5Effect of gastric electrical stimulation on cerebral blood flow in patients with gastroparesis

Introduction: Gastric electrical stimulation is a safe and efficient treatment option in medically refractory gastroparesis. Several controlled and open label trials have shown the effectiveness of gastric electrical stimulation in symptom palliation for patients with gastroparesis. Initial settings were established based on experiences in the animal model. We reviewed our experience using higher stimulator settings on symptom reduction after implantation of gastric electrical stimulators. Methods: 42 consecutive patients (38 females and 4 males, mean age 43 ± 1.8 years) with medically refractory gastroparesis were followed after implantation of gastric electrical stimulator between 1999 and 2005. The 42 patients included 29 patients with idiopathic gastroparesis and 13 patients with diabetic gastroparesis Baseline stimulation settings were 14 Hz, 5 mAmps, 330 micro‐sec pulse width, 0.1 sec on, and 5 sec off. At each follow‐up gastrointestinal symptom scores were assessed. If symptoms were unsatisfactory, stimulator settings were increased according to a standardized protocol. Dose response relationships were evaluated. Results: After a mean follow‐up of 19 ± 2.6 months, 83% of patients had higher settings (p < 0.001, Fisher's exact test). Solid gastric emptying time during this period decreased from 195 ± 45 to 106 ± 38 minutes (T1/2 value, p < 0.05, paired t‐test) along with an average weight change from 64 ± 5.3 to 62 ± 5.6 kg (p > 0.05, paired t‐test). There were 258 stimulator interrogations with a stepwise improvement in total symptom score (R² = 0.08, p < 0.001). Multivariate regression analysis showed that frequency of electric stimulation was independently associated with improvement of total symptom score (R² = 0.06, p < 0.02). A prolonged cycle on time correlated with worsening of nausea score (R² = 0.09, p < 0.02) and vomiting score (R² = 0.13, p < 0.001). A measurement of high impedance independently correlated with a low vomiting score (R² = 0.19, p < 0.0001). Though not statistically significant, most patients’ symptoms were optimized at mean current of 10 mAmps, frequency of 55 Hz, pulse width of 330 micro‐sec, cycle on time of 0.1 sec, and cycle off time of 1 sec. Conclusion: The majority of patients seem to benefit from higher than baseline settings. Optimal settings of the implanted device must be determined on an individual basis, but based upon experience in humans seem to be 10 mAmps, 55 Hz, 0.1 sec on, 0.4 to 1 sec off.