How does gastric ~80% full remission of type 2 DM after RYGB - - PDF document

Mechanisms and Use of Bariatric/Metabolic Surgery to Treat Type 2 Diabetes

David E. Cummings, M.D.

University of Washington, VA Puget Sound, Diabetes & Obesity Center of Excellence, Seattle

I am P.I. on the COSMID trial (Comparison of Surgery vs. Medicines for Indian Diabetes), funded by Johnson & Johnson

Disclosure

Roux-en-Y Gastric Bypass (RYGB) Gastric Banding

How does gastric bypass cause type 2 diabetes remission?

Gastric Bypass Reverses Diabetes

• ~80% full remission of type 2 DM after RYGB

– Buchwald meta-analysis 22,094 patients – Schauer, et al. 1,160 patients – Whitgrove, et al. 1,029 patients – Pories, et al. 608 patients – Buchwald meta-analysis 2 135,246 patients – Many others

DE Cummings

How does it work?

Swedish Obese Subjects Study

Years of Follow Up Total Body Weight Loss (%)

Gastric Bypass Gastroplasty Banding Usual Care

Commentary: Cummings DE et al. Nature Med 18:358 (2012) Sjöström L, et al. JAMA 307:56 (2012)

↑ insulin sensitivity with

weight loss undoubtedly plays an important role

Rapid Resolution of Diabetes After RYGB

• Prospective study of

1160 RYGB patients

• 240 with DM on oral

meds and/or insulin, 80% F/U

• 83% DM resolution

5 10 15 20 25 30 35 Overall <5 6-10 >10

% Diabetes Resolved Upon Initial Hospital Discharge

Duration of DM Schauer PR et al, Ann Surg 238:467

• Fast kinetics of diabetes resolution
• Glucose homeostasis improves more with RYGB

than with equal weight loss from other means Poor correlation between amount of weight lost and DM remission rates after RYGB

Evidence for Weight-Independent Anti-DM Effects

Cummings DE, Rubino F. Nature Med (in press)

Bypass Band % Weight Loss Time (months)

Percentage Weight Loss Percentage With Diabetes

Time (months)

Long-Term Follow-Up of Gastric Bypass vs. Gastric Banding

Bypass Band % With Diabetes

• C. le Roux, et al

Ann Surg 252:966 (2010)

Better Improvement in Diabetes After RYGB Than After Equivalent Weight Loss from Other Means

• RYGB vs. LAGB

– LeRoux C, et al; Pattou F, et al

• RYGB vs. Dietary Weight Loss

– LaFerrere B, et al; Meirelles K, et al

• Sleeve Gastrectomy with Proximal Intestinal Bypass vs.

Sleeve Gastrectomy Alone

– Lee WJ, et al

DE Cummings

• Fast kinetics of diabetes resolution
• Glucose homeostasis improves more with RYGB

than with equal weight loss from other means

• Inconsistent correlation after RYGB between

baseline weight or amount of weight loss and rates of DM remission, prevention, & recurrence, as well as hard clinical outcomes

Evidence for Weight-Independent Anti-DM Effects

Cummings DE, Rubino F. Nature Med (in press)

A Long-Term Study of RYGB to Treat Type 2 Diabetes in Patients With Only Mild Obesity

Cohen RV….Cummings DE. Diabetes Care 35:1420 (2012)

Prospective Study of RYGB for Type 2 DM in Patients With BMI 30–35 kg/m2

• 66 patients with diabetes

– 100% F/U up to 6 years

• BMI 30–35 kg/m2

– Mild obesity for this population

• Type 2 DM

– Confirmed with Abs, C-peptide, FHx

• Severe diabetes

– Mean duration: 13 years – 40% on insulin (the rest on oral DM meds) – HbA1c: 9.7% at start

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012) A 5 6 7 8 9 10 11 0.5 1 2 4 5 6

Years After Surgery Hemoglobin A1c (%)

Rapid & Durable Improvement in HbA1c After RYGB in BMI 30-35

100% F/U to 6 yrs

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012) 70 60 50 40 30 20 10

Number Of Patients 1% 11% 88%

Diabetes Remission in 6-Year Study of RYGB For Type 2 DM in Patients With BMI 30–35 kg/m2

No Change T2DM Improvement T2DM Remission

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

Resolved

80 90 100 110 120 130 140 6 12 24 48 60 72

Waist Circumference (cm)

80 90 100 110 120 130 140 6 12 24 48 60 72

Months After Surgery

Waist Circumference (cm)

• 50
• 45
• 40
• 35
• 30
• 25
• 20
• 15
• 10
• 5

6 12 24 48 60 72

Total Body Weight Loss (%)

• 50
• 45
• 40
• 35
• 30
• 25
• 20
• 15
• 10
• 5

6 12 24 48 60 72

Months After Surgery

Total Body Weight Loss (%)

Improved

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

A 5 6 7 8 9 10 11 0.5 1 2 4 5 6

Years After Surgery Hemoglobin A1c (%)

Do Changes in BMI Correlate With Changes in Glycemia?

∆ BMI vs. ∆ Glycemia Regression Analyses

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012) 1 2 3 4 5 6 7 2.0 4.0 6.0 8.0 10.0 12.0

No Relationship Between Change in BMI and Change in HbA1c at 6 Months

Change In HbA1c Change in BMI (kg/m2)

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

No Relationship Between Change in BMI and Change in FPG at 6 Months

Change in BMI (kg/m2)

• 40
• 20

20 40 60 80 100 120 140 2.0 4.0 6.0 8.0 10.0 12.0

Change In Fasting Plasma Glucose

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

No Relationship Between Change in BMI and Change in HbA1c at 1 Year

Change In HbA1c Change in BMI (kg/m2)

1 2 3 4 5 6 7 7.0 9.0 11.0 13.0 15.0

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

No Relationship Between Change in BMI and Change in FPG at 1 Year

Change In Fasting Plasma Glucose Change in BMI (kg/m2)

