Causal Link of Human Papillomavirus in Barrett Esophagus and Adenocarcinoma: Are We There Yet?

Fifteen to twenty percent of all human malignancies are caused by bacterial or viral pathogens, with Helicobacter pylori, high-risk human papillomavirus (hr-HPV), hepatitis B and C viruses accounting for most pathogen-related cancers. Many are modifiable with effective treatments and/or prevention strategies. Epidemiologic and basic science studies have established hr-HPV as causal in cervical, anogenital, and some oropharyngeal cancers, with viral genomes present in the majority of cervical tumors and dysplasias. Once thought to infect only squamous epithelia, hr-HPV is now also strongly associated with cervical adenocarcinoma, indicating affinity for glandular as well as squamous tissue. Against the backdrop of a dramatic rise in esophageal adenocarcinoma (OAC) incidence since the 1970s and declining cancer risk estimates within Barrett’s esophagus (BO), the authors explore hr-HPV as a potential missing etiologic link contributing to Barrett’s dysplasia (BD) and OAC.

The authors conducted a narrative review of literature from 1960 to 2022 using PubMed, MEDLINE, Scopus, and Google Scholar (with inclusion of one seminal 1881 Koch publication identified through reference review). Search terms included: human papillomavirus, HPV, viruses, viral pathogens, oncogenic viruses, HPV integration, HPV associated cancers, esophageal adenocarcinoma, esophageal cancer, esophageal tumor, esophagus, Barrett's esophagus, Barrett's metaplasia, Barrett's dysplasia, and epidemiology. Inclusion criteria were English language (or English translation), full-text availability, and relevance to the search terms. An initial 7042 articles were identified; studies focused solely on squamous cell carcinoma, gastro-esophageal junction tumors, or cardia malignancies without separate OAC data, as well as duplicates and irrelevant publications, were excluded. Ultimately, 108 articles were included. Prior reviews and meta-analyses varied in assay types, histology classification, biopsy site selection, control use, and viral load detection challenges, factors that may explain heterogeneity in HPV prevalence estimates in OAC and BO.

Narrative review methodology: Databases searched included PubMed, MEDLINE, Scopus, and Google Scholar. Timeframe: 1960–2022 (plus one 1881 Koch reference identified by citation chasing). Search terms encompassed HPV-related oncogenesis and esophageal disease (e.g., HPV, high-risk HPV, HPV integration, esophageal adenocarcinoma, Barrett’s esophagus/dysplasia). Inclusion: English or English-translated full texts relevant to the topic. Exclusion: studies restricted to esophageal squamous cell carcinoma or gastro-esophageal junction/cardia cancers without separate OAC data; duplicates; irrelevant articles. Screening reduced 7042 records to 108 included studies. The review synthesizes epidemiologic associations, molecular markers (HPV DNA, E6/E7 mRNA, p16INK4A), genomic profiles, serology, cell line data, and clinical outcomes to evaluate hr-HPV’s potential causal role in BD/OAC under Koch’s/Hill’s/Lipkin criteria.

• Meta-analyses/Prevalence: A pooled HPV prevalence of 35% (95% CI 13.2–65.7%) in OAC was reported from 5 studies (n=174). Another review of 19 studies found 13% (95% CI 2–29%). The largest meta-analysis to date (13,401 subjects; 33 studies) concluded hr-HPV 16/18 are strong co-factors in esophageal cancer, including OAC.
• Association in Barrett’s sequence and OAC: In 2013 studies, HPV DNA was detected in 31% (81/261) overall. HPV positivity differed by group: controls 18.0%, BO 22.1%, BD 68.6%, OAC 66.7. About 93% of HPV-positive were high-risk types 16/18. p16INK4A expression was higher in BD (44.1%) and OAC (44.4%) than BO (10.6%). Among HPV DNA-positive cases tested for oncogene activity, E6/E7 mRNA was detected in 40.9% of BD (9/22) and 60% of OAC (9/15), but 0% in controls (0/16) and BO (0/13) (p<0.001). High-risk HPV concentrated at the esophageal squamo-columnar junction (transformation zone) more than within lesions.
• Viral load/integration and genotype distribution: Greater HPV16/18 viral load and integration correlated with increasing disease severity along the BO–BD–OAC sequence. HPV18 was more frequent in benign lesions, whereas HPV16 predominated in BD/OAC. HPV ISH and E6/E7 mRNA signals localized to columnar-lined lesions, sparing adjacent squamous mucosa, indicating glandular tropism.
• Molecular signatures: In a cross-sectional cohort (n=218), 56 were HPV DNA+, with biologically active virus (DNA+/RNA+) only in BD/OAC (n=21). These lesions showed p16 overexpression with reduced/absent pRb and p53. Low p53 expression had the strongest association with DNA+/RNA+ lesions (OR 23.5; p=0.0029), and sequencing confirmed wild-type TP53 in most p53-low samples. The pRb low/p53 low profile most strongly associated with active HPV BD/OAC.
• Genomics: Pilot whole-exome sequencing suggested HPV+ OACs have about half the number of non-silent somatic mutations versus HPV− OACs, absent TP53 mutations in HPV+ tumors (vs frequent in HPV−), fewer driver genes, and HPV16 integration in some cases—paralleling patterns seen in HPV-driven HNSCC.
• Treatment response/outcomes: In a prospective study of 40 patients undergoing endoscopic therapy for BD/early OAC, persistent biologically active hr-HPV and p53 overexpression were mutually exclusive markers of ablation failure; HPV+ BD/IMC-OAC largely lacked TP53 mutations (wild-type TP53). Survival analysis in 142 patients with HGD/OAC showed superior DFS in HPV-positive vs HPV-negative (40.3 vs 24.1 months; p=0.003) and improved OS (43.7 vs 29.8 months; p=0.009).
• Serology: In 438 patients across the GERD–BO–BD–IMC spectrum, overall HPV seropositivity rates did not differ between cases and controls; however, among HPV16 DNA+ BD/OAC, anti-HPV16 E7 antibodies were more prevalent (11.5% vs 1.5% in controls), and among HPV18 DNA+ cases, anti-HPV18 E1 antibodies were higher (50% vs 1.5%; p=0.0002).
• Cell line data: HPV18 has been integrated in esophageal cancer cell lines with expression of E6/E7 and disruption of E2; in esophageal lines E7 preferentially targeted p130 over pRb, suggesting tissue-specific interactions.
• Causality framework: Evidence supports a probable relationship between hr-HPV and a subset of OAC when evaluated via Hill’s criteria/Lipkin’s framework: strength/consistency of association in subsets, temporality and biological gradient (viral load/integration increasing with severity), biological plausibility (transformation zone tropism), coherence/analogy (cervix/oropharynx), and supportive experimental/serologic markers.

