- Email: [email protected]
Early Outcomes of Bariatric Surgery in Patients with Metabolic Syndrome: An Analysis of the Bariatric Outcomes Longitudinal Database
Early Outcomes of Bariatric Surgery in Patients with Metabolic Syndrome: An Analysis of the Bariatric Outcomes Longitudinal Database William B Inabnet III, MD, FACS, Deborah A Winegar, PhD, Bintu Sherif, MS, Michael G Sarr, MD, FACS BACKGROUND: Metabolic syndrome (MetS) complicating obesity is endemic in the United States. STUDY DESIGN: Bariatric Outcomes Longitudinal Database, the national database for the American Society for
Metabolic and Bariatric Surgery Bariatric Surgery Center of Excellence Program, was queried to identify patients undergoing bariatric surgery from June 2007 through November 2010. MetS was defined as the presence of hypertension, diabetes, and dyslipidemia at presentation for bariatric surgery. Ninety-day and 1-year outcomes were assessed to determine early outcomes in bariatric surgery patients with MetS. RESULTS: Among 186,576 research-consented patients, 23,106 (12%) were diagnosed with MetS. Patients with MetS were more likely to be male (35% vs 20%; p ⬍ 0.0001), older (mean age 54 vs 44 years; p ⬍ 0.0001), and Caucasian (81% vs 74%; p ⬍ 0.0001). Of the 23,106 MetS patients, more underwent gastric bypass (RYGB) (62%) compared with gastric banding (32%), sleeve gastrectomy (4.5%), and biliopancreactic diversion with duodenal switch (BPD/ DS)(1.5%). MetS patients had an increase in serious complications (2.4% vs 1.0%; p ⬍ 0.0001), readmissions (6.2% vs 4.7%; p ⬍ 0.0001), and mortality (0.3% vs 0.1%; p ⬍ 0.0001) within 90 days of operation. After adjusting for sex, age, and body mass index, RYGB patients with MetS had an increased risk of 90-day serious complications compared to RYGB patients without MetS (odds ratio 1.43; 95% CI, 1.27 to 1.61; p ⬍ 0.0001). The 12-month remission rate of diabetes was least for gastric banding (28%) compared with the other procedures (RYGB 62%, sleeve gastrectomy 52%, BPD/DS 74%). CONCLUSIONS: Patients with MetS undergoing bariatric surgery showed dramatic improvement in diabetes 1-year after surgery; however, an adverse 90-day outcome was more common. (J Am Coll Surg 2012;214:550–557. © 2012 by the American College of Surgeons)
verse outcomes in patients undergoing bariatric surgery.4 Contributing factors to this perception include the severe central and visceral adiposity, which can lead to hepatomegaly and intraoperative technical challenges, the presence of moderate to severe diabetes, and the overall poor health of many patients with MetS. To date, most studies examining MetS in bariatric surgery patients are small, single institution studies that report results on 1 or 2 procedures that are preferred at the local institution.5 With the recent implementation of Bariatric Surgery Center of Excellence programs (BSCOE), which require mandatory data entry into 1 of 2 national registries, investigators are able to examine a large cohort of patients from a diverse demographic background, including community, rural-based, and urban medical centers with both private and academic affiliations. The objective of our analysis was to determine the prevalence of MetS in patients undergoing bariatric surgery at American Society for Metabolic and Bariatric Surgery (ASMBS) BSCOEs, and to
Metabolic syndrome (MetS) occurs as a constellation of metabolic abnormalities, including type 2 diabetes mellitus, hypertension, and dyslipidemia, that lead to an increased risk for cardiovascular disease.1 The true cause of MetS is not known, but there is an intertwined link between obesity, insulin resistance, and the MetS, which leads to a vicious cycle of metabolic stress.2,3 It has long been assumed that the presence of MetS is a risk factor for adDisclosure Information: Dr Winegar and Mr Sherif are salaried employees of Surgical Review Corporation. All other authors have nothing to disclose. Presented at the Southern Surgical Association 123rd Annual Meeting, Hot Springs, VA, December 2011. Received December 16, 2011; Accepted December 19, 2011. From Mount Sinai Medical Center, New York, NY (Inabnet); Surgical Review Corporation, Raleigh, NC (Winegar, Sherif ); and Mayo Clinic, Rochester, MN (Sarr). Correspondence address: William B Inabnet III, MD, Division of Metabolic, Endocrine and Minimally Invasive Surgery, Department of Surgery, Mount Sinai Medical Center, 5 East 98th St, Box 1259, New York, NY 10029. email: [email protected]
© 2012 by the American College of Surgeons Published by Elsevier Inc.
550
ISSN 1072-7515/12/$36.00 doi:10.1016/j.jamcollsurg.2011.12.019
Vol. 214, No. 4, April 2012
Abbreviations and Acronyms
ASMBS ⫽ American Society for Metabolic and Bariatric Surgery BMI ⫽ body mass index BOLD ⫽ Bariatric Outcomes Longitudinal Database BPD/DS⫽ biliopancreatic diversion with duodenal switch BSCOE ⫽ Bariatric Surgery Centers of Excellence MetS ⫽ metabolic syndrome RYGB ⫽ Roux-en-Y gastric bypass
assess the 90-day safety outcomes. In addition, a secondary aim was to determine the effectiveness of bariatric surgery on comorbidity resolution at 1 year in patients with MetS. Our hypothesis was that operative morbidity and mortality would be increased in MetS patients undergoing bariatric surgery but that the MetS would have a high rate of remission in the MetS cohort.
METHODS This study used the Bariatric Outcomes Longitudinal Database (BOLD), a national database that tracks outcomes of bariatric operations performed at ASMBS BSCOEs. The ASMBS designation as a Center of Excellence requires that centers register all patients who undergo any bariatric operation at their center and perform follow-up evaluations at specific times postoperatively. The data are monitored and audited for accuracy. Further details of the BOLD database have been previously reported.6 The study population included patients aged 18 to 75 years, with body mass index (BMI) ⬎ 35 kg/m2, who had Roux-en-Y gastric bypass (RYGB), adjustable gastric banding, sleeve gastrectomy, or biliopancreactic diversion with duodenal switch (BPD/DS) bariatric surgery performed by a BSCOE participant between June 2007 and November 2010. Preoperative data on patient age, sex, race, insurance status, last BMI before bariatric surgery, previous bariatric surgery history, and selective comorbidities were obtained from the database. Intraoperative data obtained included operative procedure, American Society of Anesthesiologists (ASA) classification, duration of the operation, and duration of the postoperative stay. The status of metabolic comorbidities was assessed at both preoperative and postoperative visits using a modified version of a scoring system developed to evaluate obesityrelated conditions in bariatric surgical patients.7 The scoring system assigns a numerical value (0 to 5) to indicate the severity of the disease as indicated and allows tracking of changes in the disease state over time (Table 1). For this study, MetS was defined as the presence hypertension, diabetes, and dyslipidemia at presentation for bari-
Inabnet et al
Bariatric Surgery for Metabolic Syndrome
551
atric surgery. Therefore, patients were categorized as having MetS if their hypertension score was ⱖ3, their diabetes score was ⱖ2, and their dyslipidemia score was ⱖ2. Patients were categorized as having sleep apnea if their score was ⱖ3. The primary safety outcomes monitored were mortality, serious complications, hospital readmissions, and reoperations that occurred within 30 and 90 days of the bariatric operation. The primary efficacy outcomes monitored were weight loss and remission of metabolic comorbidities within 12 months after bariatric surgery. Remission was defined as cessation of all medical treatment for the given condition (ie, remission of sleep apnea if continuous positive airway pressure treatment was terminated). Efficacy assessments were performed during routine follow-up, when weight and comorbidities were assessed. The visit window 9 and 15 months was used for the 12month follow-up visit and the follow-up visit closest to 12 months was used for this study. Complications entered in BOLD included all deaths and unfavorable occurrences that prolonged hospitalization, required readmission to the emergency room or hospital, or required treatment outside of standard postoperative care. A readmission was defined as any hospital or facility stay that lasted for at least 24 hours, regardless of the location within the hospital or facility and regardless of the reason for the stay. A reoperation was defined as a surgical intervention of the primary bariatric surgery required as a result of a complication. Serious complications were defined as a composite that included the following complications: death, anastomotic leakage, cardiac arrest, deep venous thrombosis, evisceration, heart failure and/or pulmonary edema, liver failure, myocardial infarction, pneumothorax, pulmonary embolism, renal failure, respiratory failure, sepsis, systemic inflammatory response, cerebrovascular accident, and bleeding requiring blood transfusion. Statistical analysis
Descriptive statistics were generated to summarize the study cohorts. Continuous variables were summarized with mean and standard deviation. Categorical variables were summarized with frequencies and percentages. Summary data were tabulated in subgroups (MetS patients or nonMetS patients). The chi-square test for categorical variables and the Kruskal-Wallis test for continuous variables were used to compare the subgroups. A nonlinear, mixed effect model was used to compute multivariate adjusted odds ratios with corresponding 95% confidence intervals, using non-MetS patients as the reference group. The fully adjusted model controlled for preoperative characteristics with hospital as a random effect. Significance level was set
552
Inabnet et al
Bariatric Surgery for Metabolic Syndrome
J Am Coll Surg
Table 1. Modified Scoring System for Obesity-Associated Comorbidities Score
Hypertension
Diabetes
Dyslipidemia
Obstructive sleep apnea
0 1
No indication of hypertension Borderline, no medication
No symptoms or evidence of diabetes Elevated fasting glucose
2
Diagnosis of hypertension, no medication Treatment with single medication Treatment with multiple medications
Diabetes, controlled with oral medication Diabetes, controlled with insulin
Not present Present, no treatment required Controlled with lifestyle change Controlled with single medication Control with multiple medications Not controlled
No symptoms Symptoms, negative sleep study Diagnosed by sleep study, no oral appliance Treated with continuous positive airway pressure Significant hypoxia or oxygen dependent
3 4 5
Diabetes, controlled with insulin and oral medication Diabetes, with severe complications
at alpha ⫽ 0.05. All statistical analyses were performed using SAS statistical software Version 9.2.
RESULTS Demographic characteristics of entire cohort
Of the study population of 186,576 patients, 23,106 (12%) had MetS at the time of bariatric surgery. Patients with MetS were more likely to be male (35% vs 20%; p ⬍ 0.0001), older (mean age 54 vs 44 years; p ⬍ 0.0001), and
Caucasian (81% vs 74%; p ⬍ 0.0001) than non-MetS patients (Table 2). There was no difference in mean BMI between the two groups (47.0 vs 46.8 kg/m2), but patients with MetS were more likely to have obstructive sleep apnea (45% vs 25%, p ⬍ 0.0001) and an American Society of Anesthesiologists category of Class III or greater (84% vs 68%, p ⬍ 0.0001). Of the 23,106 MetS patients, more underwent RYGB (62%) than adjustable gastric banding (32%), sleeve gas-
Table 2. Preoperative Demographic Characteristics of Overall Cohort Characteristic
Sex, n (%) Male Female Age at operation, y Mean (SD) Median (min, max) Race, n (%) Caucasian African American Hispanic Asian Native American Pacific Islander/Hawaiian Other Body mass index, kg/m2 Mean (SD) Median (min, max) Sleep apnea, n (%) Treatment or oral medication required ASA classification, n (%) Class I Class II Class III Class IV Class V ASA, American Society of Anesthesiologists.
Metabolic syndrome (n ⴝ 23,106)
8,207 (36) 14,899 (65) 54.1 (9.46) 55 (18, 75) 18,710 (81) 2,057 (9) 1,186 (5) 62 (0.3) 135 (0.6) 32 (0.1) 924 (4) 47.0 (7.86) 46 (35, 106)
Nonmetabolic syndrome (n ⴝ 163,470)
32,249 (20) 131,221 (80) 44.3 (11.55) 44 (18, 75) 121,263 (74) 18,806 (12) 12,362 (8) 336 (0.2) 722 (0.4) 221 (0.1) 9760 (6) 46.8 (7.81) 45 (35, 106)
Overall (n ⴝ 186,576)
40,456 (22) 146,120 (78) 45.5 (11.76) 45 (18, 75) 139,973 (75) 20,863 (11) 13,548 (7) 398 (0.2) 857 (0.5) 253 (0.1) 10684 (6) 46.9 (7.82) 45 (35, 106)
10,346 (45)
40,859 (25)
51,205 (27)
387 (2) 3,294 (14) 17,537 (76) 1,868 (8) 20 (0.1)
6,426 (4) 45,903 (28) 105,632 (65) 5,398 (3) 111 (0.1)
6,813 (4) 49,197 (26) 123,169 (66) 7,266 (4) 131 (0.1)
Vol. 214, No. 4, April 2012
Bariatric Surgery for Metabolic Syndrome
Inabnet et al
553
Table 3. Postoperative Safety Outcomes
Outcomes
30-d mortality 30-d serious complications 30-d readmission 30-d reoperation 90-d mortality 90-d serious complications 90-d readmission 90-d reoperation
Metabolic syndrome (n ⴝ 23,106) n %
Overall (n ⴝ 1,86,576) n %
173 1,941 6,847 3,880 233 2,132 9,100 6,267
0.1 1.0 3.7 2.1 0.1 1.1 4.9 3.4
52 516 1,076 605 65 558 1,427 954
Nonmetabolic syndrome (n ⴝ 163,470) n %
0.2 2.2 4.7 2.6 0.3 2.4 6.2 4.1
121 1,425 5,771 3,275 168 1,574 7,673 5,313
0.1 0.9 3.5 2.0 0.1 1.0 4.7 3.3
p Value*
⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001
*p Values for the comparison between groups were calculated using chi-square test.
trectomy (4.5%), and BPD/DS (1.5%). In contrast, among non-MetS patients, 53% underwent RYGB, followed by adjustable gastric banding (40%), sleeve gastrectomy (6%), and BPD/DS (1%). A laparoscopic approach was used equally in both groups (MetS 94%, non-MetS 95%). The mean durations of operation and hospital stay were greater in MetS patients (99 vs 73 minutes, p ⬍ 0.0001 and 2.1 vs 1.7 days, p ⬍ 0.0001, respectively). Safety outcomes data
MetS patients had a higher incidence of serious complications (2.4% vs 1.0%; p ⬍ 0.0001), readmissions (6.2% vs 4.7%; p ⬍ 0.0001), and mortality (0.3 vs 0.1%; p ⬍ 0.0001) within 30 and 90 days of operation than non-MS patients (Table 3). The incidence of serious complications and mortality was least for adjustable gastric banding and greatest for BPD/DS (p ⬍ 0.0001) (Table 4). After adjusting for sex, age, and BMI, both gastric banding and RYGB patients with MetS had an increased risk of 90-day serious complications compared with patients without MetS (gastric banding odds ratio [OR] 2.79; 95%
CI, 2.02 to 3.84, p ⬍ 0.0001) (RYGB [OR] ⫽ 1.43; 95% CI, 1.27 to 1.61; p ⬍ 0.0001) (Table 5). 12-month efficacy on comorbidity resolution
Of the entire cohort of 186,576 patients, 12-month postoperative data were available on 92,915 patients (50%), including 12,144 (53%) from the MetS group and 80,177 (49%) from the non-MetS group. In patients with MetS, remission of hypertension, diabetes, and dyslipidemia occurred in 36%, 50%, and 35% of patients, respectively; 76% of MetS patients with sleep apnea went off all forms of treatment 12 months after the bariatric operation. Among procedures, adjustable gastric banding had a lower 12month rate of remission of MetS compared with RYGB, sleeve gastrectomy, and BPD/DS (Table 6).