• 20

20 40 60 80 100 120 140 160 7.0 9.0 11.0 13.0 15.0

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

3 mo 6 mo 1 year 2 years 4 years 5 years 6 years Correl Coeff 0.065 0.001 0.195 0.097 0.081 0.003 0.016 P‐Value 0.60 0.99 0.12 0.47 0.59 0.99 0.93 Correl Coeff 0.130 0.170 0.136 0.023 0.240 0.541 0.431 P‐Value 0.30 0.17 0.28 0.86 0.10 0.001 0.017

HbA1c Fasting Glu

Lack of Correlation Between Weight Loss and Improved Glycemia

Effects on Insulin Secretion

A 5 6 7 8 9 10 11 0.5 1 2 4 5 6

Years After Surgery Hemoglobin A1c (%)

Tests of Beta-Cell Function Long After RYGB C-Peptide Testing With Standardized Meal

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

Before RYGB After RYGB

50 100 150 200 250

* * * Fasting Glucose (mg/dL)

1 2 3 4 5 6

C-Peptide (ng/mL) Post-Meal

Improved β–Cell Function for Up to 6 Years After RYGB

↑ β-cell

sensitivity to glucose by 446%

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

Effects on Roughly Estimated Insulin Sensitivity

2 4 6 8 10 12 1 2 3 4 5 6

HOMA-IR

Change in Roughly Estimated Insulin Resistance After RYGB

Years After Surgery

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

Effects on Metabolic Syndrome and Predicted CVD Risk

0.5 1 2 4 5 6 1 1 5 1 2 1 2 5 1 3 1 3 5 1 4 70 75 80 85 90

Years After Surgery Systolic Blood Pressure (mmHg) Diastolic Blood Pressure (mmHg)

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

100 120 140 160 180 200 220 0.5 1 2 4 5 6 70 90 110 130 150 170 190 210 230 60 70 80 90 100 110 120 130 140 160 20 25 30 35 40 45 50 55

Triglycerides (mg/dL) Low Density Lipoprotein (mg/dL) Years After Surgery Total Cholesterol (mg/dL) High Density Lipoprotein (mg/dL) Years After Surgery

0.5 1 2 4 5 6

Cohen RV….. Cummings DE Diabetes Care 35:1420 (2012)

10-Year Cardiovascular Risk Before vs. After RYGB

Cardio- Vascular Event Pre-Surgery (n=66) Mean Risk (%) ± SD Post-Surgery (n=66) Mean Risk (%) ± SD Absolute Risk Reduction 95% Confidence Interval

Relative Risk Reduction

P Value

CHD

35.3 ± 10.0 10.3 ± 2.6 25% 8.2–13.3

71%

0.001 Fatal CHD

26.2 ± 8.1 5.4 ± 1.9 21% 3.7–8.0

84%

0.001 Stroke

5.0 ± 0.4 2.5± 1.7 2.5% 1.7–6.0

50%

0.01 Fatal Stroke

0.7 ± 0.3 0.4 ± 0.2 0.3% 0.03–0.1

57%

0.009

Cohen RV….Cummings DE. Diabetes Care 35:1420 (2012)

Chiu M et al. Diabetes Care 34:1741, 2011

Body Mass Index (kg/m2) Diabetes Incidence per 1000 person-years

Asian Indians Have Increased Diabetes Risk at Lower BMI Levels

Asian Indian Chinese White Black

2,555 cases

WJ Lei, et al. J Gastrointest Surg 12:945

BMI

14 16 18 20 22 24 26 28 30 32 34 36 38 40 42

350 300 250 200 150 100 50

Number

>35 <2% 30–35 9% 27–30 17%

BMI & Diabetes in Taiwan NTUH

Prospective Study of RYGB for Type 2 DM in Asian Indians With BMI < 35 kg/m2

• BMI 22–35 kg/m2

– “Overweight” to “Obese” by Indian-specific WHO criteria

• Type 2 DM

– Confirmed with Abs, C-peptide, FHx

• Severe diabetes

– Mean duration: 9 years – 80% on insulin (the rest on oral DM meds) – HbA1c: 10.1%

• Other features

– Dyslipidemia: 93% – Hypertension: 60%

Shah S, Todkar J….. Cummings DE SOARD 6:332

Fasting Blood Glucose (mg/dL) Months After Surgery

50 100 150 200 250 300

1 3 6 9

80% on insulin

Gastric Bypass in Asian Indians With Type 2 DM & BMI <35 kg/m2

233

Shah S, Todkar J….. Cummings DE SOARD 6:322

Fasting Blood Glucose (mg/dL) Months After Surgery

50 100 150 200 250 300

1 3 6 9

100% 100% 100% 0% 80% % Off All DM Meds

Gastric Bypass in Asian Indians With Type 2 DM & BMI <35 kg/m2

233 89

Shah S, Todkar J….. Cummings DE SOARD 6:322 2 4 6 8 10 12 14

HbA1c (%) Months After Surgery

1 3 6 9

10.1

Gastric Bypass in Asian Indians With Type 2 DM & BMI <35 kg/m2

6.1

100% 100% 100% 0% 80% % Off All DM Meds

Shah S, Todkar J….. Cummings DE SOARD 6:332

Fasting Blood Glucose (mg/dL) Months After Surgery

50 100 150 200 250 300

1 3 6 9

Gastric Bypass in Asian Indians With Type 2 DM & BMI <35 kg/m2

9 12 24 36

100% 100% 100% 0% 80% % Off All DM Meds 100% 100% 100%

No correlations between amount of weight loss and improved glycemia

Shah S….. Cummings DE

UKPDS Risk Engine 10-Year Cardiovascular Risk Predictions (%)

Pre-Op Post-Op

2 4 6 8 10 12 14 16

Coronary Heart Disease

*

Fatal Coronary Heart Disease

2 4 6 8 10

*

Stroke

0.5 1 1.5 2 2.5 3 3.5 4

*

Fatal Stroke

0.1 0.2 0.3 0.4 0.5 0.6

*

Shah, Todkar Cummings SOARD 6:332

COSMID

Randomized Controlled Trial

Comparison Of Surgery vs. Medicines for Indian Diabetes

Shah SS, Todkar J, Kim K, Cummings DE

What about the relationship between body weight and diabetes prevention?

