Our experience at the Cleveland Clinic Florida included 13 cases of revisional LSG from 2005 to 2009 [4]. The mean operative time was 120 min (range 85–180 min) and mean hospital stay was 5.5 days (range 2–20 days). According to other series with over 400 cases, the mean operative time was 90–140 min and the mean hospital stay was 1–6 days, which were comparable to those of the primary LSG and shorter than those of the RYGB after failed GB [1, 3, 4, 12, 13, 15, 17–25]. Almost all cases were completed laparoscopically, with only 6 out of over 400 cases converted to open surgery mainly due to large incisional hernia or extensive adhesion.

Of the 13 cases performed at CCF, we had no mortality, and two major complications—a staple line leak requiring repair and drainage on postoperative day (POD) 3, and a postoperative acute gastric outlet obstruction in a patient 2 years after removal of an eroded gastric band that had to be converted to an RYGBP on POD 4. According to other studies, there was only one mortality reported out of over 400 cases due to multi-organ failure from septic shock, and overall complication rate was approximately 0–32 % [1, 3, 4, 12, 13, 15, 17–25]. A recent systematic review of the relevant articles reported that the weighted mean of complication rates of revisional LSG after failed GB was 4.1 % [28]. Commonly reported complications are listed in Table 2. Among those, the most prevalent complications were leaks, strictures, bleeding, and gastroesophageal reflux disease (GERD) [2]. Leak is one of the major complications of both primary and revisional LSG with long staple lines. The thick area around the pylorus is predisposed to leak. Esophagogastric junction is also vulnerable to leak, because an excessive traction applied during stapling the stomach leaves the tissue under tension [13]. For a revisional LSG, leak becomes a bigger problem in the upper part of the stomach due to the high probability of incomplete stapling of the thickened scar tissue around the previously banded area and compromised blood supply at the esophagogastric junction after dissecting the left crus. The ischemia or trauma during the initial procedure also contributes to a leak after the revision. Stricture is usually developed at the incisura angularis of the stomach, which would be prevented by using a bougie. Although the complication rate of LSG is lower than that of RYGB, GERD is more frequently seen after LSG than RYGB. Medical therapy with proton pump inhibitors is a treatment of choice in patients with new-onset GERD. In addition, bleeding along the staple line would be prevented by the use of staple line reinforcement with either oversewing or buttressing.

Many studies reported that the overall complication rate of revisional LSG was slightly higher compared to the primary LSG [4, 13, 17, 21, 24], although others showed no significant difference [20, 22, 25]. The possible discrepancy of complication rates between primary and revisional LSG reported by the former studies can be explained by the following technical problems of the revisional LSG: (1) difficulty of stapling the thickened scar tissue, (2) possible damage of compromised tissue when dissecting the adhesions around the previous band, and (3) compromised vascular supply to the superior part of the stomach due to dissection of the left crus. In contrast, complications occurred less frequently after LSG compared to revisional RYGB or BPD-DS. One systematic review estimated the complication rates of revisional LSG, RYGB, and BPD-DS after failed GB were 4.1 %, 10.7 %, and 24.4 %, respectively [28]. The absence of any anastomosis in case of LSG may be the reason for its being safer than malabsorptive surgeries [12, 23]. To summarize, these results support an acceptable level of safety of revisional LSG.

In our study, mean excess BMI loss at 2, 6, 12, and 18 months were 28.9, 64.2, 65.3, and 65.7 %, respectively. The results of other selected studies are summarized in Table 3. Excess weight loss and expected excess BMI loss were 42.7–53 % and 46.8–65 % at 12–14 months of follow-up, respectively. These results were maintained at 24–36 months after the procedure, although not all of the patients completed the follow-up. Obesity-related comorbidities were improved or even resolved after revisional LSG in a majority of the treated patients. According to the studies that compared the results of primary and revisional LSG, the extent of weight loss after the two procedures was not significantly different [20, 22, 25]. Furthermore, considering that sufficient weight loss was achieved after revisional LSG in the patients who had failed to lose weight after the previous restrictive procedure, LSG seems not to be a mere restrictive procedure and can be used as a valid revisional option for the failed restrictive surgery.