BARIATRIC (METABOLIC) SURGERY FOR LIFE-LONG WEIGHT CONTROL AND REMISSION OF ASSOCIATED CHRONIC DISEASES

1
BARIATRIC (METABOLIC) SURGERY FOR LIFE-LONG
WEIGHT CONTROL AND REMISSION OF ASSOCIATED
CHRONIC DISEASES A METAANALYSIS OF PUBLISHED
CLINICAL PAPERS TO-DATE

• Henry Buchwald, M.D., Ph.D.
• University of Minnesota

2
DISCLOSURES

• Henry Buchwald, MD, PhD
• Consultant or Research Support
• Ethicon Endo-Surgery
• MetaCure
• W.L. Gore

3
BARIATRIC SURGERY A REVIEW AND METAANALYSIS

• Henry Buchwald, MD, PhD, University of Minnesota
• Yoav Avidor, MD, Ethicon Endo-Surgery, Inc.
• Eugene Braunwald, MD, Harvard Medical School
• Michael D. Jensen, MD Mayo Clinic
• Walter Pories, MD, East Carolina University
• Kyle Fahbach, PhD, MetaWorks
• Karen Schoelles, MD, MetaWorks
• JAMA 20042921724-1737

4
OBJECTIVES

• Primary To determine the impact of bariatric
  surgery on 4 of the major obesity comorbidities
  diabetes, hyperlipidemia, hypertension, and
  obstructive sleep apnea.
• Secondary To determine the weight loss and
  operative mortality in the studies selected for
  review and metaanalysis.

5
METHODS

• Broad electronic search of the English-language
  literature, 1990-2003, using MEDLINE, Currents
  Contents, and the Cochran Library databases.
• Manual reference checks.
• Two levels of screening.

6
METHODS LEVEL 1 SCREENING

• Abstract review, exclusion criteria
• Publication of abstract only
• Case reports
• Letters
• Comments
• Reviews
• Animal or in vitro studies
• Fewer than 10 patients
• Follow-up less than 10 days
• Language other than English
• No surgical intervention
• Intragastric balloon therapy (experimental)

7
METHODS STATISTICAL ANALYSIS

• A random effects model was used for the
  metaanalysis (results expressed as means and
  CIs).
• Weighted means were used for a non-metaanalytic
  comparison.

8
RESULTS DATA RETRIEVAL

• 2738 citations identified

1772 studies rejected prescreening
961 studies retrieved
253 Studies rejected by screening
708 studies
572 studies for catalog only
136 studies (91 kin) qualifying for metaanalysis
9
RESULTS STUDIES SELECTED

• Total 134 primary studies (2 health care
  economics studied excluded from 136 for no
  efficacy or mortality data) (179 study groups,
  22,049 patients)
• 5 randomized controlled trials (9 study
  groups, 621 patients)
• 28 nonrandomized controlled trials (48 study
  groups, 4,613 patients)
• 101 uncontrolled case series (122 study
  groups, 16,860 patients)

10
RESULTS STUDY CHARACTERISTICS

• 56 studies North America
• 58 studies Europe
• 20 studies elsewhere

11
RESULTS PATIENT CHARACTERISTICS

• Gender 19 men
• 73 women
• 8 not reported
• Age?x 39 (range, 16-64)
• Baseline BMI?x 46.85 (range, 32.30-68.80)

12
RESULTS WEIGHT LOSS

• Total Population
• EWL 61.2 (95 CI, 58.1 - 64.4)
• ? BMI 14.2 kg/m2 (95 CI, 13.3 - 15.1)
• ? Absolute Weight 39.7 kg (95 CI, 37.2
  42.2)
• All weight loss reductions p lt 0.001

13
RESULTS WEIGHT LOSS

• Surgical Groups EWL (95 CI)
• Gastric Banding 47.5 (40.7 54.2)
• Gastric Bypass 61.6 (56.7 66.5)
• Gastroplasty 68.2 (61.5 74.8)
• Biliopancreatic Diversion/ Duodenal
  Switch 70.1 (66.3 73.9)
• All weight loss reductions p lt 0.001

14
RESULTS OPERATIVE MORTALITY (? 30 DAYS)

• Purely restrictive 0.1
• (n2,297 gastric banding, n749 gastroplasty)
• Gastric Bypass 0.5
• (n5,644)
• Biliopancreatic diversion/duodenal switch 1.1
• (n3,030)

15
RESULTS OUTCOMES DIABETES

• Total Population With Diabetes
• Resolution 76.8 (70.7-82.9)
• Resolution or Improvement 86.0 (78.4-93.7)
• Reduction FBG 13.33 mg/dL (10.81-15.86)
• All values p lt 0.01

16
RESULTS OUTCOMES DIABETES

• Surgical Groups Resolution
• Gastric Banding 47.9 (29.1-66.7)
• Gastroplasty 71.6 (55.1-88.2)
• Gastric Bypass 83.7 (77.3-90.1)
• Biliopancreatic Diversion/ Duodenal
  Switch 98.9 (96.8-100)
• All values p lt 0.01

17
RESULTS OUTCOMES HYPERLIPIDEMIA

• Patients Improved With Hyperlipidemia
• Total Population 79.3 (68.2-90.5)
• Gastric Banding 58.9 (28.2-89.6)
• Gastroplasty 73.6 (60.8-86.3)
• Gastric Bypass 96.9 (93.6-100.0)
• Biliopancreatic Diversion/ Duodenal
  Switch 99.1 (97.6-100.0)
• All values p lt 0.01

18
RESULTS OUTCOMES HYPERLIPIDEMIA

• Total Population With Hyperlipidemia
• Total Chol LDL-Chol HDL-Chol Triglycerides
• Change -33.20 mg/dL -29.34 mg/dL 2.70 mg/dL 79.65
  mg/dL
• CI 23.17-43.63 17.76-40.93 0-5.79 64.60-95.58
• p value lt0.01 lt 0.01 lt 0.1 lt0.01

19
RESULTS OUTCOMES - HYPERTENSION

• Total Population With Hypertension
• Resolution 61.7 (55.6-67.8)
• Resolution or Improvement 78.5 (70.8-86.1)
• All values plt0.01

20
RESULTS OUTCOMES OBSTRUCTIVE SLEEP APNEA

• Total Populations With Obstructive Sleep Apnea
• Resolution 85.7 (79.2-92.2)
• Resolution or Improvement 83.6 (71.8-95.4)
• Apneas or Hypopneas -33.85/hr (17.47-50.23)
• All values plt0.01

21
SUMMARY

• Review and metaanalysis of 136 studies involving
  22,094 patients.
• Results
• ?x EWL 61.2
• ?x Operative (? 30 days) Mortality 0.1-1.1
• Resolution of Diabetes 76.8
• Resolution of Hyperlipidemia 79.3
• Resolution of Hypertension 61.7
• Resolution of Obstructive Sleep Apnea 85.7

22
CONCLUSION

• A substantial majority of morbidly obese patients
  with diabetes, hyperlipidemia, hypertension,
  and/or obstructive sleep apnea have total
  resolution or marked improvement of their
  comorbid conditions after bariatric surgery.

23
EFFECTS OF BARIATRIC SURGERY ON TYPE 2 DIABETES
A SYSTEMATIC REVIEW AND METAANALYSIS

• Henry Buchwald, MD, PhD,1 Rhonda Estok, RN, BSN,2
  
• Kyle Fahrbach, PhD,2 Deirdre Banel, BS,2
• Michael D. Jensen, MD,3 Walter Pories, MD,4 John
  Bantle, MD,1 Isabella Sledge, MD, MPH2
• 1University of Minnesota, Minneapolis, MN
• 2United BioSource Corporation, Medford, MA
• 3Mayo Clinic College of Medicine, Rochester, MN
• 4East Carolina University School of Medicine,
  Greenville, NC
• Am J Med 2009122248-256

24
DIABETES METAANALYSIS

• GOAL To determine the impact of bariatric
  surgery procedures on type 2 diabetes mellitus in
  association with the weight reduction achieved.

25
DIABETES METAANALYSIS

• METHODS
• Screening of all papers published in English,
  from January 1, 1990 to April 30, 2006,
  identified through electronic searches in
  MEDLINE, Current Contents,? and the Cochran
  Library, supplemented by manual reference checks.
• All accepted studies were assigned a level of
  evidence (Centre for Evidence-Based Medicine,
  Oxford, UK), and randomized controlled trials
  were rated for quality by the Jadad scoring
  method.

26
DIABETES METAANALYSIS

• METHODS
• Restricted, maximum likelihood, random-effects
  metaanalyses (REM) were performed and
  heterogeneity was assessed using Cochrans Q
  statistic.
• Diabetes outcomes assessed for lt2 years and ?2
  years.