Surgery reduces incidence

• f new diabetes by ~80%

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45

Kaplan-Meier Cumulative Incidence

2 10 15

Follow-Up Time (years)

1,658 1,561 1,225 576 Surgery 1,771 1,513 1,076 404 Control Number at Risk

Cumulative Diabetes Incidence Over 15 Years in SOS Surgery & Control Participants Without Diabetes at Baseline

Carlsson LM…Sjöström L. New Engl J Med 367:695 (2012)

Control (392 events) Surgery (110 events)

Commentary: Rubino F & Cummings DE. Nature Rev Endo 2012

What baseline patient characteristics best predicted surgical prevention of diabetes?

Baseline Values (deciles)

20 40 60 80 100 120 140

Incidence rate per 1000 p-y (95% CI)

Serum Insulin (mU/L)

Control Surgery

20 40 60 80 100 120 140

Incidence rate per 1000 p-y (95% CI)

BMI (kg/m2)

20 40 60 80 100 120 140

Incidence rate per 1000 p-y (95% CI)

Blood Glucose (mg/dL)

Interactions Between Indicated Baseline Risk Factors and Treatment

Interaction P–Value Glucose IFG Yes/No 0.007 0.002 Interaction P–Value BMI 0.545 Interaction P–Value Insulin 0.007

Carlsson LM…Sjöström L. New Engl J Med 367:695 (2012) Commentary: Rubino F & Cummings DE. Nature Rev Endo 2012

What about the relationship between weight or ∆ weight and diabetes recurrence after initial remission?

• 4,434 patients with T2DM undergoing RYGB

– Analyzed over 13 years (retrospectively)

• Initial Remission (within 5 yrs)

– 77% off DM meds & A1c <6.5

Long-term Remission & Relapse of T2DM after RYGB

Arterburn DE et al. Obes Surg 23:93 (2013)

Recurrence of T2DM After Initial Remission

35% relapse at 5 years

Arterburn DE et al. Obes Surg 23:93 (2013)

Median duration of remission = 8.3 years

% of Initial Remitters Still in Remission Years Since Initial Remission

Relapse Not Clearly Due to Weight Regain

Arterburn DE et al. Obes Surg 23:93 (2013)

No Remission of T2DM Lowess Smoothed BMI (kg/m2) Years Since Surgery

Pre-op BMI did not predict remission or relapse

Recurrence of T2DM Prolonged Remission

RYGB

What about the relationship between body weight and hard cardiovascular

• utcomes after surgery?

Sjostrom L, et al. JAMA 307:56 (2012)

SOS 20-Year Data

Reductions in Fatal and Total Heart Attacks and Strokes

Commentary: Cummings DE, et al. Nature Med 18:358 (2012)

SOS 20-Years: Predictors of Surgical Benefit on CV Events

Baseline P Value for Baseline P Value for Feature Surgical Benefit Feature Surgical Benefit BMI .58 Gender .92 Body weight .96 Age .76 Waist–hip .73 Systolic bp .31 Waist circum .86 Diastolic bp .71 Hip circum .38 Diabetes .20 Smoking .10 Previous CVD .71 TG .93 SCORE .86 HDL .26 Metabolic synd .73 Cholesterol .28 Glucose .13 ApoB/ApoA-1 .23 Insulin <.001

Sjostrom L, et al. JAMA 307:56 (2012) Commentary: Cummings DE, et al. Nature Med 18:358 (2012)

!!

Inadequacy of BMI as a Stand-Alone Surgical Criterion SOS has shown that BMI does not predict the benefits from surgery with respect to: – Diabetes prevention – Cancer – Myocardial infarctions – Strokes – Death

SOS: NEJM 2007, JAMA 2012, NEJM 2012

….but baseline fasting insulin and/or glucose levels (reflecting insulin resistance) predict almost all of these benefits!

The benefits of surgery seem more related to effects

• n glucose homeostasis

than on body weight.

• Fast kinetics of diabetes resolution
• Glucose homeostasis improves more with RYGB

than with equal weight loss from other means

• Inconsistent correlation between amount of

weight lost and DM remission rates after RYGB

• Some intestinal bypass operations improve

diabetes with little or no weight loss

Evidence for Weight-Independent Anti-DM Effects

Cummings DE, Rubino F. Nature Med (in press)

Novel Anti-Diabetic GI Procedures

Duodenal-Jejunal Bypass Duodenal-Jejunal Bypass Sleeve Ileal Interposition

• Fast kinetics of diabetes resolution
• Glucose homeostasis improves more with RYGB

than with equal weight loss from other means

• Inconsistent correlation between amount of

weight lost and DM remission rates after RYGB

• Some intestinal bypass operations improve

diabetes with little or no weight loss

• Hints from hyperinsulinemia

Evidence for Weight-Independent Anti-DM Effects

Cummings DE, Rubino F Nature Med (in press) Service et al. NEJM 353:249 (2005)

Control Post-RYGB

Hyperinsulinemia Hypoglycemia After Gastric Bypass: Too much of a good thing for islets? Late onset: 1-26 years (typical 2-4)

Commentary: Cummings DE NEJM 353:300 (2005)

Ossabaw Pig

Cummings, Flum, Sturek

0.5 1 1.5 2 2.5 3 3.5 4

Farm Pigs RYGB Sham GJ Surgery GJD Surgery Ossabaw Controls

Average Beta‐Cell Area per Total Pancreas Area

RYGB

*

Insulin‐Positive Area per Pancreas Section Area (%)

Flum DR, Hull RL, Cummings DE

What Causes RYGB’s Weight-Independent Anti-Diabetes Effects? Changes in Gut Hormones?