DISCUSSION These data demonstrate several interesting findings in patients with morbid obesity complicated by MetS who underwent bariatric surgery. Though patients with MetS pre-
Table 4. Procedure-Specific Postoperative Safety Outcomes in Patients with Metabolic Syndrome
Outcomes
30-d mortality 30-d serious complications 30-d readmission 30-d reoperation 90-d mortality 90-d serious complications 90-d readmission 90-d reoperation
Adjustable gastric banding (n ⴝ 7,357) n %
4 63 183 84 5 67 231 134
0.1 0.9 2.5 1.1 0.1 0.9 3.1 1.8
Roux-en-Y gastric bypass (n ⴝ 14,329) n %
41 411 811 474 53 445 1094 754
0.3 2.9 5.7 3.3 0.4 3.1 7.6 5.3
Sleeve gastrectomy (n ⴝ 1,081) n %
BPD w/ duodenal switch (n ⴝ 339) n %
p Value*
3 22 43 28 3 24 51 38
4 20 39 19 4 22 51 28
⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001
0.3 2.0 4.0 2.6 0.3 2.2 4.7 3.5
*p Values for the comparison between MS groups were calculated using chi-square test and Fisher’s exact test. BPD, biliopancreatic diversion.
1.2 5.9 11.5 5.6 1.2 6.5 15.0 8.3
554
Inabnet et al
Bariatric Surgery for Metabolic Syndrome
J Am Coll Surg
Table 5. Ninety-Day Serious Complications Regression Models* Adjustable gastric banding Odds ratio (95%CI) p Value
Covariates
Sex (ref, female) Race (ref, Caucasian) BMI (each 1-kg/m2 increment) Age (each 1-year increment) MetS (ref, non-MetS) Sleep apnea (ref, no history or symptoms)
1.51 (1.13, 2.02) — 1.06 (1.05, 1.08) 1.03 (1.02, 1.04) 2.79 (2.02, 3.84) —
0.0059 NS ⬍0.0001 ⬍0.0001 ⬍0.0001 NS
Roux-en-Y gastric bypass Odds ratio (95%CI) p Value
1.49 (1.34, 1.65) — 1.04 (1.03, 1.04) 1.04 (1.03, 1.04) 1.43 (1.27, 1.61) —
⬍0.0001 NS ⬍0.0001 ⬍0.0001 ⬍0.0001 NS
*Based on non-linear mixed effect model which includes hospital as a random effects. BMI, body mass index; MetS, metabolic syndrome; NS, Not significant.
sented with a similar preoperative weight profile as nonMetS patients (mean BMI 47 kg/m2), MetS patients as a whole experienced an increased, albeit infrequent, incidence of serious complications and mortality compared with non-MetS patients. Because the mean weight was similar between the two groups, factors other than weight invariably contribute to the greater incidence of adverse outcomes in MetS patients. For patients with MetS, the incidences of serious complications and mortality 90 days after bariatric surgery were 2.4% and 0.3%, respectively, which was significantly higher than that for non-MetS patients. In another study that evaluated a large cohort of obese patients with MetS who underwent bariatric surgery at academic institutions, the morbidity (8.6%) and mortality (0.4%) were greater than in our study.8 Nevertheless, achieving a decrease in mortality and all-cause morbidity in this high risk patient population is an amazing feat and supports the utility of outcomes tracking mandated by the center of excellence programs. In our study, the MetS cohort was older and had a greater percentage of males than non-MetS patients. In general, male patients tend to have a more central adipose distribution (central obesity) than female patients, who tend to have a more peripheral or gynecoid obesity.9 The excess visceral fat of central obesity can make bariatric surgery more challenging. Because MetS leads to a heightened
state of systemic inflammation, the patient’s ability to counter the stress of bariatric surgery may be compromised and may limit the body’s response to complications. Interestingly, gastric bypass has remained the most commonly performed bariatric operation during the study time period, and more than 95% of all cases were performed via laparoscopic approach, with a 90-day mortality of less than 0.3%. Much has been written about the safety profile of the commonly performed bariatric surgery operations. Numerous studies have shown that laparoscopic adjustable gastric banding has the lowest incidence of adverse outcomes compared with the other bariatric procedures such as RYGB, sleeve gastrectomy, and BPD/DS.5,10 In our study, patients with MetS undergoing adjustable gastric banding had a lower incidence of mortality, serious complications, and readmissions compared with the other procedures; BPD/DS had the greatest incidence of adverse outcomes. The superior safety profile of adjustable gastric banding, however, was at the expense of decreased remission of hypertension, diabetes, dyslipidemia, and sleep apnea in patients with MetS. These findings are consistent with the largest meta-analyses examining obesity and diabetes remission rates after bariatric surgery.5,10 In our study, diabetes remission was achieved in only 28% of adjustable gastric banding patients at 12 months compared with more
Table 6. Twelve-Month, Procedure-Specific, Postoperative Efficacy Outcomes in Patients with Metabolic Syndrome Outcomes
Body mass index, kg/m2 Preoperative 12-month MetS remission, n (%) Hypertension Diabetes Dyslipidemia
Adjustable gastric banding (n ⴝ 4,245)
Gastric bypass (n ⴝ 7,285)
Sleeve gastrectomy (n ⴝ 406)
BPD/DS (n ⴝ 208)
45.5 (7.03) 38.5 (6.89)
47.6 (7.99) 32.4 (6.20)
48.6 (9.03) 36.1 (7.31)
51.0 (9.36) 31.8 (6.63)
800 (18.8) 1,206 (28.4) 734 (17.3)
3,267 (44.8) 4,532 (62.2) 3,271 (44.9)
143 (35.2) 211 (52.0) 139 (34.2)
110 (52.9) 154 (74.0) 135 (64.9)
n represents the number of patients with follow-up data. The visit window 9 and 15 months was used for the 12-month follow-up visit. The follow-up visit closest to 12 months was used. BPD/DS, biliopancreatic diversion with duodenal switch; MetS, metabolic syndrome.
Vol. 214, No. 4, April 2012
than 50%, 60%, and 70% for sleeve gastrectomy, RYGB, and BPD/DS patients, respectively. Although this trend in improved efficacy across procedures tends to reflect the degree of weight loss and nutrient malabsorption induced by each procedure, recent evidence suggests that other factors including hormonal changes induced by bariatric surgery may play an important role in comorbidity resolution.11,12 In the case of RYGB, increases in systemic levels of glucagon-like peptide 1 have been implicated in the effects of RYGB on both the weight loss and diabetes remission effects of the surgery.13,14 Glucagonlike peptide 1 has been associated with delayed gastric emptying, which decreases appetite and leads to weight loss over time. Glucagon-like peptide 1 also stimulates insulin secretion through its suppression of glucagon, which leads to a decrease in hyperglycemia long before significant weight loss is achieved. This study had several limitations. The data entered into BOLD were self-reported by 1,157 surgeons performing bariatric surgery at 884 hospitals participating in the BSCOE program. A portion of the data was verified by on-site inspection of patient charts before center of excellence designation and on designation renewal every 3 years as outlined by the Surgical Review Corporation and the requirements of the ASMBS centers of excellence program. During site inspection, all operations reported in BOLD were verified with the hospital surgical record; complications and readmissions reported in BOLD were verified by chart review; and 10% of total cases were selected at random for chart review. Any unreported reoperations, readmissions, deaths, transfers, or revisions found during chart review triggered a 100% chart review. Inconsistencies noted during site inspection were reported to the Bariatric Surgery Review Committee, a group of practicing surgeons responsible for reviewing all applications to the Bariatric Surgery Center of Excellence program, for clarifying the program requirements by issuing interpretive guidelines and for recommending changes to the program as needed. Surgical practices were responsible for entering all postdischarge data, including complications, even if they were managed by another health care provider. So, postdischarge complications may potentially be under-represented in the database. In addition, variations in the time lag between the occurrence of a postdischarge visit or event and when those data were entered into the database may appear as incomplete follow-up data. The use of the comorbidity severity scale rather than conventional objective measures to assess changes in the status of hypertension, diabetes, and dyslipidemia may underestimate the percentage of patients whose disease has gone into remission at a given time after bariatric surgery. For example, it is
Inabnet et al
Bariatric Surgery for Metabolic Syndrome
555
not uncommon for patients whose hypertension or increased serum cholesterol has normalized as a result of weight loss after surgery to be maintained on low-dose, antihypertensive medication or statins for an indefinite time as a precautionary measure. As defined by the comorbidity severity scale, these patients would meet the criteria of a level 3 disease state when they have substantially improved their comorbidity burden. Similarly, patients with type 2 diabetes who have resolved their requirement for insulin or decreased their dosage of oral antidiabetic therapy after bariatric surgery may still meet the criteria of a level 2 disease state and be excluded from the group of patients whose diabetes has resolved. Despite the limitations imposed by the comorbidity severity scale, a significant number of patients showed remission of MetS comorbidities using this standardized scoring method.
