Enteric tube feeding is recommended for patients with dysphagia to achieve adequate nutritional supplementation and to prevent complications such as aspiration pneumonia. In stroke patients, clinical outcomes may be better if feeding is started earlier. In the short term, nasogastric tube (NGT) feeding is acceptable, and there is no clear advantage of percutaneous endoscopic gastrostomy (PEG) over NGT feeding. However, complications including esophageal ulcer, stricture, aspiration pneumonia, and sinusitis are increased if NGT feeding continues for more than 4 weeks. Compared with NGT feeding, PEG feeding reduced treatment failures and gastrointestinal bleeding, and achieved higher feed delivery and albumin concentration [1]. Thus, PEG feeding can be recommended for dysphagia patients who require long-term nutritional support.

Technically, PEG placement is simple and safe [2]. However, it has a risk of procedure-related death and complications including bleeding, aspiration pneumonia, perforation of the aerodigestive tract, immediate or delayed gastrostomy site infection, and colocutaneous fistula formation [3]. In addition, despite the increasing number of older patients, it remains unclear whether and when PEG placement should be performed [4]. In-hospital mortality is reportedly higher in older patients, especially in those over 75 years old [5,6]. Therefore, PEG placement should be performed selectively in patients with dysphagia, according to their life expectancy. Future studies are warranted on the safety and effectiveness of PEG placement based on comorbidity, age, sex, and gastrostomy technique [7].

In this issue of Clinical Endoscopy, Sobani et al. investigated the safety of PEG tube placement in patients aged 100 years or older [8]. They reported that the success rate and in-hospital mortality were comparable in centenarian and younger patients (p>0.05), although minor complication rates were significantly higher in the centenarian patients (13.3% vs. 2.9%, p=0.022). The authors concluded that PEG tube placement may be safely attempted in carefully selected patients in this subset of the population. The study has limitations as it was a retrospective single-center study, and selection bias might be unavoidable. Nevertheless, this is the first study regarding this issue, and more studies are warranted in the future.