Cummings DE, Overduin J J Clin Invest 2007

Ghrelin

• Peptide hormone produced

primarily by stomach & proximal small intestine

• Powerfully stimulates appetite

and food intake in many species, including humans

Ghrelin

GHRELIN

 GH  ACTH & Cortisol  Epinephrine  Glucagon?  Adiponectin  Insulin Action  Insulin Secretion  Food Intake

 GLUCOSE

Counter-regulatory

Cummings DE

Human Plasma Ghrelin Levels Rise & Fall Shortly Before & After Every Meal

300 400 500 600 700 800 900

Plasma Ghrelin (pg/ml)

0600 0700 0800 0900 1000 1100 1200 1300 1400

Time of Day

B D L n = 10 Cummings, et al. Diabetes 50:1714

Plasma Ghrelin Increases After Diet-Induced Weight Loss

300 400 500 600

Plasma Ghrelin (pg/ml)

0600 0800 1000 1200 1400

Clock Time

Before Wgt Loss After Wgt Loss

L B D

n = 13 Cummings, et al. NEJM 346:1623

1600 1800 2000 2200 2400 0200 0400

Roux-en-Y Gastric Bypass

Ingested food bypasses most of the ghrelin-producing cells

Matched Obese Cummings DE, et al. NEJM 346:1623 Normal Weight

Gastric Bypass

100 200 300 400 500 600 700 800

Plasma Ghrelin (pg/ml)

0600 0800 1000 1200 1400

Clock Time

Gastric Bypass Wgt-Reduced Obese Normal Weight Detection Limit

L B D ↓ Pro-Diabetic Ghrelin Levels

Increased GLP-1 Secretion

STOMACH Ghrelin Leptin GRP, NMB PANCREAS Amylin Enterostatin Glucagon Insulin PP DUODENUM CCK, GIP ILEUM

* GLP-1 *

Oxyntomodulin PYY COLON

* GLP-1 *

Oxyntomodulin PYY JEJUNUM APOAIV

FOOD

Cummings DE & Overduin J J Clin Invest 117:13

↑ insulin secretion ↓ food intake

PANCREAS Amylin Enterostatin Glucagon Insulin PP DUODENUM CCK, GIP JEJUNUM APOAIV

FOOD

Cummings DE & Overduin J J Clin Invest 117:13

ILEUM

* GLP-1 *

Oxyntomodulin PYY COLON

* GLP-1 *

Oxyntomodulin PYY

↑ insulin secretion ↓ food intake

STOMACH Ghrelin Leptin GRP, NMB

GLP-1 50 100 150 200

• 20

20 40 60 80 10 12 14 16 18 20 Time from test meal (minutes Pre-op avg Post-op avg GLP-1 After Test Meal 500 1000 1500 2000 2500 50 100 150 200 Time from test meal (minutes pre-op avg post-op avg

Plasma GLP-1 Increases After RYGB in Humans

GLP-1 After Test Meal GLP-1 After Test Meal

Plasma GLP-1 (pM) Plasma GLP-1 (% change from Baseline)

Carlson M, Heap A, Cummings DE

 GLP-1 After RYGB

• Released in correct location to engage

neural (e.g., vagal) pathways to improve glucose homeostasis

• Associated with increased incretin effect
• Occurs immediately, lasts for years

DE Cummings

5 10 15 20 25 30 35 40 45 50 ‐15 15 30 45 60 75 90 105 120

Baseline 14 Days 60 Days

Plasma GLP-1 Increases After RYGB in Ossabaw Pigs

Plasma GLP-1 (pmol/L)

RYGB

Time from Test Meal (minutes) Baseline 14 Days 60 Days

* **

Flum DR, Cummings DE 2 4 6 8 10 12

Surgery(n=4) Necropsy(n=7)

*

GLP-1 Cells per High-Power Filed of Mucosa Before Surgery After Surgery

RYGB

Flum DR, Hull RL, Cummings DE

↑ GLP-1 Cells in Distal Ileum After RYGB in Ossabaw Pigs

Duodenal (Jejunal) Bypass: DJB

5 10 15 20 25 30 35 40 45 50 ‐15 15 30 45 60 75 90 105 120

Baseline 14 Days 60 Days

No Change in GLP-1 After DJB in Ossabaw Pigs

DJB

Flum DR, Cummings DE

Baseline 14 Days 60 Days Plasma GLP-1 (pmol/L) Time from Test Meal (minutes)

Gastro-Jejunostomy: GJ

5 10 15 20 25 30 35 40 45 50 ‐15 15 30 45 60 75 90 105 120

Baseline 14 Days 60 Days

Baseline 14 Days 60 Days Plasma GLP-1 (pmol/L) Time from Test Meal (minutes)

No Change in GLP-1 After Gastrojejunostomy in Ossabaw Pigs

Flum DR, Cummings DE

GJ

50 100 150 200 250 300 350 400 450 500 ‐15 15 30 45 60 75 90 105 120

Baseline 14 Days 60 Days

Plasma GLP-1 (pmol/L) Time from Test Meal (minutes) Baseline 14 Days 60 Days

* *

Plasma PYY Increases After RYGB in Ossabaw Pigs

Flum DR, Cummings DE

RYGB

50 100 150 200 250 300 350 400 450 500 ‐15 15 30 45 60 75 90 105 120

Baseline 14 Days 60 Days

Baseline 14 Days 60 Days Plasma GLP-1 (pmol/L) Time from Test Meal (minutes)

No Change in PYY After DJB in Ossabaw Pigs

Flum DR, Cummings DE

DJB

50 100 150 200 250 300 350 400 450 500 ‐15 15 30 45 60 75 90 105 120

Plasma PYY During Test Meal GJ

Baseline 14 Days 60 Days

Baseline 14 Days 60 Days

No Change in PYY After Gastrojejunostomy in Ossabaw Pigs

Plasma GLP-1 (pmol/L) Time from Test Meal (minutes)

Flum DR, Cummings DE

GJ

ILEUM

* GLP-1 *

PYY Oxyntomodulin COLON

* GLP-1 *

PYY Oxyntomodulin

All  after RYGB FOOD

↑ insulin

secretion

Cummings DE & Overduin J J Clin Invest 117:13

RYGB

Ileal Interposition

10-cm distal transection Vascularly intact Innervated Isoperistaltic

 GLP-1  insulin secretion

RYGB

 body weight  muscle insulin sensitivity

TIME

A Simple Model

Gastric Band

 body weight  muscle insulin sensitivity

TIME

Cummings DE, Rubino F Nature Medicine (in press 2014)

ILEUM

* GLP-1 *

PYY Oxyntomodulin COLON

* GLP-1 *

PYY Oxyntomodulin

All  after RYGB FOOD

↑ insulin

secretion

Cummings DE & Overduin J J Clin Invest 117:13

Does this explain everything?

RYGB

No. GLP‐1 loss‐of‐function experiments in humans & rodents show GLP‐1 only accounts for some

• f the glycemic effects of RYGB.

Salehi M, et al. Diabetes (2011) Kaplan LM, et al. (2012) Planas J, et al. (2012)

Hyperinsulinemic Clamp Studies Early After GI Surgery

• Biliopancreatic Diversion

– Ferrannini & Mingrone 2005 & 2006: Fast ↑ Si is not explained by weight loss, whereas slower ↑ insulin sensitivity after RYGB is explained by weight loss

• Gastric Bypass

– Campos 2009: No Δ Si at 2 weeks c/w caloric restriction – Geloneze 2010: No Δ Si at 1 month – Mingrone 2013: ↑ Si at 1 mo in DM and non‐DM – Kashyap & Schauer 2010: ↑ Si at 1 & 4 weeks – Cummings & Flum 2013: ↑ Si at 2 weeks?) No DM in RYGB groups No DM Mostly No DM All DM

DE Cummings

DM & non-DM

Upper Intestinal Nutrient Exclusion

Duodenal (Jejunal) Bypass

Rubino F Cummings DE

• No gastric restriction
• No calorie malabsorption
• No change in food intake
• No change in body weight

Duodenal (Jejunal) Bypass

F Rubino, G Mingrone, S Hu, C Otto, J Liu, D Pacheco, Y Wang, M Speck, Kindel & Tso, others

Major, durable ↑ in glucose tolerance with little or no weight loss in several rat DM models (but not in non-DM rats)

Does duodenal bypass ameliorate type 2 diabetes in humans?

Duodenal-Jejunal Bypass in Humans

(“Gastric-Sparing Gastric Bypass”)

~1.5 ft ~2.4 ft ~12–18 ft

Cohen RV, et al.

Prospective Study of DJB for Type 2 DM in Patients With BMI < 35 kg/m2

• 46 patients
• BMI 22–35 kg/m2
• Severe Diabetes

– Confirmed type 2: negative ‐GAD, ICA – Duration of DM: 2–10 years – 70% on insulin; others on oral DM meds – Mean HbA1c = 8.9%

Cohen RV, Cummings DE, et al.

Duodenal-Jejunal Bypass in 46 Patients with DM & BMI <35 kg/m2

2 4 6 8 10 12 3 6 12

Months After Surgery HbA1c (%) 8.9 6.5

0% on insulin Rx

Cohen RV, Cummings DE, et al.

70% on insulin Rx

No Relationship Between Change in Body Weight and Improvement in Glycemia

Change in BMI (kg/m2) Change in HbA1c

HbA1c Data at 1 year

Cohen RV, Cummings DE, et al.

Duodenal‐Jejunal Bypass (DJB)

F Rubino, G Mingrone, S Hu, C Otto, J Liu, D Pacheco, Y Wang, M Speck, Kindel & Tso R Cohen, A Ramos, G Ferzli, B Geloneze, & many others

Substantially improves glucose homeostasis with little or no weight loss in several rat DM models (but not in non-DM rats) and in humans

Can the same thing be accomplished with an endoscopically implantable device?

Duodenal‐Jejunal Bypass Sleeve

Substantially improves glucose homeostasis in several rat models

• f type 2 DM and in

humans, before &

• ut of proportion to

weight loss

GI Dynamics, L Kaplan, F Rubino, R Cohen, others

Food bypasses the duodenum & proximal jejunum, as it does in RYGB

Lower Intestinal vs. Upper Intestinal Hypothesis?

Duodenal (Jejunal) Exclusion Duodenal (Jejunal) Exclusion Gastro-jejunal Anastamosis

Rubino F, Cummings DE, et al., Ann Surg 244:741

OGTT GK rats

100 200 300 400 500 600 50 100 150 200 Time (min) G lu co se le vels (m g /d l)

GK DJB GK Sham GK GJ

Oral Glucose Tolerance

Rubino, Cummings, et al. Ann Surg 244:741

1

Hyperglycemic Normoglycemic?

2 weeks

Normal Oral Feeding

Hyperglycemic?

2 weeks

Gastric Feeding With G-tube 4

Normoglycemic?

