A 76-year-old man with type 2 diabetes presented with melaena and anaemia. Endoscopy found a hard 5cm bezoar with gastric ulceration. The bezoar was too hard to fragment during endoscopy so Coca-Cola was sprayed on it and prescribed orally for a week with rabeprazole. On re-endoscopy, the fragmented bezoar was removed. The man had been eating persimmons.1 Bezoars are rare solid accumulations of foreign material that are usually found in the stomach. They include phytobezoars from plant fibres, especially persimmons (Sharon fruits), medications, milk and foreign bodies, such as coins or hair (trichobezoar). In past centuries, animal bezoars were valued as a universal antidote (sadly ineffective) for poisons. In Malaysia, healers still use bezoars from porcupines.2 Bezoars most often form in patients with previous gastric surgery, ulceration or gastric retention, especially in those with diabetes. Among 3247 endoscopies performed over seven years in one US area, 14 people had bezoars among whom seven had diabetes (six with autonomic neuropathy). In the control group without bezoars, 4% had diabetes.3 Glucagon-like peptide 1 (GLP-1) receptor agonists slow gastric emptying. Of 100 people who underwent oesophagogastroduodenoscopy (OGD) before sleeve gastrectomy, four of 23 (17.4%) on GLP-1 receptor agonist treatment had gastric bezoars compared with none in 77 not using this medication.4 Coca-Cola (which contains phosphoric acid)5 has been used to treat bezoars by drinking it, nasogastric lavage and administering it into the bezoar6, 7 whilst controlling blood glucose levels. Diet Coke has also been used effectively.8 Autonomic neuropathy delays gastric emptying causing gastropathy or gastroparesis. The stomach is not just a food bag, different areas of the stomach have different functions. Food enters at the top where food is stored and decontaminated and where liquids are managed. In the body and lower part of the stomach, the antrum, food is mashed and propelled to the pylorus which acts as a sieve and closure, releasing food into the duodenum. Food in the duodenum triggers the feedback hormone system controlling gastric emptying. The stomach secretes about 1.5L of gastric acid a day with mucus and proteindigesting enzymes to break down food and kill bacteria. All these processes require full autonomic function.9 Symptoms of gastropathy include early fullness, lack of appetite, bloating, heartburn, regurgitation, epigastric pain, nausea, vomiting partly digested food (sometimes smelling feculent) and weight loss. A Swedish study found that patients with any of these symptoms were significantly more likely to have gastroparesis than asymptomatic patients. If the stomach cannot empty normally, glucose may not arrive in the bloodstream as expected so it is difficult to match with insulin. Oral glucose may not resolve hypoglycaemia as expected. Medication may be retained and released unpredictably. Comparing 53 people with gastroparesis with 30 people with normal gastric emptying, those with gastroparesis had a blunted glucose response to a standard test meal and a smaller area under the glucose curve. There was no consistent difference in the timing of peak blood glucose levels.10 Gastroparesis reduces quality of life causing depression and anxiety.11 Diagnostic tests include OGD, scintigraphy, ultrasound, gastric manometry, electrogastrography and smart pills. Gastric emptying tests are not usually required if there is clear evidence of retained food after fasting. Exclude other causes of gastroparesis, such as other endocrinopathies, renal impairment, medication, previous gastric surgery, gastritis, infections, other forms of nerve damage and depression.9 The treatment of gastropathy includes good fluid intake, frequent small, low residue, low fat meals, stopping drugs that delay gastric emptying (such as GLP-1 receptor agonists, calcium channel blockers or anticholinergics), safely controlling hyperglycaemia, dental check and psychological support.9 Among 45 people with gastroparesis treated with CSII plus CGMS, glucose control improved with reduced HbA1c and less hypoglycaemia or severe hyperglycaemia. Gastrointestinal symptoms, nutrition and