27
DIABETES METAANALYSIS

• RESULTS Data Set
• 621 studies (136 in 2004)
• 888 treatment arms (179 in 2004)
• 135,246 patients (22,094 in 2004)

28
DIABETES METAANALYSIS

• RESULTS Study Characteristics
• Study Location
• Europe 44.4
• North America 43.2
• Australia/New Zealand 3.5
• South America 3.1
• Asia 1.5
• Others 4.5

29
DIABETES METAANALYSIS

• RESULTS Patient Characteristics
• Total
• Mean Age 40.2 years
• Mean BMI 47.9 kg/m2
• Gender Male 19.9
• Female 79.6
• Type 2 Diabetes 22.3

30
DIABETES METAANALYSIS

• RESULTS
• Weight Reduction Metaanalysis EBWL
• Overall
• Total Gastric Banding Gastroplasty Gastric
  Bypass BPD/DS
• 55.9 46.2 55.5 59.5
  63.6

31
DIABETES METAANALYSIS

• RESULTS Diabetes Outcomes Metaanalysis
• Overall

Total Gastric Banding Gastroplasty Gastric Bypass BPD/DS
Resolved 78.1 56.7 79.7 80.3 95.1
Resolved or Improved 86.6 80.6 87.2 84.5 99.0
Improved 37.4 38.1 38.5 46.3 14.0
Unchanged 14.3 18.1 12.0 21.2 1.9
Worsened 1.1 3.8 0.3
Insulin (pml/L) -97.4 -68.6 -102.9 -114.6 -135.4
HbA1c () -1.3 -0.5 -1.6 -2.0
Glucose (mmol/L) -1.4 -0.6 -0.7 -1.8 -2.8
32
DIABETES METAANALYSIS

• RESULTS Diabetes Outcomes Metaanalysis
• lt 2 Years

Total Gastric Banding Gastroplasty Gastric Bypass BPD/DS
Resolved 80.3 55.0 81.4 81.6 94.0
Resolved or Improved 86.0 82.3 89.0 84.0 100.0
Improved 48.7 51.6 45.8 50.6 32.1
Unchanged 12.0 7.8 22.8 20.6 0.0
Worsened 0.4 0.37 0.3
Insulin (pml/L) -86.7 -63.0 -88.3 -116.7 -98.8
HbA1c () -1.3 -0.5 -1.6 -2.0
Glucose (mmol/L) -1.2 -0.5 -0.6 -1.8 -1.7
33
DIABETES METAANALYSIS

• RESULTS Diabetes Outcomes Metaanalysis
• ? 2 Years

Total Gastric Banding Gastroplasty Gastric Bypass BPD/DS
Resolved 74.6 58.3 77.5 70.9 95.9
Resolved or Improved 87.2 78.7 84.4 85.3 98.9
Improved 19.4 14.7 33.0 27.6 5.0
Unchanged 16.1 24.7 10.1 22.4 2.0
Worsened 2.8 5.9
Insulin (pml/L) -188.0 -76.9 -152.9 -97.8
HbA1c () -1.4 -0.8
Glucose (mmol/L) -2.0 -1.0 -1.14 -1.5 -4.3
34
DIABETES METAANALYSIS

• RESULTS Diabetes Outcomes Metaanalysis
  Diabetic Patients Only Overall

Total Gastric Banding Gastric Bypass BPD/DS
Resolved 79.3 62.7 80.5 99.4
Resolved or Improved 98.9 95.7 100.0
Improved 23.6 30.1 20.4
Unchanged 1.5 4.3 0.0 0.7
Insulin (pml/L) -98.0 -99.6 -40.2 -125.0
HbA1c () -2.1 -1.4 -2.2
Glucose (mmol/L) -4.4 -2.5 -3.9 -5.4
35
DIABETES METAANALYSIS

• RESULTS Diabetes Outcomes Metaanalysis
• Patients Resolved Associated With Weight
  Reduction Achieved and Procedure Performed

Total Gastric Banding Gastroplasty Gastric Bypass BPD/DS
EBWL 55.9 46.2 55.5 59.7 63.6
Resolved Overall 78.1 56.7 79.7 80.3 95.1
Resolved lt2 yrs 80.3 55.0 81.4 81.6 94.0
Resolved ?2 yrs 74.6 58.3 77.5 70.9 95.9
36
DIABETES METAANALYSIS

• STRENGTHS OF THE STUDY
• The data set is global, comprehensive, and as
  inclusive as possible, limited only by a
  predetermined time span.
• The selection criteria are independent of
  outcomes and, therefore, to a large extent,
  eliminate selection bias.
• The derived metabolic data are weighted by the
  number of study patients, as well as by the
  variability among studies.

