Revisional Bariatric Surgery

Background: To elucidate percent excess body weight loss (%EBWL) gained from revisional surgery and to determine resolution of comorbid conditions given the obesity epidemic. Methods: All patients undergoing revisional bariatric surgery from 2002 to 2012 were analyzed. Main outcome measures were %EBWL, resolution of comorbid conditions and complications. Descriptive statistics and paired t-tests were computed. Results: 251 cases were performed. Initial mean body mass index was 48.1 kg/m2 (+/9.4). Hypertension (32%) and treatment failure (37.5%) were most commonly reported. Mean percentage difference between reoperation and last weight and original and reoperation weight was 27.9% (29.5%), p<0.001. Proportions of all comorbid conditions decreased, but none statistically. No deaths occurred, however 22% experienced at least one complication. Conclusions: Greater %EBWL occurs between revision surgery and last follow-up. No comorbid condition decreased. Further research is needed to determine the optimal timing for revisional surgery to optimize %EBWL and resolution of comorbid condition.


Background
Obesity and its associated co-morbid complications continue to increase. Severe morbid obesity is a pandemic that affects both adults and adolescents in which surgery has proven to be the only effective means to provide for long-term weight loss and apparent resolution of comorbid conditions as compared to medical management [1][2][3][4]. As acceptance of bariatric surgical interventions has increased, the numbers of both primary and revisional surgical interventions have also increased. For example, according to the American Society of Metabolic and Bariatric Surgery (ASMBS), the number of primary bariatric procedures has increased from approximately 13,000 procedures in 1998 to 220, 000 procedures in 2008 respectively [5].
Revisional bariatric surgery entails knowledge of the primary surgical intervention, as well as expertise with regards to non-virginal operative fields. Herein, we aimed compare the percent excess body weight loss (%EBWL) and body mass index (BMI) between each bariatric procedure, as well between primary, revisional procedure, and final outcome measures. Secondary aims included identifying whether there was a particular bariatric operation that predicted a higher resolution of comorbid conditions and predicted less associated complications.

Methods
Following institutional board approval, data from a retrospective case review of all patients undergoing revisional bariatric surgical interventions, excluding port revisions of laparoscopic gastric banding procedures (LAGB), at one academic medical center were analyzed from a prospective, longitudinal database. Information from 2002 to 2012 was collected. Demographic, clinical, co-morbid conditions and complications were compared between revisional groups. Weight loss was expressed as percent excess body weight loss (%EBWL), defined as the difference between initial weight and current weight, divided by the difference of the initial weight and ideal weight. The ideal body weight and was calculated from the 1983 Metropolitan Life Insurance Company tables [6]. Percent EBWL was calculated at each revision and last clinic visit. The main outcome measure was %EBWL. Secondary outcome measures included resolution of comorbid conditions and associated complications.
For the purposes of this study, patients were grouped into 4 revision types: 1) conversion from one bariatric surgical procedure to another; 2) revision of a previous operation with no conversion; 3) reversal and 4) exploration. Overall indications for revision were divided into the following categories: 1) pain; 2) obstruction; 3) failure to thrive; 4) development of an enterocutaneous fistula; 5) treatment failure; 6) gastrointestinal issues primarily related to gastroesophageal reflux and ulcers and 7) medically or nutritional-related issues.
All patients met the criteria for bariatric surgery established by the National Institutes of Health Consensus Development Panel with patients having a body mass index (BMI) of >40 kg/m 2 , or >35 kg/ m 2 in the presence of obesity related comorbidities. All had failed to maintain weight loss by non-surgical means. All patients were ISSN: 2377-9284 required to attend a public information session given by one of the surgeons and be assessed by both a clinical psychiatrist and dietician.
Preoperative investigations were guided by the patient history and type of surgery. Routine investigations included an upper gastrointestinal barium study for all patients. The decision for each particular revisional surgical intervention was made by the patient after an extensive period of counseling and education. Revisional bariatric patients were recovered in a surgical ward by staff experienced in management of postoperative bariatric patients. All patients were reviewed in outpatient clinic at 2 weeks, as well as regularly assessed for vitamin and mineral deficiencies and treated accordingly. Descriptive statistics for continuous and categorical variables were presented, as appropriate. Paired t-tests were computed to assess outcome measure differences between original and last follow-up.
The most common comorbid conditions were hypertension (32%), obstructive sleep apnea (21%), asthma (19%) and diabetes (18%) ( Table 1). There was a decrease in the proportions of all comorbid conditions between the original and last visit, but none reached statistical significance. At the original operation, 41/224 (18.3%) had Diabetes Mellitus (DM). At first revision, 41/249 (16.5%) had DM, whereas, at last visit, 24/236 (10.2%) had DM. The procedure with the largest proportional change of DM was the laparoscopic conversion of the laparoscopic-adjusted gastric banding (LAGB surgery) to Roux-En-Y Gastric Bypass (RYGB); however, this difference did not reach statistical significance (p=0.47) ( Table 3).    Median intraoperative time was 234 minutes and median length of hospital stay was 3 days (range 2-5 days) (Table 1). Overall, there were 21.5% of cases that had perioperative complications (54/251); however, no deaths occurred. The distributions with and without complications differ according to surgery type (Fisher's exact test, p<0.001). Among those who had complications, the highest proportion of complications occurred within LRYGB and included postoperative bleeding and obstruction (Table 4).