quality of life also improved. ‘Erratic blood glucose levels can be a cause and consequence of delayed gastric emptying in patients with diabetes.’12 Drug treatment includes metoclopramide, domperidone or erythromycin (beware adverse effects). Multiple other drugs have been tried. In severe cases, nasogastric or nasojejunal tubes, electrical gastric stimulation or pacing is needed.9 Check for autonomic neuropathy elsewhere in people with diabetic gastroparesis (eg for cardiac autonomic neuropapthy (CAN), with ECG R-R ratio tests, and assess postural hypotension).13 Plan routine surgery carefully in people with diabetes (see UK guidelines).14 Note autonomic neuropathy; people with CAN are at risk of severe cardiovascular complications perioperatively (a third needed vasopressors in one study).15 Fasting before general anaesthesia will not empty a gastroparetic stomach, increasing the risk of pulmonary aspiration. A recent meta-analysis concluded that ‘patients with diabetes are at an increased risk of presenting with high-risk stomach. Point-of-care ultrasonography is an invaluable tool that can help formulate a safe and effective anaesthetic plan for patients with diabetes.’16 In 1909 in London, a 49 year old man was admitted with weakness, vomiting and diarrhoea. A year earlier, the patient was a jury member in an inquest on a decomposed body that was washed ashore. When he viewed the corpse, the smell caused immediate vomiting and diarrhoea. The patient felt that he had never recovered. On assessment, he was anaemic, and a stomach tube revealed no acid. Pernicious anaemia was diagnosed. The man was treated with a diet that included eggs, fish and meat and arsenic with dilute hydrochloric acid. His anaemia was controlled, and his condition improved. The patient reported feeling well. However, five months later, the patient was readmitted, anaemic, weak and exhausted, he also had thirst, polyuria and weight loss. Glycosuria confirmed diabetes. No treatment was possible in those days. The patient's breath began to smell of acetone, his condition deteriorated and he died five days later.17 Pernicious anaemia and type 1 diabetes are both autoimmune conditions. Among people with type 1 diabetes, 5% to 10% have autoimmune gastritis and/or pernicious anaemia.18 Autoimmune gastritis impairs gastric acid and intrinsic factor production, preventing vitamin B12 absorption and causing megaloblastic anaemia. Vitamin B12 deficiency can also cause neuropathy, confusion and cognitive dysfunction, sometimes without anaemia. Metformin also causes vitamin B12 deficiency. In Ireland, 77 of 286 people (27%) with type 2 diabetes had vitamin B12 deficiency. Among those taking metformin 33.0% had vitamin B12 deficiency, compared with 7.5% of people not taking metformin.19 A 68-year-old man with no major past illness was admitted to hospital with a blood glucose level of 56.9mmol/L (1025mg/dL), a haemoglobin of 13.8g/dL, and ketoacidosis, which was treated, with apparent recovery. The next morning, the patient suddenly collapsed with a massive melaena. Emergency OGD revealed a bleeding gastric ulcer; haemostasis was achieved. Although the patient had no abdominal symptoms, in retrospect, his admission CT scan showed a high density area in the stomach.20 An OGD study in 72 people with type 2 diabetes with few or no dyspeptic symptoms found gastric inflammation in 34 patients (gastric ulcer in 10 and gastric erosion in 21).21 Among 7232 Danish people with bleeding or perforated peptic ulcers, 730 (10.1%) had diabetes among whom the 30-day mortality of bleeding peptic ulcers was 16.6%. For perforated peptic ulcers, the 30-day mortality rates were 42.9% in patients with diabetes and 24.0% in patients without diabetes.22 Studies on the prevalence of peptic ulcer or Helicobacter in people with diabetes have reported inconsistent results.23, 24 However, a large study found that people with type 2 diabetes were more likely to have bleeding peptic ulcers (HR 1.44; 1.11–1.86) than those without diabetes when controlling for multiple confounders.25
No takes yet. Share an insight, caveat, or question.
Rowan Hillson (2024) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: