Predictors of Success after Laparoscopic Sleeve Gastrectomy

Introduction

The prevalence of obesity continues to rise worldwide. According to a report by the World Health Organization, obesity has risen from 7% (250 million people) in 1998 to 10% (500 million) in 2007, with 1.5 billion people being overweight. ¹ In the United States, there has been an even sharper increase, with 35.7% of the population suffering from obesity in 2010 compared to only 20% in 2000. ²

Bariatric surgery as a means of treatment has been shown to achieve more durable and greater weight loss than other means of therapy.^3,4 The laparascopic sleeve gastrectomy (LSG) has become increasingly popular in the field of bariatric surgery, comprising 36.3% of all bariatric surgeries performed in academic centers, and an even higher percentage in community-based hospitals, and is predicted to become the most popular form of bariatric surgery. ⁵ While it is very common, the degree of weight loss following the LSG varies greatly among individuals. The underlying mechanisms behind the improvements and the reasons for these differences are not well understood. Studies demonstrate that the rate of obesity varies with age, sex, ethnicity, race, and socioeconomic factors. ⁶ Therefore, the degree of weight loss may vary according to some or all of these factors. Various studies have compared the amount of weight loss between different ethnic groups, ⁷ different preoperative weight, ⁸ or following different types of surgery. ⁹ While these studies individually found these factors to have subtle or major differences in outcome, no study has been aimed to find predictors of better weight loss of patients undergoing the LSG.

Additionally, while socioeconomic factors were evaluated for predicting success in laparoscopic Roux en Y gastric bypass, ¹⁰ we are unaware of studies aimed at identifying predictors of success for weight loss in a large cohort undergoing LSG. In this study, we aimed to determine if postoperative weight loss can be predicted by any one or combination of various preoperative patient demographic and socioeconomic factors by examining excess weight loss at 1 year (EWL1) following LSG.

Methods

This study, following Institutional Review Board (IRB) approval, was conducted at the Chicago Institute for Advanced Bariatric Surgery (Chicago, IL). All data were collected and maintained in an IRB-approved clinical database. Our study utilized this database to identify which factors can independently predict successful postoperative weight loss after LSG.

The study included 100 consecutive patients who underwent a LSG, enrolled over a period of 36 months. All cases were performed by a single surgeon (RL) with the help of a senior surgical resident (PGY-4 or PGY-5) in a community-based teaching institution. Patients undergoing both single incision and multiport LSG were included in this study. Our single port technique involved placement of a quad port through the umbilicus, and the multiport method used a 5 mm trochar for a 5 mm laparoscope and three working trochars and a Nathanson liver retractor. Division of the gastro colic omentum started 5 cm from the pylorus and a 34 french bougie was utilized to create the sleeve. The study did not employ any inclusion or exclusion criteria. All patients met bariatric surgery indications criteria of a preoperative body mass index (BMI) of >40 kg/m² or 35–40 kg/m² with major obesity associated comorbidities (e.g., hypertension and diabetes).

The primary endpoint was percentage of EWL at 1 year (EWL1). EWL was defined as the excess weight over the ideal body weight calculated by the Metropolitan Life Weight Tables (source: Metropolitan Life Insurance Company). EWL was plotted in a normal histogram, and insufficient weight loss was defined as EWL≤−1 standard deviation (SD) from mean EWL, in accordance with previous published data.^10,11 Binary logistic regression analysis was used in both univariate and multivariate modeling to identify independent preoperative variables associated with the successful EWL after LSG surgery. Independent variables examined included 10 putative socioeconomic factors (age, gender, race, marital status, parental status, employment status, history of depression, smoking, binge eating, and preoperative BMI). For comparison, we used Pearson's chi-square test for categorical variables and Student's t-test for continuous variables. All independent variables then underwent multivariate analysis by entering them together using the backward stepwise method. The following cutoff points were used for the binary logistic regression stepwise methods: p=0.05 for entry into the model and p=0.10 for removal from the model. From these estimates, odds ratios (OR) with 95% confidence intervals (CI) were computed. The SPSS statistical software program (v11.0; SPSS, Inc., Chicago, IL) was used for all analyses. Statistical significance was set at p<0.05.

Results

Of the 100 patients enrolled, follow-up at the 1 year clinic appointment was done in 87% of the patients. The remaining patients were contacted via telephone to obtain information, which was confirmed with physician documentation from an office visit outside of our bariatric clinic, making 1 year follow-up 100%. This cohort included 13 men and 87 women, with an overall mean BMI of 48.8±8.9 (range 35–77.6). Additionally, 32 of these patients underwent a laparoendoscopic single site (LESS) sleeve gastrectomy and 68 underwent a traditional multiport sleeve gastrectomy. As Figure 1 demonstrates, mean EWL was 65.7±23.5%. Therefore, the definition for inadequate weight loss was EWL<42.2%. According to this, 18 patients had inadequate EWL at 1 year, while 82 patients were able to achieve successful weight loss. Results of the univariate analysis are demonstrated in Table 1. Preoperative BMI had a significant effect on EWL. The group with optimal weight loss had a significantly lower BMI preoperatively than the suboptimal weight group (47.49±8.44 vs. 54.18±9.41; p=0.004).

