Abstract / Summary
Background: Nasopharyngeal carcinoma (NPC) may be accompanied by systemic inflammatory changes, but routine inflammatory markers are nonspecific and have not been validated as diagnostic tests for NPC. Methods: This exploratory retrospective comparison screened 178 adults and included 72 eligible records (36 histopathologically confirmed NPC and 36 histologically confirmed benign adenoid-associated pathology); 106 records were excluded because of incomplete endoscopic or laboratory records (n = 41), hematological malignancy (n = 15), severe hepatic failure (n = 11), or a documented systemic inflammatory condition unrelated to the nasopharyngeal presentation (n = 39). CRP, erythrocyte sedimentation rate (ESR), leukocyte count, selected biochemical variables, and an investigator-derived loco-regional clinical/endoscopic severity score were compared. Prespecified sensitivity analyses included age- and sex-adjusted rank regression, restriction to the overlapping age range (39–52 years), group-stratified correlations, and partial rank correlations adjusted for diagnostic group, age, and sex. Results: The NPC group was younger than the benign group (38.1 ± 7.6 vs. 56.1 ± 8.3 years; p < 0.001). CRP, ESR, leukocyte count, uric acid, and the loco-regional score were higher in the NPC group (all p < 0.01). The principal differences persisted in the overlapping-age subset (NPC n = 17; benign n = 16; all p < 0.001) and in age- and sex-adjusted rank analyses. Pooled correlations were attenuated or changed after stratification and covariate adjustment, indicating that part of the pooled association reflected between-group separation. Exploratory ROC AUCs were 1.000 for CRP, leukocyte count, and the loco-regional score and 0.948 for ESR; these values were derived and evaluated in the same selected cohort. In separate Firth models adjusted for age and sex, CRP, leukocyte count, and the loco-regional score remained associated with NPC, whereas ESR did not (p = 0.200). Conclusions: The study identifies marked between-group differences in this selected retrospective cohort but does not establish diagnostic biomarkers or clinical cut-offs. Independent prospective evaluation, standardized sampling, and fuller control of inflammatory and tumor-related confounding are required.