DISCUSSION
The IGB positioning and retrieval were safe and mortality was absent. Early removal occurred in6.10% of the patients, mainly due to intolerance , similar to what was observed by Lopez-Nava et al (2011) [10]. Balloon leakage was observed in 32 patients (0.54%) and production of gas enlarging the balloon was observed in 12 (0.20%), both complications were rapidly diagnosed and IGBs were substituted, allowing the patients to continue the treatment. One of the patients with early IGB removal developed Wernick Korsacov Syndrome, as vomits continued for a long time, after the IGB removal and developed a bulimic syndrome that was not rapidly diagnosed because the patient abandoned the treatment with the multidisciplinary team. It is noteworthy that even in cases of hyperemesis, Wernick Korsakov Syndrome is a rare complication that can be prevented early [11].
In this series of cases there was just one esophagus perforation during balloon removal. This happened in a patient that had a high degree of intolerance after balloon implant, with a high incidence of vomiting episodes leading to intense esophagitis and fragility of the esophagus wall. There was four cases of gastric perforation that were successfully treated by endoscopy after IGB retrieval using Boston Resolution® hemoclips to close the perforation.
Weight loss has traditionally been the main outcome in IGB treatment and in our study there was a significant weight loss in all degrees of overweight. Considering the total group of patients, the mean weight loss was 19.13 kg, %TWL was 18.42, % EWL was 65.66 and BMI decreased 6.85 kg/m2. These results are similar to those observed by Nunes et al (2017) [12] (%TWL of 18.9 and decrease in BMI of 6.76 kg/m2) and are slightly superior to those observed in several other studies [13-20] and in two meta-analyses [21, 22]. This greater decrease in body weight could be partly attributed to the balloon volume. Roman et al (2004) [23], comparing patients that used IGBs of 500mL and 600mL showed a better result in weight loss with higher volume and in the present study were used volumes of 600-700mL, different from others that used IGB volumes starting at 400mL [13] or 500mL [14, 15].
The percentage of patients that did not achieve success in the treatment according to a %TWL > 10% was low in the present study (only 15%), being lower than the 31% observed in the study conducted by Stimac et al (2011) [24]. While according to a %EWL > 25% only 7% did not achieve success in IGB treatment, ranging from 1% in the pre-obesity group to 14% in grade III obesity, being below than 20-40% suggested by Dumonceau (2008) [25].
In the comparative analyses of participants according to the degree of overweight, we observed the greatest % EWL in subjects presenting pre-obesity (131%). Few studies compared the %EWL across individuals with different levels of BMI at baseline. Fernandes et al (2016) [26] also observed a significantly higher %EWL in the group with lower baseline BMI (26 – 30 kg/m2) compared to the group with higher BMI (> 30 kg/m2) (96% vs. 41%, respectively). Similarly, Nunes et al. (2017) [12] and Al-Sabah et al. (2016) [18] observed higher %EWL in the group of individuals with lower BMI at baseline. Although %EWL was greater in the subjects presenting pre-obesity, the %TWL was similar across the different degrees of overweight in our participants. This finding is similar to the obtained by Nunes et al (2017) [12] and may suggest that the greater %EWL in individuals presenting lower initial BMI is attributed to the lowest excess weight observed in these patients at the beginning of the study.
In the comparative analyses of participants according to gender, the %EWL was greater in women than in men, but the %TWL was similar. In our opinion the greater %EWL in women may also be attributed to the lower excess weight at baseline. Al-Sabah et al (2016) [18] also observed higher %EWL in women, but they did not compared the %TWL.