37
DIABETES METAANALYSIS

• WEAKNESSES OF THE STUDY
• It is impossible to extrapolate the effects on
  outcomes of missing unreported data from good
  and from bad series.
• There is limited reporting of longer-term
  diabetes data.

38
DIABETES METAANALYSIS

• SUMMARY
• Bariatric surgery causes resolution of type 2
  diabetes
• 78.1 in entire metaanalysis
• 79.3 in diabetic patients only cohort.
• Bariatric surgery causes resolution or
  improvement of type 2 diabetes
• 86.6 in entire metaanalysis
• 98.9 in diabetic patients only cohort

39
DIABETES METAANALYSIS

• SUMMARY
• The resolution and resolution or improvement of
  type 2 diabetes is associated with the degree of
  weight reduction achieved and, thereby, the
  bariatric procedure performed
• Adjustable gastric banding lt gastroplasty lt
  gastric bypass lt BPD/DS in entire metaanalysis
• Adjustable gastric banding lt gastric bypass lt
  BPD/DS in diabetic patients only cohort
• Weight and diabetes parameters showed little
  difference at lt2 years and ?2 years thus, these
  effects appear to be lasting.

40
DIABETES METAANALYSIS

• SUMMARY
• Insulin, HgA1c, and fasting glucose values
  declined substantially after bariatric surgery,
  giving laboratory metabolic substantiation to the
  clinical findings.

41
DIABETES METAANALYSIS

• CONCLUSION
• Bariatric surgery is effective therapy for type 2
  diabetes.

42
BARIATRIC SURGERY WORLDWIDE 2008

• Henry Buchwald, MD, PhD
• Professor of Surgery and Biomedical Engineering
• University of Minnesota
• Minneapolis, MN, USA
• Obes Surg 2009 191605-1611

43
RESULTS QUESTION 1
Country
44
RESULTS QUESTION 2
Country
45
RESULTS QUESTION 3
Bariatric surgery procedures worldwide
Lap.laparoscopic VBGvertical banded
gastroplasty RYGBRoux-en-Y gastric
bypass LLlong-limb
46
RESULTS 2008
47
RESULTSTRENDS 2003 to 2008
48
RESULTSTRENDS 2003 TO 2008
49
REGIONAL TREND EUROPE
Number () Operations Change
2003 2008
Total 33,771 66,769 97.7
RYGB 3,744 (11.1) 26,023 (39.0) 595.1
AGB 21,496 (63.7) 28,843 (43.2) 34.2
BPD/DS 2,061 (6.1) 3,270 (4.9) 58.7
SG 0 (0) 4,677 (7.0)
50
TRENDS 2003 TO 2008EUROPE
51
REGIONAL TREND USA/CANADA
Number () Operations Change
2003 2008
Total 103,000 220,000 113.6
RYGB 87,550 (85.0) 112,200 (51.0) 28.2
AGB 9,270 (9.0) 96,800 (44.0) 944.2
BPD/DS 4,635 (4.5) 2,200 (1.0) -52.5
SG 0 (0) 8,800 (4.0)
52
TRENDS 2003 TO 2008USA/CANADA
53
CONCLUSION

• Bariatric surgery worldwide has grown markedly in
  the past 10 years (1998 to 2008)
• Operations/year ?
• 1998 40,000
• 266
• 2003 146,301 761
• 135
• 2008 344,221

54
METABOLIC SURGERY IS THE FUTURE OF SURGERY

• Surgery has evolved from incisional to
  extirpative to reparative and reconstructive and
  now to metabolic.
• We are in the forefront of that evolution.

55
Hypothetical Female Patient With Full-Blown
Metabolic Syndrome

• Medical Therapy (Life-time diet, drugs, behavior
  modification) Outcomes
• Obesity extremely poor
• Type 2 diabetes palliative at best
• Dyslipidemia palliative at best
• Hypertension palliative at best
• Nonalcoholic steatohepatitis extremely poor
• Polycystic Ovary Syndrome extremely poor
• Metabolic/Bariatric Surgery Therapy (one
  operation, one time) Outcomes
• Obesity
• Type 2 diabetes
• Dyslipidemia Excellent
• Hypertension
• Nonalcoholic Steatohepatitis
• Polycystic Ovary Syndrome