specific to this histotype rather than a general "lung cancer" effect. The variant was associated with squamous cell carcinoma risk but not adenocarcinoma risk — evidence that the two lung cancer types have at least partly distinct genetics.
One of the two was. The 2013 discovery of rs12296850 used Han Chinese populations specifically, across three stages totalling more than 12,000 people. The 2017 study that added five more loci was predominantly European-ancestry. Whether either set of findings holds with the same effect size in the other population is not established by either study alone.
Because four of the five variants sit within one crowded region on chromosome 6 — the MHC/HLA region, which turns up repeatedly across this site's catalogue because of its unusually dense linkage disequilibrium. One real underlying association there routinely gets tagged by several nearby genes at once, so this page treats those four as most plausibly one regional signal rather than four independent discoveries. The fifth, CLPTM1L at 5p15.33, is a separate, well-established locus.
Free to reuse. This page's text is original writing from freely-available research, licensed CC BY 4.0 — reuse it, including commercially, with attribution to MyGeneLog™. It's general research-derived information, not medical advice or a diagnosis — see Terms of Use.