CONCLUSIONS Bariatric surgery can be performed safely, with overall low morbidity and mortality. Patients with MetS experience a greater incidence of early adverse outcomes compared with non-MetS patients. Although all bariatric operations led to an improvement in MetS, adjustable gastric banding led to the lowest remission rate compared with the other bariatric operations. Author Contributions
Study conception and design: Inabnet, Sarr Acquisition of data: Winegar, Sherif Analysis and interpretation of data: Inabnet, Winegar, Sherif Drafting of manuscript: Inabnet, Winegar, Sherif, Sarr Critical revision: Inabnet, Winegar, Sherif, Sarr
REFERENCES 1. Grundy SM, Brewer HB Jr, Cleeman JI, et al. Definition of metabolic syndrome: Report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition. Circulation 2004;109:433– 438. 2. McCullough AJ. Epidemiology of the metabolic syndrome in the USA. J Dig Dis 2011;12:333–340. 3. Ikramuddin S, Buchwald H. How bariatric and metabolic operations control metabolic syndrome. Br J Surg 2011;98:1339– 1341. 4. Richardson DW, Mason ME, Vinik AI. Update: metabolic and cardiovascular consequences of bariatric surgery. Endocrinol Metab Clin North Am 2011;40:81–⬎96, viii. 5. Buchwald H, Estok R, Fahrbach K, et al. Weight and type 2 diabetes after bariatric surgery: systematic review and metaanalysis. Am J Med 2009;122:248–256 e5. 6. DeMaria EJ, Pate V, Warthen M, Winegar DA. Baseline data from American Society for Metabolic and Bariatric Surgerydesignated Bariatric Surgery Centers of Excellence using the
556
7. 8. 9. 10. 11.
12.
13.
14.
Inabnet et al
Discussion
Bariatric Outcomes Longitudinal Database. Surg Obes Relat Dis 2010;6:347–355. Ali MR, Maguire MB, Wolfe BM. Assessment of obesity-related comorbidities: a novel scheme for evaluating bariatric surgical patients. J Am Coll Surg 2006;202:70–77. Varela JE, Hinojosa MW, Nguyen NT. Bariatric surgery outcomes in morbidly obese with the metabolic syndrome at US academic centers. Obes Surg 2008;18:1273–1277. Goncalves FB, Koek M, Verhagen HJ, et al. Body-mass index, abdominal adiposity, and cardiovascular risk. Lancet 2011;378: 227; author reply 228. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724–1737. Korner J, Bessler M, Inabnet W, et al. Exaggerated glucagon-like peptide-1 and blunted glucose-dependent insulinotropic peptide secretion are associated with Roux-en-Y gastric bypass but not adjustable gastric banding. Surg Obes Relat Dis 2007; 3:597–601. Korner J, Inabnet W, Febres G, et al. Prospective study of gut hormone and metabolic changes after adjustable gastric banding and Roux-en-Y gastric bypass. Int J Obes (Lond) 2009;33:786– 795. Borg CM, le Roux CW, Ghatei MA, et al. Progressive rise in gut hormone levels after Roux-en-Y gastric bypass suggests gut adaptation and explains altered satiety. Br J Surg 2006;93:210– 215. le Roux CW, Aylwin SJ, Batterham RL, et al. Gut hormone profiles following bariatric surgery favor an anorectic state, facilitate weight loss, and improve metabolic parameters. Ann Surg 2006;243:108–114.
Discussion DR BRUCE SCHIRMER (Charlottesville, VA): Dr Inabnet and his colleagues have confirmed by this study the previously held hypothesis that metabolic syndrome confers increased potential for morbidity in bariatric patients. I do congratulate them on an excellent presentation and a well-written manuscript. Some personal observations that I note include the fact that the Bariatric Outcomes Longitudinal Database (BOLD), though confirming the remarkable safety and the efficacy of all the procedures, does have a lower overall weight loss and resolution of comorbidities than have previously been reported in single-institution studies. Also, the lap band in the BOLD database is less effective at producing resolution of comorbidities than in reports from either US or international single-institution studies. However, before conclusions are drawn regarding the lap band, 2- or 3-year follow-up data would be useful, as other studies have shown continued weight loss for that operation during that time frame. My questions for Dr Inabnet are as follows: First, the 50% average follow-up of patients at 1 year is certainly disappointing for the database and well below original expectations. Do you have data on the percentage follow-up for the 30- and 90-day calculations? Are these used for morbidity calculations? Because, if not, the denominator would be lower and the percentage of complications would actually be slightly higher.
J Am Coll Surg
Second, how did you calculate the resolution of dyslipidemia? Level 2, which was considered present, was defined as controlled with lifestyle changes; level 1, absent, is defined as present but no treatment required. Differentiating between these would seem to be relatively subjective. Third, why do you think the incidences of sleep apnea, which were 45% in the metabolic group and 25% in the rest of the group, were so different? Most studies of sleep apnea have shown that the incidence seems to be related to body mass index, which was comparable for the 2 groups. Next, do you have any data on body habitus as described in your manuscript? We all know that central obesity is certainly more difficult in terms of the technical aspects of operating on patients. Do you have any data on liver disease and whether or not nonalcoholic steatohepatitis (NASH) was present in the metabolic syndrome patients or in the database? I suspect that it would be significantly increased in metabolic syndrome. Finally, given a patient with known metabolic syndrome and fatty liver disease, do you recommend a preoperative diet to lose weight? If so, have you found this effective in improving operative outcomes? DR WILLIAM RICHARDS (Mobile, AL): Dr Inabnet and colleagues are to be congratulated on an outstanding contribution to the literature and to our understanding of the impact of bariatric surgery on patients with metabolic syndrome. One of the knocks against surgical publications for years was the retrospective collection of data from a single surgeon or institution purporting superior results with a surgical procedure, while the internal medicine literature routinely fills the literature with reports of prospective trials of medications investigating thousands of patients. This paper is important because it’s another one of the studies emanating from the prospective collection of data across the US from hundreds of surgeons, and convinces me that we are ever more confident in reporting the positive results that are obtained, not just by a single superstar surgeon, but in actual practice. The bariatric surgical community in particular has come a long way since a report by Flum and his colleagues (JAMA 2005;294: 1903–1908) about the high mortality rate in Medicare beneficiaries undergoing bariatric surgery in 2005. My first question is to put the mortality and complication rates in perspective and help us understand how we have lowered mortality. Flum’s data suggested that patients greater than 75 years of age had a mortality rate 5 times greater than patients aged 65 to 74. Do your data suggest that surgeons are operating on younger patients and excluding the exorbitantly high risk elderly patients or has the center of excellence system improved outcomes while continuing to operate on these high risk 75-year-old patients with metabolic syndrome? My second question concerns the reduced improvement in diabetes compared to other studies that Dr Schirmer also pointed out. My colleague, Alfonso Torquati, and I (J GI Surg 2005;9:1112–1118) determined in a multivariate analysis of gastric bypass in diabetics that smaller waist circumference and treatment of diabetics without insulin preoperatively predicted increased resolution of diabetes after gastric bypass. Because the BOLD database characterizes the diabetics as needing insulin or not, and also has information on waist
Metabolic and Bariatric Surgery Bariatric Surgery Center of Excellence Program, was queried to identify patients undergoing bariatric surgery from June 2007 through November 2010. MetS was defined as the presence of hypertension, diabetes, and dyslipidemia at presentation for bariatric surgery. Ninety-day and 1-year outcomes were assessed to determine early outcomes in bariatric surgery patients with MetS. RESULTS: Among 186,576 research-consented patients, 23,106 (12%) were diagnosed with MetS. Patients with MetS were more likely to be male (35% vs 20%; p ⬍ 0.0001), older (mean age 54 vs 44 years; p ⬍ 0.0001), and Caucasian (81% vs 74%; p ⬍ 0.0001). Of the 23,106 MetS patients, more underwent gastric bypass (RYGB) (62%) compared with gastric banding (32%), sleeve gastrectomy (4.5%), and biliopancreactic diversion with duodenal switch (BPD/ DS)(1.5%). MetS patients had an increase in serious complications (2.4% vs 1.0%; p ⬍ 0.0001), readmissions (6.2% vs 4.7%; p ⬍ 0.0001), and mortality (0.3% vs 0.1%; p ⬍ 0.0001) within 90 days of operation. After adjusting for sex, age, and body mass index, RYGB patients with MetS had an increased risk of 90-day serious complications compared to RYGB patients without MetS (odds ratio 1.43; 95% CI, 1.27 to 1.61; p ⬍ 0.0001). The 12-month remission rate of diabetes was least for gastric banding (28%) compared with the other procedures (RYGB 62%, sleeve gastrectomy 52%, BPD/DS 74%). CONCLUSIONS: Patients with MetS undergoing bariatric surgery showed dramatic improvement in diabetes 1-year after surgery; however, an adverse 90-day outcome was more common. (J Am Coll Surg 2012;214:550–557. © 2012 by the American College of Surgeons)