2 weeks

RYGB

Human RYGB Mechanisms Study

2 3 Normal Oral Feeding Patients: severely obese type 2 diabetics Tests: meal tolerance tests FS-IVGTT hyperinsulinemic clamps (with tracers)

Gastrostomy Tube

Remove DM Meds

DE Cummings & DR Flum 0.05 0.1 0.15 0.20 0.25

Insulin Sensitivity by Clamp (M/I)

Insulin Sensitivity by Hyperinsulinemic/Euglycemic Clamp

Baseline Duodenal Passage Duodenal Exclusion 2

*

DE Cummings & DR Flum

Duodenal Exclusion 1

*

100 200 250 300 350 400 450 ‐30 ‐15 15 30 45 60 75 90 105 120 135 150 165 180

Time (minutes) Insulin (µU/mL)

Duod Passage Duod Excl 2

Meal Tolerance Test: Insulin

Baseline Duod Excl 1

Meal

DE Cummings & DR Flum

30 15 15 30 45 60 75 90 105 120 135 150 165 180

Time (minutes) Glucose (mg/dL)

Meal Tolerance Test: Glucose

Duod Passage

70 130 140 150 160 170 180 120 100 80

Baseline Duod Excl 1 Duod Excl 2

Meal

DE Cummings & DR Flum

Baseline Duod Exclusion 1

• 30

30 60

Insulin µU/mL

25 50 75 100 125 150 175 200 225 250 275 300

Duod Passage Duod Exclusion 2

Time (minutes)

Acute Insulin Response to Glucose (IVGTT)

DE Cummings & DR Flum

1

Hyperglycemic Normoglycemic?

2 weeks

Normal Oral Feeding

Hyperglycemic?

2 weeks

Gastric Feeding With G-tube 4

Normoglycemic?

2 weeks

RYGB

Human RYGB Mechanisms Study

2 3 Normal Oral Feeding Patients: severely obese type 2 diabetics Tests: meal tolerance tests FS-IVGTT hyperinsulinemic clamps (with tracers)

Gastrostomy Tube

Remove DM Meds

DE Cummings & DR Flum

Duodenal nutrient passage strongly influences insulin sensitivity, independent

• f weight change.

Novel Roles of the Gut to Regulate Insulin Sensitivity

Thaler & Cummings Nature 452:941 Relating to: Wang PY et al. Nature 452:1012

Nutrient sensing in the gut regulates food intake

Thaler & Cummings Nature 452:941 Relating to: Wang PY et al. Nature 452:1012

Nutrient sensing in the gut regulates insulin secretion

Nutrient sensing in the gut regulates insulin secretion & sensitivity

Thaler & Cummings Nature 452:941 Relating to: Wang PY et al. Nature 452:1012 Relating to: Wang PY et al. Nature 452:1012, 2008 Thaler & Cummings Nature 452:941

Drawing by Nathan Cummings 5th grade

Salinari et al PLOS ONE (Epub 2013) Glucose (mg/dL)

In vivo Insulin Tolerance Test: db/db Proteins Injected

Glucose (mg/dL)

In vivo Insulin Tolerance Test: Wild Type Proteins Injected In vitro Soleus Muscle Glucose Uptake In vitro Myoblast Glucose Uptake

Ser473-PO4

mTORC2 TSC1/TSC2

PP242

Intestinal Proteins

IRS-1 PI3K

Akt

TSC1/TSC2 Rheb/GTP mTORC1 S6K1

Thr308-PO4 Insulin

GSK3

Salinari S….Mingrone G. PLOS ONE (Epub 2013)

RYGB Mechanisms Model

 body weight  muscle insulin sensitivity  GLP-1  insulin secretion

RYGB

TIME

 body weight  muscle insulin sensitivity

Gastric Band

 intestinal nutrient sensing & metab  insulin sensitivity

Cummings DE, Rubino F Nature Medicine (in press 2014)

Weight‐Independent Anti‐Diabetes Candidate Mechanisms of RYGB

 GLP‐1 (& PYY & OXM)  Ghrelin  Amylin Duodenal factor Intestinal LCFACoA & CCK Intestinal Gluconeogenesis  Bile acids  Ceramides ∆ in Gut Microbiome  Inflamm & oxidative stress  Branched Chain AA in blood Reprogram intest glu metab

Cummings DE, Rubino F. Nature Med (in press 2014)

Where does metabolic surgery fit in diabetes care (including less obese T2DM)?

NIH Consensus Development Panel Criteria for Bariatric Surgery (1991)

BMI > 40

• r

BMI > 35 with diabetes

Surgery to treat DM in less obese patients makes sense if it improves DM through weight- independent mechanisms.

Diabetes Surgery Summit

Rome 2007

Diabetes Surgery Summit Conclusions

• Gastric bypass improves diabetes via mechanisms

beyond reduced food intake & body weight

• Gastric banding improves diabetes only via its effects
• n food intake and body weight

Rubino, Schauer, Kaplan, & Cummings, Ann Surg 2010 & Ann Rev Med 2010

• American Society for Metabolic & Bariatric Surgery
• Brazilian Society for Bariatric & Metabolic Surgery
• Italian Society for Surgery of Obesity & Metabolic Diseases
• Venezuelan Society of Bariatric & Metabolic Surgery
• Obesity & Metabolic Surgery Society of India
• International Federation for the Surgery of Obesity &

Metabolic Disorders

• Asia‐Pacific Metabolic & Bariatric Surgery Society

Societies That Changed Their Names Soon After 2007 to Include “Metabolic Surgery”

Diabetes Surgery Summit Conclusions

• Gastric bypass improves diabetes via mechanisms

beyond reduced food intake & body weight

• Gastric banding improves diabetes only via its effects
• n food intake and body weight
• Gastric bypass should be considered to treat type 2

diabetes in patients with BMI ≥ 30 kg/m2

Rubino, Schauer, Kaplan, & Cummings, Ann Surg 2010 & Ann Rev Med 2010

Distribution of T2DM According to BMI

Bays et al. Int J Clinical Prac 61:737 (2007)

Thin Normal Overweight Obese I Obese II III

>50% of patients with diabetes worldwide have BMI <35 kg/m2

Management Algorithm for Metabolic Control in Type 2 Diabetes

Basal Premixed Basal Bolus insulin Sulphonylurea Acarbose DPP‐4 inhibitor Glitazone Insulin Lifestyle Modification

• diet modification
• weight control
• physical activity

Metformin Bariatric Surgery BMI > 30 eligible & BMI > 35 prioritized *If HbA1c >7.5% despite

• ptimized conventional

therapy, especially if weight is increasing, or if other weight responsive comorbidities are not reaching target on conventional therapy. Bariatric Surgery BMI > 35 eligible BMI > 40 prioritised

Bariatric Surgical and Procedural Interventions in the Treatment of Obese Patients with Type 2 Diabetes

ADA‐EASD Consensus Algorithm for T2DM Treatment

Nathan DM, et al. Diabetes Care 32:193, 2009

Bariatric surgery not mentioned for anyone!

