Clusters of Anthropometric Features in Colorectal Cancer Patients with Synchronous Metastases and Their Association with Overall Survival and RAS Mutation.

BACKGROUND

Anthropometric traits—such as height, body mass index (BMI), and age—are shaped by genetic, nutritional, and environmental factors, which may also be influential in colorectal cancer (CRC). We investigated whether sex-adjusted variations in these parameters could be associated with different subtypes of CRC and overall survival outcomes.

METHODS

We analyzed patients with synchronous metastatic CRC enrolled in the prospective BMI-QoL study (NCT03873064). Age, height, and BMI were standardized and subjected to k-means clustering (k = 2, 25 random initializations). Principal component analysis was used to visualize cluster separation in two dimensions. “Core cluster members” were defined using sex-specific cutoffs for all three variables, excluding borderline cases. Difference in Overall survival (OS) between clusters was assessed using Kaplan–Meier and Cox proportional hazards models; clinical and molecular features were compared using chi-square and Mann–Whitney tests.

RESULTS

Two reproducible clusters emerged. Cluster 1 comprised younger, taller, normal-to-overweight patients; Cluster 2 included older, shorter, underweight patients. Core cluster 1 members (n = 23) had significantly longer OS than core cluster 2 members (n = 31) (median OS 51.1 vs. 22.0 months; HR 2.28, p = 0.03), with survival curves diverging beyond 24 months. Non-core members had intermediate OS (median 31 months). Age, height, and BMI individually were not significant OS predictors. RAS wild-type tumors were more prevalent in core cluster 1 (61%) than core cluster 2 (32%, p = 0.03). CEA levels and metastatic patterns did not differ between clusters.

CONCLUSION

Sex-adjusted anthropometric profiling identifies two prognostically distinct mCRC subtypes. The association of better survival with higher BMI, younger age and taller stature—alongside increased prevalence of RAS wild-type tumors—supports the biological relevance of these profiles. Larger prospective validation integrating anthropometric and genomic data is warranted.