The review addresses whether hr-HPV contributes causally to a subset of Barrett’s-related esophageal adenocarcinomas. OAC incidence has risen dramatically despite declining per-year cancer risk estimates within Barrett’s esophagus, implying additional risk factors. The compiled evidence aligns with a probable causal role for hr-HPV in a subset of BD/OAC: (1) epidemiologic associations and meta-analyses detect hr-HPV DNA in a proportion of OAC; (2) biologically active infections (E6/E7 mRNA+, p16 overexpression) localize to the esophageal transformation zone and dysplastic/adenocarcinoma tissues; (3) higher viral load and integration correlate with advancing disease severity; (4) HPV+ tumors exhibit molecular hallmarks seen in other HPV-driven cancers (wild-type TP53, pRb pathway aberrations, fewer genome-wide alterations), suggesting a distinct pathogenetic pathway; and (5) HPV positivity associates with improved disease-free and overall survival, mirroring prognostic patterns in HPV-driven head and neck cancers. These findings collectively address plausibility, biological gradient, consistency, specificity for site/tissue type, and analogy criteria. However, heterogeneity in detection methods, sampling, and populations, as well as long latency and multifactorial etiology, complicate definitive causality. Clinically, recognizing an HPV-driven OAC subset could enable risk stratification from BO to BD/OAC, inform surveillance priorities, personalize therapy (including potential de-escalation for HPV+ disease), and motivate preventive strategies (HPV vaccination, targeted interventions at the squamo-columnar junction).

Accumulating data support a probable relationship between high-risk HPV and a subset of Barrett’s dysplasia and esophageal adenocarcinoma. While IARC (2012) judged evidence for HPV in esophageal cancer inadequate, subsequent meta-analyses and molecular studies bolster hr-HPV (types 16/18) as a risk factor/co-factor. Further replication in larger, well-controlled studies is essential, including comprehensive genomic and proteomic profiling to define distinct molecular signatures of HPV+ OAC. Validating these signatures could refine risk stratification, guide surveillance for high-risk progressors, and enable modified treatment protocols with potential toxicity reductions. Preventive strategies, including vaccination, may reduce incidence, and therapeutic HPV vaccination could conceivably promote regression in a subset of BD. Translation to practice awaits corroboration by independent groups and prioritized evaluation by guideline bodies.

• Narrative review design; potential selection and publication biases.
• Heterogeneity across included studies in assays (nested vs non-nested PCR), histologic classification, biopsy site (often not from the squamo-columnar junction), control selection, and lack of standardized thresholds—factors likely contributing to variable HPV prevalence estimates.
• Low HPV viral load in esophageal tissues complicates detection and may yield false negatives.
• Limited sample sizes in several studies; paucity of large, multi-center, prospective cohorts with site-specific correlation of tissue HPV status and serology.
• Conflicting seroepidemiologic data and limited sensitivity/specificity of some antibody markers outside established HPV-driven sites.
• Genomic findings based on small pilot WES cohorts; need for larger whole-genome and proteomic validation to confirm distinct HPV+ OAC molecular landscapes.
• Multifactorial etiology and long latency of OAC limit ability to demonstrate temporality and fulfill stringent causality criteria; evidence is strongest for a subset rather than all OAC.