Main Problem

Relative paucity of highest quality, long‐ term outcome date from surgical compared with medical DM care

But are RCTs of surgical vs. non‐surgical care feasible?

It’s very hard to find educated patients in true equipoise about surgical vs. non-surgical options

Typical Enrollment Strategies

• Approach patients in a bariatric surgery clinic

– Likely to be biased in favor of surgery

• Approach patients in a medical clinical

– Likely to be biased in favor of non‐surgery

May be okay for observational studies, but problematic for RCTs

Our approach……

CROSSROADS Trial (an RCT)

Calorie Reduction Or Surgery: Seeking Remission for Obesity And Diabetes

Cummings, Flum, Arterburn, et al

CROSSROADS Trial

• RCT of T2DM Rx in BMI 30−40
• Standard RYGB & Medical Care OR
• Intensive Medical−Lifestyle Rx

– Aerobic exercise – Diet (low‐calorie, low‐fat) – State‐of‐the‐art DM Rx per ADA/EASD

Identify Eligible Patients From Databases

• Use administrative & clinical databases
• Meet inclusion criteria

– BMI 30−40 kg/m2 – Type 2 diabetes – Has coverage for both Rx arms

• Have no exclusion criteria

Modified from D. Arterburn

Unique, Validated Shared Decision Making Instrument

Pool of Potentially Eligible Candidates

Randomized Participants

Arterburn D….Cummings DE SOARD 2013 (Epub)

Sent invitation letter to participate in screening call (100%) Non-responders to screening call: (unable to

contact, contacted and refused , or deceased )

(41%) Contacted for screening call (59%) Eligible and interested in SDM call (23%) Ineligible for study (17%) Refused study (2%) Completed SDM call (15%) Non-responders to SDM call:

(unable to contact , contacted and refused)

(8%) At equipoise or prefers surgery and invited to RCT (6.4%) Prefers non-surgical treatment and not invited to RCT (9%) Randomized (2.4%) Ineligible after surgeon’s chart review or refused randomization (4%) Not interested in more info about weight control (17%)

Arterburn D….Cummings DE SOARD 2013 (Epub)

CROSSROADS One‐Year Results

Baseline Participant Characteristics

Surgical Medical–Lifestyle P‐Value

Age (yr) 52.0 (8.3) 54.6 (6.3) 0.4 Female sex 80.0% 58.8% 0.3 White race 80.0% 64.7% 0.4 Body weight (kg) 108.8 (14.9) 112.8 (16.5) 0.6 Height (cm) 168.4 (7.9) 174.1 (10.3) 0.1 BMI (kg/m2) 38.3 (3.7) 37.1 (3.5) 0.3 Waist circumference (cm) 121.7 (10.2) 120.8 (10.0) 0.8 Waist‐to‐hip ratio 1.0 (0.1) 1.0 (0.1) 0.8 Body fat by DEXA (%) 47.6 (5.4) 46.1 (6.4) 0.6 Body fat by BEI (%) 41.4 (6.3) 38.6 (8.2) 0.3 VO2max by ETT 19.6 (2.6) 21.1 (3.6) 0.4 HbA1C (%) 7.7 (1.0) 7.3 (0.9) 0.4 Fasting plasma glucose (mg/dL) 146.1 (47.1) 153.4 (47.7) 0.7 Fasting plasma insulin (uU/ml) 23.0 (14.7) 26.9 (19.5) 0.6 Use of insulin 60.0% 47.1% 0.5

Duration of known diabetes (yr) 11.4 (4.8) 6.8 (5.2) 0.009

Dyslipidemia 86.7% 82.4% 1.0 Cholesterol (mg/dL) Total 167.2 (37.2) 168.9 (31.1) 0.5 LDL 93.1 (28.9) 85.9 (22.7) 0.4 HDL 44.5 (11.5) 42.2 (10.8) 0.6 Triglycerides (mg/dL) 148.2 (59.4) 203.5 (102.0) 0.1 Hypertension 80.0% 94.1% 0.3 Blood pressure (mm Hg) Systolic 129.3 (20.6) 120.1 (9.6) 0.3 Diastolic 77.0 (10.2) 74.8 (7.5) 0.9

Demographics Anthropometrics & Body Composition Fitness Glycemia Lipids Blood Pressure Baseline difference biases against surgical superiority

Primary Endpoint: Diabetes Remission at One Year (HbA1c <6.0% Off All Diabetes Medications)

Percent In Remission

60% 6%

5.5 6.0 6.5 7.0 7.5 8.0 6 12

HbA1c (%) Time (months)

Med‐Lifestyle Surgical

Changes in Glycemia

Significantly greater fall in A1c with surgery

Changes in Glycemia

HbA1c (%)

Med-Lifestyle Surgery Med-Lifestyle Surgery

Fasting Glucose

* **

Changes in Glycemia

Fasting Insulin (µU/mL)

Med-Lifestyle Surgery

** * *

Med-Lifestyle Surgery

HOMA-IR

** * *

Changes in Body Composition

Body Weight (kg)

Med-Lifestyle Surgery

** * *

Med-Lifestyle Surgery

% Fat Mass (DEXA)

** * *

Changes in Lean Body Mass & Fitness

Lean Body Mass (kg, DEXA)

Med-Lifestyle Surgery

** *

2000 2200 2400 2600 2800 3000 3200 Life Surg Med-Lifestyle Surgery

VO2 Max (i.e., Fitness)

** *

Changes in Blood Pressure

Systolic BP (mm Hg)