verse outcomes in patients undergoing bariatric surgery.4 Contributing factors to this perception include the severe central and visceral adiposity, which can lead to hepatomegaly and intraoperative technical challenges, the presence of moderate to severe diabetes, and the overall poor health of many patients with MetS. To date, most studies examining MetS in bariatric surgery patients are small, single institution studies that report results on 1 or 2 procedures that are preferred at the local institution.5 With the recent implementation of Bariatric Surgery Center of Excellence programs (BSCOE), which require mandatory data entry into 1 of 2 national registries, investigators are able to examine a large cohort of patients from a diverse demographic background, including community, rural-based, and urban medical centers with both private and academic affiliations. The objective of our analysis was to determine the prevalence of MetS in patients undergoing bariatric surgery at American Society for Metabolic and Bariatric Surgery (ASMBS) BSCOEs, and to
Metabolic syndrome (MetS) occurs as a constellation of metabolic abnormalities, including type 2 diabetes mellitus, hypertension, and dyslipidemia, that lead to an increased risk for cardiovascular disease.1 The true cause of MetS is not known, but there is an intertwined link between obesity, insulin resistance, and the MetS, which leads to a vicious cycle of metabolic stress.2,3 It has long been assumed that the presence of MetS is a risk factor for adDisclosure Information: Dr Winegar and Mr Sherif are salaried employees of Surgical Review Corporation. All other authors have nothing to disclose. Presented at the Southern Surgical Association 123rd Annual Meeting, Hot Springs, VA, December 2011. Received December 16, 2011; Accepted December 19, 2011. From Mount Sinai Medical Center, New York, NY (Inabnet); Surgical Review Corporation, Raleigh, NC (Winegar, Sherif ); and Mayo Clinic, Rochester, MN (Sarr). Correspondence address: William B Inabnet III, MD, Division of Metabolic, Endocrine and Minimally Invasive Surgery, Department of Surgery, Mount Sinai Medical Center, 5 East 98th St, Box 1259, New York, NY 10029. email: [email protected]
© 2012 by the American College of Surgeons Published by Elsevier Inc.
550
ISSN 1072-7515/12/$36.00 doi:10.1016/j.jamcollsurg.2011.12.019
Vol. 214, No. 4, April 2012
Abbreviations and Acronyms
ASMBS ⫽ American Society for Metabolic and Bariatric Surgery BMI ⫽ body mass index BOLD ⫽ Bariatric Outcomes Longitudinal Database BPD/DS⫽ biliopancreatic diversion with duodenal switch BSCOE ⫽ Bariatric Surgery Centers of Excellence MetS ⫽ metabolic syndrome RYGB ⫽ Roux-en-Y gastric bypass
assess the 90-day safety outcomes. In addition, a secondary aim was to determine the effectiveness of bariatric surgery on comorbidity resolution at 1 year in patients with MetS. Our hypothesis was that operative morbidity and mortality would be increased in MetS patients undergoing bariatric surgery but that the MetS would have a high rate of remission in the MetS cohort.
METHODS This study used the Bariatric Outcomes Longitudinal Database (BOLD), a national database that tracks outcomes of bariatric operations performed at ASMBS BSCOEs. The ASMBS designation as a Center of Excellence requires that centers register all patients who undergo any bariatric operation at their center and perform follow-up evaluations at specific times postoperatively. The data are monitored and audited for accuracy. Further details of the BOLD database have been previously reported.6 The study population included patients aged 18 to 75 years, with body mass index (BMI) ⬎ 35 kg/m2, who had Roux-en-Y gastric bypass (RYGB), adjustable gastric banding, sleeve gastrectomy, or biliopancreactic diversion with duodenal switch (BPD/DS) bariatric surgery performed by a BSCOE participant between June 2007 and November 2010. Preoperative data on patient age, sex, race, insurance status, last BMI before bariatric surgery, previous bariatric surgery history, and selective comorbidities were obtained from the database. Intraoperative data obtained included operative procedure, American Society of Anesthesiologists (ASA) classification, duration of the operation, and duration of the postoperative stay. The status of metabolic comorbidities was assessed at both preoperative and postoperative visits using a modified version of a scoring system developed to evaluate obesityrelated conditions in bariatric surgical patients.7 The scoring system assigns a numerical value (0 to 5) to indicate the severity of the disease as indicated and allows tracking of changes in the disease state over time (Table 1). For this study, MetS was defined as the presence hypertension, diabetes, and dyslipidemia at presentation for bari-
Inabnet et al
Bariatric Surgery for Metabolic Syndrome
551
atric surgery. Therefore, patients were categorized as having MetS if their hypertension score was ⱖ3, their diabetes score was ⱖ2, and their dyslipidemia score was ⱖ2. Patients were categorized as having sleep apnea if their score was ⱖ3. The primary safety outcomes monitored were mortality, serious complications, hospital readmissions, and reoperations that occurred within 30 and 90 days of the bariatric operation. The primary efficacy outcomes monitored were weight loss and remission of metabolic comorbidities within 12 months after bariatric surgery. Remission was defined as cessation of all medical treatment for the given condition (ie, remission of sleep apnea if continuous positive airway pressure treatment was terminated). Efficacy assessments were performed during routine follow-up, when weight and comorbidities were assessed. The visit window 9 and 15 months was used for the 12month follow-up visit and the follow-up visit closest to 12 months was used for this study. Complications entered in BOLD included all deaths and unfavorable occurrences that prolonged hospitalization, required readmission to the emergency room or hospital, or required treatment outside of standard postoperative care. A readmission was defined as any hospital or facility stay that lasted for at least 24 hours, regardless of the location within the hospital or facility and regardless of the reason for the stay. A reoperation was defined as a surgical intervention of the primary bariatric surgery required as a result of a complication. Serious complications were defined as a composite that included the following complications: death, anastomotic leakage, cardiac arrest, deep venous thrombosis, evisceration, heart failure and/or pulmonary edema, liver failure, myocardial infarction, pneumothorax, pulmonary embolism, renal failure, respiratory failure, sepsis, systemic inflammatory response, cerebrovascular accident, and bleeding requiring blood transfusion. Statistical analysis
Descriptive statistics were generated to summarize the study cohorts. Continuous variables were summarized with mean and standard deviation. Categorical variables were summarized with frequencies and percentages. Summary data were tabulated in subgroups (MetS patients or nonMetS patients). The chi-square test for categorical variables and the Kruskal-Wallis test for continuous variables were used to compare the subgroups. A nonlinear, mixed effect model was used to compute multivariate adjusted odds ratios with corresponding 95% confidence intervals, using non-MetS patients as the reference group. The fully adjusted model controlled for preoperative characteristics with hospital as a random effect. Significance level was set
552
Inabnet et al
Bariatric Surgery for Metabolic Syndrome
J Am Coll Surg
Table 1. Modified Scoring System for Obesity-Associated Comorbidities Score
Hypertension
Diabetes
Dyslipidemia
Obstructive sleep apnea
0 1
No indication of hypertension Borderline, no medication
No symptoms or evidence of diabetes Elevated fasting glucose
2
Diagnosis of hypertension, no medication Treatment with single medication Treatment with multiple medications
Diabetes, controlled with oral medication Diabetes, controlled with insulin
Not present Present, no treatment required Controlled with lifestyle change Controlled with single medication Control with multiple medications Not controlled
No symptoms Symptoms, negative sleep study Diagnosed by sleep study, no oral appliance Treated with continuous positive airway pressure Significant hypoxia or oxygen dependent
3 4 5
Diabetes, controlled with insulin and oral medication Diabetes, with severe complications
at alpha ⫽ 0.05. All statistical analyses were performed using SAS statistical software Version 9.2.
