Pretreatment feeding‑stoma placement in advanced esophageal cancer: defining optimal patient selection criteria.

BACKGROUND

The indication of feeding-stoma creation (gastrostomy/jejunostomy) in esophageal squamous cell carcinoma (ESCC) patients and its impact on survival remain unclear. We aimed to identify patient subgroups who benefit from feeding-stoma placement before concurrent chemoradiation therapy (CCRT).

METHODS

We did a secondary analysis of prospective cohort. A total of 260 patients with advanced ESCC who underwent CCRT between April 2008 and March 2024 were included. Tumor-occupying proportion (tumor area/total lumen area) and other tumor characteristics were measured from standardized endoscopic images. Characteristics that predict post-CCRT cachexia was identified and validated as indications of feeding-stoma creation.

RESULTS

Post-CCRT cachexia developed in 60.7% of patients. Independent predictors of cachexia included longer tumor length and bigger tumor-occupying proportion. ROC curve found tumor length ≥ 6 cm (area under ROC (AUC): 0.760, P < 0.001) and tumor-occupying proportion ≥ 70% are optimal to predict post-CCRT cachexia (AUC: 0.620, P = 0.001). Sixty-two patients (23.8%) underwent pretreatment feeding-stoma creation. In the overall cohort, feeding-stoma creation was not associated with a longer progression-free survival (PFS) or overall survival (OS). In patients with tumor-occupying proportion ≥ 70%, feeding-stoma creation was associated with reduced cachexia incidence, better treatment response (OR: 2.78, P = 0.028), longer median PFS (6 vs. 4 months, P = 0.012) and OS (11 vs. 9 months, P = 0.009). In contrast, feeding-stoma creation was not associated with an improved treatment response or survival in patients with tumor-occupying proportion < 70% or in patients with different tumor length.

CONCLUSION

Tumor-occupying proportion is a novel metric that predicts cachexia and survival in advanced ESCC. Pretreatment feeding-stoma creation was associated with a better treatment response and longer survival only in patients with tumor-occupying proportion ≥ 70%, supporting its use as a selection criterion for this invasive procedure.