Jejunal feeding for Diabetic Gastroparesis (J4G study): The effects of a gastric meal with and without prior jejunal feeding on gastric function and symptoms in patients with type I diabetes with gastro-enteropathy, diabetic controls and healthy volunteers
Endocrinology, Diabetology, Metabolism
Final Report Abstract
Symptoms compatible with diabetic gastroparesis (DG) affect up to 20% patients with type I DM. Impaired gastric function is thought to be the cause of nausea, vomiting, abdominal pain and impaired glycaemic control. DG does not respond reliably to intensive insulin regimes or prokinetic medications. Jejunal nutrition (JN) is an option in patients that cannot maintain their weight. The benefits are thought to be improved nutrition and glycaemia; however, we have observed that some DG patients eat normally during and after JN. We propose that DG represents a failure of oral nutrition to “switch” the stomach from the fasted to the fed state with nutrition delivered direct to the jejunum triggering neuro-hormonal mechanisms that induces normal gastric function. The study tests the hypothesis that JN prior to a meal improves postprandial symptoms (1° outcome) and gastric function. Diabetic patients with severe symptoms (gastroparesis cardinal symptom index (GCSI) >27), diabetic controls (GCSI <14) and healthy controls (HC) entered a randomized, double blind, controlled trial. An insulin/glucose infusion-controlled glycaemia. A NJ feeding tube was placed at endoscopy with biopsies from the stomach and duodenum. Either liquid nutrient (2kcal/min) or water was infused for 60min. The Nottingham Test Meal1 was then ingested (NTM liquid: 400mL, 300kcal; solid: 12 non‑nutrient agar beads). Symptoms were documented (VAS), gastric function by MRI and the GI‑peptide response was monitored over 120min. Mixed model analysis compared response to intervention and between groups. 9 DG patients, 9 diabetic and 12 HC were recruited. There was no difference in sex, age, weight, medical history or endoscopic findings (including histology) between groups. DG patients had more psychiatric co‑morbidity and reported higher satiety, bloating and pain after NTM ingestion than diabetic and healthy controls (p<0.05). Sensations were not affected by JN in the controls; however, fullness, bloating and pain were reduced by JN in DG patients (p<0.05). Compared to water, JN induced a greater GI‑peptide response (e.g. PP, GLP‑1) and initial liquid GE was slower (gastric content volume after meal: GCV0 31±13mL higher, p = 0.019). Subsequent liquid GE was similar in both study conditions (p=0.727). Antral contraction wave (ACW) frequency was 2.7/min in health and was highest in diabetic controls (3.1/min). Solid GE was more rapid after JN than water (2 (1 to 3) beads emptied @60min) and, again, was highest in diabetic controls (3 (1 to 7) beads emptied @60min). Numerically the GI‑peptide response was less pronounced in both diabetic groups than HC; however, the difference was not significant and a correlation with postprandial symptoms or gastric function was not identified. This clinical RCT demonstrates beneficial effects of prior jejunal nutrition (JN) on fullness, bloating and pain after a 400mL test meal in diabetic patients with moderate‑severe symptoms compatible with gastroparesis (GCSI >27). Additionally, solid GE was accelerated after JN; however, this effect was not limited to DG patients and, thus, the treatment effect that improved symptoms appeared to be on visceral sensitivity and not motor function. Future studies will identify patients likely to benefit from this novel approach to treatment.