OPEN IN VIEWER

FIG. 1.

Frequency distribution of percent excess weight loss (EWL). White represents suboptimal EWL.

After univariate analysis, the data were then entered in the logistic regression analysis, and only BMI remained an independent predictor of EWL at 1 year. Preoperative BMI remained an independent predictor of success (adequate EWL at 1 year) in the multivariate logistic regression model after adjusting for covariates. We found that a unit increase in BMI is associated with 1.1 increase in the log hazard rate (HR) of having an inadequate EWL (HR 1.1; 95% CI 1–1.1; p=0.008).

Discussion

The best method to judge outcomes after bariatric surgery is still debated.^12,13 Some options include weight loss, resolution of comorbidities, or improvement in quality of life. Of these, weight loss is the most common, and can be reported as percent of preoperative weight, change in BMI, or percent EWL. ¹⁴ In our study, the primary endpoint was EWL at 1 year (EWL1), and weight loss was considered suboptimal when EWL was less than one SD from the mean EWL. A similar method was utilized by Lutfi et al. when comparing successful weight loss for patients undergoing gastric bypass. ¹⁰ The EWL achieved at 1 year (M=65.7%) was slightly superior to that of other published series.^15,16 Additionally, with our data, we were able to compare similar patient populations, as both groups underwent the same procedure by the same surgeon.

While there are many preconceived notions as to who will benefit most from weight loss surgery, no clear predictors have been presented for LSG. Studies have looked at various factors for effective weight loss in gastric bypass or gastric banding. For example, Capella demonstrated a statistically significant difference in weight loss amongst different ethnic groups for vertical gastric banding. ¹⁷ However, this correlation was not demonstrated in our cohort study, with our groups achieving similar EWL after 1 year.

Marital status has also previously been shown to predict weight loss after gastric bypass, thought to be due to single or divorced patients having more time for physical activity, also in correlation with employment status.^10,18 Yet, our study did not show these trends to be true for LSG, suggesting that these factors are not a large component of postoperative weight loss in the first year.

Behavioral factors such as smoking and binge eating have also been thought to have a negative effect on weight loss. Many consider smoking to be a barrier to successful outcomes after bariatric surgery, and have suggested that it causes a resistance to weight loss. ¹⁹ However, our study did not demonstrate this correlation. Additionally, the question is also raised as to whether those with a history of binge eating or depression will be able to achieve successful weight loss. ²⁰ Some studies have shown that preoperative binge eating had a negative effect on postsurgical BMI. ²¹ However, in our study, these behavioral factors did not seem to effect EWL at 1 year for LSG.

There are a few limitations to this study. A drawback to utilizing EWL is having weight loss as the only parameter for success. Also, those with an elevated BMI needed to lose more pounds to achieve the same EWL.^22,23 It has been shown, particularly in laparoscopic banding, that those with a BMI of >50 are known to have lower EWL compared to those patients with a BMI between 35 and 50. It has also been suggested that these patients may have better outcomes with a more aggressive malabsorptive procedure. ²⁴ Having said this, however, when looking at the number of pounds each group lost, we still found that the inadequate group patients lost significantly less weight when compared to the successful patients. Moreover, a low N and short-term follow-up are additional limitations, but our 100% follow-up allowed for full analysis of our data.

For the patients who were deemed unsuccessful, a number of different modalities were employed to help them with their weight loss goal. These patients had additional visits with the surgeon, nurse practitioner, and dietician to identify whether their lack of success was due to a behavioral or technical issue, or both. These patients received additional UGI studies to evaluate for dilated sleeves or a retained fundus. None of these patients was felt to have a technical issue, and they were given additional visits with the dietitian and on-site psychologist. We are still gathering data to evaluate if these interventions were successful in helping patients achieve a successful outcome.

Overall, many studies show varying effects of preoperative factors for other weight loss surgeries. Yet, in our study, no predictor except BMI held up in multivariate analysis, suggesting that the LSG might be an ideal initial operation for morbid obesity for patients with a BMI of <50, regardless of other preoperative demographic and socioeconomic factors.

Conclusion

The LSG is a highly successful surgery with excellent EWL in the short term. We believe it should be the operation of choice for patients with a BMI of <50.