RESULTS Demographic characteristics of entire cohort
Of the study population of 186,576 patients, 23,106 (12%) had MetS at the time of bariatric surgery. Patients with MetS were more likely to be male (35% vs 20%; p ⬍ 0.0001), older (mean age 54 vs 44 years; p ⬍ 0.0001), and
Caucasian (81% vs 74%; p ⬍ 0.0001) than non-MetS patients (Table 2). There was no difference in mean BMI between the two groups (47.0 vs 46.8 kg/m2), but patients with MetS were more likely to have obstructive sleep apnea (45% vs 25%, p ⬍ 0.0001) and an American Society of Anesthesiologists category of Class III or greater (84% vs 68%, p ⬍ 0.0001). Of the 23,106 MetS patients, more underwent RYGB (62%) than adjustable gastric banding (32%), sleeve gas-
Table 2. Preoperative Demographic Characteristics of Overall Cohort Characteristic
Sex, n (%) Male Female Age at operation, y Mean (SD) Median (min, max) Race, n (%) Caucasian African American Hispanic Asian Native American Pacific Islander/Hawaiian Other Body mass index, kg/m2 Mean (SD) Median (min, max) Sleep apnea, n (%) Treatment or oral medication required ASA classification, n (%) Class I Class II Class III Class IV Class V ASA, American Society of Anesthesiologists.
Metabolic syndrome (n ⴝ 23,106)
8,207 (36) 14,899 (65) 54.1 (9.46) 55 (18, 75) 18,710 (81) 2,057 (9) 1,186 (5) 62 (0.3) 135 (0.6) 32 (0.1) 924 (4) 47.0 (7.86) 46 (35, 106)
Nonmetabolic syndrome (n ⴝ 163,470)
32,249 (20) 131,221 (80) 44.3 (11.55) 44 (18, 75) 121,263 (74) 18,806 (12) 12,362 (8) 336 (0.2) 722 (0.4) 221 (0.1) 9760 (6) 46.8 (7.81) 45 (35, 106)
Overall (n ⴝ 186,576)
40,456 (22) 146,120 (78) 45.5 (11.76) 45 (18, 75) 139,973 (75) 20,863 (11) 13,548 (7) 398 (0.2) 857 (0.5) 253 (0.1) 10684 (6) 46.9 (7.82) 45 (35, 106)
10,346 (45)
40,859 (25)
51,205 (27)
387 (2) 3,294 (14) 17,537 (76) 1,868 (8) 20 (0.1)
6,426 (4) 45,903 (28) 105,632 (65) 5,398 (3) 111 (0.1)
6,813 (4) 49,197 (26) 123,169 (66) 7,266 (4) 131 (0.1)
Vol. 214, No. 4, April 2012
Bariatric Surgery for Metabolic Syndrome
Inabnet et al
553
Table 3. Postoperative Safety Outcomes
Outcomes
30-d mortality 30-d serious complications 30-d readmission 30-d reoperation 90-d mortality 90-d serious complications 90-d readmission 90-d reoperation
Metabolic syndrome (n ⴝ 23,106) n %
Overall (n ⴝ 1,86,576) n %
173 1,941 6,847 3,880 233 2,132 9,100 6,267
0.1 1.0 3.7 2.1 0.1 1.1 4.9 3.4
52 516 1,076 605 65 558 1,427 954
Nonmetabolic syndrome (n ⴝ 163,470) n %
0.2 2.2 4.7 2.6 0.3 2.4 6.2 4.1
121 1,425 5,771 3,275 168 1,574 7,673 5,313
0.1 0.9 3.5 2.0 0.1 1.0 4.7 3.3
p Value*
⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001
*p Values for the comparison between groups were calculated using chi-square test.
trectomy (4.5%), and BPD/DS (1.5%). In contrast, among non-MetS patients, 53% underwent RYGB, followed by adjustable gastric banding (40%), sleeve gastrectomy (6%), and BPD/DS (1%). A laparoscopic approach was used equally in both groups (MetS 94%, non-MetS 95%). The mean durations of operation and hospital stay were greater in MetS patients (99 vs 73 minutes, p ⬍ 0.0001 and 2.1 vs 1.7 days, p ⬍ 0.0001, respectively). Safety outcomes data
MetS patients had a higher incidence of serious complications (2.4% vs 1.0%; p ⬍ 0.0001), readmissions (6.2% vs 4.7%; p ⬍ 0.0001), and mortality (0.3 vs 0.1%; p ⬍ 0.0001) within 30 and 90 days of operation than non-MS patients (Table 3). The incidence of serious complications and mortality was least for adjustable gastric banding and greatest for BPD/DS (p ⬍ 0.0001) (Table 4). After adjusting for sex, age, and BMI, both gastric banding and RYGB patients with MetS had an increased risk of 90-day serious complications compared with patients without MetS (gastric banding odds ratio [OR] 2.79; 95%
CI, 2.02 to 3.84, p ⬍ 0.0001) (RYGB [OR] ⫽ 1.43; 95% CI, 1.27 to 1.61; p ⬍ 0.0001) (Table 5). 12-month efficacy on comorbidity resolution
Of the entire cohort of 186,576 patients, 12-month postoperative data were available on 92,915 patients (50%), including 12,144 (53%) from the MetS group and 80,177 (49%) from the non-MetS group. In patients with MetS, remission of hypertension, diabetes, and dyslipidemia occurred in 36%, 50%, and 35% of patients, respectively; 76% of MetS patients with sleep apnea went off all forms of treatment 12 months after the bariatric operation. Among procedures, adjustable gastric banding had a lower 12month rate of remission of MetS compared with RYGB, sleeve gastrectomy, and BPD/DS (Table 6).
