This pilot study quantified serum expression of 9 candidate microRNAs in 25 senior dogs (ages 7–17 years) classified by cachexia status (n=12 cachexia, n=13 normal) and cancer status (n=14 cancer, n=11 noncancer), using dogs as a comparative oncology model for human cachexia biomarker discovery.
Four miRNAs (miR-15a, miR-15b, miR-16, miR-140) were significantly downregulated in cachexia vs. normal dogs. miR-16 was the strongest single biomarker for cachexia (AUC=0.899; 95% CI: 0.763–1.000; sensitivity 91.7%, specificity 83.3%). A 4-miRNA panel reached AUC=0.923 but with overlapping CIs vs. miR-16 alone. In female dogs only, miR-140 was significantly lower in cancer vs. noncancer controls (P=0.045; corrected AUC=0.667).
- Very small sample (n=25), with low events-per-variable ratios in logistic regression models, risking overfitting and wide confidence intervals. - Cachexia and cancer groups overlap substantially (75% of cachectic dogs had cancer), making it hard to separate cachexia-specific from cancer-specific miRNA signals. - Retrospective, cross-sectional, single-center design with predominantly small-breed dogs limits generalizability and prevents causal inference.
Circulating miR-16 shows strong diagnostic potential for canine cachexia (AUC=0.899), and miR-140 may serve as a sex-specific signal in female dogs with hormone-sensitive cancers, but both findings require validation in larger, prospective cohorts before any clinical use. Clinicians using dogs as translational models for human cachexia research should note that miRNA direction of change (up vs. down) differs between cachexia and sarcopenia, highlighting the need for disease-context-specific biomarker interpretation.
Explore related topics