Discussion
Revisional bariatric surgery is now commonplace. In our study, it appears that greater %EBWL occurred between revision surgery and last clinic visit than between the original surgery and first reoperation. There was not a single comorbid condition that decreased statistically significantly during the study period; however, all of them had a downward trend. Overall, complications did occur; however, there was no mortalities reported. These results are not unlike contemporary research on revisional bariatric surgery. The reported incidence of reoperation in bariatric surgery ranges from 5-57%, but the quality and integrity of what is reported is sometimes questioned [3,[7][8][9]. Previous studies have grouped their patients based on the etiologies for revisional surgery [10]. However, given that a majority of our patients were referred to a tertiary weight loss center for inadequate weight loss, we chose to group our patients based on the type of revision surgery performed. The majority of our patients benefitted from either laparoscopic conversion of their primary procedure to a RYGB or revision of their primary procedure. Success, as defined by a %EBWL>50, was achieved in our cohort at the last clinic visit (%EBWL of 58.3%) and was consistent with current literature on the   effectiveness of transitioning or revising the roux-en Y gastric bypass [3,8,[10][11][12][13][14].
Unlike other contemporary literature, there appeared to be fewer revisions for gastrointestinal complaints, failure to thrive or medicallyprotracted conditions [3,15,16]. The percentages of revision cases in our cohort for technical and mechanical complications, as well as fistulas, were consistent with previously reported literature with most cases attempted laparoscopically [3,10,17]. It was interesting that fistulas were often found concurrently with treatment failure cases; however, further studies are needed to elucidate whether this plays a larger role in weight loss in our cohort. We did not necessarily see an increased rate of complications with performing these operations laprascopically, as revisional bariatric surgery is often quoted as having a complication rate of between 10-50% [8,10,18,19]. Our complication was approximately 22%, albeit, 28% of our cohort had more than one operation. We believe that this approach may be taken and does not necessarily increase the rate of complication at high-volume centers. Juxtaposed to the previously reported operative mortality rates of less than 2.5%, we had no mortalities in our series [8,10,11,13].
Surprisingly none of our comorbid medical conditions reached a statistical significant decrease following re-operations. According to one of the most recent review papers by the American Society for Metabolic and Bariatric Surgery Revision Task Force, resolution of comorbid conditions must be considered as important as and/ or more important than reduction in %EBWL and BMI [9]. Given that our study had a shorter follow-up, these conditions may become significant over time. Further long-term studies are needed to elucidate this.
Despite this study's strengths, one limitation is that given our institution is a tertiary referral facility, often we cannot control for loss to follow up, especially in the setting of a retrospective study following revision surgeries. Attrition is not uncommon following bariatric surgery and is possibly related to durable weight loss. In addition, since we are a high volume referral center, some patients choose to establish follow up care with bariatric surgeons closer to home and we do not have the long-term data on co-morbid conditions and longerterm %EBWL.

Conclusion
Revisional bariatric surgery is now a commonplace surgical intervention. Greater %EBWL occurs between revision surgery and last clinic visit than between the original surgery and reoperation. No single comorbid condition reached statistical significance during the study period; however, there was an overall decrease in the rates of hypertension, diabetes and obstructive sleep apnea. Further research is needed to determine the optimal timing for revisional bariatric surgery once initial treatment has failed to optimize excess weight loss and resolution of comorbid conditions.