DISCUSSION These data demonstrate several interesting findings in patients with morbid obesity complicated by MetS who underwent bariatric surgery. Though patients with MetS pre-
Table 4. Procedure-Specific Postoperative Safety Outcomes in Patients with Metabolic Syndrome
Outcomes
30-d mortality 30-d serious complications 30-d readmission 30-d reoperation 90-d mortality 90-d serious complications 90-d readmission 90-d reoperation
Adjustable gastric banding (n ⴝ 7,357) n %
4 63 183 84 5 67 231 134
0.1 0.9 2.5 1.1 0.1 0.9 3.1 1.8
Roux-en-Y gastric bypass (n ⴝ 14,329) n %
41 411 811 474 53 445 1094 754
0.3 2.9 5.7 3.3 0.4 3.1 7.6 5.3
Sleeve gastrectomy (n ⴝ 1,081) n %
BPD w/ duodenal switch (n ⴝ 339) n %
p Value*
3 22 43 28 3 24 51 38
4 20 39 19 4 22 51 28
⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001 ⬍0.0001
0.3 2.0 4.0 2.6 0.3 2.2 4.7 3.5
*p Values for the comparison between MS groups were calculated using chi-square test and Fisher’s exact test. BPD, biliopancreatic diversion.
1.2 5.9 11.5 5.6 1.2 6.5 15.0 8.3
554
Inabnet et al
Bariatric Surgery for Metabolic Syndrome
J Am Coll Surg
Table 5. Ninety-Day Serious Complications Regression Models* Adjustable gastric banding Odds ratio (95%CI) p Value
Covariates
Sex (ref, female) Race (ref, Caucasian) BMI (each 1-kg/m2 increment) Age (each 1-year increment) MetS (ref, non-MetS) Sleep apnea (ref, no history or symptoms)
1.51 (1.13, 2.02) — 1.06 (1.05, 1.08) 1.03 (1.02, 1.04) 2.79 (2.02, 3.84) —
0.0059 NS ⬍0.0001 ⬍0.0001 ⬍0.0001 NS
Roux-en-Y gastric bypass Odds ratio (95%CI) p Value
1.49 (1.34, 1.65) — 1.04 (1.03, 1.04) 1.04 (1.03, 1.04) 1.43 (1.27, 1.61) —
⬍0.0001 NS ⬍0.0001 ⬍0.0001 ⬍0.0001 NS
*Based on non-linear mixed effect model which includes hospital as a random effects. BMI, body mass index; MetS, metabolic syndrome; NS, Not significant.
sented with a similar preoperative weight profile as nonMetS patients (mean BMI 47 kg/m2), MetS patients as a whole experienced an increased, albeit infrequent, incidence of serious complications and mortality compared with non-MetS patients. Because the mean weight was similar between the two groups, factors other than weight invariably contribute to the greater incidence of adverse outcomes in MetS patients. For patients with MetS, the incidences of serious complications and mortality 90 days after bariatric surgery were 2.4% and 0.3%, respectively, which was significantly higher than that for non-MetS patients. In another study that evaluated a large cohort of obese patients with MetS who underwent bariatric surgery at academic institutions, the morbidity (8.6%) and mortality (0.4%) were greater than in our study.8 Nevertheless, achieving a decrease in mortality and all-cause morbidity in this high risk patient population is an amazing feat and supports the utility of outcomes tracking mandated by the center of excellence programs. In our study, the MetS cohort was older and had a greater percentage of males than non-MetS patients. In general, male patients tend to have a more central adipose distribution (central obesity) than female patients, who tend to have a more peripheral or gynecoid obesity.9 The excess visceral fat of central obesity can make bariatric surgery more challenging. Because MetS leads to a heightened
state of systemic inflammation, the patient’s ability to counter the stress of bariatric surgery may be compromised and may limit the body’s response to complications. Interestingly, gastric bypass has remained the most commonly performed bariatric operation during the study time period, and more than 95% of all cases were performed via laparoscopic approach, with a 90-day mortality of less than 0.3%. Much has been written about the safety profile of the commonly performed bariatric surgery operations. Numerous studies have shown that laparoscopic adjustable gastric banding has the lowest incidence of adverse outcomes compared with the other bariatric procedures such as RYGB, sleeve gastrectomy, and BPD/DS.5,10 In our study, patients with MetS undergoing adjustable gastric banding had a lower incidence of mortality, serious complications, and readmissions compared with the other procedures; BPD/DS had the greatest incidence of adverse outcomes. The superior safety profile of adjustable gastric banding, however, was at the expense of decreased remission of hypertension, diabetes, dyslipidemia, and sleep apnea in patients with MetS. These findings are consistent with the largest meta-analyses examining obesity and diabetes remission rates after bariatric surgery.5,10 In our study, diabetes remission was achieved in only 28% of adjustable gastric banding patients at 12 months compared with more
Table 6. Twelve-Month, Procedure-Specific, Postoperative Efficacy Outcomes in Patients with Metabolic Syndrome Outcomes
Body mass index, kg/m2 Preoperative 12-month MetS remission, n (%) Hypertension Diabetes Dyslipidemia
Adjustable gastric banding (n ⴝ 4,245)
Gastric bypass (n ⴝ 7,285)
Sleeve gastrectomy (n ⴝ 406)
BPD/DS (n ⴝ 208)
45.5 (7.03) 38.5 (6.89)
47.6 (7.99) 32.4 (6.20)
48.6 (9.03) 36.1 (7.31)
51.0 (9.36) 31.8 (6.63)
800 (18.8) 1,206 (28.4) 734 (17.3)
3,267 (44.8) 4,532 (62.2) 3,271 (44.9)
143 (35.2) 211 (52.0) 139 (34.2)
110 (52.9) 154 (74.0) 135 (64.9)
n represents the number of patients with follow-up data. The visit window 9 and 15 months was used for the 12-month follow-up visit. The follow-up visit closest to 12 months was used. BPD/DS, biliopancreatic diversion with duodenal switch; MetS, metabolic syndrome.
Vol. 214, No. 4, April 2012
than 50%, 60%, and 70% for sleeve gastrectomy, RYGB, and BPD/DS patients, respectively. Although this trend in improved efficacy across procedures tends to reflect the degree of weight loss and nutrient malabsorption induced by each procedure, recent evidence suggests that other factors including hormonal changes induced by bariatric surgery may play an important role in comorbidity resolution.11,12 In the case of RYGB, increases in systemic levels of glucagon-like peptide 1 have been implicated in the effects of RYGB on both the weight loss and diabetes remission effects of the surgery.13,14 Glucagonlike peptide 1 has been associated with delayed gastric emptying, which decreases appetite and leads to weight loss over time. Glucagon-like peptide 1 also stimulates insulin secretion through its suppression of glucagon, which leads to a decrease in hyperglycemia long before significant weight loss is achieved. This study had several limitations. The data entered into BOLD were self-reported by 1,157 surgeons performing bariatric surgery at 884 hospitals participating in the BSCOE program. A portion of the data was verified by on-site inspection of patient charts before center of excellence designation and on designation renewal every 3 years as outlined by the Surgical Review Corporation and the requirements of the ASMBS centers of excellence program. During site inspection, all operations reported in BOLD were verified with the hospital surgical record; complications and readmissions reported in BOLD were verified by chart review; and 10% of total cases were selected at random for chart review. Any unreported reoperations, readmissions, deaths, transfers, or revisions found during chart review triggered a 100% chart review. Inconsistencies noted during site inspection were reported to the Bariatric Surgery Review Committee, a group of practicing surgeons responsible for reviewing all applications to the Bariatric Surgery Center of Excellence program, for clarifying the program requirements by issuing interpretive guidelines and for recommending changes to the program as needed. Surgical practices were responsible for entering all postdischarge data, including complications, even if they were managed by another health care provider. So, postdischarge complications may potentially be under-represented in the database. In addition, variations in the time lag between the occurrence of a postdischarge visit or event and when those data were entered into the database may appear as incomplete follow-up data. The use of the comorbidity severity scale rather than conventional objective measures to assess changes in the status of hypertension, diabetes, and dyslipidemia may underestimate the percentage of patients whose disease has gone into remission at a given time after bariatric surgery. For example, it is
Inabnet et al
Bariatric Surgery for Metabolic Syndrome
555
not uncommon for patients whose hypertension or increased serum cholesterol has normalized as a result of weight loss after surgery to be maintained on low-dose, antihypertensive medication or statins for an indefinite time as a precautionary measure. As defined by the comorbidity severity scale, these patients would meet the criteria of a level 3 disease state when they have substantially improved their comorbidity burden. Similarly, patients with type 2 diabetes who have resolved their requirement for insulin or decreased their dosage of oral antidiabetic therapy after bariatric surgery may still meet the criteria of a level 2 disease state and be excluded from the group of patients whose diabetes has resolved. Despite the limitations imposed by the comorbidity severity scale, a significant number of patients showed remission of MetS comorbidities using this standardized scoring method.