Med-Lifestyle Surgery Med-Lifestyle Surgery

Diastolic BP (mm Hg)

P = 0.05

*

Changes in Blood Lipids

Total Cholesterol

Med-Lifestyle Surgery Med-Lifestyle Surgery

LDL Cholesterol

Changes in Blood Lipids

HDL Cholesterol

Med-Lifestyle Surgery

* *

P = 0.08 Med-Lifestyle Surgery

Triglycerides

* *

Average # of Metabolic Meds Taken at 1 Year

1 2 3 4 Diabetes HTN Lipids Metab Syndr Med-Lifestyle Surgery

On GERD Meds 30% 0%

CROSSROADS: One‐Year Results

• Surgery Was Better than Meds‐Lifestyle for:

– ↓ HbA1c & diabetes meds usage – Diabetes remission

– ↓ fasting insulin (but not fasting glucose)

– ↓ HOMA‐IR

– ↓ body weight

– ↓ % body fat (by DEXA and BIA) – ↓ systolic blood pressure

• Interventions Were Similar for:

– ↓ diastolic blood pressure – ↓ cholesterol

– ↓ LDL – ↑ HDL (trend for surgical superiority) – ↓ triglycerides

CROSSROADS: One‐Year Results

• No major events in either group
• Substantially more hypoglycemia reported in

the medical group

• One inpatient admission for alcohol abuse in

a surgical patient

Adverse Events

RCT Interventions Compared N Baseline BMI (kg/m2) Follow-Up Main Findings Ikramuddin, et al RYGB Intensive medical-lifestyle care 120 With T2DM 30–40 1 year Achieved composite goal for T2DM, hypertension, and dyslipidemia RYGB: 49% Medical: 19% (O.R. 4.8) Schauer, et al RYGB VSG Intensive medical care 150 With T2DM 27–43 1 year HbA1c <6.0% RYGB: 42% VSG: 37% Medical: 12% Mingrone, et al RYGB BPD Conventional medical care 60 With T2DM ≥35 2 years HbA1c <6.5% off diabetes medications RYGB: 75% (O.R. 7.5 vs. meds) BPD: 95% (O.R. 9.5 vs. meds) Medical: 0% Dixon, et al LAGB Conventional medical care 60 With T2DM 30–40 2 years HbA1c <6.2% off diabetes medications LAGB: 73% Medical: 13% (O.R. 5.5) O’Brien, et al LAGB Supervised lifestyle intervention 50 No T2DM >35 2 years >50% excess body weight loss LAGB: 84% Lifestyle: 12%

Published RCTs of Surgical vs. Medical and/or Lifestyle Interventions for Diabetes & Obesity

Cummings DE, Cohen RV Lancet DM Endo (2014)

5 RCTs

• All 4 ops now in use
• Various med-lifestyles

440 pts BMI down to 27 1–2 yrs

Conclusions from This & Other Recent RCTs

Bariatric/metabolic surgery is more effective than medical-lifestyle interventions for weight loss, glycemic control, DM remission, & improvements in other CVD risk factors, with acceptable complications, for at least 1-2 years, including in patients with a BMI < 35.

Management Algorithm for Metabolic Control in Type 2 Diabetes

Basal Premixed Basal Bolus insulin Sulphonylurea Acarbose DPP‐4 inhibitor Glitazone Insulin Lifestyle Modification

• diet modification
• weight control
• physical activity

Metformin Bariatric Surgery BMI > 30 eligible & BMI > 35 prioritized *If HbA1c >7.5% despite

• ptimized conventional

therapy, especially if weight is increasing, or if other weight responsive comorbidities are not reaching target on conventional therapy. Bariatric Surgery BMI > 35 eligible BMI > 40 prioritised

Bariatric Surgical and Procedural Interventions in the Treatment of Obese Patients with Type 2 Diabetes

• Type 2 Diabetes Patient Population

– SLIMM‐T2D (Goldfine, Lautz, et al.) – Triabetes & Triabetes‐2 (Courcoulas, et al.) – CROSSROADS (Cummings, Flum, et al.) – Stampede II (Schauer, Kirwan, et al.) – IDeaLS (Clark, Brancati, et al.) – SOLID (Sarwer, et al.)

• Obstructive Sleep Apnea Patient Population

– ABC Trial (Patel, et al.)

NIH-Sponsored Ongoing RCTs

Cummings DE, Cohen RV Lancet DM Endo (in press)

• Type 2 Diabetes Patient Population

– SLIMM‐T2D (Goldfine, Lautz, et al.) – Triabetes & Triabetes‐2 (Courcoulas, et al.) – CROSSROADS (Cummings, Flum, et al.) – Stampede II (Schauer, Kirwan, et al.) – IDeaLS (Clark, Brancati, et al.) – SOLID (Sarwer, et al.)

• Obstructive Sleep Apnea Patient Population

– ABC Trial (Patel, et al.)

NIH-Sponsored Ongoing RCTs

Cummings DE, Cohen RV Lancet DM Endo (2014)

Proposed UO1 consortium for long-term F/U

Are we ready for a mega‐RCT of surgery vs. medical‐lifestyle care for T2DM with hard endpoints?

I think so.

Contributors

• U. Washington

– Joost Overduin – Karen Foster‐Schubert – Scott Frayo – Ian Townsend – Dave Flum – Allison Rhodes – Skye Stewart – Leon Chan – Jessica Kuzma – David Arterburn – Emily Westbrook – Jon Purnell – Scott Weigle Funding: NIH/NIDDK

Elsewhere

– Francesco Rubino (Cornell) – Ricardo Cohen (Sao Paulo, Brazil) – Phil Schauer (Cleveland Clinic) – Carel LeRoux (U. London) – Shashank Shah (Pune, India) – Jaya Todkar (Pune, India) – Lars Sjostrom (Sweden) – G.I. Dynamics (Boston) – Jens Holst (U. Copenhagen)