CONCLUSIONS Bariatric surgery can be performed safely, with overall low morbidity and mortality. Patients with MetS experience a greater incidence of early adverse outcomes compared with non-MetS patients. Although all bariatric operations led to an improvement in MetS, adjustable gastric banding led to the lowest remission rate compared with the other bariatric operations. Author Contributions
Study conception and design: Inabnet, Sarr Acquisition of data: Winegar, Sherif Analysis and interpretation of data: Inabnet, Winegar, Sherif Drafting of manuscript: Inabnet, Winegar, Sherif, Sarr Critical revision: Inabnet, Winegar, Sherif, Sarr
REFERENCES 1. Grundy SM, Brewer HB Jr, Cleeman JI, et al. Definition of metabolic syndrome: Report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition. Circulation 2004;109:433– 438. 2. McCullough AJ. Epidemiology of the metabolic syndrome in the USA. J Dig Dis 2011;12:333–340. 3. Ikramuddin S, Buchwald H. How bariatric and metabolic operations control metabolic syndrome. Br J Surg 2011;98:1339– 1341. 4. Richardson DW, Mason ME, Vinik AI. Update: metabolic and cardiovascular consequences of bariatric surgery. Endocrinol Metab Clin North Am 2011;40:81–⬎96, viii. 5. Buchwald H, Estok R, Fahrbach K, et al. Weight and type 2 diabetes after bariatric surgery: systematic review and metaanalysis. Am J Med 2009;122:248–256 e5. 6. DeMaria EJ, Pate V, Warthen M, Winegar DA. Baseline data from American Society for Metabolic and Bariatric Surgerydesignated Bariatric Surgery Centers of Excellence using the
556
7. 8. 9. 10. 11.
12.
13.
14.
Inabnet et al
Discussion
Bariatric Outcomes Longitudinal Database. Surg Obes Relat Dis 2010;6:347–355. Ali MR, Maguire MB, Wolfe BM. Assessment of obesity-related comorbidities: a novel scheme for evaluating bariatric surgical patients. J Am Coll Surg 2006;202:70–77. Varela JE, Hinojosa MW, Nguyen NT. Bariatric surgery outcomes in morbidly obese with the metabolic syndrome at US academic centers. Obes Surg 2008;18:1273–1277. Goncalves FB, Koek M, Verhagen HJ, et al. Body-mass index, abdominal adiposity, and cardiovascular risk. Lancet 2011;378: 227; author reply 228. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery: a systematic review and meta-analysis. JAMA 2004;292:1724–1737. Korner J, Bessler M, Inabnet W, et al. Exaggerated glucagon-like peptide-1 and blunted glucose-dependent insulinotropic peptide secretion are associated with Roux-en-Y gastric bypass but not adjustable gastric banding. Surg Obes Relat Dis 2007; 3:597–601. Korner J, Inabnet W, Febres G, et al. Prospective study of gut hormone and metabolic changes after adjustable gastric banding and Roux-en-Y gastric bypass. Int J Obes (Lond) 2009;33:786– 795. Borg CM, le Roux CW, Ghatei MA, et al. Progressive rise in gut hormone levels after Roux-en-Y gastric bypass suggests gut adaptation and explains altered satiety. Br J Surg 2006;93:210– 215. le Roux CW, Aylwin SJ, Batterham RL, et al. Gut hormone profiles following bariatric surgery favor an anorectic state, facilitate weight loss, and improve metabolic parameters. Ann Surg 2006;243:108–114.
Discussion DR BRUCE SCHIRMER (Charlottesville, VA): Dr Inabnet and his colleagues have confirmed by this study the previously held hypothesis that metabolic syndrome confers increased potential for morbidity in bariatric patients. I do congratulate them on an excellent presentation and a well-written manuscript. Some personal observations that I note include the fact that the Bariatric Outcomes Longitudinal Database (BOLD), though confirming the remarkable safety and the efficacy of all the procedures, does have a lower overall weight loss and resolution of comorbidities than have previously been reported in single-institution studies. Also, the lap band in the BOLD database is less effective at producing resolution of comorbidities than in reports from either US or international single-institution studies. However, before conclusions are drawn regarding the lap band, 2- or 3-year follow-up data would be useful, as other studies have shown continued weight loss for that operation during that time frame. My questions for Dr Inabnet are as follows: First, the 50% average follow-up of patients at 1 year is certainly disappointing for the database and well below original expectations. Do you have data on the percentage follow-up for the 30- and 90-day calculations? Are these used for morbidity calculations? Because, if not, the denominator would be lower and the percentage of complications would actually be slightly higher.
J Am Coll Surg
Second, how did you calculate the resolution of dyslipidemia? Level 2, which was considered present, was defined as controlled with lifestyle changes; level 1, absent, is defined as present but no treatment required. Differentiating between these would seem to be relatively subjective. Third, why do you think the incidences of sleep apnea, which were 45% in the metabolic group and 25% in the rest of the group, were so different? Most studies of sleep apnea have shown that the incidence seems to be related to body mass index, which was comparable for the 2 groups. Next, do you have any data on body habitus as described in your manuscript? We all know that central obesity is certainly more difficult in terms of the technical aspects of operating on patients. Do you have any data on liver disease and whether or not nonalcoholic steatohepatitis (NASH) was present in the metabolic syndrome patients or in the database? I suspect that it would be significantly increased in metabolic syndrome. Finally, given a patient with known metabolic syndrome and fatty liver disease, do you recommend a preoperative diet to lose weight? If so, have you found this effective in improving operative outcomes? DR WILLIAM RICHARDS (Mobile, AL): Dr Inabnet and colleagues are to be congratulated on an outstanding contribution to the literature and to our understanding of the impact of bariatric surgery on patients with metabolic syndrome. One of the knocks against surgical publications for years was the retrospective collection of data from a single surgeon or institution purporting superior results with a surgical procedure, while the internal medicine literature routinely fills the literature with reports of prospective trials of medications investigating thousands of patients. This paper is important because it’s another one of the studies emanating from the prospective collection of data across the US from hundreds of surgeons, and convinces me that we are ever more confident in reporting the positive results that are obtained, not just by a single superstar surgeon, but in actual practice. The bariatric surgical community in particular has come a long way since a report by Flum and his colleagues (JAMA 2005;294: 1903–1908) about the high mortality rate in Medicare beneficiaries undergoing bariatric surgery in 2005. My first question is to put the mortality and complication rates in perspective and help us understand how we have lowered mortality. Flum’s data suggested that patients greater than 75 years of age had a mortality rate 5 times greater than patients aged 65 to 74. Do your data suggest that surgeons are operating on younger patients and excluding the exorbitantly high risk elderly patients or has the center of excellence system improved outcomes while continuing to operate on these high risk 75-year-old patients with metabolic syndrome? My second question concerns the reduced improvement in diabetes compared to other studies that Dr Schirmer also pointed out. My colleague, Alfonso Torquati, and I (J GI Surg 2005;9:1112–1118) determined in a multivariate analysis of gastric bypass in diabetics that smaller waist circumference and treatment of diabetics without insulin preoperatively predicted increased resolution of diabetes after gastric bypass. Because the BOLD database characterizes the diabetics as needing insulin or not, and also has information on waist